Peer-Reviewed Journal Tracking and Analyzing Disease Trends pages 1557–1716

EDITOR-IN-CHIEF
D. Peter Drotman

Managing Senior Editor EDITORIAL BOARD

Polyxeni Potter, Atlanta, Georgia, USA
Dennis Alexander, Addlestone, Surrey, UK
Associate Editors Timothy Barrett, Atlanta, Georgia, USA
Paul Arguin, Atlanta, Georgia, USA Barry J. Beaty, Ft. Collins, Colorado, USA
Charles Ben Beard, Ft. Collins, Colorado, USA Martin J. Blaser, New York, New York, USA
Ermias Belay, Atlanta, Georgia, USA Christopher Braden, Atlanta, Georgia, USA
David Bell, Atlanta, Georgia, USA Arturo Casadevall, New York, New York, USA
Sharon Bloom, Atlanta, GA, USA Kenneth C. Castro, Atlanta, Georgia, USA
Mary Brandt, Atlanta, Georgia, USA Louisa Chapman, Atlanta, Georgia, USA
Corrie Brown, Athens, Georgia, USA Thomas Cleary, Houston, Texas, USA
Charles H. Calisher, Ft. Collins, Colorado, USA Vincent Deubel, Shanghai, China
Michel Drancourt, Marseille, France Ed Eitzen, Washington, DC, USA
Paul V. Effler, Perth, Australia Daniel Feikin, Baltimore, Maryland, USA
David Freedman, Birmingham, Alabama, USA Anthony Fiore, Atlanta, Georgia, USA
Peter Gerner-Smidt, Atlanta, Georgia, USA Kathleen Gensheimer, Cambridge, Massachusetts, USA
Stephen Hadler, Atlanta, Georgia, USA Duane J. Gubler, Singapore
Nina Marano, Atlanta, Georgia, USA Richard L. Guerrant, Charlottesville, Virginia, USA
Martin I. Meltzer, Atlanta, Georgia, USA Scott Halstead, Arlington, Virginia, USA
David Morens, Bethesda, Maryland, USA David L. Heymann, London, UK
J. Glenn Morris, Gainesville, Florida, USA Charles King, Cleveland, Ohio, USA
Patrice Nordmann, Paris, France Keith Klugman, Atlanta, Georgia, USA
Tanja Popovic, Atlanta, Georgia, USA Takeshi Kurata, Tokyo, Japan
Didier Raoult, Marseille, France S.K. Lam, Kuala Lumpur, Malaysia
Pierre Rollin, Atlanta, Georgia, USA Stuart Levy, Boston, Massachusetts, USA
Ronald M. Rosenberg, Fort Collins, Colorado, USA John S. MacKenzie, Perth, Australia
Dixie E. Snider, Atlanta, Georgia, USA Marian McDonald, Atlanta, Georgia, USA
Frank Sorvillo, Los Angeles, California, USA John E. McGowan, Jr., Atlanta, Georgia, USA
David Walker, Galveston, Texas, USA Tom Marrie, Halifax, Nova Scotia, Canada
J. Todd Weber, Atlanta, Georgia, USA Philip P. Mortimer, London, UK
Fred A. Murphy, Galveston, Texas, USA
Founding Editor
Joseph E. McDade, Rome, Georgia, USA Barbara E. Murray, Houston, Texas, USA
P. Keith Murray, Geelong, Australia
Senior Associate Editor, Emeritus Stephen M. Ostroff, Harrisburg, Pennsylvania, USA
Brian W.J. Mahy, Bury St. Edmunds, Suffolk, UK Richard Platt, Boston, Massachusetts, USA
Copy Editors Claudia Chesley, Karen Foster, Thomas Gryczan, Gabriel Rabinovich, Buenos Aires, Argentina
Jean Michaels Jones, Carol Snarey, P. Lynne Stockton Mario Raviglione, Geneva, Switzerland
David Relman, Palo Alto, California, USA
Production Carrie Huntington, Ann Jordan, Shannon O’Connor, Connie Schmaljohn, Frederick, Maryland, USA
Reginald Tucker
Tom Schwan, Hamilton, Montana, USA
Editorial Assistant Christina Dzikowski Ira Schwartz, Valhalla, New York, USA
Social Media/Communications Sarah Logan Gregory Tom Shinnick, Atlanta, Georgia, USA
Bonnie Smoak, Bethesda, Maryland, USA
Emerging Infectious Diseases is published monthly by the Centers for Disease Rosemary Soave, New York, New York, USA
Control and Prevention, 1600 Clifton Road, Mailstop D61, Atlanta, GA 30333, P. Frederick Sparling, Chapel Hill, North Carolina, USA
USA. Telephone 404-639-1960, fax 404-639-1954, email eideditor@cdc.gov. Robert Swanepoel, Pretoria, South Africa
The opinions expressed by authors contributing to this journal do not neces- Phillip Tarr, St. Louis, Missouri, USA
sarily reflect the opinions of the Centers for Disease Control and Prevention or Timothy Tucker, Cape Town, South Africa
the institutions with which the authors are affiliated. Elaine Tuomanen, Memphis, Tennessee, USA
All material published in Emerging Infectious Diseases is in the public do-
John Ward, Atlanta, Georgia, USA
main and may be used and reprinted without special permission; proper citation,
however, is required. Mary E. Wilson, Cambridge, Massachusetts, USA
Use of trade names is for identification only and does not imply endorsement
∞ Emerging Infectious Diseases is printed on acid-free paper that meets the requirements
by the Public Health Service or by the U.S. Department of Health and Human
of ANSI/NISO 239.48-1992 (Permanence of Paper)
Services.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 October 2012
On the Cover WU and KI Polyomaviruses in
Mori Sosen (1747–1821) Respiratory Samples from
Monkey Performing the Sanbasō Dance Allogeneic Hematopoietic Cell
(Dated 1800, the first day of the Monkey Year) Transplant Recipients ...................... 1580
Scroll painting, ink on paper
(49.5 cm x 115.6 cm)
J. Kuypers et al.
Pacific Asia Museum, Pasadena, California, USA, Routine testing for these viruses in immuno-
Gift of Mr. and Mrs. Bruce Ross compromised patients is not recommended.
www.pacificasiamuseum.org

About the Cover p. 1711 Wild Birds and Urban Ecology

of Ticks and Tick-borne
Pathogens, Chicago, Illinois,
Research 2005–2010.......................................... 1589
S.A. Hamer et al.
The rare introduction but successful
establishment of ticks and pathogens poses a
Methicillin-Resistant significant health risk.
Staphylococcus aureus
Sequence Type 239-III, Spread of Influenza Virus A (H5N1)
Ohio, 2007–2009 ............... 1557 Clade 2.3.2.1 to Bulgaria in
S.-H. Wang et al. Common Buzzards ........................... 1596
Identification of virulent strains emphasizes the A. Marinova-Petkova et al.
need for molecular surveillance. Detection of this highly pathogenic clade in
p. 1578 Europe poses a health threat to both humans
and poultry.

Epidemiology of Dengue Outbreaks in High-Income

Foodborne Norovirus Area, Kaohsiung City, Taiwan,
Outbreaks, United States, 2003–2009.......................................... 1603
2001–2008 ......................... 1566 C.-H. Lin et al.
A.J. Hall et al. Cases are distributed in a clustered pattern, and
Interventions should focus on commercial food p. 1623 elderly persons have the highest risk for illness
handlers and production of commodities eaten and death.
raw.
Nontuberculous Mycobacteria
in Household Plumbing as
Constant Transmission Possible Cause of Chronic
Properties of Variant Rhinosinusitis ................................... 1612
Creutzfeldt-Jakob Disease in W.S. Tichenor et al.
5 Countries ........................................ 1574
Patients with treatment-resistant rhinosinusitis
A.B. Diack et al. should have cultures performed for NTM; also,
Current diagnostic criteria should be sufficient sinuses should not be irrigated with tap water.
to detect new cases.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 i
Autochthonous and Dormant
Cryptococcus gattii Infections
in Europe ........................................... 1618 October 2012
F. Hagen et al.
Dormant infections can be reactivated many
1662 Visceral Leishmaniasis in
years after having been acquired on another Rural Bihar, India
continent. E. Hasker et al.

1665 Circulation of Influenza A(H1N1)

Dispatches pdm09 Virus in Pigs, Réunion
1625 Echinococcus multilocularis in Island
Urban Coyotes, Alberta, Canada E. Cardinale et al.
S. Catalano et al.
1669 Powassan Virus Encephalitis,
1629 Orthobunyavirus Antibodies in Minnesota
p. 1648 J. Birge and S. Sonnesyn
Humans, Yucatan Peninsula,
Mexico
B.J. Blitvich et al. 1672 Influenza Virus Infection in
Nonhuman Primates
1633 Tetanus as Cause of Mass Die-off E.A. Karlsson et al.
of Captive Japanese Macaques,
Japan, 2008 1676 Human Polyomaviruses
T. Nakano et al. in Children Undergoing
Transplantation, United States,
1636 Human Infection with Candidatus 2008–2010
Neoehrlichia mikurensis, China E.A. Siebrasse et al.
H. Li et al.
1680 Preventing Maritime Transfer of
1640 Anthroponotic Enteric Parasites Toxigenic Vibrio cholerae
in Monkeys in Public Park, China N.J. Cohen et al.
J. Ye et al.

p. 1669
Another Dimension
1644 Schmallenberg Virus as Possible
Ancestor of Shamonda Virus 1684 A Natural History of Infective
K.V. Goller et al. Endocarditis, Preceded by
Decompensated Chronic Liver
1647 Monkey Bites among US Military Disease and Severe Community-
Members, Afghanistan, 2011 acquired Pneumonia
L.E. Mease and K.A. Baker N.L. Merridew

1650 Human Parvovirus 4 in Nasal and Letters

Fecal Specimens from Children,
Ghana 1686 Trypanososma brucei
J.F. Drexler et al. rhodesiense Sleeping Sickness,
Uganda
1654 Hepatitis E Virus Seroprevalence
among Adults, Germany 1687 Rickettsia felis in Aedes
M.S. Faber et al. albopictus Mosquitoes, Libreville,
Gabon
1658 Scarlet Fever Epidemic, Hong
Kong, 2011 1689 Bartonella spp. Infection Rate and
E.Y.Y. Luk et al. B. grahamii in Ticks

ii Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 1704 Duffy Phenotype and
Plasmodium vivax Infections in
October 2012 Humans and Apes, Africa
1690 Human Parvovirus 4 Viremia in
Young Children, Ghana 1705 Rickettsia parkeri and
Candidatus Rickettsia andeanae
1692 Multidrug-Resistant Salmonella in Gulf Coast Ticks, Mississippi
enterica, Democratic Republic of
the Congo 1707 Attributing Cause of Death
for Patients with Clostridium
1694 Co-Circulation and Persistence difficile Infection
of Genetically Distinct Saffold
Viruses, Denmark 1708 Characterization of
Mycobacterium orygis
p. 1673
1696 Pathogenic Leptospira spp. in
Bats, Madagascar and Union of 1709 Epsilonproteobacteria in
the Comoros Humans, New Zealand
p. 1703

1698 West Nile Virus About the Cover

Meningoencephalitis Imported
1711 Human minus Three Pieces of
into Germany
Hair

1700 Scarlet Fever Outbreak,

Etymologia
Hong Kong, 2011
1635 Tetanus
1702 Hand, Foot, and Mouth Disease
Caused by Coxsackievirus A6

Conference Summaries/
Reports Online Only
wwwnc.cdc.gov/eid/articles/online-reports.htm

Manuscripts submitted for online publication may

include illustrations and relevant links.

More information on online only

requirements at
wwwnc.cdc.gov/eid/pages/author-resource-center.htm

Submit manuscripts at
wwwnc.cdc.gov/eid/pages/submit-manuscript.htm

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 iii
iv Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, Octoer 2012
 Methicillin-Resistant
Staphylococcus aureus Sequence
Type 239-III, Ohio, USA, 2007–20091
Shu-Hua Wang, Yosef Khan, Lisa Hines, José R. Mediavilla, Liangfen Zhang, Liang Chen,
Armando Hoet, Tammy Bannerman, Preeti Pancholi, D. Ashley Robinson, Barry N. Kreiswirth,
and Kurt B. Stevenson, for the Prevention Epicenter Program of the Centers for Disease Control
and Prevention

Medscape, LLC is pleased to provide online continuing medical education (CME) for this journal article, allowing clinicians the
opportunity to earn CME credit.
This activity has been planned and implemented in accordance with the Essential Areas and policies of the Accreditation
Council for Continuing Medical Education through the joint sponsorship of Medscape, LLC and Emerging Infectious Diseases.
Medscape, LLC is accredited by the ACCME to provide continuing medical education for physicians.
TM
Medscape, LLC designates this Journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit(s) . Physicians
should claim only the credit commensurate with the extent of their participation in the activity.
All other clinicians completing this activity will be issued a certificate of participation. To participate in this journal CME activity:
(1) review the learning objectives and author disclosures; (2) study the education content; (3) take the post-test with a 70% minimum
passing score and complete the evaluation at www.medscape.org/journal/eid; (4) view/print certificate.

Release date: September 20, 2012; Expiration date: September 20, 2013

Learning Objectives
Upon completion of this activity, participants will be able to:
• Distinguish the most common strains of methicillin-resistant Staphylococcus aureus (MRSA) in the United States
• Assess the clinical characteristics of infection with MRSA ST239-III
• Analyze the treatment and prognosis of MRSA ST239-III infection
• Evaluate molecular characteristics of MRSA ST239-III.

CME Editor
Thomas J. Gryczan, MS, Technical Writer/Editor, Emerging Infectious Diseases. Disclosure: Thomas J. Gryczan, MS, has
disclosed no relevant financial relationships.

CME Author
Charles P. Vega, MD, Health Sciences Clinical Professor; Residency Director, Department of Family Medicine, University of
California, Irvine. Disclosure: Charles P. Vega, MD, has disclosed no relevant financial relationships.

Authors
Disclosures: Shu-Hua Wang, MD, MPH; Yosef Khan, MBBS, PhD; Jose R. Mediavilla, MBS, MPH; Liangfen Zhang, MD,
PhD; Liang Chen, PhD; Armando Hoet, PhD; Tammy Bannerman; D. Ashley Robinson, PhD; Barry N. Kreiswirth, PhD; and
Kurt B. Stevenson, MD, MPH, have disclosed no relevant financial relationships. Lisa Hines, RN, CIC, has disclosed the following
relevant financial relationships: owns stock, stock options, or bonds from Kimberly Clark Corp., General Electric Co., Medtronic Inc.,
Stryker Corp. TEVA Pharmaceutical Industries. Preeti Pancholi, PhD, has disclosed the following relevant financial relationships:
served as an advisor or consultant for Abbott; served as a speaker or a member of a speakers bureau for Abbott, Nanosphere;
received grants for clinical research from Cepheid, Abbott, Quidel, QIAGEN, Nanosphere.

Author affiliations: The Ohio State University Wexner Medical Methicillin-resistant Staphylococcus aureus (MRSA) is
Center, Columbus, Ohio, USA (S.H. Wang, Y. Khan, L. Hines, A. a human pathogen that has diverse molecular heterogeneity.
Hoet, P. Pancholi, K.B. Stevenson); University of Medicine and Most MRSA strains in the United States are pulsed-field gel
Dentistry of New Jersey, Newark, New Jersey, USA (J.R. Mediavilla, electrophoresis USA100 sequence type (ST) 5 and USA300
L. Chen, B.N. Kreiswirth); University of Mississippi Medical Center, ST8. Infections with MRSA ST239-III are common and
Jackson, Mississippi, USA (L. Zhang, D.A. Robinson); and The
found during health care–associated outbreaks. However,
this strain has been rarely reported in the United States.
Ohio Department of Health Laboratories, Reynoldsburg, Ohio, USA
(T. Bannerman) Presented in part at the 48th Annual Meeting of the Infectious
1

Diseases Society of America, Vancouver, British Columbia,

DOI: http://dx.doi.org/10.3201/eid1810.120468 Canada, October 21–24, 2010.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1557
RESEARCH

As part of a study supported by the Prevention Epicenter Program study evaluating the transmission of MRSA between
Program of the Centers for Disease Control and Prevention hospitals in Ohio, USA, molecular analysis was performed
(Atlanta, GA, USA), which evaluated transmission of MRSA on a group of clinical MRSA isolates collected from The
among hospitals in Ohio, molecular typing identified 78 Ohio State Health Network (OSHN). The OSHN consists of
(6%) of 1,286 patients with MRSA ST239-III infections. The Ohio State University (OSU) Wexner Medical Center
Ninety-five percent (74/78) of these infections were health
(WMC), which is a tertiary care medical center, and 7 smaller
care associated, and 65% (51/78) of patients had histories
of invasive device use. The crude case-fatality rate was
community hospitals located 30–120 miles from OSU. All
22% (17/78). Identification of these strains, which belong to MRSA specimens from community hospitals and selected
a virulent clonal group, emphasizes the need for molecular OSU MRSA blood isolates and isolates from patients
surveillance. residing in the catchment areas of the outreach hospitals
were prospectively collected during March 2009–February
2010 for genotyping by using a repetitive element PCR (rep-
taphylococcus aureus is a major human pathogen that
S possesses multiple toxins and virulence mechanisms
(1). Antimicrobial drug resistance in S. aureus has added to
PCR). Among archived MRSA isolates from OSUWMC
from January 2007 through February 2009, only a selection
of isolates was chosen for genotyping (not a randomized
the complexity of treating serious infections caused by this sampling). The total number of isolates and the total number
bacteria, and methicillin-resistant S. aureus (MRSA) appears of ST239 from each time period was collected.
to have greater virulence than methicillin-susceptible
strains (2,3). Most MRSA strains in the United States are Data Collection
pulsed-field gel electrophoresis (PFGE) types USA100 and We performed medical record reviews for 1,286
USA300, corresponding to multilocus sequence typing patients. Patient demographic characteristics (presence of
(MLST) ST5 and ST8, respectively (4). MRSA belonging health care–related risk factors during the preceding 12
to MLST ST239 and harboring staphylococcal cassette months, presence of an invasive device during the previous
chromosome mec (SCCmec) type III (MRSA ST239- 7 days, and concurrent conditions) were collected. Patient-
III) are associated with infections in health care settings, level data, including addresses for geocoding, were entered
outbreaks, increased resistance to antimicrobial drugs, and into a secure database within the OSUWMC Information
capacity for invasive disease (5–7). Warehouse.
MRSA ST239-III has a history of successful
dissemination in many regions, leading to a diverse array Classification of MRSA Infections
of regionally prevalent clones. These clones include the All MRSA cases were classified into 3 categories on
Brazilian; British Epidemic 1, 4, 7, 9, and 11; Canadian the basis of accepted epidemiologic definitions (11). The
Epidemic 3/Punjab; Czech; Eastern Australian 2 and 3; first category was health care–associated, defined as a
Georgian; Hungarian; Lublin; Nanjing/Taipei (ST241); culture obtained >48 hours after admission. The second
Portuguese; and Vienna clones (8,9). Although it is category was health care–associated community onset,
common worldwide, MRSA ST239-III has not played defined as a culture obtained <48 hours after admission
any predominant role in the United States; infections with with identified health care–associated risk factors. The
MRSA ST239-III have been rarely reported in the United third category was community-associated, defined as a
States since the 1990s (9–13). Recently, only 2 reports culture obtained <48 hours after admission without health
of this strain in the United States involving sporadic care–associated risk factors. Health care–associated risk
nasal colonization and bloodstream infections have been factors comprised presence of an invasive device, history of
published (13,14). MRSA infection or colonization, surgery, hospitalization,
In this study, we describe clinical epidemiologic dialysis, or residence in a long-term care facility in the 12
characteristics and molecular analysis of clinical infections months preceding the culture.
with MRSA ST239-III in the midwestern United States. Outcomes for MRSA infection were categorized
Identification of a strain from such a virulent clonal group as cure (complete resolution after antimicrobial drug
in the United States with wide dissemination in other parts treatment); failure (persistence of infection and change in
of the world represents a potential public health concern. antimicrobial drug regimen); relapse (resolution of infection
after complete treatment with subsequent development of
Methods new symptoms); recurrent (redevelopment of MRSA at
same or other site >2 weeks after completion of treatment
Sampling Method for initial MRSA infection); indeterminate (unknown
As part of Centers for Disease Control and Prevention outcome); and death (death <30 days after diagnosis of
(CDC) (Atlanta, GA, USA)–sponsored Prevention Epicenter MRSA infection because of any cause or during the same

1558 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 MRSA Sequence Type 239-III, Ohio, USA, 2007–2009

hospitalization). Destination after hospital discharge, such used because we were not analyzing for an outbreak (17).
as home or skilled nursing facility, was also noted. Thus, interpretations of possibly or probably related PFGE
subtypes between strains obtained by PFGE band patterns
Drug Susceptibility Testing were not made. Each PFGE band difference was classified
The respective OSHN Clinical Microbiology as a unique PFGE pattern.
Laboratories initially identified all MRSA isolates by using
standard microbiological methods. Antimicrobial drug spa Typing
susceptibility testing was performed at each institution, spa typing (18) was performed on all MRSA isolates
and results were interpreted according to Clinical and by using eGenomics software (www.egenomics.com)
Laboratory Standards Institute break point guidelines (15). as described (19); Ridom spa types were subsequently
At OSUWMC, antimicrobial drug susceptibility testing assigned by using the SpaServer website (www.spaserver.
was performed by using the automated Micro-Scan method ridom.de).
(Siemens Diagnostics, Sacramento, CA, USA), and only
constitutive clindamycin testing was performed. Linezolid SCCmec Typing
MIC >4 mg/L were confirmed by using the Etest method SCCmec typing was performed on all MRSA isolates
(bioMérieux, Marcy l’Etoile, France). by using a described multiplex real-time PCR (20). This
PCR is specific for 2 essential gene complexes (ccr and
Genotyping mec) found in all SCCmec elements.
The MRSA ST239 III isolates were genotyped initially
by using rep-PCR, followed by PFGE, staphylococcal dru Typing
protein A sequencing (spa typing), SCCmec typing, and MRSA isolates were also characterized by sequencing
mec-associated direct repeat unit (dru) typing. Selected the hypervariable dru repeat region within the SCCmec
isolates were also characterized by MLST and single- element (21) and using DruID software (9). New dru types
nucleotide polymorphism (SNP) typing. Detection of genes were submitted to www.dru-typing.org.
encoding Panton–Valentine leukocidin (PVL), toxic shock
syndrome toxin (TSST), arginine catabolic mobile element MLST and PVL, TSST, ACME, and Mupirocin
(ACME), and high-level mupirocin resistance (mupA) was Resistance
also performed. Brief descriptions of each testing method MLST was performed on representative isolates as
are outlined below. described (22) by using the MLST database (http://saureus.
mlst.net). PCR-based detection of PVL (23), TSST (24),
rep-PCR ACME (25), and mupA (26) was performed on all isolates
The DiversiLab System (bioMérieux, Durham, as described.
NC, USA) was used for rep-PCR analysis according to
described methods (16). Isolates belonging to designated SNP Typing
rep-PCR clusters shared >95% similarity. In addition, A panel of 43 SNPs for describing the global population
comparison of matching patterns in the DiversiLab System structure of MRSA ST239-III (9) was used to identify the
library was initially used to infer the PFGE and SCCmec haplotypes of 22 isolates. These SNPs were typed by using
types, which were later validated by using appropriate Golden-Gate Genotyping Assay (Illumina, San Diego, CA,
testing methods. The numeric classification system used USA) and conventional Sanger sequencing.
for rep-PCR analysis is unique to OSUWMC.
Statistical Analysis
PFGE All patient demographic, clinical, and molecular typing
The PulseNet protocol for molecular subtyping of data were aggregated in tabular format, and descriptive
S. aureus was followed. Salmonella enterica serotype statistics were generated by using SAS version 9.2 (SAS
Braenderup DNA was digested with XbaI (Roche, Institute Inc., Cary NC, USA), for demographic and risk
Indianapolis, IN, USA) and used as the normalization factor history. A significant difference between ST239 and
standard for gel analysis. S. aureus chromosomal DNA was other strains (US300, US100, and all other strains) was
digested with SmaI (Roche). Fragments were separated in examined by using χ2 tests for categorical variables and
a clamped homogenous electric field mapper unit (Bio-Rad t-tests for continuous variables. An α level of 0.5 was used.
Laboratories, Hercules, CA, USA). Fingerprint images
were analyzed by using Bionumerics software version Human Subjects Protection
4.61 (Applied Maths NV, Sint-Martens-Latem, Belgium). We obtained approval for this study from the OSU
The traditional classification of PFGE subtypes was not Office of Responsible Research Practices’ Biomedical

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1559
RESEARCH

Institutional Review Board. The OSU Information For linezolid, 69 (97%) of 71 isolates were susceptible
Warehouse has established honest broker status with the and 2 blood isolates with MICs >4 mg/L were classified as
OSU Institutional Review Board, enabling storage of uninterpretable.
fully identifiable data and presentation of patient data to
investigators in a coded format that maintains patient Molecular Typing of MRSA ST239-III Isolates
confidentiality. Eight rep-PCR patterns (6, 19, 22, 42, 53, 78, 79, and
97) were detected among MRSA ST239-III isolates, and 39
Results (50%) isolates had rep-PCR pattern 6. Similarly, isolates
were distributed among 9 PFGE patterns (A, B, C, D, E,
Patient Characteristics F, G, H, and I), and 58 (74%) had PFGE pattern A. The
Of 1,286 clinical MRSA isolates, 78 (6%) were rep-PCR and PFGE patterns are shown in Figure 1. PFGE
identified as MRSA ST239-III; 71 (91%) were obtained pattern F is <80% similar with other PFGE subtypes in
from OSU and 7 (9%) from community hospitals. the dendrogram and was identified as an unknown PFGE
Seven (2%) of 397 isolates were obtained from outreach type by the CDC database. However, molecular testing
community hospitals, 37 (6%) of 613 from OSUWMC, and confirmed it to be MRSA ST239-III: MLST 239, SCCmec
34 (12%) of 276 from OSUWMC archives. type III, spa type 3(t037), dru type dt15b, and SNP
These strains were first recognized by rep-PCR as haplotype H9.
possible SCCmecA type III isolates from the DiversiLab Four spa types, including eGenomics spa types 3
System library with imputed Brazilian PFGE types. (Ridom t037), 314 (t363), 121 (t421), and 1256 (t631)
Additional molecular typing identified MRSA ST239-III in were identified, and 74 (95%) isolates were classified as
a clade different from that containing the Brazilian strains. spa type 3 (t037). In contrast, isolates were grouped into
Demographics and clinical characteristics of 78 13 dru types (dt2c, dt6n, dt9g, dt12i, dt13k, dt14 g, dt15b,
patients infected with MRSA ST239-III are shown in the dt15h, dt15i, dt15j, dt15k, dt16a, and dt23a), and 60 (77%)
Table. Among patients with MRSA ST239-III infections, isolates were classified as dt15b. MLST typing of selected
46 (59%) were male, 65 (83%) were white, 31 (40%) representative isolates confirmed 13 to be MRSA ST239-
disabled, and 29 (37%) were retired. Seventy-four (95%) III (2-3-1-1-4-4-3), 3 as ST1801 (2-3-1-1-4-19-3), and 1
had health care–associated isolates and 51 (65%) had as ST2017 (2-3-1-1-4-19-227). ST1801 is a single-locus
histories of use of invasive devices within the previous variant of MRSA ST239-III, and ST2017 is a novel single-
7 days. Distribution of specimen types include 37 (47%) locus variant of ST1801 and a double-locus variant of
bloodstream infections (BSIs), 19 (24%) lower respiratory MRSA ST239. All were SCCmec type III, and the ccrC
tract infections (LRTIs), 11 (14%) skin and soft tissue locus was not detected. The genes that encode PVL, TSST,
infections, and 11 (14%) other infections. Seventeen (22%) ACME, or high-level mupirocin resistance were not found
patients died, and 18 (23%) patients had treatment failures, in any isolates.
recurring infections, or relapses. Only 32% of the patients A total of 33 unique genotypic combinations from 78
were discharged home. patients infected with MRSA ST239-III were represented
Comparison of clinical characteristics of MRSA among isolates with complete SCCmec, rep-PCR, PFGE,
ST239-III with those of PFGE types USA100 and USA300 spa, and dru data. This information and year of isolation
and other non–ST239 infections are shown in the Table. are shown in Figure 2. The genotype cluster sizes ranged
Most ST239 isolates were health care–associated MRSA from 2 to 18 isolates; 23 of the isolates were unique. In
and had characteristics and concurrent conditions similar contrast, SNP typing of 22/78 isolates with a panel of
to those associated with USA100. MRSA ST239-III did 43 SNPs showed them to be indistinguishable from each
not have virulent determinants often associated with health other. All isolates tested belonged to haplotype 9 (H9)
care–associated or community-associated MRSA strains, within MRSA ST239-III clade II (Figure 3). This clade
such as PVL, TSST, ACME, or mupA. was composed of isolates from many continents, including
many from sources in Asia (9).
Characteristics of MRSA ST239-III Isolates
Resistance was observed to clindamycin (76/76, Discussion
100%), moxifloxacin (47/47, 100%), gentamicin (74/77, MRSA ST239-III has demonstrated epidemic potential
96%), tetracycline (70/74, 95%), and trimethoprim– worldwide, and identification of this strain in the United
sulfamethoxazole (62/77, 81%). All MRSA ST239-III States might represent a major public health concern.
isolates were susceptible to vancomycin. For daptomycin, Although the precise time frame of emergence of this
64 (98%) of 65 isolates were susceptible, and 1 blood isolate strain in Ohio is unknown, it appears to have been present
with an MIC of 2 mg/L was classified as uninterpretable. before the study because it was identified in archived

1560 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 MRSA Sequence Type 239-III, Ohio, USA, 2007–2009

specimens dating back to 2007. The origin of the MRSA >6% if complete sampling was conducted is unknown.
ST239-III clonal group has been dated to the mid-20th Furthermore, we found no epidemiologic evidence of
century in 2 studies (9,27). Thus, the Ohio strain might clustering of cases consistent with an outbreak of infection
have been introduced into the region anytime during the with this strain. The large number of isolates at OSUWMC
past 40 years. Because of incomplete sampling during the merely reflects its role as a large referral center.
study, the transmission dynamics of Ohio MRSA ST239 Clinically, MRSA ST239-III is primarily health care
III are uncertain. Whether the prevalence would have been associated, causes major outbreaks, shows increased

Table. Characteristics of 1,286 patients with MRSA ST239-III and non–MRSA ST239-III infections, Ohio, USA, 2007–2009*
ST239, USA100, USA300, All other,
Characteristic n = 78 n = 481 p value n = 574 p value n = 153 p value
Outreach hospital isolates 7 (9) 81 (17) NA 262 (46) NA 47 (31) NA
Patient demographics
Mean age, y (range) 58 (19–90) 61 (18–99) 0.06 43 (18–92) <0.0001 49 (18–94) 0.001
Male sex 46 (59) 259 (54) 0.39 311 (54) 0.42 80 (52) 0.33
White race 65 (83) 393 (82) 0.85 441 (77) 0.61 126 (82) 0.77
Medical history
Diabetes 33 (42) 149 (31) 0.047 116 (20) <0.0001 35 (23) 0.002
Chronic lung disease 26 (33) 117 (24) 0.09 74 (13) <0.0001 17 (11) <0.0001
Renal failure 19 (24) 93 (19) 0.39 37 (6) <0.0001 17 (11) 0.008
Malignancy 13 (17) 97 (20) 0.47 48 (8) 0.02 30 (20) 0.58
Health care–associated risk factors, past 12 mo
Hospitalization 58 (74) 296 (62) 0.03 148 (26) <0.0001 58 (38) <0.0001
Use of invasive device 35 (45) 168 (35) 0.09 66 (12) <0.0001 35 (23) 0006
Surgery 35 (45) 179 (37) 0.19 79 (14) <0.0001 33 (22) 0.0002
History of MRSA infection 26 (33) 77 (16) 0.0003 82 (14) <0.0001 29 (19) 0.015
Stay in long-term care facility 22 (28) 155 (32) 0.47 34 (6) <0.0001 14 (9) 0.0002
Hemodialysis 15 (19) 47 (10) 0.068 18 (3) <0.0001 10(7) 0.015
Other 10 (13) 29 (6) 0.03 40 (7) 0.06 7 (5) 0.02
Invasive devices <7 d before infection
Central venous catheter 25 (32) 156 (32) 0.94 61 (11) <0.0001 32 (21) 0.06
Foley catheter 17 (22) 99 (21) 0.8 55 (10) 0.001 17 (11) 0.03
Hemodialysis 13 (17) 49 (10) 0.02 19 (3) <0.0001 8 (5) 0.008
Mechanical ventilation 14 (18) 85 (18) 0.95 35 (6) 0.0002 11 (7) 0.012
Drainage tubes 8 (10) 50 (10) 0.97 10 (2) <0.0001 5 (3) 0.02
Total parenteral nutrition 8 (10) 22 (5) 0.04 4 (1) <0.0001 0 0.0001
Other 20 (26) 71 (15) 0.016 39 (7) <0.0001 15 (10) 0.0015
Classification of MRSA infection
Health care–associated 32 (41) 197 (41) 0.97 63 (11) <0.0001 37 (24) 0.01
Health care–associated community onset 42 (54) 214 (44) 0.12 113 (20) <0.0001 49 (32) 0.0012
Community-associated 4 (5) 70 (15) 0.02 398 (69) <0.0001 67 (44) <0.0001
Outcome
Cure 23 (29) 149 (31) 0.81 94 (16) 0.005 38 (25) 0.49
Failure 3 (4) 7 (1.5) 0.14 12 (2) 0.33 5 (3) 0.81
Relapse 4 (5) 7 (1.5) 0.03 2 (1) 0.0001 1 (1) 0.022
Recurrent 11 (14) 45 (9) 0.2 36 (6) 0.012 3 (2) 0.0002
Indeterminate 20 (26) 192 (40) 0.015 408 (71) 0.001 97 (63) 0.0001
Death 17 (22) 81 (17) 0.29 22 (4) 0.001 9 (6) 0.0003
No. patients admitted 74 429 0.005 310 0.0003 111 0.003
Admitting service
Intensive care unit 14 (19) 58 (14) 0.2 29 (9) 0.019 9 (8) 0.047
Medicine service 31 (42) 238 (55) 0.03 203 (66) 0.0002 75 (68) 0.0007
Surgery service 20 (27) 117 (27) 0.95 59 (19) 0.12 20 (18) 0.126
Other specialty care unit† 9 (12) 16 (4) 0.002 19 (6) 0.073 7 (6) 0.16
Destination after discharge
Home 20 (27) 146 (34) 0.24 209 (67) <0.0001 60 (54) 0.0003
Another facility, long-term care center, or 35 (47) 204 (48) 0.97 79 (24) <0.0001 40 (36) 0.12
rehabilitation center
Median duration of hospitalization,‡ d (range) 16 (1–143) 11 (0–136) 0.07 6 (0–169) <0.0001 9 (1–124) 0.01
*Values are no. (%) unless otherwise indicated. MRSA, methicillin-resistant Staphylococcus aureus; ST, sequence type; NA, not applicable. The p-values
were generated to assess whether differences in demographic and risk factor history existed between ST239 and other strains (USA100, USA300, all
other strains). Ȥ2 tests were used for categorical variables and t-tests were used for continuous variables.
†Other specialty care unit admitting services included bone marrow, cardiology, hematology, surgical oncology, transplant, and obstetrics and
gynecology.
‡Duration of hospitalization and days to MRSA culture–positive included only inpatients; outpatient visits such as clinic or emergency department visits
were excluded from the analysis. Duration of hospitalization was calculated from the date of the current hospital admission to the date of discharge.
Hospitalization days at another institution before transfer were not included.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1561
RESEARCH

47% (9/19) respectively. In our study population, MRSA

ST239-III had a higher case-fatality rate than did USA300
(4%) and USA100 (17%) (Table).
Increased illness and death for pulmonary disease
might be caused by an enhanced ability of MRSA ST239-III
strains to produce biofilms and adhere to airway epithelial
cells, providing a biologically plausible explanation for its
potential to cause pneumonia and central line–associated
BSIs (28,29). The association of MRSA ST239-III isolates
with pulmonary infections has also been reported in South
Korea (30). Moreover, in a multicenter study involving 21
hospitals in Beijing, where MRSA ST239-III is especially
prevalent, 61% of MRSA LRTI cases were attributed to
MRSA ST239-III (31). The London outbreak strain of
MRSA ST239-III, and most examined MRSA ST239-
III strains from the People’s Republic of China, belong
to the same clade (clade II) as the Ohio strain of MRSA
ST239-III, demonstrating the potential for this lineage
to disseminate and cause disease (9,27,32). Because of
incomplete collection of all MRSA strains causing LRTIs,
whether MRSA ST239-III is the prevalent strain in MRSA
pneumonias is unknown. A total of 17% (203/1,206) of the
study isolates were pulmonary infections and 9% (19/222)
of the pneumonias were caused by MRSA ST239-III.
PFGE is the standard for MRSA outbreak investigations.
Figure 1. Methicillin-resistant Staphylococcus aureus sequence
However, we chose a rapid PCR–based method that could
type 239-III isolates, Ohio, USA, 2007–2009, based on A) repetitive serve as a potential screening surveillance tool to identify
element PCR (rep-PCR) and B) pulsed-field gel electrophoresis. outbreaks of a particular strain. Although many institutions
Virtual gel results are shown for 8 DiversiLab System (bioMérieux, might be using PFGE for outbreak investigations to identify
Durham, NC, USA) rep-PCR patterns. Pattern numbers assigned clonal clusters, methods for PFGE subtype classification
are unique to the Ohio State University Wexner Medical Center.
are often not standardized between laboratories (33,34).
To maximize uniformity and enable comparisons of PFGE
classifications between institutions, the standardized CDC
resistance to antimicrobial drugs, and demonstrates protocols for performing and analyzing PFGE results
a capacity for causing invasive disease (5–7,28,29). were used. The initial CDC PFGE analysis classified our
Comparison of characteristics of patients infected with isolates as the Brazilian clone. However, SNP typing
MRSA ST239-III with those infected with PFGE USA300, unambiguously placed the Ohio isolates within a different
PFGE USA100, and other strains showed that MRSA clade than the Brazilian clone (Figure 3) (9). Because PFGE
ST239-III is similar to USA100 in terms of health care– typing cannot define specific lineages for MRSA ST239-III,
associated risk factors and outcomes (Table). CDC has changed the reporting of MRSA ST239-III strains
Concern for the invasive potential of MRSA ST239-III from Brazilian subtype to ST239. Laboratories will need
was illustrated during a 2-year intensive care unit (ICU) to perform additional testing to identify specific clades or
outbreak in London in which patients infected with MRSA clonal groups. Classification of the 78 MRSA ST239-III
ST239-III were more likely than patients with non–ST239 isolates by the traditional PFGE outbreak method would
strains to show development of bacteremia (47% vs 13%; identify only 2 PFGE subtypes instead of the 8 observed
p<0.001) and have culture-positive vascular access device patterns. PFGE subtype 1 would consist of isolates A–E,
infections (59% vs 26%; p<0.001) (29). In our study, 19% G, and H, and subtype 2 would consist of isolate F (Figure
(14/74) of patients infected with MRSA ST239-III strains 1, panel B).
were admitted to ICU, compared with 11% (96/850) of Classification issues also exist with respect to the
patients infected with non–ST239 strains (Table). Among DiversiLab system. Five of the 8 initial rep-PCR patterns
MRSA ST239-III ICU admissions, 50% (7/14) were BSIs matched isolates in the DiversiLab library, which likewise
and 50% (7/14) were LRTIs. The crude mortality rate categorized them as the Brazilian clone. In contrast, 3
for persons with BSIs and LRTIs was 22% (8/37) and other rep-PCR patterns did not match any isolates in

1562 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 MRSA Sequence Type 239-III, Ohio, USA, 2007–2009

Figure 2. Molecular characteristics of 78 methicillin-

resistant Staphylococcus aureus sequence type 239-III
isolates, Ohio, USA, 2007–2009. SCCmec, staphylococcal
cassette chromosome mec; spa, staphylococcal protein
A; rep-PCR, repetitive element PCR; PFGE, pulsed-field
gel electrophoresis; dru, direct repeat unit; dt, dru type.
The 33 unique genotypic combinations are contrasted by
black and gray shading.

the DiversiLab library. Additional molecular tests were dataset. Ninety-four percent (63/67) of the MRSA ST239-
performed to identify these isolates as MRSA ST239-III. III isolates were resistant to all 3 drugs (clindamycin,
Because of the novel identification of MRSA ST239- tetracycline, and gentamicin) (35). Use of phenotypic
III strains, we performed a battery of molecular tests to drug susceptibility patterns might alert infection control
better classify the strains. Although the same dru types practitioners that they are dealing with a MRSA ST239-III
might be found in unrelated MRSA lineages and different clone. Additional molecular testing, such as spa typing in
staphylococcal species (32), variation at the dru locus conjunction with SCCmec typing, can then be performed to
within the SCCmec III element is consistent with MRSA confirm the strain type.
ST239-III phylogeny (9). Because use of this typing method We have genotyped and geocoded 1,286 MRSA
is relatively new, further study of its phylogenetic utility is isolates in the CDC Prevention Epicenter Study by using
needed (21). The feasibility of population-based genome- rep-PCR. The lack of a clear association between MRSA
wide SNP datasets has been demonstrated for MRSA ST239-III molecular genotype patterns could be caused
ST239-III by Harris et al. (27). Cluster groupings from by the retrospective nature of this study, lack of complete
these PFGE, rep-PCR, and dru methods were not identical, population sampling, and the inability to gather social
suggesting that the molecular tests are not completely network history. The rep-PCR and PFGE methods might
interchangeable. Of the various molecular methods used be of limited use in evaluating geographic clustering at the
to index genomic variation in MRSA, dru typing and rep- scale studied here unless specific lineages of the reference
PCR appeared to be more discriminatory. strains are included in the respective databases. Because
The publication of only 2 recent reports of MRSA of the clonality exhibited by MRSA ST239-III, even a
ST239-III in the United States (13,14) might have relatively small panel of 43 SNPs is able to identify the
resulted from inadequate national surveillance or low same major phylogenetic lineages that are identified
transmissibility of the strain. Institutions might not be by genome-wide SNPs. A more detailed genome-wide
able to perform genotyping on all their isolates because SNP analysis might be required to resolve geographic
of lack of appropriate laboratory facilities or resources. clustering. Additional social networking for determining
An alternative method of surveillance for MRSA ST239- spatial, temporal, and geographic relationships of patients
III might be evaluation of the drug susceptibility pattern. at the medical institutions studied is under way to identify
Most of the MRSA ST239-III strains in our study were potential nosocomial interhospital and intrahospital
resistant to clindamycin, tetracycline, and gentamicin. In transmission.
a subset analysis, the phenotypic patterns of antimicrobial Recent identification of a strain of MRSA ST239-III
drug susceptibility and MRSA genotype prediction were in the midwestern United States is a major public health
determined for 798 MRSA isolates from the OSHN concern. Globally, this strain has demonstrated increased

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1563
RESEARCH

SCCmec typing; Gregory Fosheim for assistance with PFGE

and SCCmec typing; Jennifer Santangelo, David Newman, and
Kelly Kent for assistance with the Web entry portal and data
management; Ruchi Tiwari for rep-PCR typing; Eric Brandt for
PFGE typing; Melissa Abley and Brandon Palinski for MLST
typing; and Rachael Tumin for phenotypic analysis.
S.-H.W. was supported by OSU Centers for Clinical
Translational Science and the Marie Davis Bremer Award and
by Award KL2 RR02574 from the National Centers for Research
Resources. D.A.R. was supported by a grant from the American
Heart Association and National Institutes of Health grant
GM080602. S.-H.W., Y.K, L.H. P.P., and K.B.S. were supported
by CDC Prevention Epicenter program grant (U01 CI000328).
Dr Wang is an assistant professor of medicine at The Ohio
State University in Columbus, Ohio. Her research interests are
molecular genotyping, geocoding, and social network analysis
for evaluation of transmission of MRSA and Mycobacterium
tuberculosis.
Figure 3. Single-nucleotide polymorphism (SNP) haplotype map
showing position of methicillin-resistant Staphylococcus aureus References
sequence type 239-III (MRSA ST239-III) isolates, Ohio, USA,
2007–2009, within the global population structure of the MRSA 1. Lowy FD. Staphylococcus aureus infections. N Engl J Med.
ST239-III clonal group. Circles indicate distinct haplotypes, as 1998;339:520–32. http://dx.doi.org/10.1056/NEJM199808203390
defined by using a panel of 43 SNPs (9). Sizes of circles indicate 806
relative frequency of different haplotypes. Arrows indicate 2. Cosgrove SE, Sakoulas G, Perencevich EN, Schwaber MJ,
haplotype 5 (H5), which includes the Brazilian clone, and haplotype Karchmer AW, Carmeli Y. Comparison of mortality associated with
9 (H9), which includes the 22 MRSA ST239-III isolates from Ohio. methicillin-resistant and methicillin-susceptible Staphylococcus
Relationships between haplotypes were determined by using aureus bacteremia: a meta-analysis. Clin Infect Dis. 2003;36:53–9.
maximum-parsimony analysis (9). http://dx.doi.org/10.1086/345476
3. Melzer M, Eykyn SJ, Gransden WR, Chinn S. Is methicillin-
resistant Staphylococcus aureus more virulent than methicillin-
susceptible S. aureus? A comparative cohort study of British
virulence and widespread dissemination. Because of patients with nosocomial infection and bacteremia. Clin Infect Dis.
2003;37:1453–60. http://dx.doi.org/10.1086/379321
an inadequate national surveillance system, temporal
4. McDougal LK, Steward CD, Killgore GE, Chaitram JM, McAllister
emergence and dissemination of the strain in the United SK, Tenover FC. Pulsed-field gel electrophoresis typing of oxacillin-
States is uncertain. The MRSA ST239-III strains in Ohio resistant Staphylococcus aureus isolates from the United States:
have molecular heterogeneity and geographic diversity. establishing a national database. J Clin Microbiol. 2003;41:5113–
20. http://dx.doi.org/10.1128/JCM.41.11.5113-5120.2003
The SNP-based data suggest that the strains also show
5. Feil EJ, Nickerson EK, Chantratita N, Wuthiekanum V, Srisomang
clonal diversity. Continued surveillance is warranted P, Cousind R, et al. Rapid detection of the pandemic methicillin-
because this MRSA ST239-III strain, similar to related resistant Staphylococcus aureus clone ST 239, a dominant strain in
strains worldwide, exhibits increased antimicrobial drug Asian hospitals. J Clin Microbiol. 2008;46:1520–2. http://dx.doi.
org/10.1128/JCM.02238-07
resistance, capacity for causing invasive disease, potential
6. Holden MT, Lindsay JA, Corton C, Quail MA, Cockfield JD,
for causing outbreaks, and resulting in illnesses and Pathak S, et al. Genome sequence of a recently emerged, highly
deaths. MRSA ST239-III strains might have the potential transmissible, multi-antibiotic- and antiseptic-resistant variant of
to become established locally and disseminate among methicillin-resistant Staphylococcus aureus, sequence type 239
(TW). J Bacteriol. 2010;192:888–92. http://dx.doi.org/10.1128/
health care institutions as reported in other regions. Our
JB.01255-09
study underscores the value of molecular surveillance, 7. Aires de Sousa M, Conceição T, Simas C, de Lencastre H.
including traditional fingerprinting methods and newer Comparison of genetic backgrounds of methicillin-resistant and
sequence-based typing methods, as a critical component in -susceptible Staphylococcus aureus isolates from Portuguese
hospitals and the community. J Clin Microbiol. 2005;43:5150–7.
understanding the evolving epidemiology of MRSA.
http://dx.doi.org/10.1128/JCM.43.10.5150-5157.2005
8. Monecke S, Coombs G, Shore AC, Coleman DC, Akpaka P,
Acknowledgments Borg M, et al. A field guide to pandemic, epidemic and sporadic
We thank the OSU Epicenter Program Team, OSHN, for clones of methicillin-resistant Staphylococcus aureus. PLoS ONE.
2011;6:e17936. http://dx.doi.org/10.1371/journal.pone.0017936
assistance with data collection; CDC for assistance with initial

1564 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 MRSA Sequence Type 239-III, Ohio, USA, 2007–2009

9. Smyth DS, McDougal LK, Gran FW, Manoharan A, Enright MC, 23. Saïd-Salim B, Mathema B, Braughton K, Davis S, Sinsimer D, Eisner
Song JH, et al. Population structure of a hybrid clonal group of W, et al. Differential distribution and expression of Panton-Valentine
methicillin-resistant Staphylococcus aureus, ST239-MRSA-III. leucocidin among community-acquired methicillin-resistant
PLoS ONE. 2010;5:e8582. http://dx.doi.org/10.1371/journal. Staphylococcus aureus strains. J Clin Microbiol. 2005;43:3373–9.
pone.0008582 http://dx.doi.org/10.1128/JCM.43.7.3373-3379.2005
10. de Lencastre H, Severina EP, Roberts RB, Kreiswirth BN, Tomasz A. 24. Johnson WM, Tyler SD, Ewan EP, Ashton FE, Pollard DR, Rozee
Testing the efficacy of a molecular surveillance network: methicillin- KR. Detection of genes for enterotoxins, exfoliative toxins, and
resistant Staphylococcus aureus (MRSA) and vancomycin-resistant toxic shock syndrome toxin 1 in Staphylococcus aureus by the
Enterococcus faecium (VREF) genotypes in six hospitals in the polymerase chain reaction. J Clin Microbiol. 1991;29:426–30.
metropolitan New York City area. The BARG Initiative Pilot Study 25. Diep BA, Stone GG, Basuino L, Graber CJ, Miller A, des Etages
Group. Bacterial Antibiotic Resistance Group. Microb Drug Resist. SA, et al. The arginine catabolic mobile element and staphylococcal
1996;2:343–51. http://dx.doi.org/10.1089/mdr.1996.2.343 chromosomal cassette mec linkage: convergence of virulence
11. Klevens RM, Morrison MA, Nadle J, Petit S, Gershman K, Ray S, and resistance in the USA300 clone of methicillin-resistant
et al. Invasive methicillin-resistant Staphylococcus aureus infections Staphylococcus aureus. J Infect Dis. 2008;197:1523–30. http://
in the United States. JAMA. 2007;298:1763–71. http://dx.doi. dx.doi.org/10.1086/587907
org/10.1001/jama.298.15.1763 26. Anthony RM, Connor AM, Power EG, French GL. Use of the
12. Chung M, Dickinson G. de Lencastre H, Tomasz A. International polymerase chain reaction for rapid detection of high-level
clones of methicillin-resistant Staphylococcus aureus in two mupirocin resistance in staphylococci. Eur J Clin Microbiol Infect
hospitals in Miami, Florida. J Clin Microbiol. 2004;42:542–7. http:// Dis. 1999;18:30–4. http://dx.doi.org/10.1007/s100960050222
dx.doi.org/10.1128/JCM.42.2.542-547.2004 27. Harris SR, Feil EJ, Holden MT, Quail MA, Nickerson EK, Chantratita
13. Tenover FC, Tickler IA, Goering RV, Kreiswirth BN, Mediavilla JR, N, et al. Evolution of MRSA during hospital transmission and
Persing DH. Characterization of nasal and blood culture isolates of intercontinental spread. Science. 2010;327:469–74. http://dx.doi.
methicillin-resistant Staphylococcus aureus from patients in United org/10.1126/science.1182395
States hospitals. Antimicrob Agents Chemother. 2012;56:1324–30. 28. Amaral MM, Coelho LR, Flores RP, Souza RR, Silva-Carvalho
http://dx.doi.org/10.1128/AAC.05804-11 MC, Teixeira LA, et al. The predominant variant of the Brazilian
14. Mermel LA, Eells SJ, Acharya MK, Cartony JM, Dacus D, Fadem epidemic clonal complex of methicillin-resistant Staphylococcus
S, et al. Quantitative analysis and molecular fingerprinting of aureus has an enhanced ability to produce biofilm and to adhere to
methicillin-resistant Staphylococcus aureus nasal colonization and invade airway epithelial cells. J Infect Dis. 2005;192:801–10.
in different patient populations: a prospective, multicenter study. http://dx.doi.org/10.1086/432515
Infect Control Hosp Epidemiol. 2010;31:592–7. http://dx.doi. 29. Edgeworth JD, Yadegarfar G, Pathak S, Batra R, Cockfield JD,
org/10.1086/652778 Wyncoll D, et al. An outbreak in an intensive care unit of a strain
15. Clincal and Laboratory Standards Institute. Performance standards of methicillin-resistant Staphylococcus aureus sequence type 239
for antimicrobial susceptibility testing. 19th informational associated with an increased rate of vascular access device-related
supplement, vol 29, no. 3. M100–S19. Wayne (PA): The Institute; bacteremia. Clin Infect Dis. 2007;44:493–501. http://dx.doi.
2009. org/10.1086/511034
16. Healy M, Huong J, Bittner T, Lising M, Frye S, Raza S, et al. 30. Cho DT, Cha HY, Chang HH, Kim SW, Chung JM, Kim J, et al.
Microbial DNA typing by automated repetitive-sequence-based Risk factors for specific methicillin-resistant Staphylococcus aureus
PCR. J Clin Microbiol. 2005;43:199–207. http://dx.doi.org/10.1128/ clones in a Korean hospital. J Antimicrob Chemother. 2006;57:1122–
JCM.43.1.199-207.2005 7. http://dx.doi.org/10.1093/jac/dkl114
17. Tenover FC, Arbeit RD, Goering RV, Mickelsen PA, Murray BE, 31. Li DZ, Chen YS, Yang JP, Zhang W, Hu CP, Li JS, et al. Preliminary
Persing DH, et al. Interpreting chromosomal DNA restriction molecular epidemiology of the Staphylococcus aureus in lower
patterns produced by pulsed-field gel electrophoresis: criteria for respiratory tract infections: a multicenter study in China. Chin Med
bacterial strain typing. J Clin Microbiol. 1995;33:2233–9. J (Engl). 2011;124:687–92.
18. Shopsin B, Gomez M, Montgomery SO, Smith DH, Waddington M, 32. Smyth DS, Wong A, Robinson DA. Cross-species spread of SCCmec
Dodge DE, et al. Evaluation of protein A gene polymorphic region IV subtypes in staphylococci. Infect Genet Evol. 2011;11:446–53.
DNA sequencing for typing of Staphylococcus aureus strains. J Clin http://dx.doi.org/10.1016/j.meegid.2010.12.005
Microbiol. 1999;37:3556–63. 33. Van Bellum A, Struelens M, de Visser A, Verbrugh H, Tibayrenc
19. Mathema B, Mediavilla J, Kreiswirth BN. Sequence analysis of the M. Role of genomic typing in taxonomy, evolutionary genetics,
variable number tandem repeat in Staphylococcus aureus protein and microbial epidemiology. Clin Microbiol Rev. 2001;14:547–60.
A gene: spa typing. Methods Mol Biol. 2008;431:285–305. http:// http://dx.doi.org/10.1128/CMR.14.3.547-560.2001
dx.doi.org/10.1007/978-1-60327-032-8_22 34. Olive DM, Bean P. Principles and applications of methods for
20. Chen L, Mediavilla JR, Oliveira DC, Willey BM, de Lencastre H, DNA-based typing of microbial organisms. J Clin Microbiol.
Kreiswirth BN. Multiplex real-time PCR for rapid staphylococcal 1999;37:1661–9.
cassette chromosome mec typing. J Clin Microbiol. 2009;47:3692– 35. Tumin R, Wang SH, Pancholi P, Khan Y, Hines L, Stevenson K.
706. http://dx.doi.org/10.1128/JCM.00766-09 Relationship of social and medical factors to other antimicrobial
21. Goering RV, Morrison D, Al-Doori Z, Edwards GF, Gemmell resistance among methicillin-resistant Staphylococcus aureus
CG. Usefulness of mec-associated direct repeat unit (dru) typing strains. In: Abstracts of the 139th American Public Health
in the epidemiological analysis of highly clonal methicillin- Association Annual Meeting; Washington, DC, Oct 29–Nov 2,
resistant Staphylococcus aureus in Scotland. Clin Microbiol 2011. Washington (DC): American Public Health Association; 2011.
Infect. 2008;14:964–9. http://dx.doi.org/10.1111/j.1469-0691.2008. Abstract 246417.
02073.x
22. Enright MC, Day NP, Davies CE, Peacock SJ, Spratt BG. Multilocus Address for correspondence: Shu-Hua Wang, The Ohio State University
sequence typing for characterization of methicillin-resistant and
Wexner Medical Center, N-1120 Doan Hall, 410 West 10th Ave,
methicillin-susceptible clones of Staphylococcus aureus. J Clin
Microbiol. 2000;38:1008–15. Columbus, OH 43210, USA; email: shu-hua.wang@osumc.edu

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1565
RESEARCH

Epidemiology of Foodborne
Norovirus Outbreaks, United States,
2001–2008
Aron J. Hall, Valerie G. Eisenbart, Amy Lehman Etingüe, L. Hannah Gould, Ben A. Lopman,
and Umesh D. Parashar

Medscape, LLC is pleased to provide online continuing medical education (CME) for this journal article, allowing clinicians the
opportunity to earn CME credit.
This activity has been planned and implemented in accordance with the Essential Areas and policies of the Accreditation
Council for Continuing Medical Education through the joint sponsorship of Medscape, LLC and Emerging Infectious Diseases.
Medscape, LLC is accredited by the ACCME to provide continuing medical education for physicians.
TM
Medscape, LLC designates this Journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit(s) . Physicians
should claim only the credit commensurate with the extent of their participation in the activity.
All other clinicians completing this activity will be issued a certificate of participation. To participate in this journal CME activity:
(1) review the learning objectives and author disclosures; (2) study the education content; (3) take the post-test with a 70% minimum
passing score and complete the evaluation at www.medscape.org/journal/eid; (4) view/print certificate.

Release date: September 17, 2012; Expiration date: September 17, 2013

Learning Objectives
Upon completion of this activity, participants will be able to:
• Describe general characteristics and outcomes of US norovirus outbreaks, based on an analysis of data reported
during 2001-2008 to the CDC Foodborne Disease Outbreak Surveillance System
• Describe sources of US norovirus outbreaks, based on an analysis of data reported during 2001-2008 to the CDC
Foodborne Disease Outbreak Surveillance System
• Describe recommended interventions to reduce the frequency and effects of foodborne norovirus outbreaks, based on
an analysis of data reported during 2001-2008 to the CDC Foodborne Disease Outbreak Surveillance System

CME Editor
Carol E. Snarey, MA, Technical Writer/Editor, Emerging Infectious Diseases. Disclosure: Carol E. Snarey, MA, has disclosed no
relevant financial relationships.

CME Author
Laurie Barclay, MD, freelance writer and reviewer, Medscape, LLC. Disclosure: Laurie Barclay, MD, has disclosed no relevant
financial relationships.

Authors
Disclosures: Aron J. Hall, DVM, MSPH; Valerie G. Eisenbart, DVM; Amy Lehman Etingüe, DVM; L. Hannah Gould, PhD; Ben
A. Lopman, PhD; and Umesh D. Parashar, MBBS, have disclosed no relevant financial relationships.

Noroviruses are the leading cause of foodborne illness Prevention during 2001–2008. On average, 365 foodborne
in the United States. To better guide interventions, we norovirus outbreaks were reported annually, resulting in
analyzed 2,922 foodborne disease outbreaks for which an estimated 10,324 illnesses, 1,247 health care provider
norovirus was the suspected or confirmed cause, which visits, 156 hospitalizations, and 1 death. In 364 outbreaks
had been reported to the Foodborne Disease Outbreak attributed to a single commodity, leafy vegetables (33%),
Surveillance System of the Centers for Disease Control and fruits/nuts (16%), and mollusks (13%) were implicated most
commonly. Infected food handlers were the source of 53%
Author affiliations: Centers for Disease Control and Prevention, of outbreaks and may have contributed to 82% of outbreaks.
Most foods were likely contaminated during preparation
Atlanta, Georgia, USA (A.J. Hall, V.G. Eisenbart, A. Lehman
and service, except for mollusks, and occasionally, produce
Etingüe, L.H. Gould, B.A. Lopman, U.D. Parashar); University of
was contaminated during production and processing.
Illinois, Urbana, Illinois, USA (V.G. Eisenbart); and North Carolina
Interventions to reduce the frequency of foodborne norovirus
State University, Raleigh, North Carolina, USA (A. Lehman Etingüe) outbreaks should focus on food workers and production of
DOI: http://dx.doi.org/10.3201/eid1810.120833 produce and shellfish.

1566 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Foodborne Norovirus Outbreaks

N oroviruses, the leading cause of foodborne illness in

the United States, are responsible for an estimated
58% of all domestically acquired foodborne illness from
resulting in markedly increased recognition of norovirus
illnesses and outbreaks. Using more recent surveillance
data, we therefore sought to robustly describe foodborne
known agents (1). The estimated 5.5 million annual norovirus outbreaks, including temporal, geographic, and
foodborne norovirus illnesses in the United States, which demographic trends, and attribution to specific foods,
constitute those associated with recognized outbreaks settings, and contamination factors.
and those considered sporadic, result annually in 15,000
hospitalizations and 150 deaths and cost ≈$2 billion in health Methods
care expenses and lost productivity (1,2). Noroviruses are
also the leading cause of foodborne disease outbreaks Data Source
reported in the United States, accounting for about half Since 1973, the Centers for Disease Control and
of all foodborne outbreaks in which an etiologic agent is Prevention (CDC) has collected data on foodborne
identified (3–5). Classified into the genus Norovirus within disease outbreaks from state, local, and territorial health
the family Caliciviridae, noroviruses are a genetically departments through the Foodborne Disease Outbreak
diverse group of nonenveloped, single-stranded RNA Surveillance System (FDOSS) (20). A foodborne disease
viruses, comprising at least 5 genogroups (GI–GV) and >35 outbreak is defined as >2 similar illnesses resulting from
genotypes (6). Since 2001, noroviruses within the GII.4 ingestion of a common food. Data collected for each
genotype have caused most viral gastroenteritis outbreaks outbreak included outbreak characteristics (e.g., dates,
worldwide (7). number of ill persons, locations, etiologic agents), case-
Noroviruses can be spread through a variety of means, patient characteristics (e.g., demographic characteristics,
including direct person-to-person transmission through symptoms, health care seeking, and whether the illness
the fecal–oral route; ingestion of aerosolized vomitus; resulted in death), setting of food preparation, contributing
and indirect transmission through contaminated surfaces, factors, and the implicated food vehicle(s) (21). Outbreaks
food, or water. Norovirus has a low infectious inoculum of norovirus infection are considered laboratory confirmed
(>18 viral particles) and is shed copiously by ill persons if stool or vomitus specimens from >2 ill persons are
(105–1011 viral copies per gram of feces), which enables its positive for norovirus by reverse transcription PCR,
rapid and efficient spread (8–10). Noroviruses also remain enzyme immunoassay, or electron microscopy (6).
infectious on surfaces for as long as 2 weeks and in water Norovirus may be implicated as the suspected etiologic
for >2 months (11,12) and are resistant to many common agent in the absence of laboratory confirmation when
disinfectants (13,14). Foods can be contaminated with reasonable clinical or epidemiologic evidence exists, such
noroviruses at any point along the farm-to-fork continuum, as the previously validated Kaplan criteria (22). Data were
although the most frequent pathways are thought to be extracted for all foodborne outbreaks in which norovirus
through an infected food handler or exposure to water was either a suspected or confirmed etiologic agent and the
contaminated with fecal matter (e.g., surface water used for first illness occurred during 2001–2008.
produce irrigation or water containing sewage discharge
where shellfish grow) (15–17). Bivalve mollusks, such as Data Analysis
oysters, bioaccumulate noroviruses in their body through Frequencies of outbreaks and outbreak-related
filtration and selective binding mechanisms and therefore illnesses, health care provider visits, hospitalizations, and
are readily contaminated when they are grown in harvesting deaths were calculated. Annual variations were analyzed
areas contaminated with human feces (18). by grouping data into seasonal years from July–June and
Attribution of norovirus disease to specific foods and comparing data with information from the 2 adjacent years.
increasing understanding of the various contamination Seasonal trends and variations in reporting by state were
pathways that result in disease can help identify potential also assessed. Because data on health care provider visits,
targets for interventions. Although most foodborne hospitalizations, deaths, age groups, and sex were not
norovirus disease in the United States is not outbreak always reported (i.e., they were provided for 63%, 66%,
associated, outbreaks provide the most robust information 69%, 82%, and 89% of all reported illnesses, respectively),
about the foods that cause illness and the factors contributing the relative proportions of illnesses by age and outcome
to their contamination. The last published description of from the outbreaks that included such data were extrapolated
foodborne norovirus outbreaks in the United States was to all reported outbreak-associated illnesses. For example,
based on surveillance data from 1991 through 2000, before among the 68,452 outbreak-associated illnesses that
molecular diagnostic tools were widely available (19). Since included data on age, 22,301 (32.5%) occurred in persons
that time, norovirus diagnostics have become incorporated >50 years of age; this proportion (32.5%) was applied to
more routinely into public health outbreak investigations, the total number of outbreak-associated illnesses reported

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1567
RESEARCH

(82,591) to yield the estimated number of outbreak- Results

associated illnesses among those >50 years of age (26,872).
Rates of reported outbreaks and outbreak-associated Outbreak Characteristics
illnesses per 1,000,000 person-years were calculated by During 2001–2008, a total of 9,206 foodborne
dividing the average annual number of these illnesses by disease outbreaks were reported in the United States. For
the corresponding US intercensal estimate at the midpoint 2,922 (46%) of the 6,355 outbreaks with a known cause,
of the study period, July 2004 (23). norovirus was the confirmed or suspected cause, an average
Food vehicles implicated in outbreaks were classified of 365 foodborne norovirus outbreaks annually (1.2
on the basis of a categorization hierarchy of 17 mutually outbreaks/1,000,000 person-years). Of these, 1,683 (58%)
exclusive commodity groups (24). If a food contained outbreaks were laboratory confirmed, primarily on the
a single contaminated ingredient or all ingredients basis of reverse transcription PCR. Genogroup information,
belonged to a single commodity, it was classified into reported in most confirmed norovirus outbreaks only since
that commodity. Food vehicles that contained ingredients 2006, was available for 648 norovirus outbreaks, among
from multiple commodities were classified as complex. which 111 (17%) were caused by genogroup I (GI) and 518
Outbreaks were not attributed to any of the commodities if (80%) were caused by genogroup II (GII); 19 (3%) involved
an implicated food vehicle could not be assigned to one of both GI and GII noroviruses. Sixteen (0.5%) norovirus
these commodities, multiple food vehicles were implicated, outbreaks also involved other suspected or confirmed
or no specific food vehicle was implicated. etiologic agents, including nontyphoidal Salmonella
Factors contributing to contamination and methods spp. (n = 4), adenovirus (n = 2), Bacillus cereus (n = 2),
of food preparation were reported according to standard Campylobacter spp. (n = 2), Clostridium perfringens (n =
categorization schemes (21). Contributing factors were 2), Listeria monocytogenes (n = 1), Shigella spp. (n = 1),
not mutually exclusive and were further grouped into and Vibrio parahaemolyticus (n = 1).
the following categories: food handler contact, cross- Foodborne norovirus outbreaks were reported from 49
contamination during preparation, contaminated raw states (all but Vermont) and the District of Columbia, with
product, and insufficient cooking and/or heating. Outbreak substantial state-to-state variation in the rate of reported
reports also indicated whether a food handler was outbreaks (Figure 1). The highest per capita rates of reported
specifically implicated as the source of contamination outbreaks were in Minnesota (6.2/1,000,000 person-years)
(e.g., handled the implicated food while symptomatic) and Oregon (6.1/1,000,000 person-years); the lowest
rather than simply being a potential contributor. For reported rates were in Texas (0.25/1,000,000 person-years)
implicated foods that could be classified into 1 of the 17 and Kentucky (0.27/1,000,000 person-years). The greatest
commodities, the likely point of contamination (POC) number of outbreaks was reported by California (n = 526,
was assessed on the basis of a combination of contributing 18%). Residents from multiple states were affected in 138
factors and whether a food handler was implicated as (5%) outbreaks, of which 8 (6%) involved exposures to
the source of contamination. POC was categorized to contaminated foods distributed to multiple states.
distinguish production or processing (i.e., a raw product
contaminated from the environment or obtained from a
polluted source) versus preparation or service (i.e., vehicle
handled by infected food worker or cross-contamination
during preparation). Those outbreaks with insufficient or
conflicting information reported regarding the likely POC
were classified as unknown.
The settings of food preparation in outbreaks were
classified into the following categories: commercial (e.g.,
restaurant, grocery store, caterer), institutional (e.g., school,
nursing home, prison), private (e.g., home, church, picnic),
and other (including unknown); multiple settings could be
selected in a given outbreak. Differences in median number
of illnesses associated with outbreaks in different settings
were assessed by using Wilcoxon rank-sum tests. χ2 tests Figure 1. Total number and rate of reported foodborne norovirus
were used to evaluate trends among categorical variables. outbreaks per 1,000,000 person-years by affected states, United
States, 2001–2008. Number given in each state indicates the total
Analyses were performed by using SAS v9.2 (SAS Institute
number of outbreaks over the 8-year study period; shaded boxes in
Inc., Cary, NC, USA) and Epi Info v3.4.3 (CDC, Atlanta, key indicate the reported rate by quartiles. Multistate outbreaks are
GA, USA). assigned as outbreaks to each state involved.

1568 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Foodborne Norovirus Outbreaks

Outbreaks were slightly more frequent in the winter implicated food; among these outbreaks, food handler
months than in the rest of the year, with 957 (33%) contact with ready-to-eat food was identified in 725 (82%),
outbreaks occurring during December–February; however, consumption of a contaminated raw product in 111 (13%),
the seasonal pattern varied somewhat from year to year cross-contamination during preparation in 109 (12%), and
(Figure 2). The temporal pattern of outbreaks in which inadequate cooking or heating in 28 (3%). A food handler
norovirus was confirmed by laboratory testing was similar was specifically implicated as the source of contamination
to that of outbreaks in which norovirus was suspected as in 473 (53%) outbreaks. This determination was reportedly
the etiologic agent. Although no consistent secular trend made on the basis of laboratory and epidemiologic evidence
was observed over the 8-year study period, the number of (33%), epidemiologic evidence only (45%), laboratory
outbreaks in 2006–07 (n = 442) was 24% higher than the evidence only (4%), and prior experience only (17%); no
average number of outbreaks during the 2 adjacent seasonal reason for implicating a food handler was given for 2 of
years of 2005–06 (n = 359) and 2007–08 (n = 352). these outbreaks. No significant differences in contributing
On average, 10,324 reported illnesses were associated factors were identified between different settings where
with foodborne norovirus outbreaks each year (Table foods were prepared.
1). These included an estimated 1,247 (12%) health Among the 364 outbreaks with a single, simple
care provider visits, 156 (1.5%) hospitalizations, and 1 food implicated, the most frequent commodities were
(0.01%) death annually. Most (80%) outbreak-associated leafy vegetables (33%), fruits/nuts (16%), and mollusks
illnesses affected adults >20 years of age, and 56% of (13%), although all commodities except sprouts were
illnesses affected women. Children <5 years of age had a implicated in at least 1 outbreak (Figure 3). Information
significantly lower rate of foodborne norovirus outbreak– was available to indicate the likely POC in 191 (52%)
associated illness (11/1,000,000 person-years) than did all of these outbreaks, among which contamination during
other age groups combined (37/1,000,000 person-years) preparation or service was more frequent (85%, n = 162)
(p<0.001). Among outbreaks for which data were available, than was contamination during production or processing
the median incubation period was 33 hours (n = 2,348), and (15%, n = 29) (p<0.001). All 26 mollusk-associated
the median attack rate was 61% (n = 1,099). outbreaks in which the likely POC could be determined
Norovirus outbreaks involved foods prepared most often were caused by contamination during production or
in commercial settings (83%) and less frequently in private processing. Production or processing contamination was
(11%), institutional (8%), and other (12%) settings (Table 2). also indicated in 3 of the 109 outbreaks associated with
Outbreaks involving institutional settings were significantly either leafy greens or fruits/nuts in which a POC could be
larger in terms of total number of illnesses (median 36 determined. Outbreaks involving all other commodities
illnesses/outbreak) than those in other settings (median 15 resulted either from contamination during preparation
illnesses/outbreak) (p<0.001). The most commonly reported or service or from an unknown POC. A traceback was
settings were restaurants or delicatessens (62%), caterers performed in 8 outbreaks, all of which involved mollusks;
(11%), and private homes (10%). 3 of these resulted in a product recall.

Food Attribution Discussion

At least 1 food vehicle was implicated in 1,298 This study demonstrates the predominant etiologic
(44%) outbreaks, of which 534 (41%) involved a complex role of norovirus in foodborne disease outbreaks in the
food, 364 (28%) involved a simple food that could be United States, with an average of 1 foodborne norovirus
classified into one of the 17 commodities, 279 (21%)
involved multiple foods, and 121 (9%) involved a single
food item that could not be classified. The reasons for
implicating specific foods included statistical evidence
from epidemiologic investigations (55%); laboratory
evidence, such as identification of the agent in the food
(2%); compelling supportive information (30%); other
data (0.1%); and prior experience in the absence of specific
evidence (5%). For 8% of implicated foods, no reasons
were given. Of the 813 outbreaks that involved complex or
multiple foods, 328 (40%) implicated sandwiches, salads, Figure 2. Number of reported foodborne norovirus outbreaks by
month of first illness onset, United States, 2001–2008. Outbreaks
or other foods eaten raw or lightly cooked.
are confirmed as caused by norovirus if fecal or vomitus specimens
Factors that may have contributed to contamination from >2 persons are positive for the virus by reverse transcription
were provided in 886 (68%) of outbreaks with at least 1 PCR, electron microscopy, or enzyme immunoassay.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1569
RESEARCH

Table 1. Estimated annual number and rate of reported illnesses underscoring the need to better understand and control
associated with foodborne norovirus outbreaks by age, sex, and endemic norovirus disease as a means of foodborne disease
outcome, United States, 2001–2008 prevention. Contact with food handlers during preparation
Estimated annual outbreak-
associated illnesses*
was cited in 82% of outbreaks as a possible contributor to
Characteristic No. (%) Rate† contamination and was specifically implicated as the source
Age group, y of contamination in 53% of outbreaks. These proportions
<5 215 (2) 10.6 likely reflect lack of positive evidence and therefore may
5–19 1,827 (18) 30.3
20–49 4,923 (48) 39.8 be underestimated because of disincentives for reporting
>50 3,359 (33) 37.7 illness and asymptomatic infections among food handlers
Sex (25,26). Most foods implicated in norovirus outbreaks were
F 5,819 (56) 39.0
M 4,505 (44) 31.3
prepared in restaurants, delicatessens, and other commercial
Outcome settings, which suggests that these are key locations for
Health care provider visit 1,247 (12) 4.3 intervention. Steps to curtail contamination of ready-to-eat
Hospitalization 156 (1.5) 0.5
foods by food handlers in these settings include adherence
Death 1 (0.01) 0.002
Total Illnesses 10,324 (100) 35.2 to appropriate recommendations for hand washing and use
*Proportions of illnesses by age, sex, and outcomes among outbreaks for of gloves; compliance with policies to prevent ill staff from
which such data were reported were extrapolated to all reported foodborne
norovirus outbreak–associated illnesses.
working; and presence of a certified kitchen manager, as
†Reported rate per 1,000,000 person-years calculated by dividing the recommended by the Food Code of the US Food and Drug
number of illnesses by the corresponding US intercensal estimate at the
study period midpoint (July 2004) (23).
Administration (27).
With the exception of mollusk-associated outbreaks,
outbreak reported every day, and highlights the frequency we rarely identified contamination during production
of norovirus contamination of raw and other ready-to-eat or processing in this analysis, although it is likely
foods. Fresh produce, primarily leafy vegetables and fruits, underrecognized in norovirus outbreaks. Indeed, the point
was implicated in over half of all outbreaks that could be of contamination could not be determined in ≈50% of single-
classified into a single commodity; ready-to-eat foods that commodity outbreaks, many of which may be additional
contain fresh produce, such as sandwiches and salads, were instances of contamination before preparation. Oysters are
frequently implicated complex vehicles. Mollusks, which well documented as food vehicles that are prone to norovirus
are also commonly served raw or undercooked, were also contamination during production (17,28), but examples of
implicated frequently. Food handler contact with raw and contaminated produce have also been reported, including
ready-to-eat foods was identified as the most common raspberries and lettuce, which are likely contaminated from
scenario resulting in foodborne norovirus outbreaks, irrigation waters (16,29). In a recent multicountry study,
Table 2. Settings of food preparation in foodborne norovirus outbreaks, United States, 2001–2008*
Setting† No. (%) outbreaks Median no. cases/outbreak (IQR)
Commercial 2,432 (83) 14 (7–26)
Restaurant or delicatessen 1,824 (62) 12 (6–23)
Caterer 313 (11) 25 (16–40)
Banquet facility 110 (4) 30 (19–59)
Grocery store 105 (4) 13 (9–18)
Commercial product served without further preparation 42 (1) 20 (11–30)
Wedding reception 22 (1) 25 (18–35)
Fair, festival, other temporary/mobile service 16 (1) 21 (6–50)
Institutional 248 (8) 36 (19–71)
School 91 (3) 43 (29–92)
Nursing home 66 (2) 35 (19–60)
Camp 29 (1) 34 (18–53)
Noncafeteria workplace 18 (1) 23 (13–56)
Prison or jail 15 (1) 58 (23–137)
Workplace cafeteria 12 (0.4) 22 (18–78)
Hospital 11 (0.4) 53 (16–72)
Day care center 4 (0.1) 31 (19–65)
Office setting 2 (0.1) 18 (15–21)
Private 336 (11) 16 (10–26)
Private home 296 (10) 15 (9–25)
Church, temple, or other place of worship 31 (1) 23 (12–40)
Picnic 9 (0.3) 19 (12–32)
Other/Unknown 267 (9) 28 (14–49)
All outbreaks 2,922 (100) 14 (8–30)
*IQR, interquartile range.
†Multiple locations may be implicated in a given outbreak.

1570 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Foodborne Norovirus Outbreaks

Figure 3. Commodity and point of contamination implicated in reported norovirus outbreaks involving simple foods (consisting of a single
commodity; n = 364), United States, 2001–2008. Point of contamination was classified as unknown if insufficient or conflicting information
was provided in outbreak report.

norovirus RNA was detected in 28%–50% of samples from prioritizations, and limited resources are key challenges to
leafy green vegetables and in 7%–34% of samples from improving the knowledge base for effective prevention and
soft red fruit (e.g., strawberries and raspberries) obtained control of foodborne norovirus disease.
from retail markets or directly from processing companies Although overall reporting of foodborne norovirus
(30). These findings demonstrate that fresh produce outbreaks has vastly improved since the 1990s (19), great
often comes in contact with norovirus during production. disparity remains among states. A 25-fold difference
However, because the available detection methods cannot in population-based rates of foodborne norovirus
distinguish infectious virus from noninfectious genomic outbreaks was identified between the highest and lowest
material, the specific risks posed to public health from reporting states. Although these differences may be due
this potential contamination of produce remain unclear. in part to true variations in the incidence of foodborne
Outbreak investigations can provide learning opportunities norovirus outbreaks, they also suggest varying degrees
to possibly link such contamination with human illness of underreporting. Thus, rates of outbreaks and outbreak-
through tracebacks and root-cause analyses to determine associated illnesses captured through surveillance
when, where, and how foods became contaminated. represent reporting rates, not true incidence, and likely
Unfortunately, no tracebacks were performed during underestimate the true incidence. Capacity of state and
produce-associated outbreaks included in this analysis, local health departments to investigate foodborne disease
despite evidence in some of these outbreaks that suggested outbreaks varies widely, with the most notable limitations
possible contamination before preparation. being lack of dedicated personnel and delayed notification
No specific food vehicle was implicated in 56% of outbreaks (31). Efforts to better understand the gaps
of the foodborne norovirus outbreaks included in this in foodborne outbreak response, including laboratory,
analysis, a substantially higher proportion than that for epidemiologic, and environmental health capacity, may
foodborne outbreaks with bacterial etiologic agents (3– ultimately inform strategies to overcome the challenges of
5). This discrepancy underscores the challenges in food limited public health resources (32).
attribution of norovirus outbreaks, given the potential for The increasing trend in the 1990s in the number of
multiple transmission pathways, contamination of multiple foodborne norovirus outbreaks reported appears to have
vehicles by an ill food handler, and time lags in reporting leveled off, likely reflecting the availability of diagnostic
by citizen complaint. However, the discrepancy may testing for norovirus at almost all public health laboratories
also reflect the relative deprioritization of investigating across US states. However, a substantial increase was
suspected norovirus outbreaks. Misperceptions may exist observed during 2006–07, contemporaneous with the
that foodborne norovirus outbreaks result only from local emergence of 2 new GII type 4 (GII.4) norovirus variants
contamination events and afford little opportunity for (7). Interestingly, the surge in foodborne outbreaks during
further prevention and/or control. Such decisions regarding the 2006–07 epidemic season appeared less pronounced than
investigations are often made in the face of scant public that observed among norovirus outbreaks in general (7,33),
health resources, inadequate staffing, and competing most of which result from direct person-to-person spread (34),
priorities (31). Overcoming these misperceptions, as well sporadic norovirus-associated hospitalizations and

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1571
RESEARCH

deaths in the United States (35,36). Furthermore, foodborne Acknowledgments

norovirus outbreaks did not increase in association with We gratefully acknowledge the state and local health
emergence of a new GII.4 variant in the 2002–03 seasonal year, department staff who submitted data on outbreak investigations
which has been previously correlated with other indicators to CDC. We also thank Karen Herman, Kelly Walsh, and Dana
of increased norovirus activity (7,35,36). These observations Cole for assistance with data extraction and colleagues at the
coupled with the paucity of genotype information reported US Department of Agriculture, Food and Drug Administration,
directly through FDOSS underscores the need for and and Food Safety and Inspection Service for their review and
importance of CaliciNet, the recently implemented national comments on this manuscript.
laboratory surveillance network for norovirus outbreaks
This project was supported in part by Agriculture and Food
(37). Integration of sequence-based genotyping data with
Research Initiative Competitive Grant no. 2011-68003-30395
epidemiologic data will enable timely recognition of the
from the US Department of Agriculture, National Institute of
impact of emergent noroviruses on foodborne disease and
Food and Agriculture.
potentially identify links between outbreaks due to widely
distributed food vehicles. Additionally, beginning in 2009, This study was presented in part at the 4th International
US data on enteric disease outbreaks associated with any Conference on Caliciviruses, Santa Cruz, Chile, October 16–19,
mode of transmission, i.e., foodborne and nonfoodborne, 2010; the International Association for Food Protection Annual
are systematically collected through a single comprehensive Meeting, Milwaukee, Wisconsin, USA, July 21–August 3, 2011;
system, the National Outbreak Reporting System (NORS), and the 7th Annual OutbreakNet Conference, Long Beach,
allowing for attribution of all norovirus outbreaks by mode California, USA, September 19–22, 2011.
of transmission and setting (38).
Dr Hall is an epidemiologist with the Viral Gastroenteritis
Despite the demonstrated prevalence and effects of
Team in the Division of Viral Diseases, National Center for
these infections, relatively few evidence-based interventions
Immunization and Respiratory Diseases, CDC. His research
exist for preventing and controlling foodborne norovirus
interests focus on all aspects of the epidemiology of noroviruses
disease. This study provides a comprehensive analysis
and other agents of viral gastroenteritis.
of US foodborne norovirus outbreaks, highlighting both
potential targets for interventions and key remaining data
gaps. Fresh produce, mollusks, and ready-to-eat foods were References
identified as the primary vehicles of foodborne norovirus
1. Scallan E, Hoekstra RM, Angulo FJ, Tauxe RV, Widdowson M-A,
disease. Continued focus on food handlers in retail Roy SL, et al. Foodborne illness acquired in the United States—
settings is warranted, although the role and mechanisms major pathogens. Emerg Infect Dis. 2011;17:7–15. http://dx.doi.
of contamination during production, particularly as they org/10.3201/eid1701.09-1101p1
relate to fresh produce, remain unclear. To that end, 2. Batz MB, Hoffmann S, Morris JG Jr. Ranking the risks: the 10
pathogen-food combinations with the greatest burden on public
further development and validation of analytic methods health: University of Florida; 2011 [cited 2012 Jul 26]. http://www.
for the detection of norovirus in foods, particularly those rwjf.org/files/research/72267report.pdf
that can quantitatively distinguish contamination levels 3. Centers for Disease Control and Prevention. Surveillance for food-
correlating with infection risk and distinguish infectious borne disease outbreaks—United States, 2006. MMWR Morb Mor-
tal Wkly Rep. 2009;58:609–15.
from noninfectious virus, are critically needed (32). 4. Centers for Disease Control and Prevention. Surveillance for food-
Likewise, there is ongoing need to define the duration of borne disease outbreaks—United States, 2007. MMWR Morb Mor-
infectiousness of ill food handlers and to elucidate the role tal Wkly Rep. 2010;59:973–9.
of asymptomatic food handlers in norovirus transmission, 5. Centers for Disease Control and Prevention. Surveillance for food-
borne disease outbreaks—United States, 2008. MMWR Morb Mor-
issues that have been hampered by the inability to culture tal Wkly Rep. 2011;60:1197–202.
human noroviruses in vitro. Other areas needing attention 6. Centers for Disease Control and Prevention. Updated norovirus
are improving understanding of the roles of various outbreak management and disease prevention guidelines. MMWR
norovirus genotypes in different foods, identifying the Recomm Rep. 2011;60(RR-3):1–18.
7. Zheng DP, Widdowson MA, Glass RI, Vinje J. Molecular epidemi-
environmental antecedents leading to food contamination, ology of genogroup II–genotype 4 noroviruses in the United States
and the role of other transmission modes in propagating between 1994 and 2006. J Clin Microbiol. 2010;48:168–77. http://
norovirus spread. Initiatives to improve outbreak dx.doi.org/10.1128/JCM.01622-09
surveillance and response can help address some of these 8. Teunis PF, Moe CL, Liu P, Miller SE, Lindesmith L, Baric RS, et al.
Norwalk virus: how infectious is it? J Med Virol. 2008;80:1468–76.
data gaps (35,37–39), but only if public health agencies http://dx.doi.org/10.1002/jmv.21237
have adequate resources. Finally, as the development of 9. Atmar RL, Opekun AR, Gilger MA, Estes MK, Crawford SE, Neill
candidate norovirus vaccines continues to progress (40), FH, et al. Norwalk virus shedding after experimental human infec-
their possible public health utility in specific populations, tion. Emerg Infect Dis. 2008;14:1553–7. http://dx.doi.org/10.3201/
eid1410.080117
such as food handlers, should be assessed.

1572 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Foodborne Norovirus Outbreaks

10. Aoki Y, Suto A, Mizuta K, Ahiko T, Osaka K, Matsuzaki Y. Dura- 26. Todd EC, Greig JD, Bartleson CA, Michaels BS. Outbreaks where
tion of norovirus excretion and the longitudinal course of viral load food workers have been implicated in the spread of foodborne dis-
in norovirus-infected elderly patients. J Hosp Infect. 2010;75:42–6. ease. Part 5. Sources of contamination and pathogen excretion from
http://dx.doi.org/10.1016/j.jhin.2009.12.016 infected persons. J Food Prot. 2008;71:2582–95.
11. Cheesbrough JS, Barkess-Jones L, Brown DW. Possible pro- 27. US Food and Drug Administration. Food code, 2005 [cited 2012 Jul
longed environmental survival of small round structured viruses. 26]. http://www.fda.gov/Food/FoodSafety/RetailFoodProtection/
J Hosp Infect. 1997;35:325–6. http://dx.doi.org/10.1016/S0195- FoodCode/FoodCode2009/default.htm
6701(97)90230-9 28. Le Guyader FS, Atmar RL, Le Pendu J. Transmission of viruses
12. Seitz SR, Leon JS, Schwab KJ, Lyon GM, Dowd M, McDaniels M, through shellfish: when specific ligands come into play. Curr Opin Vi-
et al. Norovirus infectivity in humans and persistence in water. Appl rol. 2012;2:103–10. http://dx.doi.org/10.1016/j.coviro.2011.10.029
Environ Microbiol. 2011;77:6884–8. http://dx.doi.org/10.1128/ 29. Ethelberg S, Lisby M, Bottiger B, Schultz AC, Villif A, Jensen T, et
AEM.05806-11 al. Outbreaks of gastroenteritis linked to lettuce, Denmark, January
13. Park GW, Barclay L, Macinga D, Charbonneau D, Pettigrew CA, 2010. Euro Surveill. 2010;15: pii: 19484..
Vinje J. Comparative efficacy of seven hand sanitizers against mu- 30. Baert L, Mattison K, Loisy-Hamon F, Harlow J, Martyres A, Leb-
rine norovirus, feline calicivirus and GII.4 norovirus. J Food Prot. eau B, et al. Review: norovirus prevalence in Belgian, Canadian and
2010;73:2232–8. French fresh produce: a threat to human health? Int J Food Micro-
14. Park GW, Sobsey MD. Simultaneous comparison of murine norovi- biol. 2011;151:261–9. http://dx.doi.org/10.1016/j.ijfoodmicro.
rus, feline calicivirus, coliphage MS2, and GII.4 norovirus to evalu- 2011.09.013
ate the efficacy of sodium hypochlorite against human norovirus 31. Centers for Disease Control and Prevention. Food safety epidemi-
on a fecally soiled stainless steel surface. Foodborne Pathog Dis. ology capacity in state health departments—United States, 2010.
2011;8:1005–10. http://dx.doi.org/10.1089/fpd.2010.0782 MMWR Morb Mortal Wkly Rep. 2011;60:1701–4.
15. Lynch M, Painter J, Woodruff R, Braden C. Surveillance for food- 32. Biggerstaff GK, Wright JG, Mitchell JR. FoodCORE: foodborne
borne-disease outbreaks—United States, 1998–2002. MMWR Sur- diseases centers for outbreak response enhancement—improving
veill Summ. 2006;55:1–42. foodborne disease outbreak response capacity in state and local
16. Falkenhorst G, Krusell L, Lisby M, Madsen SB, Bottiger B, Molbak health departments. In: 2012 International Conference on Emerg-
K. Imported frozen raspberries cause a series of norovirus outbreaks ing Infectious Diseases program and abstracts book, March 11–14,
in Denmark, 2005. Euro Surveill. 2005;10:E050922.2. Atlanta, Georgia, USA [cited 2012 Jul 26]. http://www.iceid.org/
17. Westrell T, Dusch V, Ethelberg S, Harris J, Hjertqvist M, Jourdan-da images/iceid_2012_finalprogram_final.pdf
Silva N, et al. Norovirus outbreaks linked to oyster consumption in 33. Centers for Disease Control and Prevention. Norovirus activ-
the United Kingdom, Norway, France, Sweden and Denmark, 2010. ity—United States, 2006–2007. MMWR Morb Mortal Wkly Rep.
Euro Surveill. 2010;15: pii: 19524. 2007;56:842–6.
18. Maalouf H, Schaeffer J, Parnaudeau S, Le Pendu J, Atmar RL, 34. Yen C, Wikswo ME, Lopman BA, Vinje J, Parashar UD, Hall AJ.
Crawford SE, et al. Strain-dependent norovirus bioaccumulation in Impact of an emergent norovirus variant in 2009 on norovirus out-
oysters. Appl Environ Microbiol. 2011;77:3189–96. http://dx.doi. break activity in the United States. Clin Infect Dis. 2011;53:568–71.
org/10.1128/AEM.03010-10 http://dx.doi.org/10.1093/cid/cir478
19. Widdowson MA, Sulka A, Bulens SN, Beard RS, Chaves SS, Ham- 35. Lopman BA, Hall AJ, Curns AT, Parashar UD. Increasing rates of
mond R, et al. Norovirus and foodborne disease, United States, gastroenteritis hospital discharges in US adults and the contribution
1991–2000. Emerg Infect Dis. 2005;11:95–102. http://dx.doi. of norovirus, 1996–2007. Clin Infect Dis. 2011;52:466–74. http://
org/10.3201/eid1101.040426 dx.doi.org/10.1093/cid/ciq163
20. Centers for Disease Control and Prevention. Foodborne disease out- 36. Hall AJ, Curns AT, McDonald LC, Parashar UD, Lopman BA. The
break surveillance. September 2011 [cited 2012 Jul 13]. http://www. roles of Clostridium difficile and norovirus among gastroenteritis-
cdc.gov/outbreaknet/surveillance_data.html associated deaths in the United States, 1999–2007. Clin Infect Dis.
21. Centers for Disease Control and Prevention. Electronic Foodborne 2012;55:216–23. http://dx.doi.org/10.1093/cid/cis386
Outbreak Reporting System: investigation of a foodborne outbreak 37. Vega E, Barclay L, Gregoricus N, Williams K, Lee D, Vinje J. Novel
(CDC form 52.13). November 2004 [cited 2012 Feb 22]. http:// surveillance network for norovirus gastroenteritis outbreaks, United
www.cdc.gov/outbreaknet/toolkit/efors_form.pdf States. Emerg Infect Dis. 2011;17:1389–95.
22. Turcios RM, Widdowson MA, Sulka AC, Mead PS, Glass RI. Re- 38. Manikonda KL, Wikswo ME, Roberts VA, Richardson L, Gould LH,
evaluation of epidemiological criteria for identifying outbreaks of Yoder JS, et al. The national outbreak reporting system: preliminary
acute gastroenteritis due to norovirus: United States, 1998–2000. results of the first year of surveillance for multiple modes of trans-
Clin Infect Dis. 2006;42:964–9. http://dx.doi.org/10.1086/500940 mission—United States, 2009. In: 2012 International Conference on
23. National Center for Health Statistics. Postcensal estimates of the Emerging Infectious Diseases program and abstracts book, March
resident population of the United States for July 1, 2000–July 1, 11–14, Atlanta, Georgia, USA [cited 2012 Jul 26]. http://www.iceid.
2009, by year, county, age, bridged race, Hispanic origin, and sex org/images/iceid_2012_finalprogram_final.pdf
(Vintage 2009). June 20, 2010 [cited 2011 May 11]. http://www.cdc. 39. Centers for Disease Control and Prevention. National Voluntary En-
gov/nchs/nvss/bridged_race.htm vironmental Assessment Information System (NVEAIS). December
24. Painter JA, Ayers T, Woodruff R, Blanton E, Perez N, Hoekstra RM, 2011 [cited 2012 April 3]. http://www.cdc.gov/nceh/ehs/EHSNet/
et al. Recipes for foodborne outbreaks: a scheme for categorizing and resources/nveais.htm
grouping implicated foods. Foodborne Pathog Dis. 2009;6:1259–64. 40. Atmar RL, Bernstein DI, Harro CD, Al-Ibrahim MS, Chen WH,
http://dx.doi.org/10.1089/fpd.2009.0350 Ferreira J, et al. Norovirus vaccine against experimental human
25. Phillips G, Tam CC, Rodrigues LC, Lopman B. Prevalence and Norwalk virus illness. N Engl J Med. 2011;365:2178–87. http://
characteristics of asymptomatic norovirus infection in the commu- dx.doi.org/10.1056/NEJMoa1101245
nity in England. Epidemiol Infect. 2010;138:1454–8. http://dx.doi.
org/10.1017/S0950268810000439 Address for correspondence: Aron J. Hall, Centers for Disease Control
and Prevention, 1600 Clifton Rd NE, Mailstop A34, Atlanta, GA 30333,
USA; email: ajhall@cdc.gov

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1573
RESEARCH

Constant Transmission Properties

of Variant Creutzfeldt-Jakob
Disease in 5 Countries
Abigail B. Diack, Diane Ritchie, Matthew Bishop, Victoria Pinion, Jean-Philippe Brandel,
Stephane Haik, Fabrizio Tagliavini, Cornelia Van Duijn, Ermias D. Belay, Pierluigi Gambetti,
Lawrence B. Schonberger, Pedro Piccardo, Robert G. Will,1 and Jean C. Manson1

Variant Creutzfeldt-Jakob disease (vCJD) has been condition in humans. vCJD was first reported in the United
reported in 12 countries. We hypothesized that a common Kingdom in 1996 (1) and was most likely caused by dietary
strain of agent is responsible for all vCJD cases, regardless exposure to contaminated products from cattle that had
of geographic origin. To test this hypothesis, we inoculated bovine spongiform encephalopathy (BSE). The similarity
strain-typing panels of wild-type mice with brain material from between BSE and vCJD was shown by experimental
human vCJD case-patients from France, the Netherlands,
transmission of the 2 diseases into standard panels of
Italy, and the United States. Mice were assessed for clinical
disease, neuropathologic changes, and glycoform profile;
inbred wild-type mouse lines (RIII, C57BL, and VM [1,2])
results were compared with those for 2 reference vCJD and into FVB mice (3).
cases from the United Kingdom. Transmission to mice The strain properties of vCJD and BSE have been
occurred from each sample tested, and data were similar extensively characterized in these sets of mice by using
between non-UK and UK cases, with the exception of the a combination of the order in which each strain of mouse
ranking of mean clinical incubation times of mouse lines. dies of disease (incubation period rankings), distribution of
These findings support the hypothesis that a single strain brain vacuolation at the terminal stage of disease (lesion
of infectious agent is responsible for all vCJD infections. profiles), distribution pattern of abnormal prion protein
However, differences in incubation times require further (PrPSc) in the brain, and glycosylation pattern of PrPSc as
subpassage in mice to establish any true differences in assessed by Western blot analysis. BSE and vCJD produce
strain properties between cases.
highly reproducible and similar incubation period rankings
and neuropathology, which indicates that they are the same

V ariant Creutzfeldt-Jakob disease (vCJD) is an acquired

transmissible spongiform encephalopathy (TSE),
or prion disease, that results in a fatal neurodegenerative
strain in the RIII, C57BL, and VM mice, whereas in the
FVB mice, both diseases exhibit the same pattern of PrPSc
deposition (3–5).
During 1996–2011, a total of 176 cases of definite or
Author affiliations: The Roslin Institute, Easter Bush, Scotland,
probable vCJD were reported in the United Kingdom; the
UK (A.B. Diack, V. Pinion, J.C. Manson); University of Edinburgh,
number of deaths peaked at 28 in 2000. Since 2006, deaths
Edinburgh, Scotland, UK (D. Ritchie, M. Bishop, R.G. Will); Cellule
from vCJD have leveled off to 2–5 per year (6). Although
Nationale de Référence des Maladies de Creutzfeldt-Jakob,
vCJD was originally restricted to the United Kingdom, 49
Universite Pierre et Marie Curie-Paris 6, INSERM, and CNRS,
cases have been reported in 11 other countries, including
Paris, France (J.-P. Brandel, S. Haik); Fdazione IRCCS Istituto
some outside Europe, bringing the worldwide total
Neurologico Carlo Besta, Milan, Italy (F. Tagliavini); Erasmus
number of cases to 225. Most cases outside the United
University Medical School, Rotterdam, the Netherlands (C. Van
Kingdom have occurred in France, which is believed to
Duijn); Centers for Disease Control and Prevention, Atlanta,
be related to the large volume of beef imports; ≈60% of
Georgia, USA (E.D. Belay, L.B. Schonberger); Case Western
UK beef exports to Europe were destined for France (7).
Reserve University, Cleveland, Ohio, USA (P. Gambetti); and Food
This possible relationship is borne out by a peak in vCJD
and Drug Administration, Rockville, Maryland, USA (P. Piccardo)

DOI: http://dx.doi.org/10.3201/eid1810.120792 1
Joint senior authors.

1574 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Transmission Properties of vCJD

cases in France 5 years after a similar peak in cases in the Materials and Methods
United Kingdom, which fits in well with the timing of an
increase in beef imports from the United Kingdom during CJD Inocula
1985–1995 (7). Frozen brain tissue consisting of ≈3 g of frontal cor-
More recently, comparative studies of vCJD cases tex from each of 4 vCJD case-patients originating from
from France and the United Kingdom have shown evidence the Netherlands, Italy, France, and the United States was
from clinical, epidemiologic, pathologic, and biochemical available for transmission studies (Table). An additional
analyses that a common strain of agent may be responsible similar sample of cerebellum from the vCJD case-patient
for vCJD infection in both countries (8). In Europe, active from Italy was also studied to assess any differences in
surveillance to monitor BSE cases was implemented in transmission characteristics between different regions of
2001 (6,9). Although it appears that exports of meat, cattle, the brain. Tissue samples were homogenized at 10% (wt/
or both from the United Kingdom may have played a major vol) concentration in sterile physiologic saline and stored
role in the incidence of vCJD cases in other countries (10), at −20°C until use. Ethical consent for the use of these
indigenous BSE from before 2001 or another unidentified materials for research was obtained and approved by the
source may have caused some vCJD infections. Lothian National Health Service Board Research Ethics
To determine whether vCJD cases in different Committee (reference: 2000/4/157).
countries have been caused by the same infectious agent,
samples from 4 vCJD case-patients from the Netherlands, Experimental Animals
Italy, France, and the United States were made available for Two lines of mice expressing Prn-pa (RIII, C57BL)
strain typing analysis using a standard panel of wild-type and 1 line expressing Prn-pb (VM) were used for the
mouse lines. Each sample was methionine homozygous at transmission experiments. The Prn-pa and Prn-pb alleles
codon 129 of the prion gene (Table). None of the patients have a major influence on the incubation period of disease,
had received blood or organ donations. Two patients (from with each TSE strain having a distinct and reproducible
Italy and the United States) were treated with quinacrine. incubation period ranking with each of the Prn-p genotypes
All case-patients showed classic clinical characteristics of (11). Mice were anesthetized with halothane and inoculated
vCJD; postmortem examination confirmed the diagnosis. with brain homogenate by a combination of intracerebral
We conducted transmission studies in mice using brain (i.c.) (0.02 mL) and intraperitoneal (0.1 mL) routes to
samples from these 4 vCJD case-patients and compared ensure efficient uptake of the agent. Because the quantity
transmission characteristics of all 4 cases to those of 2 of material available was limited, mice received only i.c.
historical cases of vCJD in the United Kingdom. inoculations (0.02 mL) from the case-patient from the
United States.
Mice were scored weekly for signs of clinical
neurologic disease from 100 days as described by
Table. Demographic and clinical features of case-patients with variant CJD from the Netherlands, France, Italy, and United States and
2 reference case-patients from the United Kingdom*
The United United Kingdom
Characteristic Netherlands France Italy States 1 2
Case-patient sex F F F F M M
Case-patient age at illness onset, y 24 36 25 22 24 35
Case-patient age at death, y 26 37 27 24 25 36
Disease duration, mo 19 14 27 32 14 12
Early psychiatric symptoms Yes Yes Yes Yes Yes Yes
Persistent painful sensory symptoms Yes No Yes No Yes No
Ataxia Yes Yes Yes Yes Yes Yes
Myoclonus, dystonia, or chorea Yes Yes Yes Yes Yes Yes
Dementia Yes Yes Yes Yes Yes Yes
No typical appearance of sporadic CJD on EEG Yes Yes No Yes Yes Yes
Bilateral symmetric pulvinar high signal on MRI scan Yes No Yes Yes Yes No
of brain
Positive tonsil biopsy result ND Yes Yes Yes ND ND
Treatment No specific No specific Quinacrine Quinacrine No specific No specific
treatment treatment treatment treatment
History of travel to or residence in the United No No No Yes† Yes Yes
Kingdom
Codon 129MM Yes Yes Yes Yes Yes Yes
Type 2B PrP Yes Yes Yes Yes Yes Yes
*CJD, Creutzfeldt-Jakob disease; EEG, electroencephalogram; MRI, magnetic resonance imaging; ND, not done; PrP, prion protein.
†Born in United Kingdom in 1979, moved to United States in 1992.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1575
RESEARCH

Fraser and Dickinson (12). Mice were killed by cervical Results

dislocation whether due to TSE or other nonspecific disease All 5 vCJD brain isolates transmitted successfully to
and the brain removed at postmortem. Brains were cut the wild-type mouse panel with the appearance of clinical
sagitally; half was snap-frozen for biochemical analysis, and pathologic signs associated with prion disease. Mean
and the remaining half was fixed in 10% formol saline for incubation periods and mouse line ranking order were
histologic analysis. Experiments were approved by The calculated and compared with reference data from the
Roslin Institute’s Ethical Review committee and were United Kingdom. The inocula from the Netherlands, Italy
conducted according to the regulations of the UK Home (cortex and cerebellum), and the United States showed
Office Animals (Scientific Procedures) Act 1986. variation in the temporal order of appearance of clinical
symptoms in each of the mouse strains when compared with
Scoring of Vacuolation the UK reference cases (Figure 1). These 3 brain inocula
After fixation, mouse brains were treated for 1.5 hours showed the shortest incubation periods in the RIII mice,
in 98% formic acid to reduce the infective titer for safety occurring at ≈365 days postinoculation (dpi), followed by
reasons. Brains were then cut transversely into 4 sections VM at ≈450 dpi; the longest was found in the C57BL mice
(hind brain, midbrain, forebrain, and hippocampus/ (≈465 dpi). The ranking of incubation time differed with
thalamus) and embedded in paraffin. Samples underwent brain inocula from France, with the C57BL mice dying of
hematoxylin and eosin staining, and each mouse was disease before the VM mouse line, similar to the pattern
scored for degree of vacuolation (ranked 1–5, with 1 the observed in the UK reference cases. The exact timing of the
least affected) found in 9 standard gray matter regions and appearance of clinical symptoms varied, with wide ranges
3 white matter regions. Scores for >6 mice were averaged observed for each line of mouse within and between the
to produce a mean lesion profile for each of these areas different inocula; incubation periods were 301–463 days in
(12). RIII mice, 361–553 days in VM mice, and 335–553 days
in C57BL mice.
PrP Immunohistochemical Analysis
Paraffin-embedded tissue sections were pretreated Lesion Profiles
to aid antigen retrieval by autoclaving for 10 min at Sufficient numbers of wild-type mice scored positive
121°C. Sections were then immersed in formic acid for for the presence of TSE-associated vacuolation to enable
10 min to enhance PrP labeling. PrP immunostaining was lesion profiles to be generated from their mean scores.
performed by using the monoclonal anti-PrP antibody 6H4 Similar patterns of vacuolation distribution were seen for
(Prionics, Schlieren-Zurich, Switzerland) at a dilution of all vCJD isolates, including those taken from either the
1:10,000 overnight at room temperature. The secondary cortex or cerebellum (case-patient from Italy) (Figure 2).
antibody was biotinylated rabbit anti-mouse (Jackson Mouse strains of the same PrP genotype showed close
ImmunoResearch Laboratories, Inc., West Grove, PA, similarities in lesion profiles with Prn-pa mice (RIII and
USA) used at 1:400 for 1 h. The Vectastain Elite ABC C57BL), showing moderate gray matter vacuolation of the
Kit (Vector Laboratories, Burlingame, CA, USA) was medulla, hypothalamus, and septum, with C57BL mice
used, and visualization of antibody binding was through additionally exhibiting mild white matter vacuolation of
deposition of 3,3′-diaminobenzidine chromogen. the basal cerebellar peduncle. Prn-pb mice (VM) showed
mild to moderate gray matter vacuolation of the medulla,
Western Blot Analysis of PrPSc superior colliculus, thalamus, and septum. Some variability
Frozen brain samples were homogenized in 0.9% was found in the intensity of vacuolation dependent on the
saline solution to yield a 10% suspension. This material vCJD isolate, most notably in the VM mice and particularly
was cleared by centrifugation and the supernatant treated in the superior colliculus, hypothalamus, and thalamic
with 50 μg/mL proteinase K for 1 h at 37°C (for detailed regions.
methods, see Head et al. [13]). The digested product
was denatured and then loaded onto a 10% Bis/Tris Immunohistochemical Analysis of PrP
NuPAGE Novex gel (Invitrogen, Paisley, UK). After A total of 9 mice per inoculum, corresponding to 3
electrophoresis, the gel was blotted onto polyvinylidene animals per genotype, were analyzed in a blinded experiment
fluoride membranes. Detection of prion protein used the design. Animals from each group showed variability in
enhanced chemiluminescence ECL+ technique (Amersham the amount of PrP accumulation (Figure 3). Fine punctate
Biosciences, Little Chalfont, UK) with primary antibody deposits were the most consistently observed pattern of
6H4 at 1:30,000 and an anti-mouse IgG peroxidase-linked PrP accumulation in mice from all genotypes. However,
secondary antibody (Amersham Biosciences) at 1:60,000. in numerous instances, PrP plaques were also seen in the
Images were captured on radiographic film. corpus callosum and subventricular area and, eventually, in

1576 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Transmission Properties of vCJD

strain of agent is present in each inoculate. C57BL mice

showed more variability in PrPSc deposition, which may
be a result of the transmission of the vCJD agent across
a species barrier. Each brain isolate produced a type 2B–
like glycoform profile on biochemical analysis; similar to
results from previous studies, the unglycosylated fragment
appeared as a doublet (14).
We observed differences in the timing of the onset of
clinical signs of disease for each of the non-UK cases, which
may be because of differences in the infectious titer of the
different vCJD isolates. We also identified a change in the
Figure 1. Comparison of variant Creutzfeldt-Jakob disease
incubation periods from 5 sources in wild-type mice. Data
show mean incubation period ± SEM. i.c., intracerebral; i.p.,
intraperitoneal; UK, United Kingdom.

the brain parenchyma of the cerebrum and cerebellum. VM

mice showed a similar pattern of PrP deposition with each
inoculum and consistently showed PrP plaques. RIII mice
appeared to show less staining with the accumulation of
PrP forming fine punctate, coarse, pericellular, and plaque
deposits. In the C57BL line, PrP deposition appeared more
variable within and between inocula.

Biochemical Analysis of PrP

Biochemical analysis of PrPSc by Western blot produced
a type 2B–like gel mobility and glycosylation profile in
all mouse strains (Figure 4). This profile is characterized
by the predominance of the diglycosylated form of PrPSc
and an ≈19-kDa unglycosylated fragment, similar to that
seen in all vCJD human brain tissue examined to date.
The unglycosylated fragment appeared as a doublet with a
distinct upper and lower band and the relative intensity of
the 2 bands varying with mouse Prn-p genotype; the lower
band appeared more intense in Prn-pa mice and the higher
band more intense in Prn-pb mice. The biochemical profile
was identical for all 5 brain isolates and identical to the
pattern observed in the UK cases.

Discussion
We used transmission studies with wild-type mice to
define the strain characteristics of geographically distinct
cases of vCJD to establish whether a common strain
of agent is responsible for vCJD cases from 5 different
Figure 2. Lesion profile comparison of variant Creutzfeldt-Jakob
countries. Transmission properties such as TSE-associated disease cases show similarities in vacuolar pathology levels and
vacuolation, PrPSc deposition, glycosylation profile, and regional distribution in mouse brains. Wild-type mouse lines RIII
mobility of PrP all show strong similarities between vCJD (A), C57 (B), and VM (C) are shown. Data show mean lesion
cases from the Netherlands, Italy, France, and the United profile ± SEM (n>6). G1–G9, gray matter scoring regions: G1,
States and with reference cases from the United Kingdom. dorsal medulla; G2, cerebellar cortex; G3, superior colliculus; G4,
hypothalamus; G5, thalamus; G6, hippocampus; G7, septum; G8,
The distribution of TSE vacuolation and PrPSc retrosplenial and adjacent motor cortex; G9, cingulate and adjacent
deposition in the brains of the RIII and VM mice was similar motor cortex. W1–W3, white matter scoring regions: W1, cerebellar
for each brain isolate examined and indicates that the same white matter; W2, mesencephalic tegmentum; W3, pyramidal tract.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1577
RESEARCH

in brain area; we have established similar transmission

characteristics between brain regions and have aimed
in this study to use identical brain regions for each non-
UK case. Differences in inoculation route are also not
an explanation for the alteration in ranking. Differences
were not found between the case from the United States,
which was inoculated i.c. only, and the cases from the
Netherlands and Italy, which were inoculated i.c. and
intraperitoneally.
The infectivity titer of the inoculum may have played
a role in alteration of the incubation periods. However,
Ritchie et al. (14) suggested that changes in titer would
affect all mouse strains and would not result in a change in
the ranking of the mouse strains. Experimental transmission
of vCJD from the case from the United States, in which
lower volumes were inoculated, shows the same temporal
pattern as the cases from the Netherlands and Italy.
Variability in incubation time is often associated with
the primary transmission of a TSE agent between species
but often stabilizes on subsequent mouse-to-mouse passage,
Figure 3. Immunohistochemical detection of abnormal prion protein while variability within groups decreases. Therefore,
(PrPSc) in the hippocampus and thalamus of RIII, C57, and VM wild- variations observed in this study do not necessarily point
type mice after inoculation with variant Creutzfeldt-Jakob disease
brain tissue. Scale bars = 500 μm. The anti–prion protein detection
to strain variation between the cases of vCJD. However,
antibody used was 6H4. to rule out this possibility, further transmission studies
will be conducted. Moreover, passage of cases of vCJD
occurring during the past 15 years will be undertaken in
incubation period ranking of the mouse strains between the a comparative study to establish whether genetic drift
inoculum from France and other locations (Figure 1). For in mouse lines or strain variation in vCJD underlies the
each brain isolate, the RIII mice were the first line to die of differences observed in this study. Two of the brain isolates
disease, consistent with previous transmissions from vCJD
cases in the United Kingdom.
Furthermore, our study has shown that the ≈100-
day difference in incubation period between the RIII and
C57BL mice (both Prn-pa), which is characteristic of
the BSE strain, has been maintained in the experimental
transmission of the vCJD worldwide cases (15). The C57BL
and VM mice have resulted in close incubation period
ranges that can overlap, as was demonstrated previously
(14,15). For inocula from the non-UK cases, incubation
periods appeared in the order RIII, VM, C57BL, whereas
the inocula from France showed incubation periods in the
order RIII, C57BL, VM, as do the historic UK cases. This
change in the order could be attributed to genetic drift in
the mouse lines used over the course of the historical and
Figure 4. Western blot analysis of brain extracts from RIII (A) and
current studies or variations in the strain of TSE agent. VM (B) wild-type mice inoculated with variant Creutzfeldt-Jakob
The VM mice used in this study have shown disease (vCJD) brain tissue. Lane M, positive control; lane 1, human
substantially shortened incubation periods compared with vCJD brain homogenate (UK origin) showing the typical abnormal
those from in earlier studies, which may have caused the prion protein (PrPSc) type 2B; lane 2, United Kingdom; lane 3, The
alteration in ranking. This alteration in incubation time Netherlands; lane 4: Italy (cortex); lane 5, Italy (cerebellum); lane 6,
France; lane 7, United States; lane 8, human sporadic Creutzfeldt-
ranking was also identified recently in a comparable Jakob disease brain homogenate showing the typical PrPSc type 1.
transmission study involving a more recent UK vCJD case Type 2B and 1 differ in mobility of the unglycosylated band (≈19 kDa
(M. Bishop et al., unpub. data). Moreover, the difference and ≈20 kDa, respectively). All samples were treated with proteinase
in time ranking is unlikely to be attributable to variation K. The anti–prion protein detection antibody used was 6H4.

1578 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Transmission Properties of vCJD

inoculated (from Italy and the United States) were from 4. Dickinson AG, Meikle VMH, Fraser H. Identification of a gene
patients who had been treated with quinacrine. However, which controls the incubation period of some strains of scra-
pie agent in mice. J Comp Pathol. 1968;78:293–9. http://dx.doi.
this treatment appears to have had no effect on incubation org/10.1016/0021-9975(68)90005-4
periods, vacuolation profiles, PrPSc deposition patterns, or 5. Bruce MB, Will RG, Fraser H. Comparison of the biological charac-
the glycosylation and mobility of PrP. teristics of BSE and CJD in mice. In: Iqbal K, Swaab DF, Winblad B,
The similarities in lesion profiles, biochemistry, Wisniewski HM, editors. Alzheimer’s disease and related disorders.
Chichester (UK): John Wiley; 1999. p. 553–60.
and immunohistochemistry between this series of vCJD 6. The National Creutzfeldt-Jakob Disease Research & Surveillance
transmission studies support the hypothesis that a single Unit. CJD figures [cited 2010 Nov 9]. http://www.cjd.ed.ac.uk/fig-
strain of infectious agent is responsible for all vCJD cases, ures.htm
regardless of geographic origin, which would suggest that 7. Chadeau-Hyam M, Alperovitch A. Risk of variant Creutzfeldt-Jakob
disease in France. Int J Epidemiol. 2005;34:46–52. http://dx.doi.
current diagnostic criteria for vCJD are sufficient to detect org/10.1093/ije/dyh374
cases in all countries at this time. Still to be determined 8. Brandel JP, Heath CA, Head MW, Levavasseur E, Knight R,
is whether the differences in incubation period rankings Laplanche JL, et al. Variant Creutzfeldt-Jakob disease in France and
in some cases represent changes in strain phenotype over the United Kingdom: evidence for the same agent strain. Ann Neu-
rol. 2009;65:249–56. http://dx.doi.org/10.1002/ana.21583
time, which could affect future diagnosis. 9. Bird SM. European Union’s rapid TSE testing in adult cattle and sheep:
implementation and results in 2001 and 2002. Stat Methods Med Res.
Acknowledgments 2003;12:261–78. http://dx.doi.org/10.1191/0962280203sm331ra
We thank Wun-Ju Shieh for the collection and provision 10. Sanchez-Juan P, Cousens SN, Will RG, van Duijn CM. Source
of variant Creutzfeldt-Jakob disease outside United Kingdom.
of brain tissue from the case-patient in the United States. We Emerg Infect Dis. 2007;13:1166–9. http://dx.doi.org/10.3201/
also thank Irene McConnell, the Animal Facility staff of the eid1308.070178
Neurobiology Division, Sandra Mack, and the Pathology staff 11. Bruce ME, McConnell I, Fraser H, Dickinson AG. The disease
for sectioning the mouse brains and assessing levels of TSE characteristics of different strains of scrapie in Sinc congenic
mouse lines: implications for the nature of the agent and host con-
vacuolation. trol of pathogenesis. J Gen Virol. 1991;72:595–603. http://dx.doi.
org/10.1099/0022-1317-72-3-595
This project was funded by Neuroprion and the Department
12. Fraser H, Dickinson AG. The sequential development of the brain le-
of Health (England). sion of scrapie in three strains of mice. J Comp Pathol. 1968;78:301–
11. http://dx.doi.org/10.1016/0021-9975(68)90006-6
The views expressed in the publication are those of the 13. Head MW, Ritchie D, Smith N, McLoughlin V, Nailon W, Samad
authors and not those of the Department of Health (England). The S, et al. Peripheral tissue involvement in sporadic, iatrogenic, and
findings and conclusions in this article have not been formally variant Creutzfeldt-Jakob disease: an immunohistochemical, quanti-
disseminated by the Food and Drug Administration and should tative, and biochemical study. Am J Pathol. 2004;164:143–53. http://
dx.doi.org/10.1016/S0002-9440(10)63105-7
not be construed to represent any Administration determination 14. Ritchie DL, Boyle A, McConnell I, Head MW, Ironside JW, Bruce
or policy. ME. Transmissions of variant Creutzfeldt-Jakob disease from
brain and lymphoreticular tissue show uniform and conserved bo-
Dr Diack is a research fellow at the Roslin Institute, vine spongiform encephalopathy–related phenotypic properties
University of Edinburgh, UK. Her research interests focus on on primary and secondary passage in wild-type mice. J Gen Virol.
prion diseases, in particular strain characterization and modeling 2009;90:3075–82. http://dx.doi.org/10.1099/vir.0.013227-0
15. Bruce M, Chree A, McConnell I, Foster J, Pearson G, Fraser H.
of human diseases.
Transmission of bovine spongiform encephalopathy and scrapie to
mice: strain variation and the species barrier. Philos Trans R Soc
Lond B Biol Sci. 1994;343:405–11. http://dx.doi.org/10.1098/
References rstb.1994.0036
1. Will RG, Ironside JW, Hornlimann B, Zeidler M. A new variant
of Creutzfeldt-Jakob disease in the UK. Lancet. 1996;347:921–5. Address for correspondence: Jean C. Manson, The Roslin Institute and
http://dx.doi.org/10.1016/S0140-6736(96)91412-9 Royal (Dick) School of Veterinary Studies, University of Edinburgh,
2. Bruce ME, Will RG, Ironside JW, McConnell I, Drummond D, Sut- Easter Bush, Midlothian EH25 9RG, Scotland, UK: email: jean.manson@
tie A, et al. Transmissions to mice indicate that ‘new variant’ CJD is roslin.ed.ac.uk
caused by the BSE agent. Nature. 1997;389:498–501. http://dx.doi.
org/10.1038/39057
3. Hill AF, Desbruslais M, Joiner S, Sidle KC, Gowland I, Collinge Use of trade names is for identification only and does not
J, et al. The same prion strain causes vCJD and BSE. Nature. imply endorsement by the Public Health Service or by the US
1997;389:448–50, 526. http://dx.doi.org/10.1038/38925 Department of Health and Human Services.

Search past issues of EID at www.cdc.gov/eid

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1579
RESEARCH

WU and KI Polyomaviruses in
Respiratory Samples from
Allogeneic Hematopoietic
Cell Transplant Recipients
Jane Kuypers,1 Angela P. Campbell,1 Katherine A. Guthrie, Nancy L. Wright, Janet A. Englund,
Lawrence Corey, and Michael Boeckh

Data are limited regarding 2 new human polyoma- and from patients co-infected with a respiratory virus
viruses, KI polyomavirus (KIPyV) and WU polyomavirus (3–11). However, KIPyV and WUPyV were detected at
(WUPyV), in immunocompromised patients. We used real- similar rates in specimens from symptomatic patients and
time PCR to test for these and 12 respiratory viruses in in persons without respiratory symptoms (6,8,12,13), sug-
2,732 nasal wash samples collected during the first year gesting that these viruses might not cause respiratory ill-
after allogeneic hematopoietic cell transplantation from 222
ness in immunocompetent children.
patients. Specimens were collected weekly until day 100;
then at least every 3 months. One year after hematopoietic
Two other human polyomaviruses, BK and JC, cause
cell transplantation, the cumulative incidence estimate was mild or asymptomatic primary infections early in life, fol-
26% for KIPyV and 8% for WUPyV. Age <20 years predict- lowed by persistent, subclinical infections in healthy per-
ed detection of KIPyV (hazard ratio [HR] 4.6) and WUPyV sons (14–16). However, these viruses can reactivate, pri-
(HR 4.4), and detection of a respiratory virus in the previ- marily from the kidney, bone marrow, and lymphoid tissue,
ous 2 weeks predicted KIPyV detection (HR 3.4). Sputum and cause serious disease in immunocompromised patients
production and wheezing were associated with detection (14–16). Similarly, reactivation of KIPyV and WUPyV
of KIPyV in the past week and WUPyV in the past month. from lymphoid tissue was described among immunosup-
There were no associations with polyomavirus detection pressed persons with AIDS, although clinical consequenc-
and acute graft versus host disease, cytomegalovirus re- es of reactivation were not examined (17).
activation, neutropenia, lymphopenia, hospitalization, or
Because KIPyV and WUPyV are frequently detected
death.
in association with respiratory symptoms, inhalation is sus-
pected as a potential route of transmission. If KIPyV and

I n 2007, two new human polyomaviruses, KI polyoma-

virus (KIPyV) and WU polyomavirus (WUPyV), were
identified in respiratory specimens from patients with re-
WUPyV are respiratory pathogens, they may be more like-
ly to cause respiratory illness in immunocompromised per-
sons, either during primary infection or reactivation. Few,
spiratory illness (1,2). Since then, KIPyV and WUPyV mostly retrospective, studies on the prevalence of KIPyV
have been frequently detected in respiratory specimens, and WUPyV in respiratory specimens have included large
especially those from children with respiratory symptoms numbers of immunocompromised patients (18–21). No
prospective data are available that comprehensively de-
Author affiliations: University of Washington, Seattle, Washington,
scribe the incidence, symptoms, risk factors, and outcomes
USA (J. Kuypers, A.P. Campbell, N.L. Wright, J.A. Englund, L. Co-
associated with detection of KIPyV and WUPyV in respi-
rey, M. Boeckh); Fred Hutchinson Cancer Research Center, Seattle
ratory specimens from hematopoietic cell transplant (HCT)
(J. Kuypers, A.P. Campbell, K.A. Guthrie, J.A. Englund, L. Corey,
recipients. This question is particularly relevant with the
M. Boeckh); and Seattle Children’s Hospital, Seattle (A.P. Camp-
increasing use of multiplex PCR panels for detection of re-
bell, J.A. Englund)

DOI: http://dx.doi.org/10.3201/eid1810.120477 1
These authors contributed equally to this article.

1580 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 WU and KI Polyomaviruses in HCT Recipients

spiratory viruses, especially in samples from immunocom- values >40 were considered negative. Viral copies per mil-
promised patients. liliter were determined by using standard curves generat-
To investigate whether respiratory detection of these ed by PCR amplification of 10-fold dilutions of plasmids
new polyomaviruses is associated with specific outcomes containing amplicon sequences ranging in concentration
in patients after HCT, a real-time PCR specific for KIPyV from 10 to 1 ×107 copies/reaction. To validate this PCR, a
and WUPyV DNA was developed and used to examine subset of KIPyV-positive, WUPyV-positive, KIPyV-neg-
nasal wash specimens collected prospectively from HCT ative, and WUPyV-negative samples was blindly retested
recipients with and without respiratory symptoms for 1 by using published assays for detection of KIPyV (23) and
year after transplantation. Clinical data and standardized WUPyV (4).
symptom surveys obtained at each specimen collection The in-house PCR had a sensitivity of 5–10 DNA cop-
were analyzed to determine associations between respira- ies/PCR, which provided a sensitivity of 500–1,000 cop-
tory KIPyV and WUPyV detection and illness. ies/mL. This PCR did not detect JC or BK virus DNA. A
subset of 397 samples, including 31 positive for KIPyV
Methods and 27 positive for WUPyV by the PCR, was retested by
using published real-time PCRs. Samples with discordant
Patients and Collection of Specimens KIPyV results included 3 positive by the in-house PCR and
Combined nasopharyngeal wash (or swab) and oro- negative by the alternate PCR (23) and 1 negative by the
pharyngeal swab samples were collected weekly beginning in-house PCR and positive by the alternate PCR. Samples
1–2 weeks before transplantation until day 100; then every with discordant WUPyV results included 4 positive by the
1–3 months for ≤1-year after transplantation from alloge- in-house PCR and negative by the alternate PCR (4) and 1
neic HCT recipients enrolled in a prospective surveillance negative by the in-house PCR and positive by the alternate
study approved by the Institutional Review Board at Fred PCR. Samples with discordant results had <5,000 copies/
Hutchinson Cancer Research Center (Seattle, WA, USA) mL, indicating low levels of DNA near the assay limits of
(22). Participants provided written informed consent. Ad- detection.
ditional specimens were collected when respiratory symp-
toms were reported. Patients had ≥1 specimen collected Detection of Respiratory Virus
during January 2006–December 2007. Participants com- Samples were tested for 12 respiratory viruses by
pleted surveys weekly for 1 year and reported any of 11 re- using a multiplexed panel of real-time, TaqMan reverse
spiratory or 4 systemic symptoms. Clinical and laboratory transcription PCRs. Assays to detect respiratory syncytial
data were obtained from medical records. virus; human metapneumovirus; influenza virus A; para-
influenza viruses 1, 2, and 3; adenoviruses; coronaviruses;
Detection of KIPyV and WUPyV DNA rhinoviruses; and bocavirus were performed as described
and PCR Validation (24–30). Assays to detect influenza virus B and parainflu-
An in-house, duplex, real-time TaqMan PCR was de- enza virus 4 were performed by using the same reagents
veloped, which was specific for viral protein 2–3 and vi- and thermocycling conditions as the other respiratory vi-
ral protein 1 genes of KIPyV and WUPyV, respectively. rus assays but by substituting the specific primer and probe
Ten microliters of extracted sample were added to the PCR sets (Table 1). The sensitivity of each assay was 1,000 viral
master mixture containing KIPyV and WUPyV primers copies/mL. Throughout this report, respiratory virus refers
and probes (Table 1). Samples with PCR cycle threshold to any of these 12 viruses.

Table 1. Primers and probes used for detecting KIPyV, WUPyV, and 2 other viruses by real-time RT-PCR and PCR in HCT recipients*
Target Amplicon PCR concentration,
Virus gene size, bp Function Sequence/label, 5co3c nmol/L
KIPyV VP2–3 74 Forward primer CTATCCCTGAATACCAGTTGGAAAC 425
Reverse primer GTATGACGCGACAAGGTTGAAG 425
Probe FAM-TTCCGGGCATCCCAGACTGGC-BHQ1 125
WUPyV VP1 75 Forward primer AACCAGGAAGGTCACCAAGAAG 300
Reverse primer TCTACCCCTCCTTTTCTGACTTGT 300
Probe HEX-CAACCCACAAGAGTGCAAAGCCTTCC-BHQ1 75
Influenza B Matrix 76 Forward primer CACAATTGCCTACCTGCTTTCA 250
Reverse primer CCAACAGTGTAATTTTTCTGCTAGTTCT 250
Probe VIC-CTTTGCCTTCTCCATCTT-MGBNFQ 100
Parainfluenza NP 94 Forward primer TGCCAAATCGGCAATAAACA 250
type 4 Reverse primer GGCTCTGGCAGCAATCATAAG 250
Probe VIC-TGATTCTGCATTGATGTGG-MGBNFQ 100
*KIPyV, KI polyomavirus; WUPyV, WU polyomavirus; RT-PCR, reverse transcription PCR; HCT, hematopoietic cell transplantation; VP, viral protein; NP,
nucleoprotein.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1581
RESEARCH

Statistical Analysis analyzed as functions of recent KIPyV or WUPyV detec-

The probability of detecting KIPyV or WUPyV DNA tion by using logistic and linear regression models. Virus
in ≥1 nasal wash sample from transplantation to 1-year af- detection within the last week was defined as a positive
ter transplantation was estimated by using cumulative in- sample at the current or last study contact and a lag time
cidence curves. Patient records were censored at 14 days ≤14 days. Virus detection within the last month was ana-
past the time of last eligible respiratory sample, death, or 1 lyzed for wheezing and cough because these symptoms
year after transplantation, whichever occurred first. Death may persist, and was defined as a positive sample at the
before 1 year and within 14 days after the last sample of a current or any of the last 4 study contacts, as long as those
patient was treated as a competing risk for detection. contacts occurred within 45 days. Models were adjusted for
The association of KIPyV and WUPyV detection with detection of respiratory virus within the last week, day rela-
respiratory virus positivity was estimated by using a logis- tive to transplantation, age, conditioning regimen, donor
tic regression model. Differences in KIPyV and WUPyV type, and acute GvHD. Measurements based on multiple
copies/mL according to patient age at transplantation and patient contacts were entered as repeated measures and
co-occurrence of a respiratory virus were assessed by us- adjusted for possible correlation between values within a
ing linear regression. For both model types, robust standard person by using generalized estimating equations.
errors were calculated to account for correlation within re- For regression models, p values were obtained by us-
peated measures for the same person. ing the Wald test; no adjustments were made for multiple
Cox regression models were fit to evaluate potential comparisons. Two-sided p values <0.05 were considered
risk factors for detection of KIPyV or WUPyV in the first significant. Analyses were performed by using SAS ver-
year after transplantation, including age, sex, disease risk sion 9.0 (SAS Institute, Cary, NC, USA).
(standard or high) (29,31), stem cell source, donor type,
conditioning regimen (myeloablative versus nonmyeloab- Results
lative), donor and recipient cytomegalovirus (CMV) se-
rostatus, grades 2–4 graft versus host disease (GvHD), Patient Characteristics and Detection of KIPyV
and respiratory virus detection. Donor CMV serostatus and WUPyV
was defined as negative for cord blood recipients. Acute A total of 2,732 nasal wash specimens were collected
GvHD and respiratory virus detection were treated as time- from 222 eligible patients at weekly or longer intervals
dependent covariates; respiratory virus detection was set to (median 7 days, range 4–155 days). The cohort had a me-
1 when ≥1 respiratory virus had been detected in the previ- dian of 13 specimens per patient (range 1–30 specimens).
ous 2 weeks and 0 otherwise. Cox regression models were Patients ranged in age from 9.6 months to 75.2 years (me-
fit to evaluate risk factors for KIPyV or WUPyV detection dian 51.5 years) and remained in the study a median of 111
in the first 100 days after transplantation. days (range 1–365 days) after HCT. A respiratory virus
CMV reactivation was modeled as a time-dependent was detected in 17% of specimens; 51% of patients had
indicator defined as any antigenemia or positive PCR val- ≥1 sample positive for any respiratory virus during their
ue, or antigenemia >10 cells per slide or >100 copies/mL time in the study. Clinical characteristics of the cohort are
by PCR. Neutropenia was modeled as a time-dependent provided in Table 2.
covariate set to 1 when the absolute neutrophil count was KIPyV and WUPyV DNA was detected in 203 (7%)
<500 cells/mm3. When values were missing, the indica- specimens from 49 (22%) patients and in 35 (1%) speci-
tor for the previous day was carried forward. Lymphope- mens from 15 (7%) patients, respectively. Two patients
nia was modeled similarly with 2 thresholds, 100 and 300 were positive for both viruses; 1 patient was positive con-
cells/mm3. currently in the same specimen, and an additional patient
We evaluated detection of KIPyV or WUPyV as pre- was positive for 1 virus in different specimens. Among the
dictors of clinical outcomes, including diagnosis of grades 62 WUPyV-positive or KIPyV-positive patients, 1 patient
2–4 acute GvHD, CMV reactivation, neutropenia, lym- provided only 1 specimen during the study. At 1 year af-
phopenia, and hospitalization within 100 days; and viral ter HCT, cumulative incidence estimates were 26% (95%
symptoms, increased liver transaminase and total biliru- CI 20%–33%) for KIPyV and 8% (95% CI 4%–12%) for
bin levels, and death within 1 year. For time-to-event out- WUPyV (Figure 1). KIPyV and WUPyV were first detected
comes, patient records were censored at date of last contact a median of 45 days (range 1–187 days) and 43 days (range
or death, and KIPyV and WUPyV detection were treated 4−46 days) after transplantation, respectively. KIPyV-pos-
as time-dependent covariates in Cox regression models. itive or WUPyV-positive specimens were detected in every
Potential confounders for these models included age, sex, month, and there was no apparent seasonality (Figure 2). A
donor type, stem cell source, and conditioning regimen. respiratory virus was detected in 32% of KIPyV-positive
Outcomes with multiple occurrences over time were and 14% of WUPyV-positive specimens. Rhinoviruses and

1582 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 WU and KI Polyomaviruses in HCT Recipients

Table 2. Characteristics of 222 HCT recipients tested for KIPyV or positive for KIPyV and a respiratory virus than in speci-
WUPyV* mens in which KIPyV was the only virus detected (medians
No. (%) recipients 6.35 vs. 4.25, respectively, p<0.001). In contrast, WUPyV-
Infected with Not
All, n = either virus, infected,
positive specimens with a respiratory virus co-pathogen
Characteristic 222 n = 62 n = 160 had significantly lower viral copy numbers than specimens
Age, y positive for WUPyV virus alone (medians 2.84 vs. 3.62,
<20 23 (10) 18 (29) 5 (3) respectively, p = 0.01). Patient age at transplantation was
20–39 37 (17) 14 (23) 23 (14)
40–59 99 (45) 17 (27) 82 (51) not correlated with KIPyV or WUPyV copies/mL.
>60 63 (28) 13 (21) 50 (31)
Sex Risk Factors for Detection of KIPyV and WUPyV
F 87 (39) 23 (37) 64 (40)
M 135 (61) 39 (63) 96 (60) Time in study and number of specimens collected did
Underlying disease risk† not differ among patients according to polyomavirus detec-
Standard 142 (64) 40 (65) 102 (64) tion. Age <20 years was a significant predictor of KIPyV
High 80 (36) 22 (35) 58 (36)
and WUPyV detection in multivariable models (hazard ra-
Stem cell source
Bone marrow 28 (13) 12 (19) 16 (10) tio [HR] 4.6, 95% CI 2.5–8.6, p<0.001; HR 4.4, 95% CI
Peripheral blood 179 (81) 47 (76) 132 (83) 1.5–12.8, p = 0.007; respectively). All 10 patients <12 years
Cord blood 15 (7) 3 (5) 12 (8) of age and 18 (78%) of 23 patients <20 years of age were
CMV serostatus
D+/R+ 56 (25) 17 (27) 39 (24) positive for KIPyV or WUPyV compared with 44 (22%)
D+/R– 81 (36) 22 (35) 59 (37) of 199 patients ≥20 years of age (Figure 4). Detection of a
D–/R+ 16 (7) 4 (6) 12 (8) respiratory virus within the last 2 weeks was a significant
D–/R– 69 (31) 19 (31) 50 (31)
Donor match
predictor of KIPyV detection (HR 3.4, 95% CI 1.8–6.4,
Related–matched 77 (35) 23 (37) 54 (34) p<0.001) in a model adjusted for age, transplantation type,
Related–mismatched 9 (4) 6 (10) 3 (2) and donor type. CMV reactivation, neutropenia, and lym-
Unrelated–cord blood 15 (7) 3 (5) 12 (8)
phopenia were not associated with detection of KIPyV and
Unrelated–matched 97 (44) 24 (39) 73 (46)
Unrelated–mismatched 24 (11) 6 (10) 18 (11) WUPyV within the first 100 days after transplantation.
Conditioning regimen
Myeloablative 128 (58) 41 (66) 87 (54) Associations between KIPyV or WUPyV
Nonmyeloablative 94 (42) 21 (34) 73 (46)
Acute graft-versus-host disease Detection and Symptoms
Grade 0 or 1 78 (35) 30 (48) 48 (30) Detection of KIPyV within the past week was signifi-
Grade 2–4 144 (65) 32 (52) 112 (70) cantly associated with sputum production (OR 1.7, 95%
*HCT, hematopoietic cell transplantation; KIPyV, KI polyomavirus;
WUPyV, WU polyomavirus; CMV, cytomegalovirus; D, donor; R, recipient. CI 1.0–2.9, p = 0.04) (Table 3). WUPyV detection within
†The underlying disease stage of a patient was categorized as low, the past month was significantly associated with wheezing
intermediate, or high risk (29,31).
(OR 3.1, 95% CI 1.2–8.1, p = 0.02). Limiting the analysis
to 20 patients with high levels of KIPyV detection within
coronaviruses accounted for 78% of respiratory virus co- the past week (>5 log10 copies/mL) showed a significant
detections. Respiratory and KIPyV viruses tended to co-
occur (odds ratio [OR] 2.4, 95% CI 1.2–5.1, p = 0.02).
Twenty-one (43%) KIPyV-positive and 9 (60%)
WUPyV-positive patients had 1 positive specimen. Posi-
tive episodes with detection in ≥4 consecutive specimens
were seen in 18 (37%) KIPyV-positive patients, includ-
ing 6 with 9–19 consistently positive specimens (Figure 3,
panel A) and in 2 (13%) WUPyV-positive patients, includ-
ing 1 with 11 consecutive specimens (Figure 3, panel B).
The maximum number (log10 copies/mL) of KIPyV and
WUPyV virus per positive episode ranged from 2.55 to
10.58 (median 5.31) and 2.57–9.02 (median 3.08), respec-
tively. Patients with >1 KIPyV-positive specimen had an
average of >3 logs higher maximum viral log10 copies/mL
(median 6.56) than KIPyV-positive patients with 1 positive
Figure 1. Cumulative incidence of KI polyomavirus (KIPyV) and
specimen (median 2.86, p<0.001).
WU polyomavirus (WUPyV) detection after transplantation in 222
If we considered all positive samples, the number of hematopoietic cell transplantation recipients. Cumulative incidence
KIPyV copies/mL was significantly higher in specimens of human rhinovirus is shown for comparison.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1583
RESEARCH

Figure 2. Number of samples positive

for KI polyomavirus (black bars) and
WU polyomavirus (white bars) by month
of first detection in hematopoietic cell
transplantation recipients.

association with sputum production (OR 2.0, 95% CI 1.1– rus was associated with the 1-year mortality rate in in an
3.8, p = 0.03). Analysis of symptoms as a function of KIP- adjusted model.
yV positivity among 17 patients with a respiratory virus
detected within the past week showed a similar association Discussion
between sputum production and KIPyV detection (OR 2.7, Our large longitudinal surveillance study provides a
95% CI 1.1–6.6, p = 0.03). rigorous evaluation of KIPyV and WUPyV detection in
upper respiratory tract specimens from HCT recipients.
Associations between KIPyV or WUPyV Detection One-year cumulative incidence estimates were high, 26%
and Clinical Outcomes and 8% for KIPyV and WUPyV, respectively, and there
Longitudinal analyses showed significant relationships were prolonged episodes of detection for ≥4 weeks in 37%
between WUPyV detection and a lower risk for lymphope- of patients positive for KIPyV and 13% of patients positive
nia defined by <300 cells/mm3 (OR 0.3, 95% CI 0.1–0.6, for WUPyV. These numbers are comparable to those of our
p = 0.001) and grades 2–4 acute GvHD (HR 3.1, 95% CI report of rhinoviruses and coronaviruses, the most com-
1.3–7.7, p = 0.01) in a model adjusted for donor type and mon respiratory virus types detected in the same patient
stem cell source among 136, 27, and 6 patients given di- population, in which we found day 100 estimates of 22%
agnoses of grades 2, 3, and 4 acute GvHD, respectively. and 11%, respectively, and detection for ≥1 month in 44%
However, this association was based on few cases; only 5 and 37%, respectively (29). In comparison, cross-sectional
patients had WUPyV detected before the GvHD diagnosis. studies testing specimens from immunocompetent children
No relationships between detection of KIPyV or WUPyV with acute respiratory tract illnesses found the prevalence
and CMV reactivation, risk for hospitalization ≤100 days of KIPyV and WUPyV ranged from 0% to 2.8% and from
after transplantation, and mean values for alanine amino- 2% to 7.1% (1–13,21).
transferase, aspartate aminotransferase, and total bilirubin A recent study of pediatric hematology/oncology
levels were found in multivariable models. patients and immunocompetent persons found a higher
Nineteen bronchoalveolar lavage (BAL) samples from KIPyV mean viral load in respiratory tract specimens from
13 KIPyV-positive patients and 57 BAL samples from 37 the immunocompromised group, suggesting potential for
KIPyV-negative and WUPyV-negative patients were test- increased pathogenicity in this population (21). A study
ed for KIPyV and WUPyV. Collection of BAL samples in adult HCT recipients tested sequential nasopharyngeal
occurred within 2 weeks of a positive nasal wash sample aspirates from 31 asymptomatic patients, in which KIPyV
in only 4 of the KIPyV-positive patients, including 3 with and WUPyV were detected in 1 each of 126 samples (19).
collection on the same day. Only 1 of 76 BAL samples was A cross-sectional study that tested specimens from 45 HCT
positive for KIPyV (6.3 log10 copies/mL). This sample was recipients with respiratory illness detected KIPyV in 8
collected from a 2-year-old child on the same day as the (17.8%), similar to our findings, but with no WUPyV de-
KIPyV-positive nasal wash sample (165 days after trans- tected (19,20). Respiratory viruses were often co-detected
plantation); CMV was also isolated in culture and the pa- in our cohort, 32% with KIPyV and 14% with WUPyV,
tient received treatment for CMV pneumonia. Of 66 deaths consistent with high rates of co-infection reported by others
within 1 year of transplantation, 12 occurred in patients (1–3,6,10,21,32).
positive for KIPyV, 4 in patients positive for WUPyV, and In our study, young age was a risk factor for detection
1 in a patient positive for KIPyV and WUPyV. Neither vi- of KIPyV or WUPyV, which has been reported (4,7,32,33).

1584 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 WU and KI Polyomaviruses in HCT Recipients

of immune suppression. Prolonged detection episodes may

represent long-term viral shedding after new acquisition of
virus, especially in young patients, or reactivation or stimu-
lation of replication of persistent virus in the respiratory
tract because of immune suppression or a recent or con-
comitant respiratory virus infection. A site of persistence
for these new polyomaviruses has not yet been identified,
although human tonsils have been postulated as a reservoir
and reactivation has been demonstrated in lymph nodes and
spleen (17,36,37).
The longitudinal design of our study and specimen
collection during symptomatic and asymptomatic periods
enabled us to thoroughly evaluate symptoms associated
with KIPyV and WUPyV detection in HCT recipients. We
found that KIPyV detection in the past week was associated
with sputum production and WUPyV detection in the past
month was associated with wheezing. Limiting the analy-
sis to patients with high viral copy numbers of KIPyV in
their nasal wash sample did not provide additional associa-
tions. Likewise, co-detection of KIPyV and a respiratory
virus was not associated with symptoms other than sputum
production, suggesting that infection with KIPyV did not
exacerbate symptoms caused by the respiratory virus. Of
50 patients who underwent bronchoscopy for workup of
Figure 3. Detection of A) KI polyomavirus (KIPyV) DNA in 28 possible lower respiratory illness and had a BAL specimen
hematopoietic cell transplantation (HCT) recipients with >2 KIPyV-
positive specimens and B) WU polyomavirus (WUPyV) DNA in
available for testing, we detected KIPyV in 1 sample from
6 HCT recipients with >2 WUPyV-positive specimens, with and a child who also had CMV pneumonia. Overall, our analy-
without detection of a respiratory virus by day after transplantation. ses provide some support for these viruses as respiratory
Each line represents 1 patient in order of age (KIPyV-positive pathogens. However, we did not observe severe respira-
patients 1–10 and WUPyV-positive patients 1 and 2 are <20 tory illness or death that could be conclusively attributed to
years of age). Circles indicate specimen collection. Gray indicates
detection of only KIPyV or WUPyV, black indicates detection
polyomavirus detection.
of KIPyV or WUPyV and a respiratory virus, and white indicates In seeking other biologically plausible and clinically
negative results for KIPyV or WUPyV. relevant associations for these new viruses, we found no
association of KIPyV and WUPyV detection in nasal wash-

The seroprevalence of antibodies to KIPyV and WUPyV

for children 5–20 years of age was similar to that for adults
(34,35), suggesting that primary exposure occurs in child-
hood. We also found that detection of a respiratory virus
within the previous 2 weeks was an independent risk factor
for detection of KIPyV. Compared with samples in which
only KIPyV DNA was detected, higher viral copy num-
bers of KIPyV were detected in samples also positive for
a respiratory virus, perhaps because of stimulation of viral
replication during a respiratory virus infection. Our cur-
rent data do not provide evidence that this is a biologically
meaningful difference.
Prolonged detection and younger age associated with
a higher prevalence of KIPyV and WUPyV were also re-
ported in a longitudinal study of KIPyV and WUPyV in
Figure 4. Proportion of hematopoietic cell transplantation recipients
children (33). As with BK and JC polyomaviruses, KIPyV in each age group with samples positive for KI polyomavirus (black
and WUPyV might persist for life after primary infection in bars) and WU polyomavirus (white bars). Values above bars are
undetermined sites and become reactivated during periods percentages.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1585
RESEARCH

Table 3. Respiratory and systemic symptoms associated with KIPyV or WUPyV detection within the prior week or month in HCT
recipients*
KIPyV† WUPyV†
Symptom OR (95% CI) p value OR (95% CI) p value
Within prior week
Runny nose 1.3 (0.7–2.4) 0.39 1.2 (0.5–3.2) 0.66
Sinus congestion 0.8 (0.5–1.3) 0.34 1.4 (0.6–3.5) 0.46
Postnasal drip 0.8 (0.4–1.8) 0.65 1.2 (0.5–3.0) 0.72
Shortness of breath 0.9 (0.4–1.8) 0.73 0.9 (0.3–2.9) 0.83
Sputum production 1.7 (1.0–2.9) 0.04 1.5 (0.5–4.8) 0.47
Pharyngitis 1.0 (0.6–1.7) 0.99 2.1 (0.8–5.4) 0.12
Sneezing 1.2 (0.7–2.0) 0.54 1.3 (0.6–2.9) 0.48
Watery eyes 1.5 (0.8–2.6) 0.17 1.6 (0.4–6.2) 0.51
Cough 1.1 (0.6–2.0) 0.68 1.4 (0.6–3.5) 0.47
Wheezing 1.1 (0.5–2.5) 0.82 2.1 (0.8–5.9) 0.15
Fever 0.9 (0.5–1.5) 0.59 1.3 (0.4–3.7) 0.64
Headache 0.6 (0.4–1.1) 0.08 0.7 (0.2–2.8) 0.58
Myalgia 0.6 (0.3–1.1) 0.08 1.1 (0.3–3.5) 0.92
Diarrhea 0.8 (0.5–1.3) 0.40 0.7 (0.3–1.8) 0.47
Within prior month
Cough 1.0 (0.6–1.8) 0.91 1.6 (0.7–3.5) 0.26
Wheezing 1.0 (0.5–2.2) 0.99 3.1 (1.2–8.1) 0.02
*Values are for 211 of 222 patients with symptom survey data. Ear pain was too rare for analysis. KIPyV, KI polyomavirus; WUPyV, WU polyomavirus;
HCT, hematopoietic cell transplantation. OR, odds ratio.
†Each result was adjusted for the following covariates: detection of the other polyomavirus (KIPyV or WUPyV), detection of a respiratory virus, day
relative to transplantation, age at transplantation, stem cell source, donor type, and grades 2–4 acute graft-versus-host-disease.

es with CMV reactivation, increased liver enzyme levels, upper and lower respiratory tract symptoms. At this time,
hospitalization, or neutropenia in the first 100 days after we do not recommend routine testing for these viruses in
transplantation. Detection of WUPyV was associated with immunocompromised patients or inclusion in multiplexed
a lower risk for lymphopenia. respiratory virus PCR panels. Further investigations exam-
We did not test other samples, such as blood, urine, ining other specimen types, such as BAL, blood, and urine,
or feces, for KIPyV and WUPyV DNA. Previous studies from a larger patient cohort over a longer period of time
reported detection of these viruses in plasma, serum, and may be necessary to elucidate the role of these viruses in
peripheral blood samples from patients infected with HIV- highly immunocompromised patients.
1, healthy blood donors, and children and in urine samples
from children (1,2,38) and immunocompetent and immu- Acknowledgments
nocompromised adults (18,38). WUPyV was detected in We thank Cheryl Callais and other members of the study
serum and feces of children whose nasopharyngeal aspirate team for providing attention to study details; Chris Davis for pro-
samples contained high viral loads (39). Both viruses were viding database services; and Terry Stevens-Ayers, Reggie Sam-
detected in fecal samples from HCT recipients (18,40), and poleo, and Rohit Shankar for providing laboratory expertise.
an association was found between KIPyV and diarrhea (40).
This study was supported by National Institutes of Health
However, we did not find an association between virus and
(grants CA 18029, HL081595, K23HL091059, L40AI071572,
diarrhea. Another limitation is that although we report pro-
and K24HL093294). A.P.C. was supported by a MedImmune Pe-
longed, uninterrupted KIPyV and WUPyV detection, we
diatric Fellowship Grant Award, a Pediatric Infectious Diseases
did not perform genetic analyses to confirm whether these
Society Fellowship Award funded by MedImmune, the Seattle
detections represent the same or different viral subtypes.
Children’s Center for Clinical and Translational Research, and
In conclusion, we detected KIPyV or WUPyV in one
Clinical and Translational Science Award grant ULI RR025014.
third of allogeneic HCT recipients during the first year af-
ter transplantation. Prolonged detection episodes and high Dr Kuypers is a senior research scientist in the Molecular
viral copy numbers in respiratory specimens were also ob- Virology Laboratory, Department of Laboratory Medicine, Uni-
served. However, we did not observe many associations versity of Washington, Seattle. Her primary research interests are
with acute respiratory symptoms. We found that detection epidemiology and molecular detection of respiratory viruses.
of a respiratory virus was a risk factor for KIPyV detec-
tion and that concurrent detection of the polyomaviruses
References
and respiratory viruses was common. Although it has been
suggested that KIPyV and WUPyV might be major patho- 1. Allander T, Andreasson K, Gupta S, Bjerkner A, Bogdanovic G,
gens in immunosuppressed patients, we did not find a clear Persson MA, et al. Identification of a third human polyomavirus. J
role for these viruses as respiratory pathogens with classic Virol. 2007;81:4130–6. http://dx.doi.org/10.1128/JVI.00028-07

1586 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 WU and KI Polyomaviruses in HCT Recipients

2. Gaynor AM, Nissen MD, Whiley DM, Mackay IM, Lambert SB, Wu 19. Debiaggi M, Canducci F, Brerra R, Sampaolo M, Marinozzi MC,
G, et al. Identification of a novel polyomavirus from patients with Parea M, et al. Molecular epidemiology of KI and WU polyomavi-
acute respiratory tract infections. PLoS Pathog. 2007;3:e64. http:// ruses in infants with acute respiratory disease and in adult hemato-
dx.doi.org/10.1371/journal.ppat.0030064 poietic stem cell transplant recipients. J Med Virol. 2010;82:153–6.
3. Bialasiewicz S, Whiley DM, Lambert SB, Wang D, Nissen MD, http://dx.doi.org/10.1002/jmv.21659
Sloots TP. A newly reported human polyomavirus, KI virus, is 20. Mourez T, Bergeron A, Ribaud P, Scieux C, de Latour RP, Tazi A,
present in the respiratory tract of Australian children. J Clin Virol. et al. Polyomaviruses KI and WU in immunocompromised patients
2007;40:15–8. http://dx.doi.org/10.1016/j.jcv.2007.07.001 with respiratory disease. Emerg Infect Dis. 2009;15:107–9. http://
4. Le BM, Demertzis LM, Wu G, Tibbets RJ, Buller R, Arens MQ, et dx.doi.org/10.3201/1501.080758
al. Clinical and epidemiologic characterization of WU polyomavirus 21. Rao S, Garcea RL, Robinson CC, Simões EA. WU and KI poly-
infection, St. Louis, Missouri. Emerg Infect Dis. 2007;13:1936–8. omavirus infections in pediatric hematology/oncology patients with
http://dx.doi.org/10.3201/eid1312.070977 acute respiratory tract illness. J Clin Virol. 2011;52:28–32. http://
5. Norja P, Ubillos I, Templeton K, Simmonds P. No evidence for an dx.doi.org/10.1016/j.jcv.2011.05.024
association between infections with WU and KI polyomaviruses 22. Kuypers J, Campbell AP, Cent A, Corey L, Boeckh M. Comparison of
and respiratory disease. J Clin Virol. 2007;40:307–11. http://dx.doi. conventional and molecular detection of respiratory viruses in hema-
org/10.1016/j.jcv.2007.09.008 topoietic cell transplant recipients. Transpl Infect Dis. 2009;11:298–
6. Han TH, Chung JY, Koo JW, Kim SW, Hwang ES. WU poly- 303. http://dx.doi.org/10.1111/j.1399-3062.2009.00400.x
omavirus in children with acute lower respiratory tract infections, 23. Bialasiewicz S, Whiley DM, Lambert SB, Gould A, Nissen MD,
South Korea. Emerg Infect Dis. 2007;13:1766–8. http://dx.doi. Sloots TP. Development and evaluation of real-time PCR assays
org/10.3201/eid1311.070872 for the detection of the newly identified KI and WU polyoma-
7. Ren L, Gonzalez R, Xie Z, Zhang J, Liu C, Li J, et al. WU and viruses. J Clin Virol. 2007;40:9–14. http://dx.doi.org/10.1016/j.
KI polyomavirus present in the respiratory tract of children, but not jcv.2007.07.015
in immunocompetent adults. J Clin Virol. 2008;43:330–3. http:// 24. Foulongne V, Olejnik Y, Perez V, Elaerts S, Rodiere M, Segon-
dx.doi.org/10.1016/j.jcv.2008.08.003 dy M. Human bocavirus in French children. Emerg Infect Dis.
8. Wattier RL, Vazquez M, Weibel C, Shapiro ED, Ferguson D, Landry 2006;12:1251–3. http://dx.doi.org/10.3201/eid1708.060213.
ML, et al. Role of human polyomaviruses in respiratory tract disease 25. Kuypers J, Wright N, Corey L, Morrow R. Detection and quantifica-
in young children. Emerg Infect Dis. 2008;14:1766–8. http://dx.doi. tion of human metapneumovirus in pediatric specimens by real-time
org/10.3201/eid1411.080394 RT-PCR. J Clin Virol. 2005;33:299–305. http://dx.doi.org/10.1016/j.
9. Hormozdi DJ, Arens MQ, Le BM, Buller RS, Agapov E, Storch GA. jcv.2004.11.023
KI polyomavirus detected in respiratory tract specimens from pa- 26. Kuypers J, Wright N, Ferrenberg J, Huang ML, Cent A, Corey L, et
tients in St. Louis, Missouri. Pediatr Infect Dis J. 2010;29:329–33. al. Comparison of real-time PCR assays with fluorescent-antibody
10. Neske F, Blessing K, Ullrich F, Prottel A, Wolfgang Kreth H, assays for diagnosis of respiratory virus infections in children. J Clin
Weissbrich B. WU polyomavirus infection in children, Germa- Microbiol. 2006;44:2382–8. http://dx.doi.org/10.1128/JCM.00216-
ny. Emerg Infect Dis. 2008;14:680–1. http://dx.doi.org/10.3201/ 06
eid1404.071325 27. Kuypers J, Wright N, Morrow R. Evaluation of quantitative and
11. Yuan XH, Jin Y, Xie ZP, Gao HC, Xu ZQ, Zheng LS, et al. Preva- type-specific real-time RT-PCR assays for detection of respiratory
lence of human KI and WU polyomaviruses in children with acute syncytial virus in respiratory specimens from children. J Clin Virol.
respiratory tract infection in China. J Clin Microbiol. 2008;46:3522– 2004;31:123–9. http://dx.doi.org/10.1016/j.jcv.2004.03.018
5. http://dx.doi.org/10.1128/JCM.01301-08 28. Lu X, Holloway B, Dare RK, Kuypers J, Yagi S, Williams JV, et
12. Abed Y, Wang D, Boivin G. WU polyomavirus in children, Cana- al. Real-time reverse transcription–PCR assay for comprehensive
da. Emerg Infect Dis. 2007;13:1939–41. http://dx.doi.org/10.3201/ detection of human rhinoviruses. J Clin Microbiol. 2008;46:533–9.
eid1312.070909 http://dx.doi.org/10.1128/JCM.01739-07
13. van de Pol AC, Wolfs TF, Jansen NJ, Kimpen JL, van Loon AM, 29. Milano F, Campbell AP, Guthrie KA, Kuypers J, Englund JA,
Rossen JW. Human bocavirus and KI/WU polyomaviruses in pediat- Corey L, et al. Human rhinovirus and coronavirus detection
ric intensive care patients. Emerg Infect Dis. 2009;15:454–7. http:// among allogeneic hematopoietic stem cell transplantation re-
dx.doi.org/10.3201/eid1503.081203 cipients. Blood. 2010;115:2088–94. http://dx.doi.org/10.1182/
14. Boothpur R, Brennan DC. Human polyoma viruses and disease blood-2009-09-244152
with emphasis on clinical BK and JC. J Clin Virol. 2010;47:306–12. 30. Kuypers J, Martin ET, Heugel J, Wright N, Morrow R, Englund
http://dx.doi.org/10.1016/j.jcv.2009.12.006 JA. Clinical disease in children associated with newly described
15. Tan CS, Koralnik IJ. Progressive multifocal leukoencephalopathy coronavirus subtypes. Pediatrics. 2007;119:e70–6. http://dx.doi.
and other disorders caused by JC virus: clinical features and patho- org/10.1542/peds.2006-1406
genesis. Lancet Neurol. 2010;9:425–37. http://dx.doi.org/10.1016/ 31. Koc S, Leisenring W, Flowers ME, Anasetti C, Deeg HJ, Nash RA,
S1474-4422(10)70040-5 et al. Therapy for chronic graft-versus-host disease: a randomized
16. zur Hausen H. Novel human polyomaviruses: re-emergence of a trial comparing cyclosporine plus prednisone versus prednisone
well known virus family as possible human carcinogens. Int J Can- alone. Blood. 2002;100:48–51. http://dx.doi.org/10.1182/blood.
cer. 2008;123:247–50. http://dx.doi.org/10.1002/ijc.23620 V100.1.48
17. Sharp CP, Norja P, Anthony I, Bell JE, Simmonds P. Reactivation 32. Abedi Kiasari B, Vallely PJ, Corless CE, Al-Hammadi M, Klapper
and mutation of newly discovered WU, KI, and Merkel cell carcino- PE. Age-related pattern of KI and WU polyomavirus infection. J Clin
ma polyomaviruses in immunosuppressed individuals. J Infect Dis. Virol. 2008;43:123–5. http://dx.doi.org/10.1016/j.jcv.2008.05.003
2009;199:398–404. http://dx.doi.org/10.1086/596062 33. van der Zalm MM, Rossen JW, van Ewijk BE, Wilbrink B, van
18. Bialasiewicz S, Whiley DM, Lambert SB, Nissen MD, Sloots TP. Esch PC, Wolfs TF, et al. Prevalence and pathogenicity of WU and
Detection of BK, JC, WU, or KI polyomaviruses in faecal, urine, KI polyomaviruses in children, the Netherlands. Emerg Infect Dis.
blood, cerebrospinal fluid and respiratory samples. J Clin Virol. 2008;14:1787–9. http://dx.doi.org/10.3201/eid1411.080464
2009;45:249–54. http://dx.doi.org/10.1016/j.jcv.2009.05.002 34. Kean JM, Rao S, Wang M, Garcea RL. Seroepidemiology of hu-
man polyomaviruses. PLoS Pathog. 2009;5:e1000363. http://dx.doi.
org/10.1371/journal.ppat.1000363

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1587
RESEARCH

35. Nguyen NL, Le BM, Wang D. Serologic evidence of frequent hu- 39. Neske F, Blessing K, Prottel A, Ullrich F, Kreth HW, Weissbrich
man infection with WU and KI polyomaviruses. Emerg Infect Dis. B. Detection of WU polyomavirus DNA by real-time PCR in na-
2009;15:1199–205. http://dx.doi.org/10.3201/eid1508.090270 sopharyngeal aspirates, serum, and stool samples. J Clin Virol.
36. Babakir-Mina M, Ciccozzi M, Bonifacio D, Bergallo M, Costa C, 2009;44:115–8. http://dx.doi.org/10.1016/j.jcv.2008.12.004
Cavallo R, et al. Identification of the novel KI and WU polyoma- 40. Babakir-Mina M, Ciccozzi M, Alteri C, Polchi P, Picardi A, Gre-
viruses in human tonsils. J Clin Virol. 2009;46:75–9. http://dx.doi. co F, et al. Excretion of the novel polyomaviruses KI and WU in
org/10.1016/j.jcv.2009.06.009 the stool of patients with hematological disorders. J Med Virol.
37. Kantola K, Sadeghi M, Lahtinen A, Koskenvuo M, Aaltonen LM, 2009;81:1668–73. http://dx.doi.org/10.1002/jmv.21559
Mottonen M, et al. Merkel cell polyomavirus DNA in tumor-free
tonsillar tissues and upper respiratory tract samples: implications for Address for correspondence: Jane Kuypers, Molecular Virology,
respiratory transmission and latency. J Clin Virol. 2009;45:292–5.
University of Washington, 1616 Eastlake Ave E, Suite 320, Seattle, WA
http://dx.doi.org/10.1016/j.jcv.2009.04.008
38. Bialasiewicz S, Whiley DM, Lambert SB, Jacob K, Bletchly C, 98102, USA; email: kuypers@u.washington.edu
Wang D, et al. Presence of the newly discovered human polyomavi-
ruses KI and WU in Australian patients with acute respiratory tract
infection. J Clin Virol. 2008;41:63–8. http://dx.doi.org/10.1016/j.
jcv.2007.11.001

1588 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Wild Birds and Urban Ecology of
Ticks and Tick-borne Pathogens,
Chicago, Illinois, USA, 2005–2010
Sarah A. Hamer, Tony L. Goldberg, Uriel D. Kitron, Jeffrey D. Brawn, Tavis K. Anderson,
Scott R. Loss, Edward D. Walker, and Gabriel L. Hamer

Bird-facilitated introduction of ticks and associated tries in Europe most likely occurred through the movement
pathogens is postulated to promote invasion of tick-borne of migratory birds (1). Infected wild birds also contribut-
zoonotic diseases into urban areas. Results of a longitu- ed to the spread of West Nile virus (WNV) across North
dinal study conducted in suburban Chicago, Illinois, USA, America (2). Thus, models of interseasonal connectivity
during 2005–2010 show that 1.6% of 6,180 wild birds cap- among areas used by migratory birds can be used to fore-
tured in mist nets harbored ticks. Tick species in order of
cast disease spread (3).
abundance were Haemaphysalis leporispalustris, Ixodes
dentatus, and I. scapularis, but 2 neotropical tick species
Over finer spatial scales, the patterns of bird use by
of the genus Amblyomma were sampled during the spring blood-feeding vectors affect the prevalence of vector-borne
migration. I. scapularis ticks were absent at the beginning of pathogens. Host variation impacts the survival of vectors
the study but constituted the majority of ticks by study end that feed on birds rather than on other vertebrates (4), and
and were found predominantly on birds captured in areas avian species exhibit differential reservoir competency for
designated as urban green spaces. Of 120 ticks, 5 were vector-borne pathogens (5). In combination, these factors
infected with Borrelia burgdorferi, spanning 3 ribotypes, influence disease risk; for example, just a few avian species
but none were infected with Anaplasma phagocytophilum. that are heavily fed upon by mosquitoes and highly com-
Results allow inferences about propagule pressure for in- petent for WNV apparently drive most WNV transmission
troduction of tick-borne diseases and emphasize the large (6). Furthermore, host association of strains might help
sample sizes required to estimate this pressure.
maintain pathogen diversity in some vector-borne diseases
systems for which birds play critical roles (7).

W ild birds can affect zoonotic disease risk to hu-

mans, wildlife, and domestic animals through their
mobility and influence on the distribution and abundance
Urban environments may promote pathogen trans-
mission through increased host contact rates, high rates
of pathogen introduction (i.e., propagule pressure), and
of pathogens and vectors. Most notably, avian migration warmer microclimates that are favorable to pathogens and
allows for rapid transcontinental transportation of novel vectors (8). These effects, in turn, may elevate disease risk
pathogens and vectors that may seed new disease foci in to high-density urban human populations. Across gradients
receptive environments. For example, the spread of highly of urbanization, the incidence of some zoonotic pathogens
pathogenic avian influenza into and throughout most coun- has been found to be highest in urban cores (9). Reduced
species richness in urban areas may contribute to elevated
Author affiliations: Michigan State University, East Lansing, Michi-
risk for diseases that are caused by multihost pathogens
gan, USA (S.A. Hamer, E.D. Walker, G.L. Hamer); Texas A&M Uni-
with generalist vectors (10), although the associations be-
versity, College Station, Texas, USA (S.A. Hamer, G.L. Hamer);
tween biodiversity and disease risk are variable (11).
University of Wisconsin, Madison, Wisconsin, USA (T.L. Goldberg,
In humans, Lyme disease and anaplasmosis caused
T.K. Anderson); Emory University, Atlanta, Georgia, USA (U.D.
by infection with the bacteria Borrelia burgdorferi and
Kitron); University of Illinois, Urbana, Illinois, USA (J.D. Brawn, S.R.
Anaplasma phagocytophilum, respectively, are the 2 most
Loss); and Smithsonian Migratory Bird Center, Washington, DC,
common tick-borne diseases in the midwestern and north-
USA (S.R. Loss)
eastern United States, and both are emerging among human
DOI: http://dx.doi.org/10.3201/eid1810.120511 and canine populations (12,13). In eastern North America,

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1589
RESEARCH

both pathogens are maintained in blacklegged tick (Ixodes 1, 2, or 3 (23) and IGS subtype by comparing them with the
scapularis)–rodent cycles (14,15). We investigated the role 25 major B. burgdorferi IGS subtypes (21,24). The outer
of birds in the urban ecology of tick-borne zoonotic dis- surface protein C (ospC) genotype was inferred on the basis
eases. Our objectives were to 1) ascertain the prevalence of the linkage disequilibrium between IGS locus and ospC
of tick parasitism of birds in residential and urban green locus (21,22).
spaces in southwestern suburban Chicago, Illinois, USA,
during a 6-year period; 2) estimate the infection prevalence Statistical Analyses
of Borrelia spp. and A. phagocytophilum in ticks removed Logistic regression was used to assess the variation in
from birds; and 3) characterize the diversity of pathogens in tick infestations among years. We used 2- and 3-sample
ticks removed from birds by using genetic methods. tests for equality of proportions to assess the effects of site
category, sex, and age on the prevalence of tick infesta-
Materials and Methods tions. The Wilson interval with continuity correction was
used to estimate the 95% binomial CIs for infection preva-
Bird Capture lence data. Minimum infection prevalence (i.e., assuming 1
During May–October 2005–2010, birds were captured positive larva/pool) was used for tests conducted on pooled
at 20 field sites in southwestern suburban Chicago (Cook larvae. Statistical analyses were performed by using Pro-
County; 87°44′ W, 41°42′ N; Figure). Field sites were cat- gram R (R Foundation for Statistical Computing, Vienna,
egorized as residential sites (n = 14) or urban green spaces Austria).
(n = 6) and have been described in detail (6). We used 8–10
mist nets (Avinet, Dryden, NY, USA) to capture birds at Results
7–15 sites per year ≈1 morning per site every 1.5 weeks
(2005–2007) or every 3 weeks (2008–2010). For each Bird Captures
captured bird, we recorded species, sex, age class (hatch We recorded 6,180 total captures, comprising 5,506 in-
year and after hatch year), and weight, and we attached a dividual birds (10.9% recaptures) and 78 species (Table 1).
numbered leg band before release. All birds were checked
for ticks by blowing apart feathers and inspecting the skin,
especially around the ears, head, and vent. Ticks were re-
moved and preserved in 70% ethanol. Migratory status of
each avian species was assigned (16). Fieldwork was car-
ried out with approvals from animal care review boards at
Michigan State University and University of Illinois.

Detection and Typing of Borrelia spp.

and A. phagocytophilum
Ticks were identified morphologically to species and
stage; a subset was subjected to PCR and sequencing for
confirmation (17). All ticks were tested for pathogens, ex-
cept for 2 specimens that were deposited in the US National
Tick Collection (housed at Georgia Southern University,
Statesboro, GA, USA) for molecular identification and
vouchering. Total DNA from ticks was extracted by using Figure. Field sites used for sampling birds in southwest suburban
a DNeasy Blood and Tissue Kit (QIAGEN, Valencia, CA, Chicago, Illinois, USA, 2005–2010. Sites consist of residential
USA) with modifications as described (18). Nymphal ticks areas (numbered sites) and urban green spaces (lettered sites).
were extracted individually, whereas same-species larvae Two residential sites not shown on the map (21 and 22) are ≈20 km
from the same individual animal were pooled. All ticks north of this region. Box in inset map indicates location of sampling
area. Main map shows the landscape gradient of impervious
were tested for the presence of B. burgdorferi sensu stricto surfaces (National Land Cover Database 2001, US Geological
and A. phagocytophilum by using a quantitative PCR tar- Survey, Sioux Falls, SD, USA): dark gray areas are those with a low
geting the 16S rRNA gene (19) and PCR targeting the p44 proportion of impervious cover (urban green spaces, e.g., forest
gene (20), respectively. preserves, parks, cemeteries, riparian buffers); light gray areas
B. burgdorferi–positive tick samples were typed by and white areas are those with a high proportion of impervious
cover (areas with high density of buildings, residential housing, and
DNA sequencing of both strands of the 16S–23S rRNA in- roads). EC, Evergreen Cemetery; PHN, Palos Hills Natural; PL,
tergenic spacer (IGS) region (21); strains were identified, Pleasure Lake; WW, Wolfe Wildlife Refuge; HS, Holy Sepulchre
using updated nomenclature (22), to ribosomal spacer type Cemetery; SC, Saint Casimir Cemetery.

1590 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Urban Ecology of Ticks and Tick-borne Pathogens

Table 1. Birds sampled for presence of ticks in southwestern suburban Chicago, Illinois, USA, 2005–2010*
No. birds infested with
Haemaphysalis Ixodes
Migratory Total no. Proportion leporispalustris dentatus I. scapularis
Bird status examined infested Larvae Nymphs Larvae Larvae Nymphs
American goldfinch B, M 363
American redstart† B, M 38 0.03
American robin B, M 1,049 0.01 2 4 1 4 2
Baltimore oriole B, M 31
Barn swallow B, M 7
Black and white warbler NB, M 9
Black-capped chickadee B, NM 25
Blue jay B, M 22 0.09 2
Brown-headed cowbird B, M 65
Brown thrasher B, M 12
Cedar waxwing B, M 16
Chipping sparrow B, M 24
Common grackle B, M 105 0.03 2 1
Common yellowthroat B, M 8
Dark-eyed junco NB, M 8
Downy woodpecker B, M 50
Eastern wood-pewee B, M 5
Empidonax spp. flycatchers B, M 27
European starling B, M 141 0.01 1
Fox sparrow NB, M 5
Gray catbird B, M 429 0.01 3 3
Gray-cheeked thrush NB, M 18 0.11 1 1
Hermit thrush B, M 5
House finch B, M 157
House sparrow B, NM 2,097 0.01 25 4
House wren B, M 57 0.02 1
Indigo bunting B, M 19
Least flycatcher B, M 5
Lincoln's sparrow NB, M 5
Magnolia warbler NB, M 19
Mourning dove B, M 63
Mourning warbler NB, M 5
Nashville warbler NB, M 7
Northern cardinal B, NM 311 0.04 9 3 1
Northern flicker B, M 10
Northern waterthrush NB, M 44
Orchard oriole B, M 4
Ovenbird B, M 41 0.10 4
Palm warbler NB, M 6
Red-eyed vireo B, M 11
Red-winged blackbird B, M 191 0.01 1 2
Song sparrow B, M 228 0.07 13 6 1
Swainson's thrush‡ NB, M 131 0.08 4 4 1 1
Tennessee warbler NB, M 9
Tree swallow B, M 14
Veery B, M 8
Warbling vireo B, M 35
White-crowned sparrow NB, M 11
White-throated sparrow NB, M 61 0.02 1
Willow flycatcher B, M 63
Wilson's warbler NB, M 8
Yellow warbler B, M 34
Yellow-bellied flycatcher NB, M 6 0.17 1
Yellow-rumped warbler NB, M 26
All 6,197§ 0.02 64 28 6 6 5
*Empidonax spp. flycatchers that could not be identified are considered at the genus level. Numbers of birds infested by larvae and nymphs of 3 tick
species are indicated. Common names conform to species as specified by the American Ornithologist Union. B, confirmed breeding in Chicago region; M,
migratory; NB, non-breeder in Chicago region; NM, non-migratory. Blank spaces mean none infested.
†One American redstart infested with a single Amblyomma longirostre nymph.
‡One Swainson's thrush infested with a single A. nodosum larva.
§This total includes 49 unlisted captured birds from the following species: American woodcock, American tree sparrow, black-billed cuckoo, black-
throated blue warbler, blackpoll warbler, brown creeper, Carolina wren, Canada warbler, Eastern towhee, Eurasian collared–dove, great crested
flycatcher, golden-crowned kinglet, hairy woodpecker, killdeer, marsh wren, olive-sided flycatcher, red-breasted nuthatch, rose-breasted grosbeak, ruby-
crowned kinglet, savannah sparrow, scarlet tanager, swamp sparrow, white-breasted nuthatch, and wood thrush. The sample size for each of these
species was <5, and none of the birds harbored ticks.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1591
RESEARCH

Five species comprised 67% of all captures: Passer do- 6-year study (z value = -1.6, df = 6178, p = 0.109), the
mesticus (house sparrow), Turdus migratorius (Ameri- proportion of infested birds that harbored I. scapularis in-
can robin), Dumetella carolinensis (gray catbird), Spinus creased significantly from 0 to 80% (z value = 3.873, df =
tristis (American goldfinch), and Cardinalis cardinalis 96, p = 0.0001), and I. scapularis comprised >90% of ticks
(northern cardinal). Among all captured birds, 27.3% removed from birds in the final year of the study. Of the
were known males, 21.3% known females, and 51.3% 10 I. scapularis–infested birds, the majority (8) came from
of unknown sex. The age class was after hatch year for urban green spaces (0.28% I. scapularis infestation preva-
53.1%, hatch year for 41.8%, and unknown for 5.1% of lence across all green spaces), and the minority (2) came
the birds. Similar numbers of birds were captured from from residential sites (0.06% prevalence; z value = 2.2, p =
residential sites (3,326, 53.8%) and urban green spaces 0.03). Information about the timing of I. scapularis infesta-
(2,854, 46.2%). Approximately 2× the number of birds tion combined with the species and age of the avian host
were captured per year in 2005–2007 (1,455 ± 45) as in provides evidence for local (Chicago area) acquisition of
2008–2010 (605 ± 159) due to higher mist netting efforts ticks and for migratory importation of ticks from the north
in the initial 3 years of the study. and the south (Table 2).

Tick Prevalence Tick Infection with B. burgdorferi

We removed 357 ticks from 97 individual birds (1 bird and A. phagocytophilum
with ticks was caught twice), yielding an overall tick in- A total of 120 tick samples were tested for pathogens.
festation prevalence of 1.6% (Table 1). Ticks were usually No ticks tested positive for A. phagocytophilum infection.
located beneath the auricular feathers within the skin of the Five samples tested positive for B. burgdorferi infection: 3
ear canal and second most commonly located in the rictus of 6 I. scapularis nymphs (50%, 95% CI 14.0%–86.1%), 1
of the bill and in the skin of the orbital region. Infested of 22 I. scapularis larval pools (minimum infection preva-
birds were collected at 17 of the 20 field sites (11/14 resi- lence 4.5%), and 1 of 34 H. leporispalustris nymphs (2.9%,
dential sites, 6/6 urban green spaces). Birds with the highest 95% CI 0.2%–17.1%) (Table 3). All 5 positive tick sam-
prevalence of infestation (>7% of captures infested) were ples were from unique after–hatch year birds of 4 species
song sparrows (Melospiza melodia), Swainson’s thrushes (American robin, blue jay, red-winged blackbird [Agelaius
(Catharus ustulatus), blue jays (Cyanocitta cristata), oven- phoeniceus], Swainson’s thrush) at 4 field sites, including
birds (Seiurus aurocapilla), gray-cheeked thrushes (Catha- urban green spaces and residential sites. B. burgdorferi
rus minimus), and yellow-bellied flycatchers (Empidonax 16S–23S rRNA IGS sequences were obtained from all 3
flaviventris) (Table 1). I. scapularis nymphs and represented 3 IGS ribotypes (2,
Most ticks were of 3 species: Haemaphysalis leporis- 28, and 14; GenBank accession nos. JQ868562–JQ868564)
palustris (87.4% of all ticks), Ixodes dentatus (4.8%), and within ribosomal spacer type 2 and 3; inferred ospC geno-
I. scapularis (7.8%). Morphologic and molecular iden- types were H, T, and A3, respectively (Table 3).
tifications were congruent for all 21 birds subjected to
both methods of identification (GenBank accession nos. Discussion
JQ868565–JQ868585). Overall, 1.3%, 0.1%, and 0.2% The presence of B. burgdorferi–infected I. scapularis
of birds were infested with H. leporispalustris, I. denta- ticks on migratory and residential birds in the Chicago re-
tus, and I. scapularis, respectively (Table 1). In addition, gion reflects the continued invasion and establishment of
a single Amblyomma nodosum larva was removed from an this tick and pathogen across the Midwest. In Illinois, as in
after–hatch year Swainson’s thrush on May 17, 2005, and many other areas of North America (25), there is growing
a single A. longirostre nymph was removed from an af- public health concern over the emergence of Lyme disease
ter–hatch year American redstart (Setophaga ruticilla) on (26); although, the statewide incidence in Illinois over the
May 18, 2005. The 2 ticks were found on birds captured study period (1.1 cases/100,000 persons) was an order of
at site HS (see Figure) during the spring migration. They magnitude lower than that which characterizes the Lyme
were identified genetically and vouchered at the US Na- disease–endemic regions in the northeastern United States
tional Tick Collection but not tested for pathogens. (27). Our study provides evidence of established local pop-
The number of ticks on infested birds ranged from 1 ulations of I. scapularis ticks in Chicago that may be sup-
to 23 (median 2 ticks). Of the infested birds, 47% harbored plemented by importation of I. scapularis ticks from other
1 tick and 20% harbored >5 ticks. H. leporispalustris lar- populations to the north or south by migratory birds. The
vae accounted for the greatest tick loads (average 4.3 ticks/ Chicago region is a natural corridor for migratory birds,
bird). Of 98 parasitized birds, 11 (11.2%) were infested and the risk for tick and pathogen introduction is likely to
with >1 life stage of tick or >1 tick species. Although the be elevated on migratory flyways because of seasonal con-
overall prevalence of infested birds did not change over the centrations of birds.

1592 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Urban Ecology of Ticks and Tick-borne Pathogens

Table 2. Demographic information about 10 avian hosts infested with Ixodes scapularis ticks in southwestern suburban Chicago,
Illinois, USA, 2005–2010*
I. scapularis Presumed I. scapularis
Bird Date of capture Age Site, category stage (quantity) acquisition
American robin 2007 Jul 18 AHY 1, residential L (9); N (1) Local
American robin 2009 Aug 18 HY PL, green space L( 2) Local
American robin 2010 Jun 22 AHY PHN, green space N( 2) Local
American robin 2010 Jul 13 AHY PL, green space L (1) Local
American robin 2010 Jul 26 HY PL, green space L (8) Local
Blue jay 2009 Jun 15 AHY PHN, green space N( 1) Local
Blue jay 2009 Jun 15 AHY PHN, green space N (1) Local
Gray-cheeked thrush 2010 Sep 16 HY PHN, green space N( 1) Migratory (from north)
Northern cardinal 2007 Aug 16 HY 13, residential L( 1) Local
Swainson’s thrush 2006 May 23 AHY WW, green space L (1) Migratory (from south)
*AHY, after hatch year; L, larva; N, nymph. HY, hatch year; PL, Pleasure Lake; PHN, Palos Hills Natural; WW, Wolfe Wildlife Refuge.

We detected a B. burgdorferi–positive I. scapularis Other ticks commonly found on birds in Chicago are I.
larval pool from a Swainson’s thrush. Given the absence dentatus and H. leporispalustris ticks, both of which feed
of transovarial transmission in the I. scapularis tick, this almost exclusively on rabbits and birds. I. dentatus ticks
finding demonstrates that the Swainson’s thrush can be an are enzootic vectors of B. burgdorferi in regions where I.
infectious reservoir host. On the basis of a limited sample scapularis ticks do not occur (24). H. leporispalustris ticks
(n = 6), we determined that birds in Chicago harbored B. transmit Francisella tularensis and spotted-fever group
burgdorferi–infected I. scapularis nymphs at a prevalence rickettsiae among wildlife (32). In our study, H. leporis-
(14.0%–86.1%) consistent with that reported for questing palustris ticks had a wide geographic presence across most
nymphs and ticks from birds in Michigan (18), Minnesota residential sites and were most commonly found on house
(28), and Canada (29). All 3 B. burgdorferi IGS ribotypes sparrows, including 7 hatch-year birds, implying local ac-
present within nymphs in this study have been associated quisition in the residential neighborhoods. Neither I. den-
with host-seeking nymphs in Lyme disease–endemic areas tatus nor H. leporispalustris ticks regularly infest humans.
of the midwestern and northeastern United States; 2 of the We document the presence of 2 neotropical tick spe-
3 ribotypes were previously detected in larvae removed cies, A. longirostre and A. nodosum, on birds migrating
from birds (30). Two of the ospC types (H and A) pre- north through Chicago. We note that other species of
sumed present in the collected ticks were among the 4 most neotropical Amblyomma ticks have been recovered in the
invasive genotypes (I, A, H, B) from a study of B. burgdor- spring on migrant birds in southern Canada (33). A. longi-
feri isolates from humans in New York (31). The presence rostre and A. nodosum ticks are widely distributed in the
of avian reservoirs and I. scapularis nymphs infected with neotropical region, and are vectors of Rickettsia ambly-
B. burgdorferi strains capable of causing disseminated hu- ommii (34) (which may cause rickettsiosis in humans in
man disease supports the possibility that reported cases of North America) (35), R. bellii, and R. parkeri (36). In the
human Lyme disease in Chicago residents may result from United States, R. parkeri is a newly recognized cause of
local exposure to infected I. scapularis ticks. Although human disease, and a high prevalence of infection (>40%
none of the ticks removed from birds were positive for A. in adults) has been associated with growing populations
phagocytophilum, the growing I. scapularis tick population of Gulf Coast ticks (A. maculatum) (37). Migrant birds
in the region raises the possibility that infection with this from the neotropics likely account for many imports of
pathogen could become an emerging health concern. engorged neotropical ticks and associated pathogens in

Table 3. Prevalence of Borrelia burgdorferi infection in ticks removed from birds, by site of origin and date of capture, southwest
suburban Chicago, Illinois, USA, 2005–2010*
Larva Nymph
No. pools % Infected Birds with infected No. % Infected Birds with infected IGS strain ospC
Tick species (no. larvae) (MIP) larvae, site, date tested (95% CI) nymphs, site, date (RST group) strain
Haemaphysalis 65 (277) 0 NA 34 2.9 RWBL, SC site, NA NA
leporispalustris (0.2–17.1) 2007 Jun 6
Ixodes dentatus 6 (17) 0 NA 0 NA NA NA NA
I. scapularis 6 (22) 16.7 (4.5) SWTH, WW site, 6 50 AMRO, 1 site, 2007 2 (2); 28 (3); H, T,
2006 May 23 (14.0–86.1) Jul 18; AMRO, 14 (2) A3
PHN site, 2010 Jun
22; BLJA, PHN site,
2009 Jun 15
*MIP, minimum infection prevalence; IGS, B. burgdorferi 16S-23S rRNA intergenic spacer ribotype; RST, ribosomal spacer type 1, 2, or 3; ospC, inferred
outer surface protein C allele based on linkages reported by Travinsky et al. (23); NA, not applicable; RWBL, Red-winged blackbird; SC, Saint Casimir
Cemetery; SWTH, Swainson’s thrush; WW, Wolfe Wildlife Refuge; AMRO, American robin; PHN, Palos Hills Natural; BLJA, Blue jay.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1593
RESEARCH

North America each spring, but a lack of environmental 3. Peterson AT, Andersen MJ, Bodbyl-Roels S, Hosner P, Nyari A,
receptivity (host or climatic limitations) has likely pre- Oliveros C, et al. A prototype forecasting system for bird-borne
disease spread in North America based on migratory bird move-
vented establishment. ments. Epidemics. 2009;1:240–9. http://dx.doi.org/10.1016/j.
Data from our large sampling effort show that the dis- epidem.2009.11.003
persal of I. scapularis ticks, B. burgdorferi, and neotropi- 4. Keesing F, Brunner J, Duerr S, Killilea M, LoGiudice K, Schmidt
cal vector ticks is a rare but detectable event. We sampled K, et al. Hosts as ecological traps for the vector of Lyme dis-
ease. Proc Biol Sci. 2009;276:3911–9. http://dx.doi.org/10.1098/
several thousand birds and detected I. scapularis ticks on rspb.2009.1159
<0.2% and neotropical ticks on <0.05%. However, the 5. Ginsberg HS, Buckley PA, Balmforth MG, Zhioua E, Mitra S,
rarity of infestations does not mean that infestation is bio- Buckley FG. Reservoir competence of native North American
logically insignificant. Despite the positive relationship be- birds for the Lyme disease spirochete, Borrelia burgdorferi. J
Med Entomol. 2005;42:445–9. http://dx.doi.org/10.1603/0022-
tween propagule pressure and invasion success, some suc- 2585(2005)042[0445:RCONNA]2.0.CO;2
cessful species invasions, especially those of arthropods, 6. Hamer GL, Chaves LF, Anderson TK, Kitron UD, Brawn JD, Ruiz
can be initiated by a very small number of individuals (38). MO, et al. Fine-scale variation in vector host use and force of in-
Low propagule pressure but successful invasion may occur fection drive localized patterns of West Nile virus transmission.
PLoS ONE. 2011;6:e23767. http://dx.doi.org/10.1371/journal.
when the environment is receptive to the particular species pone.0023767
of ticks and pathogens being dispersed. Indeed, during our 7. Kurtenbach K, De Michelis S, Etti S, Schafer SM, Sewell HS, Brade
study, other researchers showed an increase in the occur- V, et al. Host association of Borrelia burgdorferi sensu lato—the key
rence of B. burgdorferi–infected adult I. scapularis ticks role of host complement. Trends Microbiol. 2002;10:74–9. http://
dx.doi.org/10.1016/S0966-842X(01)02298-3
in northwestern Chicago, confirming our prediction (26). 8. Bradley CA, Altizer S. Urbanization and the ecology of wild-
Such scenarios of rare introduction but successful estab- life diseases. Trends Ecol Evol. 2007;22:95–102. http://dx.doi.
lishment of ticks and pathogens pose a major risk for the org/10.1016/j.tree.2006.11.001
health of humans, wildlife, and domestic animals in urban 9. Hamer SA, Lehrer E, Magle SB. Wild birds as sentinels for mul-
tiple zoonotic pathogens along an urban to rural gradient in greater
environments worldwide. Chicago, Illinois. Zoonoses Public Health. 2012;59:355–64. http://
dx.doi.org/10.1111/j.1863-2378.2012.01462.x
Acknowledgments 10. Keesing F, Belden LK, Daszak P, Dobson A, Harvell CD, Holt
We thank Diane Ghode, Patrick Kelly, Bethany Krebs, and RD, et al. Impacts of biodiversity on the emergence and transmis-
sion of infectious diseases. Nature. 2010;468:647–52. http://dx.doi.
Timothy Thompson for field support; Marilyn Ruiz and Jean org/10.1038/nature09575
Tsao for assisting with project infrastructure and thoughtful dis- 11. Randolph SE, Dobson AD. Pangloss revisited: a critique of
cussions; and Lorenza Beati for assisting with tick identification the dilution effect and the biodiversity-buffers-disease para-
and archiving. We thank the Village of Oak Lawn, Illinois, for digm. Parasitology. 2012;139:847–63. http://dx.doi.org/10.1017/
S0031182012000200
providing field laboratory facilities; other Illinois municipalities 12. Chomel B. Tick-borne infections in dogs—an emerging infectious
(Evergreen Park, Palos Hills, Burbank, Alsip, Indian Head Park) threat. Vet Parasitol. 2011;179:294–301. http://dx.doi.org/10.1016/j.
and the City of Chicago; and private homeowners for allowing vetpar.2011.03.040
us to conduct this research. We also thank the 2 anonymous peer 13. Dumler JS, Choi KS, Garcia-Garcia JC, Barat NS, Scorpio DG,
Garyu JW, et al. Human granulocytic anaplasmosis and Anaplas-
reviewers whose comments helped to improve our manuscript. ma phagocytophilum. Emerg Infect Dis. 2005;11:1828–34. http://
dx.doi.org/10.3201/eid1112.050898
This project was funded through National Science Founda-
14. Telford SR, Dawson JE, Katavolos P, Warner CK, Kolbert CP,
tion Ecology of Infectious Disease Grants 0429124 and 0840403. Persing DH. Perpetuation of the agent of human granulocytic eh-
rlichiosis in a deer tick-rodent cycle. Proc Natl Acad Sci U S A.
Sarah A Hamer is a veterinary ecologist and assistant profes- 1996;93:6209–14. http://dx.doi.org/10.1073/pnas.93.12.6209
sor in the Veterinary Integrative Biosciences Department at Texas 15. Barbour AG, Fish D. The biological and social phenomenon of
A&M University. Her research interests include the ecology, evo- Lyme disease. Science. 1993;260:1610–6. http://dx.doi.org/10.1126/
lution, and epidemiology of wildlife and vector-borne and zoo- science.8503006
16. Walk JW, Ward MP, Benson TJ, Deppe JL, Lischka SA, Bailey SD,
notic diseases. et al. Illinois birds: a century of change. Champaign (IL): Illinois
Natural History Survey; 2010.
17. Poucher KL, Hutcheson HJ, Keirans JE, Durden LA, Black WC.
References Molecular genetic key for the identification of 17 Ixodes species of
the United States (Acari: Ixodidae): a methods model. J Parasitol.
1. Kilpatrick AM, Chmura AA, Gibbons DW, Fleischer RC, Marra 1999;85:623–9. http://dx.doi.org/10.2307/3285734
PP, Daszak P. Predicting the global spread of H5N1 avian influen- 18. Hamer SA, Tsao JI, Walker ED, Hickling GJ. Invasion of the
za. Proc Natl Acad Sci U S A. 2006;103:19368–73. http://dx.doi. Lyme disease vector Ixodes scapularis: implications for Borrelia
org/10.1073/pnas.0609227103 burgdorferi endemicity. EcoHealth. 2010;7:47–63. http://dx.doi.
2. Rappole JH, Hubálek Z. Migratory birds and West Nile vi- org/10.1007/s10393-010-0287-0
rus. J Appl Microbiol. 2003;94(Suppl):47S–58S. http://dx.doi.
org/10.1046/j.1365-2672.94.s1.6.x

1594 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Urban Ecology of Ticks and Tick-borne Pathogens

19. Tsao JI, Wootton JT, Bunikis J, Luna MG, Fish D, Barbour AG. An 31. Wormser GP, Brisson D, Liveris D, Hanincova K, Sandigursky S,
ecological approach to preventing human infection: vaccinating wild Nowakowski J, et al. Borrelia burgdorferi genotype predicts the ca-
mouse reservoirs intervenes in the Lyme disease cycle. Proc Natl pacity for hematogenous dissemination during early Lyme disease. J
Acad Sci U S A. 2004;101:18159–64. http://dx.doi.org/10.1073/ Infect Dis. 2008;198:1358–64. http://dx.doi.org/10.1086/592279
pnas.0405763102 32. Sonenshine DE. Ticks of Virginia. Blacksburg (VA): Virginia Poly-
20. Holden K, Boothby JT, Anand S, Massung RF. Detection of Bor- technic Institute and State University, College of Agriculture and
relia burgdorferi, Ehrlichia chaffeensis, and Anaplasma phagocyto- Life Sciences; 1979.
philum in ticks (Acari: Ixodidae) from a coastal region of California. 33. Scott JD, Lee M-K, Fernando K, Durden LA, Jorgensen DR, Mak
J Med Entomol. 2003;40:534–9. http://dx.doi.org/10.1603/0022- S, et al. Detection of Lyme disease spirochete, Borrelia burgdor-
2585-40.4.534 feri sensu lato, including three novel genotypes in ticks (Acari: Ix-
21. Bunikis J, Garpmo U, Tsao J, Berglund J, Fish D, Barbour AG. Se- odidae) collected from songbirds (Passeriformes) across Canada.
quence typing reveals extensive strain diversity of the Lyme bor- J Vector Ecol. 2010;35:124–39. http://dx.doi.org/10.1111/j.1948-
reliosis agents Borrelia burgdorferi in North America and Borrelia 7134.2010.00068.x
afzelii in Europe. Microbiology. 2004;150:1741–55. http://dx.doi. 34. Ogrzewalska M, Uezu A, Jenkins CN, Labruna MB. Effect of forest
org/10.1099/mic.0.26944-0 fragmentation on tick infestations of birds and tick infection rates by
22. Travinsky B, Bunikis J, Barbour AG. Geographic differences in Rickettsia in the Atlantic forest of Brazil. EcoHealth. 2011;8:320–
genetic locus linkages for Borrelia burgdorferi. Emerg Infect Dis. 31. http://dx.doi.org/10.1007/s10393-011-0726-6
2010;16:1147–50. http://dx.doi.org/10.3201/eid1607.091452 35. Apperson CS, Engber B, Nicholson WL, Mead DG, Engel J, Yabs-
23. Liveris D, Gazumyan A, Schwartz I. Molecular typing of Borrelia ley MJ, et al. Tick-borne diseases in North Carolina: is “Rickettsia
burgdorferi sensu lato by PCR-restriction fragment length polymor- amblyommii” a possible cause of rickettsiosis reported as Rocky
phism analysis. J Clin Microbiol. 1995;33:589–95. Mountain spotted fever? Vector-borne Zoonotic Dis. 2008;8:597–
24. Hamer SA, Hickling GJ, Sidge JL, Rosen ME, Walker ED, Tsao JI. 606. http://dx.doi.org/10.1089/vbz.2007.0271
Diverse Borrelia burgdorferi strains in a bird tick cryptic cycle. Appl 36. Ogrzewalska M, Pacheco RC, Uezu A, Richtzenhain LJ, Fer-
Environ Microbiol. 2011;77:1999–2007. http://dx.doi.org/10.1128/ reira F, Labruna MB. Rickettsial infection in Amblyomma nodo-
AEM.02479-10 sum ticks (Acari: Ixodidae) from Brazil. Ann Trop Med Parasitol.
25. Diuk-Wasser MA, Hoen AG, Cislo P, Brinkerhoff R, Hamer SA, 2009;103:413–25. http://dx.doi.org/10.1179/136485909X451744
Rowland M, et al. Human risk of infection with Borrelia burgdorferi, 37. Fornadel CM, Zhang X, Smith JD, Paddock CD, Arias JR, Norris
the Lyme disease agent, in eastern United States. Am J Trop Med DE. High rates of Rickettsia parkeri infection in Gulf Coast ticks
Hyg. 2012;86:320–7. http://dx.doi.org/10.4269/ajtmh.2012.11-0395 (Amblyomma maculatum) and identification of “Candidatus Rickett-
26. Jobe DA, Nelson JA, Adam MD, Martin SA. Lyme disease in urban sia andeanae” from Fairfax County, Virginia. Vector Borne Zoonotic
areas, Chicago. Emerg Infect Dis. 2007;13:1799–800. http://dx.doi. Dis. 2011;11:1535–9. http://dx.doi.org/10.1089/vbz.2011.0654
org/10.3201/eid1311.070801 38. Simberloff D. The role of propagule pressure in biological inva-
27. Centers for Disease Control and Prevention. Summary of notifiable sions. Annu Rev Ecol Evol Syst. 2009;40:81–102. http://dx.doi.
diseases. MMWR Morb Mortal Wkly Rep. 2011;58:1–100. org/10.1146/annurev.ecolsys.110308.120304
28. Weisbrod AR, Johnson RC. Lyme disease and migrating birds in the
Saint Croix River Valley. Appl Environ Microbiol. 1989;55:1921–4. Address for correspondence: Sarah A. Hamer, Veterinary Integrative
29. Ogden NH, Lindsay LR, Hanincova K, Barker IK, Bigras-Poulin M,
Bioscience Department, 107-A VMA Building, Texas A&M University,
Charron DF, et al. Role of migratory birds in introduction and range
expansion of Ixodes scapularis ticks and of Borrelia burgdorferi and College Station, TX 77843-4458, USA; email: shamer@cvm.tamu.edu
Anaplasma phagocytophilum in Canada. Appl Environ Microbiol.
2008;74:1780–90. http://dx.doi.org/10.1128/AEM.01982-07
30. Brinkerhoff RJ, Bent SJ, Folsom-O’Keefe CM, Tsao K, Hoen AG,
The opinions expressed by authors contributing to this
Barbour AG, et al. Genotypic diversity of Borrelia burgdorferi
journal do not necessarily reflect the opinions of the Centers for
strains detected in Ixodes scapularis larvae collected from North
Disease Control and Prevention or the institutions with which
American songbirds. Appl Environ Microbiol. 2010;76:8265–8.
the authors are affiliated.
http://dx.doi.org/10.1128/AEM.01585-10

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1595
RESEARCH

Spread of Influenza Virus A (H5N1)

Clade 2.3.2.1 to Bulgaria in
Common Buzzards
Atanaska Marinova-Petkova, Georgi Georgiev, Patrick Seiler, Daniel Darnell, John Franks,
Scott Krauss, Richard J. Webby, and Robert G. Webster

On March 15, 2010, a highly pathogenic avian influenza a second outbreak a week later in Kowloon Park, also in
virus was isolated from the carcass of a common buzzard Hong Kong (4). Since then, cases of HPAIVs (H5 and H7
(Buteo buteo) in Bulgaria. Phylogenetic analyses of the subtypes) in wild birds have been reported often.
virus showed a close genetic relationship with influenza In May 2005, a massive HPAIV (H5N1) outbreak
virus A (H5N1) clade 2.3.2.1 viruses isolated from wild occurred in wild aquatic birds in Qinghai Lake, western
birds in the Tyva Republic and Mongolia during 2009–2010.
People’s Republic of China; 6,184 gulls, geese, great
Designated A/common buzzard/Bulgaria/38WB/2010,
this strain was highly pathogenic in chickens but had low
cormorants, and ruddy shelducks died (5). The lake is a
pathogenicity in mice and ferrets and no molecular markers staging area for migratory waterfowl, and the scientific
of increased pathogenicity in mammals. The establishment community’s fear that the virus would spread during
of clade 2.3.2.1 highly pathogenic avian influenza viruses migration (6) was realized when the so-called Qinghai-
of the H5N1 subtype in wild birds in Europe would increase like influenza virus A (H5N1) clade 2.2 spread to western
the likelihood of health threats to humans and poultry in the Siberia in Russia and then to many countries in Asia, the
region. Middle East, Europe, and Africa at the end of 2005 and
during 2006, killing poultry flocks and wild birds.
The 2008 classification system used to describe the
W ild aquatic birds are considered natural reservoirs
of all known influenza virus subtypes (1). Highly
pathogenic avian influenza viruses (HPAIVs) usually cause
evolution and diversification of the HPAIVs (H5N1) that
emerged from the A/goose/Guangdong/96 lineage (7)
asymptomatic infections in waterfowl. Compared with that was updated in 2011. Phylogenetic analysis of all isolated
for poultry, the number of reported outbreaks of HPAIVs influenza (H5N1) viruses showed that some of the 10
in wild birds (aquatic or terrestrial) before 2002 was low; first-order clades (0–9) had stopped circulating in 2008
1 influenza A (H5N3) outbreak occurred in wild common or earlier (clades 0, 3, 4, 5, 6, 8, 9), as had some second-
terns (Sterna hirundo) in South Africa in 1961 (2), and H7 and third-order groups of clade 2. Meanwhile, clades 1,
subtype HPAIV was isolated from a Saker falcon (Falco 2.1.3, 2.2, 2.2.1, 2.3.2, 2.3.4, and 7 continued to evolve
cherrug) in Italy in 2000 (3). In December 2002, a die-off rapidly (8). Clade 2.3.2 is widely distributed in Asia,
of aquatic waterfowl caused by an HPAIV (H5N1) occurred particularly in China, Hong Kong, Korea, Vietnam, Laos,
in Penfold Park in Hong Kong. That event was followed by Bangladesh, Nepal, Mongolia, and the Tyva Republic; it
is also distributed in eastern Europe, mainly in Romania
Author affiliations: Regional Diagnostic Laboratory on Avian
and Bulgaria (8,9). Tyva is part of the Siberian Federal
Influenza and Newcastle Disease in Birds, Varna, Bulgaria (A.
District of Russia, which is located north of Mongolia.
Marinova-Petkova); National Diagnostic and Research Veterinary
The remaining circulating influenza (H5N1) clades have
Medical Institute, Sofia, Bulgaria (A. Marinova-Petkova, G.
specific geographic locations: clade 1 circulates in southern
Georgiev); and St. Jude Children’s Research Hospital, Memphis,
Vietnam and Cambodia; clade 2.1.3 in Indonesia; clade 2.2
TN, USA (P. Seiler, D. Darnell, J. Franks, S. Krauss. R.J. Webby,
in India and Bangladesh; clade 2.2.1 in Egypt; clade 2.3.4
R.G. Webster)
in China, Hong Kong, Vietnam, Thailand, and Laos; and
DOI: http://dx.doi.org/10.3201/eid1810.120357 clade 7 in China and Vietnam.

1596 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Influenza Virus A (H5N1) Clade 2.3.2.1, Bulgaria

Before 2006, no avian influenza outbreaks in poultry MrBayes program (18) to apply a Bayesian method. The
had been reported in Bulgaria; 4 cases of HPAIV (H5N1) model of nucleotide substitution that best fits our data was
clade 2.2 were confirmed in dead swans found in 4 regions selected by using the Modeltest 3.7 program (19).
of the country early that year (10). On March 15, 2010,
the carcass of a common buzzard (Buteo buteo) containing Antigenic Characterization and Pathogenicity Studies
HPAIV (H5N1) was found at St. Konstantin and Helena We assessed the antigenic relationship of A/common
Black Sea Resort in Bulgaria and submitted to the Regional buzzard/Bulgaria/38WB/2010 with other influenza
Diagnostic Laboratory on Avian Influenza (Varna, (H5N1) viruses by performing HI assays with a panel of
Bulgaria). The virus was characterized as clade 2.3.2.1. postinfection ferret antiserum against viruses from clades
2.3.2, 2.3.2.1, and 2.3.4; the panel was produced at St.
Materials and Methods Jude Children’s Research Hospital (Memphis, TN, USA;
Table). All animal studies were conducted in United
Virus Isolation and Initial Characterization States Department of Agriculture–approved biosafety
Pooled lung, trachea, liver, cecal tonsil, and gizzard level 3 enhanced facilities at St. Jude Children’s Research
tissue from a common buzzard were injected into Hospital.
embryonated chicken eggs, and an avian influenza virus (A/
common buzzard/Bulgaria/38WB/2010) was isolated. The Pathogenicity Tests in Chickens
isolate was subjected to hemagglutination inhibition (HI) The intravenous pathogenicity index test was conducted
assays, as specified in the World Organisation for Animal according to the OIE Manual of Diagnostic Tests and
Health (OIE) Manual of Diagnostic Tests and Vaccines for Vaccines for Terrestrial Animals 2011 (11) to determine
Terrestrial Animals 2011 (11). The 50% egg infectious dose whether A/common buzzard/Bulgaria/38WB/2010 is
(EID50) was assayed in 10-day-old embryonated chicken pathogenic in chickens. A natural route of infection study
eggs after a 40-hour incubation at 35°C and calculated was performed in five 8-week-old specific pathogen–free
according to Reed and Muench (12). chickens. A/common buzzard/Bulgaria/38WB/2010 (106
EID50 in 0.5 mL) was administered to each bird as follows:
Sequence Analysis 0.1 mL in the nares, 0.1 mL in the trachea, 0.2 mL in the
Viral RNA was extracted from virus-containing throat, and 1 drop in each eye. The chickens were examined
allantoic fluid by using the MagMax-96 AI/ND RNA daily for clinical signs of disease.
extraction kit (Applied Biosystems/Ambion, Austin, TX,
USA). Reverse transcription PCR was performed by Pathogenicity Tests in Mice and Ferrets
using a universal set of primers (13,14); whole-genome The 50% mouse lethal dose was determined to assess
sequencing was performed by using an Illumina Genome the pathogenicity of the buzzard influenza (H5N1) virus in
Analyzer (Illumina, San Diego, CA, USA), as described by mammals. The experiment was conducted as described by
Ducatez et al. (15). Boon et al. (20), and the titer was calculated according to
Reed and Muench (12).
Phylogenetic Analysis We also performed a pathogenicity and transmissibility
We performed a phylogenetic analysis of all gene study in a ferret model (21). We collected nasal washes
segments of A/common buzzard/Bulgaria/38WB/2010. from all ferrets (i.e., inoculated ferrets, naive direct-contact
Sequences were retrieved from the National Center for ferrets, and naive respiratory droplet–contact ferrets)
Biotechnology Information influenza virus database (16) every other day and nasal swabs from only the donors
and aligned by using the ClustalW tool in MEGA4 (17). (i.e., inoculated ferrets) on day 4 postinfection. EID50 was
Phylogenetic relationships were estimated by using the calculated according to Reed and Muench (12).

Table. Hemagglutination inhibition titers used to compare the antigenicity of influenza A(H5N1) virus strain A/common
buzzard/Bulgaria/38WB/2010 with other common strains*
Strain (clade)
A/common A/Muscovy A/duck/Laos/ A/Japanese
magpie/Hong Kong/ duck/Vietnam/ 3295/2006 white-eye/Hong
Strain 5052/2007 (2.3.2.1) 1455/2006 (2.3.2) (2.3.4) Kong/1038/2006 (2.3.4)
A/common magpie/Hong Kong/5052/2007 160 160 <40 <40
A/duck/Laos/3295/2006 <40 <40 160 80
A/Japanese white-eye/Hong Kong/1038/2006 <40 40 160 320
A/Muscovy duck/Vietnam/1455/2006 80 80 <40 <40
A/common buzzard/Bulgaria/38WB/2010 80 80 <40 <40
*Values represent titers that are the reciprocal of the lowest dilution of ferret antisera that inhibited hemagglutination caused by 4 hemagglutination units
of the virus.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1597
RESEARCH

Results

Gross Pathologic Findings

The fresh common buzzard carcass that was initially
found was in good condition, without any discharge or other
clinical signs of disease. Necropsy revealed a sufficient
quantity of yellow subcutaneous and abdominal fat tissue.
Gross pathologic changes were not observed in the lungs,
trachea, heart, liver, spleen, gizzard, stomach, pancreas, or
intestines.

Antigenic Characterization and

Pathogenicity in Chickens
Figure. Virus titers in nasal washes or nasal swabs (*) of individual
In the HI test, A/common buzzard/Bulgaria/38WB/2010 donor ferrets inoculated with A/common buzzard/38WB/2010.
had a titer of 80 to antiserum from influenza (H5N1) clades EID50, 50% egg infectious dose.
2.3.2 and 2.3.2.1 (Table). The intravenous pathogenicity
index of A/common buzzard/Bulgaria/38WB/2010 was
scored 3.0 because all 10 birds were found dead within
24 hours after inoculation. The natural route of infection However, none of the antigenic sites in HA1 of the H5
study resulted in 2 chickens being found dead on day HA were mutated in comparison with A/whooper swan/
2 postinfection, and the other 3 were sick. On day 3 Mongolia/2/2009, A/great crested grebe/Qinghai/1/2009, A/
postinfection, the remaining 3 chickens were found dead. grebe/Tyva/3/2009, A/bar-headed goose/Qinghai/1/2010,
The clinical signs observed before death were cloudy eyes, and other closely related strains from clade 2.3.2.1 available
cyanosis of the exposed skin and wattles, edema of the in GenBank (site 1: amino acids 136–141; site 2: 152–153;
face, and diarrhea. site 3: 124–129, H5 numbering).
The receptor-binding pocket of HA1 retained residues
Pathogenicity in Mice and Ferrets E186, Q222, and G224, which preferentially bind to avian
The 50% mouse lethal dose of A/common buzzard/ α-2,3-NeuAcGal receptors (23). All conserved residues
Bulgaria/38WB/2010 was 30 EID50 (EID50 of the virus was within the HA receptor–binding domains of A/common
109/mL). All 3 donor ferrets were shedding virus by day 5 buzzard/Bulgaria/38WB/2010 were identical to those of
postinfection; however, only 1 was still shedding virus by other clade 2.3.2 HA sequences available in GenBank
day 7 postinfection (Figure). Virus titers were not detected (online Technical Appendix Figure).
in the nasal washes of contact ferrets, indicating that A/
common buzzard/Bulgaria/38WB/2010 is not transmissible Neuraminidase
by direct contact or respiratory droplets. All ferrets used in Compared with that of A/goose/Guangdong/1/96,
this study were healthy during the experiment and showed the amino acid sequence of A/common buzzard/
no clinical signs of disease; they were alert, playful, and Bulgaria/38WB/2010s neuraminidase had a 20-residue
eating and drinking normally. deletion in the stalk region (residues 49–68), which was
thought to be required for influenza viruses to adapt from
Molecular Characterization of A/common buzzard/ wild aquatic birds to domestic chickens (24). This deletion
Bulgaria/38WB/2010 causes a loss of the N terminal NQS glycosylation site
Genomic mutations identified in A/common buzzard/ (positions 50–52). Residues E119, H275, or N295 (N1
Bulgaria/38WB/2010 are listed in the online Technical numbering) were not mutated, which suggests sensitivity
Appendix Table (wwwnc.cdc.gov/EID/pdfs/12-0357- to oseltamivir and zanamivir (25,26).
Techapp.pdf). The genome sequences of A/common
buzzard/Bulgaria/38WB/2010 (H5N1) have been deposited Polybasic Proteins
in GenBank (accession nos. CY110851–CY110858). The A/common buzzard/Bulgaria/38WB/2010s
PB1 gene sequence contained 3 synonymous mutations,
Hemagglutinin G204A, A750G, and G2052A; G204A in PB1’s open
The virus’s hemagglutinin (HA) cleavage site had the reading frame (ORF) caused a unique R37Q mutation in the
polybasic amino acid sequence PQRERRRKRGLF, which polybasic (PB) 1-F2 protein. The PB1-F2 protein generated
is characteristic of HPAIVs (22). The cleavage site also by the ORF plus 1 of the virus’s PB1 genes is a 90-aa
had a K329 deletion (H5 numbering used throughout). polypeptide. According to Zell et al. (27), PB1-F2 proteins

1598 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Influenza Virus A (H5N1) Clade 2.3.2.1, Bulgaria

consisting of >78 aa are intact and functional. The PB1-F2 phylogenetic analysis of the remaining genes using the same
N66S mutation, which is characteristic of increased viral group of viruses that had been used to make the HA tree.
pathogenicity and contributed to the high lethality of the In the N1, PB1, PB2, PA, NS, and NP phylogenetic trees,
1918 pandemic influenza virus (28), was not observed in A/ A/common buzzard/Bulgaria/38WB/2010 clustered with
common buzzard/Bulgaria/38WB/2010. However, several the other subtype H5N1 viruses from clade 2.3.2.1 (data not
PB2 mutations were present (online Technical Appendix shown). In the M gene tree, the Bulgarian subtype H5N1
Table). virus clustered with the clade 2.3.4 subtypes from Guangxi,
Hunan, Fujian, Shantou, and Hong Kong that were isolated in
Nonstructural Proteins 2005, 2006, and 2008; all other subtype H5N1 viruses from
The nonstructural (NS) 1 protein of A/common clade 2.3.2.1 clustered in a separate group (online Technical
buzzard/Bulgaria/38WB/2010 is encoded by 230 aa. This Appendix Figure). The evolutionary distance between the 2
sequence differs from other closely related NS sequences groups of isolates in the M tree is long, indicating that the M
in the National Center for Biotechnology Information gene of A/common buzzard/Bulgaria/38WB/2010 originates
database in that it lacks the 5-aa deletion (residues 80–84) from a non–clade 2.3.2 ancestor.
that became common in HPAIV (H5N1) NS sequences
after they were first observed during 2001 in poultry in Discussion
Hong Kong. The day after we received the HPAIV (H5N1)–
containing common buzzard carcass, officials in Romania
Matrix Proteins notified the OIE of an outbreak of an HPAIV (H5N1)
The M gene–encoded ORF of matrix (M) 1 protein in Letea Village (Danube Delta); 47 backyard chickens
consists of 252 aa, and that of M2 consists of 97 aa. were found dead. These reports are considered to be the
Residues 26L, 27V, 30A, 31S, and 34G of M2’s introduction of HPAIV (H5N1) clade 2.3.2.1 into Europe.
transmembrane region indicate that A/common buzzard/ Furthermore, the European Reference Laboratory on Avian
Bulgaria/38WB/2010 is an amantadine-sensitive strain Influenza and Newcastle Disease (Weybridge, UK) found
(29). that the HA gene sequence of the Bulgarian isolate is
99.9% similar to that of the Romanian isolates from Letea
Nucleoprotein Village (34), confirming that both viruses are derived from
This virus’s nucleoprotein (NP) contained 4 aa a common source, which is most likely wild birds.
substitutions, V33I, R293K, N395T and S450N. Normally, St. Konstantin and Helena Black Sea Resort, where
equine and avian influenza NPs have V33 (30). I33 is the common buzzard carcass was found, is located on
characteristic of human and swine influenza NPs (30,31), the border of Batova (43°21′11″N, 27°57′33″E), which is
and V33I is a common amino acid substitution found in the a complex of habitats typical for woodland bird species,
pandemic viruses from 1918, 1957, 1968, 1977, and 2009 waterfowl, and poultry. This area is defined as a bottleneck
(32). K293 is also unique to human viruses (32,33). The migration site of global importance because 3 flows of
NP of A/common buzzard/Bulgaria/38WB/2010 and that migratory birds meet over the Batova River Valley (35).
of the closely related A/whooper swan/Mongolia/1/2010 Each spring (March–April), migratory waterfowl from
virus contain N450, a substitution found in North American Africa, Bosphorus, and the Dardanelle Straits stop at the
bird strains; however, Eurasian strains typically contain lakes along the coast of the Black Sea in Bulgaria on their
S450. way north (36). Their next resting spot is the Danube
Delta, where the Romanian outbreak occurred and <250
Phylogenetic Analysis km from where the Bulgarian buzzard carcass was found.
To determine the place of A/common buzzard/ Thus, if we are correct in our hypothesis that birds from
Bulgaria/38WB/2010 in the modern classification of Via Pontica are the hosts that carried the HPAIV (H5N1)
influenza (H5N1) viruses, we phylogenetically analyzed to Bulgaria and Romania, the virus most likely came from
HA nucleotide sequences of viruses representing all clades Tyva and Mongolia, where its ancestral viruses had been
that had been reported as of the end of 2010. The HA of isolated, to Via Pontica by using >2 other overlapping
the Bulgarian isolate clustered with subtypes from clade flyways. Although this is the most likely scenario, it is still
2.3.2.1 that were isolated in Mongolia and Tyva in 2009 unconfirmed. To confirm this hypothesis, the pathogenicity
and 2010 and originated from the A/black headed gull/ of A/common buzzard/Bulgaria/38WB/2010 in ducks and
Tyva/115/2009 and A/great crested grebe/Qinghai/1/2009 its possible transmission among them must be defined,
strains (online Technical Appendix Figure). which will require additional biologic studies.
To assess possible reassortment in the A/common To trace the possible routes of introduction of
buzzard/Bulgaria/38WB/2010 genome, we performed HPAIVs (H5N1) into Bulgaria, the veterinary authorities

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1599
RESEARCH

and ornithologists from the Bulgarian Society for the spread the virus over a long distance. Additionally, the lack
Protection of Birds organized continuous monitoring of of poultry farms within 10 km of the area where the buzzard
the bird areas along the Black Sea coast, near the Danube carcass was found may partially explain why no outbreak
River and around the Ogosta Dam and Lom River. From occurred. As part of a regular avian influenza surveillance
January 1, 2010, through April 30, 2010, a total of 812 plan, we tested 1,709 cloacal and fecal samples from mule
cloacal, fecal, and tissue samples from wild birds collected ducks that were collected monthly during January 1, 2010–
from these areas were tested for avian influenza virus; 269 April 30, 2010, from 64 farms in 5 regions of Bulgaria
samples were collected after March 15, 2010. All samples (Plovdiv, Pazardjik, Stara Zagora, Haskovo, and Dobrich).
tested were negative for HPAIV (H5N1). Five carcasses of No notifiable avian influenza viruses were isolated from
common buzzards found in different areas were submitted any sample.
to the Regional Diagnostic Lab on Avian Influenza after Since clade 2.3.2 was first isolated from a dead
March 15, 2010; only 1 was carrying HPAIV (H5N1). Chinese pond heron in Hong Kong in 2004, it has spread
Common buzzards are considered territorial birds geographically and evolved genetically. A new fourth-
that usually do not migrate long distances. A 3-year study order clade, 2.3.2.1, was recently identified, and A/
conducted in southern England showed that local radio- common buzzard/Bulgaria/38WB/2010 was classified
tagged common buzzards forage within 1 km of their in this clade (8). The question that arises now that clade
nests during their first winter; most of the birds that do not 2.3.2.1 has spread from Asia to Europe is whether it can
disperse make only brief excursions before they opt for a cause a scenario similar to that caused by clade 2.2 from
stay-at-home strategy, and most of those that disperse return 2005–2006, when HPAIV (H5N1) killed millions of birds
to their natal area during the following breeding season in Asia, Europe, and Africa. Although no new HPAIV
(37). Long-term ornithologic studies conducted during (H5N1)–related events have been reported in Europe since
1979–2005 in Bulgaria, however, showed unambiguously March 2010, some of the aspects of the 2.3.2.1 clade make
that extensive autumn and spring migrations of common it difficult to predict the consequences of the clade’s arrival
buzzards (as many as 42,100 birds) occur on the western on the continent. For example, this clade is already widely
Black Sea Via Pontica flyway (35,38). Kostadinova et al. distributed in Asia and is being perpetuated in many wild
reported ≈6 pairs of common buzzards breeding in the bird species, which is a prerequisite for long-distance
Batova habitat, but during autumn and spring migrations, distribution through migration. Wild bird species infected
ornithologists counted as many as 19,712 individuals of the with HPAIV (H5N1) from clade 2.3.2.1 include gray
species in the area (35). In most cases, common buzzards herons, peregrine falcons, and great egrets in Hong Kong;
migrate singly or in loose flocks with other raptors, lesser whooper swans, ruby shelducks, and bar-headed geese in
spotted eagles, white pelicans, or black storks (38). Mongolia; and grebes and black-headed gulls in Tyva (8).
Migration of common buzzards in different parts of The potential of clade 2.3.2.1 HPAIV (H5N1) to cause
Europe appears to depend on the local climate; buzzards an outbreak is heightened because vaccines currently in use
from northern Europe fly to the western Black Sea area do not efficiently protect poultry flocks from a strain of this
during the winter season, whereas buzzards from Bulgaria clade that was recently identified in Vietnam (40). Now that
fly south. Tracking bands belonging to common buzzards clade 2.3.2.1 has spread to Europe, implementing active
from Finland, Romania, and Israel have been found in surveillance plans in all high-risk areas and monitoring
Bulgaria (39). The buzzard infected with HPAIV (H5N1) the wild birds in the region will play key roles in early
in Bulgaria was not banded; however, even if it had been detection of incidences of HPAIV (H5N1) infection and in
a migrant, the habitat nearest to St. Konstantin and Helena prevention of outbreaks. The expansion of the geographic
Black Sea Resort that provides different food sources is distribution of HPAIV (H5N1) in wild birds and poultry
Batova, with an area of 38,132.8 ha (35). The migration of and the virus’s repeated interspecies transmission to
the common buzzards suggests that these birds are capable humans make this virus a substantial pandemic threat.
of spreading pathogens over long distances.
Our results show that chickens are highly susceptible to Acknowledgments
influenza virus A/common buzzard/Bulgaria/38WB/2010 We thank Mariette Ducatez, Pamela McKenzie, and James
(H5N1) and that the virus is highly pathogenic in them. Knowles for technical support and Cherise Guess for excellent
Mammals appear not to be susceptible. Although buzzards editorial assistance. We gratefully acknowledge the contributions
can serve as intermediate hosts of HPAIV (H5N1) between of Ivaylo Ivanov and the Bulgarian Society for the Protection of
migratory birds and poultry, the lack of gross pathologic Birds, Georgi Stoyanov and the Birds of Prey Protection Society,
findings in the buzzard carcass we examined indicates that Bulgaria, and Miglena Ivanova and the Regional Inspectorate of
the bird died shortly after infection. Thus, in this case, the Environment and Waters, Varna, Bulgaria.
buzzard could not have served as a reservoir of infection to

1600 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Influenza Virus A (H5N1) Clade 2.3.2.1, Bulgaria

This work was supported by National Institutes of Health 14. Zhou B, Donnelly ME, Scholes DT, St George K, Hatta M,
contract no. HHSN266200700005C. Kawaoka Y, et al. Single reaction genomic amplification accelerates
sequencing and vaccine production for classical and swine origin
Dr Marinova-Petkova is a research associate at the Regional human influenza A viruses. J Virol. 2009;83:10309–13. http://dx.doi.
org/10.1128/JVI.01109-09
Diagnostic Laboratory on Avian Influenza and Newcastle Disease
15. Ducatez MF, Hause B, Stigger-Rosser E, Darnell D, Corzo C, Juleen
in Birds, Varna, Bulgaria, and a doctoral student at the National K, et al. Multiple reassortment between pandemic (H1N1) 2009 and
Diagnostic and Research Veterinary Medical Institute, Sofia, endemic influenza viruses in pigs, United States. Emerg Infect Dis.
Bulgaria. Her research interests include ecology and evolution of 2011;17:1624–9. http://dx.doi.org/10.3201/eid1709.110338
16. National Center for Biotechnology Information. Influenza Virus
influenza viruses at the animal–human interface.
Sequence Database [cited 2012 Jan 5]. http://www.ncbi.nlm.nih.
gov/genomes/FLU/Database/nph-select.cgi?go=database
17. Tamura K, Dudley J, Nei M, Kumar S. MEGA4: Molecular
References Evolutionary Genetics Analysis (MEGA) software version 4.0.
Mol Biol Evol. 2007;24:1596–9. http://dx.doi.org/10.1093/molbev/
1. Webster RG, Bean WJ, Gorman OT, Chambers TM, Kawaoka
msm092
Y. Evolution and ecology of influenza A viruses. Microbiol Rev.
18. Huelsenbeck JP, Ronquist F. MRBAYES: Bayesian inference
1992;56:152–79.
of phylogeny. Bioinformatics. 2001;17:754–5. http://dx.doi.
2. Becker WB. The isolation and classification of Tern virus: influenza
org/10.1093/bioinformatics/17.8.754
virus A/tern/South Africa/1961. J Hyg (Lond). 1966;64:309–20.
19. Posada D, Crandall K. MODELTEST: testing the model of DNA
http://dx.doi.org/10.1017/S0022172400040596
substitution. Bioinformatics. 1998;14:817–8. http://dx.doi.
3. Magnino S, Fabbi M, Moreno A, Sala G, Lavazza A, Ghelfi E, et al.
org/10.1093/bioinformatics/14.9.817
Avian influenza virus (H7 serotype) in a saker falcon in Italy. Vet
20. Boon ACM, deBeauchamp J, Krauss S, Rubrum A, Webb AD,
Rec. 2000;146:740.
Webster RG, et al. Cross-reactive neutralizing antibodies directed
4. Ellis TM, Bousfield RB, Bissett LA, Dyrting KC, Luk GSM,
against pandemic H1N1 2009 virus are protective in a highly
Tsim ST, et al. Investigation of outbreaks of highly pathogenic
sensitive DBA/2 mouse influenza model. J Virol. 2010;84:7662–7.
H5N1 avian influenza in waterfowl and wild birds in Hong Kong
http://dx.doi.org/10.1128/JVI.02444-09
in late 2002. Avian Pathol. 2004;33:492–505. http://dx.doi.
21. Yen HL, Liang CH, Wu CY, Forrest HL, Ferguson A, Choy KT,
org/10.1080/03079450400003601
et al. Hemagglutinin-neuraminidase balance confers respiratory-
5. Chen H, Li Y, Li Z, Shi J, Shinya K, Deng G, et al. Properties and
droplet transmissibility of the pandemic H1N1 influenza virus in
dissemination of H5N1 viruses isolated during an influenza outbreak
ferrets. Proc Natl Acad Sci U S A. 2011;108:14264–9. http://dx.doi.
in migratory waterfowl in western China. J Virol. 2006;80:5976–83.
org/10.1073/pnas.1111000108
http://dx.doi.org/10.1128/JVI.00110-06
22. Horimoto T, Kawaoka Y. Reverse genetics provides direct evidence
6. Webster RG, Peiris M, Chen H, Guan Y. H5N1 outbreaks and
for a correlation of hemagglutinin cleavability and virulence of an
enzootic influenza. Emerg Infect Dis. 2006;12:3–8. http://dx.doi.
avian influenza A virus. J Virol. 1994;68:3120–8.
org/10.3201/eid1201.051024
23. Ha Y, Stevens DJ, Skehel JJ, Wiley DC. X-ray structures of H5 avian
7. World Health Organization/World Organisation for Animal Health/
and H9 swine influenza virus hemagglutinins bound to avian and
Food and Agriculture Organization H5N1 Evolution Working
human receptor analogs. Proc Natl Acad Sci U S A. 2001;98:11181–
Group. Toward a unified nomenclature system for highly pathogenic
6. http://dx.doi.org/10.1073/pnas.201401198
avian influenza virus (H5N1) [conference summary]. Emerg Infect
24. Matrosovich M, Zhou N, Kawaoka Y, Webster RG. The surface
Dis. 2008;14:e1.
glycoproteins of H5 influenza viruses isolated from humans,
8. World Health Organization/World Organisation for Animal Health/
chickens, and wild aquatic birds have distinguishable properties. J
Food and Agriculture Organization H5N1 Evolution Working
Virol. 1999;73:1146–55.
Group. Continued evolution of highly pathogenic avian influenza
25. Abed Y, Baz M, Boivin G. Impact of neuraminidase mutations
A (H5N1): updated nomenclature. Influenza Other Respi Viruses.
conferring influenza resistance to neuraminidase inhibitors in the N1
2012;6:1–5. http://dx.doi.org/10.1111/j.1750-2659.2011.00298.x
and N2 genetic backgrounds. Antivir Ther. 2006;11:971–6.
9. European Commission Directorate-General for Health and
26. Ilyushina NA, Seiler JP, Rehg JE, Webster RG, Govorkova EA. Effect
Consumers. Animal Disease Notification System. Annual report
of neuraminidase inhibitor–resistant mutations on pathogenicity
2010 [cited 2012 Jan 5]. http://ec.europa.eu/food/animal/diseases/
of clade 2.2 A/Turkey/15/06 (H5N1) influenza virus in ferrets.
adns/adns_report2010_en.pdf
PLoS Pathog. 2010;6:e1000933. http://dx.doi.org/10.1371/journal.
10. European Commission Directorate-General for Health and
ppat.1000933
Consumers. Animal Disease Notification System. Annual report
27. Zell R, Krumbholz A, Wutzler P. Influenza A virus PB1–F2
2006 [cited 2012 Jan 5]. http://ec.europa.eu/food/animal/diseases/
gene [letter]. Emerg Infect Dis. 2006;12:1607–9. http://dx.doi.
adns/adns_report2006_en.pdf
org/10.3201/eid1210.060511
11. World Organisation for Animal Health. Avian influenza. In: Manual
28. Conenello GM, Zamarin D, Perrone LA, Tumpey T, Palese P.
of diagnostic tests and vaccines for terrestrial animals. 2011 [cited
A single mutation in the PB1–F2 of H5N1 (HK/97) and 1918
2012 Jan 5]. http://www.oie.int/fileadmin/Home/eng/Health_
influenza A viruses contributes to increased virulence. PLoS Pathog.
standards/tahm/2.03.04_AI.pdf
2007;3:1414–21. http://dx.doi.org/10.1371/journal.ppat.0030141
12. Reed LJ, Muench H. A simple method for estimating fifty percent
29. Hay AJ, Wolstenholme AJ, Skehel JJ, Smith MH. The molecular
endpoints. Am J Hyg. 1938;27:493–7.
basis of the specific anti-influenza action of amantadine. EMBO J.
13. Hoffmann E, Stech J, Guan Y, Webster RG, Perez DR. Universal
1985;4:3021–4.
primer set for the full-length amplification of all influenza A
30. Reid AH, Fanning TG, Janczewski TA, Lourens RM, Taubenberger
viruses. Arch Virol. 2001;146:2275–89. http://dx.doi.org/10.1007/
JK. Novel origin of the 1918 pandemic influenza virus nucleoprotein
s007050170002
gene. J Virol. 2004;78:12462–70. http://dx.doi.org/10.1128/
JVI.78.22.12462-12470.2004

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1601
RESEARCH

31. Chen GW, Chang SC, Mok CK, Lo YL, Kung YN, Huang JH, et 36. Birdlife International. Mediterranean/Black Sea Flyway factsheet.
al. Genomic signatures of human versus avian influenza A viruses. 2012 [cited 2012 Jan 5]. http://www.birdlife.org/datazone/userfiles/
Emerg Infect Dis. 2006;12:1353–60. http://dx.doi.org/10.3201/ file/sowb/flyways/5_Mediterranean-_Black_Sea_Factsheet.pdf
eid1209.060276 37. Walls SS, Kenward RE. Movements of radio-tagged common
32. Pan C, Cheung B, Tan S, Li C, Li L, Liu S, et al. Genomic signature buzzards Buteo buteo in their first year. Ibis. 1995;137:177–82.
and mutation trend analysis of pandemic (H1N1) 2009 influenza A http://dx.doi.org/10.1111/j.1474-919X.1995.tb03237.x
virus. PLoS ONE. 2010;5:e9549. http://dx.doi.org/10.1371/journal. 38. Michev T, Profirov L, Nyagolov K, Dimitrov M. Br Birds.
pone.0009549 2011;104:16–37.
33. Shaw M, Cooper L, Xu X, Thompson W, Krauss S, Guan Y, et 39. Nankinov D, Dalaklieva S, Nikolov B, Djingova M. Bird ringing
al. Molecular changes associated with the transmission of avian bulletin no 11: Bulgarian report for 1987–1988. Sofia (Bulgaria):
influenza A H5N1 and H9N2 viruses to humans. J Med Virol. Bulgarian Ornithological Centre, BAS; 2008. p. 98.
2002;66:107–14. http://dx.doi.org/10.1002/jmv.2118 40. Food and Agriculture Organization. EMPRES. H5N1 HPAI global
34. Reid SM, Shell WM, Barboi G, Onita I, Turcitu M, Cioranu R, et overview July–September 2011. Issue no. 29 [cited 2012 Jan 5].
al. First reported incursion of highly pathogenic notifiable avian http://www.fao.org/docrep/015/al913e/al913e00.pdf
influenza A H5N1 viruses from clade 2.3.2 into European poultry.
Transbound Emerg Dis. 2011;58:76–8. http://dx.doi.org/10.1111/ Address for correspondence: Robert G. Webster, Department of Infectious
j.1865-1682.2010.01175.x
Diseases, MS 330, St. Jude Children’s Research Hospital, 262 Danny
35. Kostradinova I, Gramatikov M, eds. Important bird areas in Bulgaria
and Natura 2000. Bulgarian Society for the Protection of Birds, Thomas Pl, Memphis, TN 38105-3678, USA; email: robert.webster@
conservation series, book 11. Sofia (Bulgaria): The Society; 2007. p. stjude.org
1–639.

1602 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Dengue Outbreaks in High-Income
Area, Kaohsiung City, Taiwan,
2003–2009
Chia-Hsien Lin, Karin L. Schiøler, Martin R. Jepsen, Chi-Kung Ho, Shu-Hua Li,
and Flemming Konradsen

Kaohsiung City, a modern metropolis of 1.5 million annual outbreaks of variable scales, resulting in ≈6,800
persons, has been the focus of dengue virus activity in confirmed cases (3).
Taiwan for several decades. The aim of this study was Cocirculation of >2 of the 4 dengue virus serotypes
to provide a temporal and spatial description of dengue (DENV-1–4) has been reported in Kaohsiung City, and
virus epidemiology in Kaohsiung City by using data for all the molecular characteristics of the serotypes have been
laboratory-confirmed dengue cases during 2003–2009. We
well documented for several epidemics, indicating the
investigated age- and sex-dependent incidence rates and
the spatiotemporal patterns of all cases confirmed through
possible origin and transmission dynamics of the causative
passive or active surveillance. Elderly persons were at strains (2–9). The spatiotemporal patterns of disease
particularly high risk for dengue virus–related sickness transmission during the 2002 DENV-2 epidemic also have
and death. Of all confirmed cases, ≈75% were detected been investigated, and findings indicate several possible
through passive surveillance activities; case-patients mechanisms by which the virus might have dispersed after
detected through active surveillance included immediate being introduced into the population (10,11). Furthermore,
family members, neighbors, and colleagues of confirmed Lin et al. (12) examined the relationship between disease-
case-patients. Changing patterns of case clustering could related illness and death and the distribution of primary and
be due to the effect of unmeasured environmental and secondary infections for dengue virus cases reported across
demographic factors. Taiwan during 2002–2007.
We provide a detailed description of the epidemiology

D engue virus disease or dengue virus–like disease has

circulated in southern Taiwan since the late nineteenth
century (1); transmission initially occurred as intermittent
of dengue virus infection in Kaohsiung City during 2003–
2009; the description is based on routine disease and
vector surveillance data provided by the Department of
epidemics with intervals of several years to decades (2–6). Health, Kaohsiung City Government. The temporal case
However, for the past decade, dengue virus epidemics have distribution is compared with available climate data and
occurred annually in Taiwan, and the main focus of activity the index of peridomestic adult vectors, Aedes aegypti and
has been in Kaohsiung City, a modern metropolis of 1.5 Ae. albopictus mosquitoes, and case characteristics are
million persons. Kaohsiung City was the epicenter of the examined across the age and sex of patients and across
2002 dengue virus epidemic, which with 2,820 confirmed the surveillance method (active vs. passive). In addition,
cases and several hundred cases of dengue hemorrhagic the study provides an assessment of spatiotemporal case
fever (DHF), was one of the largest ever recorded in clusters, identifying possible hot spots for dengue virus
Taiwan (3). During 2002–2011, Kaohsiung City has had transmission in Kaohsiung City during the 7-year study
period.
Author affiliations: University of Copenhagen, Copenhagen,
Denmark (C.-H. Lin, K.L. Schiøler, M.R. Jepsen, F. Konradsen);
Materials and Methods
National Taiwan University, Taipei, Taiwan (C.-H. Lin); Kaohsiung
City Government, Kaohsiung City, Taiwan (C.-K. Ho, S.-H. Li);
Study Area
Kaohsiung Medical University, Kaohsiung City (C.-K. Ho); and
Kaohsiung City, located on the southwestern coast
National Sun Yat-Sen University, Kaohsiung City (S.-H. Li)
of Taiwan (22°38′N, 120°16′E), has a tropical monsoon
DOI: http://dx.doi.org/10.3201/eid1810.111929 climate (13) (Figure 1). During the study period, January

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1603
RESEARCH

Prevention Act in 1999 by the Taiwan Center for Disease

Control (Taiwan CDC). The system ensured reporting
and laboratory confirmation of all suspected cases of
dengue virus infection identified through passive or active
surveillance activities, using the probable case definition
of the World Health Organization (WHO) (16). Passive
surveillance involved mandatory reporting of probable
dengue virus infection cases to Taiwan CDC within 24
hours after a patient sought medical assistance at any of the
city’s 1,747 health facilities, including public health centers,
general practitioner clinics, and public and private hospitals.
Passive surveillance activities also involved school-based
reporting of febrile students and self-reporting to health
authorities by citizens with probable dengue virus infection
(3). Active surveillance included fever checkpoints at the
airport and screening by the district public health nurse
of immediate contacts (e.g., family members, colleagues,
and neighbors) of persons with confirmed dengue virus
Figure 1. Kaohsiung City, Taiwan (22°38′N, 120°16′E), indicating infection (3,17). Basic patient information was collected
the 11 districts and 463 administrative units (Li) of the city and the and blood samples were obtained from all suspected case-
main entry points for international travel and commerce. Insets patients and sent to the Fifth Branch Office of Taiwan CDC
show location of Taiwan in Southeast Asia (box) and of Kaohsiung in Kaohsiung City.
City in Taiwan (gray shading).
A patient was confirmed to have dengue virus infection
if 1) dengue virus RNA was detected in a serum sample
2003–December 2009, the highest annual temperatures by real-time reverse transcription PCR, 2) dengue virus–
were during June–August (monthly mean temperature specific antibodies were detected in single serum samples
range 28.7°C–30.5°C), and the lowest temperatures were by IgM- or IgG-capture ELISA, or 3) a >4-fold increase
recorded during January and February (mean temperature in IgG ELISA titers was detected in paired acute- and
18.4°C and 20.4°C, respectively). The highest monthly convalescent-phase serum samples (3). In August 2008,
accumulative precipitations occurred during June–August detection of dengue virus nonstructural protein 1 by use of
(range 901.5–1,229.3 mm), whereas there was almost no a rapid diagnostic test (Platelia Dengue NS1 Ag assay; Bio-
precipitation during November–February (14) (Figure 2). Rad, Marnes la Coquette, France) was included as a fourth
Kaohsiung City is a major port and industrial diagnostic method (3). DHF, including dengue shock
metropolis and the most densely populated urban center syndrome (DSS), was distinguished from dengue fever
in Taiwan (1.5 million persons within a total area of 150 by the presence of 1) hemorrhagic tendencies, including
km2). The city is divided into 11 districts; a Li is the a positive tourniquet test result and bleeding from the
smallest administrative unit within these districts. The
overall number of Lis decreased from 463 to 459 in 2006.
On average, the population density of a Li is 150–65,000
persons/km2 (Figure 1). Piped water is available for 99%
of the city households, and household waste is removed
daily throughout the city by the Environmental Protection
Bureau, Kaohsiung City Government. The city has an
international airport, which is a major access point for
tourists and foreign workers, many of whom are employed
in the commercial port and the industrial zones of the city.
Most of the 15,000 foreign workers who arrive in the city
each year are citizens from neighboring countries, such as
the Philippines, Indonesia, Vietnam, and Thailand (15).
Figure 2. Monthly average temperature, rainfall, and adult index
Dengue Surveillance and Laboratory Diagnosis (AI) for Aedes aegypti and Ae. albopictus mosquitoes, Kaohsiung
The disease surveillance system in use during the study City, Taiwan, 2003–2009. AI was calculated as number of adult
period was introduced under the Communicable Disease female mosquitoes captured per number of inspected premises.

1604 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Dengue Outbreaks in High-Income Area

mucosa, gastrointestinal tract (hematemesis, hematuria, or autocorrelation by using global statistics and actual cluster
melena) or other locations; 2) thrombocytopenia (<100,000 location by using the local indicator of spatial association
cells/mm3); or 3) plasma leakage (3). The severity of DHF (LISA). Anselin’s LISA provided the local version of
was not further classified. All laboratory tests and most Moran’s I, used here to compare mean incidence rates for
of the incurred medical expenses were covered by the each Li and its neighboring Lis (19). The mapped LISA
National Health Insurance. results indicated how spatial autocorrelation varied over
the study region according to 5 categories: 1) hot spot,
Patient Data denoting a high-incidence Li surrounded by high-incidence
During the study period, January 2003–December Lis; 2) high-value outlier, denoting a high-incidence Li
2009, patient data for all laboratory-confirmed cases were surrounded by low-incidence Lis; 3) low-value outlier,
provided by the Department of Health, Kaohsiung City denoting a low-incidence Li surrounded by high-incidence
Government. The data included the registered home address, Lis; 4) cold spot, denoting a low-incidence Li surrounded
sex, date of birth, date of manifestation onset, surveillance by low-incidence Lis; 5) not significant, denoting no spatial
methods (active or passive), and reported clinical autocorrelation presented.
manifestations (fever, anorexia, headache, arthralgia, rash, All epidemiologic and temporal analyses were
myalgia, thirst, diarrhea, nausea, pruritus, vomiting, retro- performed by using Excel 2002 (Microsoft, Redmond, WA,
orbital pain, and hemorrhagic manifestations). USA) and R-2.7.2 for Windows (http://cran.r-project.org/
bin/windows/base/old/2.7.2/). Spatial analyses were done
Vector Index by using ArcGIS 9.2 (ESRI, Redlands, California, USA).
Vector surveillance activities by the Department of
Health, Kaohsiung City Government, were initiated in 2005 Results
by using specially trained personnel. The Li was used as the During January 2003–December 2009, Taiwan CDC
surveying unit in which 50–100 households were randomly recorded 2,087 laboratory-confirmed cases of dengue virus
selected for inspection of Ae. aegypti and Ae. albopictus infection in Kaohsiung City. The cases were detected by
mosquito infestation (3). Adult Aedes mosquitoes were passive and active surveillance activities. Of the confirmed
captured indoors and outdoors with hand-nets at 8:30–11:30 cases, 98.7% (2,060) were classified as dengue fever and
AM or 1:30–4:30 PM (3). Capture activities were completed 1.3% (27) as DHF/DSS. The 7-year fatality rate for patients
for all rooms, including the basement, within a maximum with DHF/DSS was 25.9% (7/27).
of 10 minutes for each inspected premise. The adult index
was calculated as the number of adult female mosquitoes Temporal Case Distribution
captured divided by number of inspected premises. Most (96.9%) of the confirmed cases of dengue virus
infection were recorded during epidemics occurring during
Epidemiologic Analysis July–December of each year. The interannual variations
Incidence rates and clinical manifestations were in outbreak scale were considerable, ranging from 45
calculated for age-specific groups and sex by using the year- confirmed patients in 2004 to 766 in 2006. A dominant
end population data for each study year as the denominator. serotype was evident during each epidemic, representing
The z test was applied for incidence rate comparison. The >80.0% of cases confirmed by virus detection (real-time
2-sample t test was used for the comparison of the average reverse transcription PCR) in a given year (Figure 3).
number of clinical manifestations in patients detected The annual onset of epidemic activity generally
in the passive versus the active surveillance system. The coincided with the peak in monsoon rainfall and temperature
threshold of statistical significance was 0.05. levels (Figures 2 and 3). The epidemic peaked within 1–3
months after the onset of the epidemic, and all activity
Spatial Analysis ceased at the end of the monsoon season. Vector data for
Spatial patterns of dengue incidence in each Li 2005–2009 showed that the peak of the adult mosquito
were assessed by use of global and local indices. The population followed the peak of the monsoon rainfall, with
global spatial pattern was measured by using Moran’s I, a lag period of 1–2 months that corresponded to disease
an index of spatial autocorrelation coefficient, yielding activity (Figures 2 and 3).
only 1 summary statistic for the whole study area. The
theoretical range of Moran’s I was from −1 to 1; the Characteristics of Case-Patients
value around 0 provided the indication of spatial random
distribution. Higher positive values implied a stronger Age
clustering pattern, and lower negative values represented The median age of patients with confirmed dengue
a stronger dispersion tendency (18). We determined partial virus infection was 46 years (range 4 months–95 years).

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1605
RESEARCH

The average number of reported disease manifestations

was significantly higher among cases detected through
passive than through active surveillance (n = 5.1 vs. 3.7;
p<0.001), and each clinical manifestation was reported
more frequently through the passive than the active system,
with the exception of pruritus (Table 2). In both surveillance
systems, the number of reported manifestations was lowest
for persons 0–4, 65–74, and >74 years of age: average of
2.2, 1.9, and 1.2 manifestations, respectively, when detected
through active surveillance (Table 3). These numbers of
reported clinical manifestations are lower than the number
required to meet the WHO criterion for probable dengue
Figure 3. Epidemic curve of confirmed cases of dengue virus virus infection.
(DENV) infection (N = 2,087), by week of onset, Kaohsiung City,
Taiwan, 2003–2009. Predominant serotypes (DENV-1–3) and Spatial Analysis
numbers of confirmed cases are shown.
The global level of spatial autocorrelation for the
dengue virus infection incidence rates across the Lis of
Kaohsiung City was significant for each epidemic (range
The average age-specific incidence rate was lowest 0.03–0.14, Moran’s I; p<0.001) (Table 4), indicating
among persons <5 years of age (4.5/100,000 persons) and a significant positive spatial autocorrelation within the
increased steadily by age group (12.0, 14.3, 15.3, 17.9, city for each epidemic year. The type and area of local
and 25.3/100,000 persons for case-patients 5–14, 15–24, clustering, as determined by LISA, were identified for each
25–34, 35–44, and 45–54 years of age, respectively) until epidemic. In general, hot-spot Lis with a high incidence of
peaking in persons 55–64 years of age (37.9/100,000 infection did not overlap for consecutive years; however,
persons) (Table 1). The pattern of observed age-specific certain hot spots recurred or were adjacent to other hot spots
incidence rates was uniform across all epidemics during for the epidemics of 2004, 2006, 2008, and 2009. These hot
the study period (data not shown). spots overlapped with clusters of high residential density.
Among 27 patients with confirmed cases of DHF, Some dengue hot spots were also observed for areas of
fatality was highest among those >74 years of age (3/5, low population density throughout the study period, except
60.0%) followed by those 65–74 years of age (3/9, 33.3%) for 2004 and 2007. Half of the high-value outliers were
and 55–64 years of age (1/5, 20.0%). No fatalities occurred observed in low population clusters (Figure 5).
among other age groups (Table 1).
Discussion
Sex Evidence from Kaohsiung City shows that dengue
For both sexes, persons 55–64 and 65–74 years of virus infection occurs in persons of all ages; however,
age had the highest and second highest incidence rates of the incidence of infection increases notably in persons of
confirmed dengue virus infection (Table 1). Overall, the increasing age, and elderly persons are at especially high
incidence rate for the female population was slightly, but risk for DHF/DSS and death. Most case-patients detected
not significantly, higher than that for the male population by the active surveillance system were contacts of case-
(19.9 vs. 19.4/ 100,000 persons; p = 0.624). Among patients detected by the passive surveillance system.
persons 15–24 years of age, the incidence rate for the Changes in disease hot spots were noted between successive
male population was significantly higher than that for the years of the study.
female population (17.1 vs. 11.1/100,000 persons; p = The findings in this study support previous reports
0.002). of a clear correlation between precipitation, temperature,
and the occurrence of dengue virus epidemics in Taiwan
Case Detection by Active and Passive Surveillance (20,21). For Kaohsiung City, the study shows that the
The active surveillance system detected 538 (25.8%) annual onset of dengue virus epidemics coincides with
of the confirmed cases during 2003–2009. Of these cases, the time of peak rainfall and temperature. Dengue virus
520 (96.7%) were in household members, neighbors, or outbreaks coincide with the seasonal increase of adult
colleagues of confirmed case-patients, and 18 were detected vectors in the peridomestic environment, but there seems
by the fever-screening system at the airport. The highest to be no correlation between the level of vector abundance
proportions of cases detected through active surveillance and the scale of outbreaks. The same can be concluded
were among children 0–4 and 5–14 years of age (Figure 4). for the meteorologic variables assessed in this study. This

1606 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Dengue Outbreaks in High-Income Area

Table 1. Age- and sex-specific incidence rates of confirmed cases of dengue virus infection, Kaohsiung City, Taiwan, 2003–2009*
Incidence rate/100,000
Age group, y, sex No. cases persons p value† No. DHF cases No. fatal cases Fatality rate, %
0–4
M 14 5.7 0.184 2 0 0
F 7 3.1 0 0 0
All 21 4.5 2 0 0
5–14
M 90 12.7 0.497 0 0 0
F 74 11.3 1 0 0
All 164 12.0 1 0 0
15–24
M 134 17.1 0.002‡ 2 0 0
F 82 11.1 2 0 0
All 216 14.3 4 0 0
25–34
M 144 15.9 0.582 0 0 0
F 135 14.7 0 0 0
All 279 15.3 0 0 0
35–44
M 157 18.2 0.968 1 0 0
F 161 17.7 0 0 0
All 318 17.9 1 0 0
45–54
M 189 23.4 0.097 0 0 0
F 234 27.1 0 0 0
All 423 25.3 0 0 0
55–64
M 167 35.0 0.142 1 0 0
F 210 41.1 4 1 25.0
All 377 37.9 5 1 20.0
65–74
M 96 35.0 0.352 5 2 40.0
F 118 39.4 4 1 25.0
All 214 37.3 9 3 33.3
>74
M 44 21.9 0.418 5 3 60.0
F 31 18.1 0 0 0
All 75 20.2 5 3 60.0
Total
M 1,035 19.9 0.624 16 5 31.3
F 1,052 19.4 11 2 18.2
All 2,087 19.6 27 7 25.9
*DHF, dengue hemorrhagic fever.
†Difference in incidence between male and female population for the age group given.
‡Level of significance p<0.05.

observation is in line with the general understanding that those with which substantial immigration, tourism, and trade
the extent of dengue virus epidemics may be influenced by relations are maintained (4,7,8).
a variety of factors, including the level of herd immunity to The age distribution of confirmed dengue case-patients
the circulating serotype(s); the virulence of the circulating in Kaohsiung City was consistent throughout the study
strain(s); and the effect of human-vector contact exerted period. Children <5 years of age had the lowest disease
by human behavior, specific climatic phenomena, and incidence, and persons 55–64 and 65–74 years of age had
prevention and control operations. the highest incidence of confirmed cases, including cases of
All 4 dengue virus serotypes were identified in DHF and dengue virus–related deaths.
Kaohsiung City, and at least 2 serotypes cocirculated during The low incidence of dengue virus infection among
each outbreak of the study period. DENV-1 and DENV-3 the youngest age group fits well with descriptions of mild
were clearly predominant during 3 outbreaks each; DENV-2 or mainly asymptomatic dengue virus infection in younger
detection was limited, and DENV-4 detection was negligible. children (22). However, it is often suggested that the lack
Specific information on the circulating genotypes or strains of vocal ability among small children plays a factor in the
was not available for assessment. However, recent studies health-seeking behavior of their caretakers. In either case,
suggest that most dengue virus outbreaks in Taiwan can be one would expect a relatively larger group of 0- to 4-year-
attributed to the importation of novel dengue virus strains old children than older persons to be identified through the
from neighboring Southeast Asian countries, in particular active surveillance system. In fact, the 0- to 4-year-old age

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1607
RESEARCH

The case-fatality rate of 26% for older age groups in

this study is far greater than the expected average of <1%
reported by WHO (23). However, our findings correspond
with those in previous reports from Taiwan (12,24–26),
where underlying chronic diseases more commonly
observed among older persons (e.g., hypertension, chronic
renal insufficiency [uremia], or diabetes mellitus) have
been identified as possible risk factors for severe and fatal
DHF (27,28). The association between age, the presence
of underlying disease, and severe dengue virus infection
and related death has also been reported from Cuba (29,30).
However, these findings have been disputed by findings
from Singapore, where elderly patients did not exhibit more
signs or symptoms of dengue virus infection or have higher
death rates despite having a greater incidence of underlying
diseases (31). Information about the presence of underlying
diseases was not available for our study.
Overall in Kaohsiung City, the incidence of dengue
Figure 4. Age-specific distribution of case-patients with confirmed
dengue virus infection (N = 2,087) detected by passive and active virus infection in the female population was slightly
surveillance systems, Kaohsiung City, Taiwan, 2003–2009. Cases higher than that in the male population. This finding was
detected through passive surveillance were suspected dengue not statistically significant, but it was in agreement with
virus infections reported by health care facilities to Taiwan Center findings from previous studies conducted in Taiwan and
for Disease Control; cases detected through active surveillance
contrary to findings from several other Asian countries
were reported from airport screenings and by community contacts
of case-patients. where dengue is reported more frequently among the
male population (32–35). These data suggest that the risk
for exposure to dengue virus in Kaohsiung City is shared
between sexes. It may also suggest that the combination
group had the second highest proportion of cases detected of passive and active surveillance activities eliminates
through active surveillance (Figure 4); however, the potential differences in health care–seeking behavior or
overall detection rate remained far below that for all other health care access, as has been suggested for other Asian
age groups, suggesting generally lower rates of dengue countries with lower reported rates of cases among the
virus exposure. This finding could be attributed to specific female population (34).
behavioral aspects of urban living in a high-income setting, More than 25% of the confirmed dengue virus cases
such as Kaohsiung City, where, young children spend most were detected through the active surveillance system.
of their day in enclosed air-conditioned environments at Most (96.7%) of these cases involved household members,
their home, day care center, or preschool. Hence, vector neighbors, or colleagues of case-patients detected through
exposure may be substantially lower for young Taiwanese the passive surveillance system. The 0- to 4-year-old and
children than for most of their peers in other Southeast 5- to 14-year-old age groups had the highest proportion of
Asian countries. cases (33.3% and 44.2%, respectively) detected through

Table 2. Comparison of reported clinical manifestations in persons with confirmed dengue virus infection (N = 2,087) detected through
passive or active surveillance, Kaohsiung City, Taiwan, 2003–2009
Manifestation No. passive (%), n = 1,549 No. active (%), n = 538 p value*
Fever 1,509 (97.4) 392 (72.9) <0.001
Anorexia 916 (59.1) 218 (40.5) <0.001
Headache 829 (53.5) 224 (41.6) <0.001
Arthralgia 765 (49.4) 166 (30.9) <0.001
Rash 760 (49.1) 209 (38.8) <0.001
Myalgia 754 (48.7) 175 (32.5) <0.001
Thirst 694 (44.8) 168 (31.2) <0.001
Diarrhea 486 (31.4) 121 (22.5) <0.001
Nausea 447(28.9) 95 (17.7) <0.001
Pruritus 324 (20.9) 120 (22.3) 0.4941
Vomiting 296 (19.1) 53 (9.9) <0.001
Retro-orbital pain 182 (11.7) 43 (8.0) <0.001
Hemorrhagic manifestations 96 (6.2) 11 (2.0) <0.001
*Proportional difference in reporting of specific manifestation by passive and active surveillance.

1608 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Dengue Outbreaks in High-Income Area

Table 3. Age-specific frequencies of reported clinical manifestations and average reported number of manifestations for confirmed
dengue cases (N = 2,087) detected through passive or active surveillance, Kaohsiung City, Taiwan, 2003–2009
Age group, y
Variable 0–4 5–14 15–24 25–34 35–44 45–54 55–64 65–74 >74
Passive surveillance system
No. cases 14 92 161 220 226 310 297 172 57
Average no. manifestations 3.4 5.0 5.7 5.9 5.5 5.1 4.9 4.5 3.6
Manifestation, no.
Fever 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.9 0.9
Anorexia 0.4 0.7 0.6 0.6 0.6 0.6 0.6 0.6 0.6
Headache 0.1 0.5 0.7 0.6 0.6 0.6 0.5 0.4 0.3
Arthralgia 0.0 0.2 0.5 0.6 0.6 0.5 0.5 0.5 0.3
Rash 0.6 0.7 0.6 0.6 0.5 0.4 0.4 0.3 0.3
Myalgia 0.1 0.3 0.5 0.6 0.6 0.5 0.5 0.4 0.3
Thirst 0.3 0.3 0.4 0.5 0.5 0.5 0.5 0.4 0.2
Diarrhea 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3
Nausea 0.0 0.3 0.4 0.4 0.3 0.3 0.2 0.3 0.2
Pruritus 0.1 0.3 0.3 0.3 0.3 0.2 0.2 0.1 0.0
Vomiting 0.4 0.3 0.2 0.2 0.2 0.2 0.2 0.2 0.2
Retro-orbital pain 0.0 0.1 0.2 0.2 0.2 0.1 0.1 0.0 0.0
Hemorrhagic manifestations 0.0 0.0 0.1 0.1 0.0 0.1 0.1 0.1 0.1
Active surveillance system
No. cases 7 72 55 59 92 113 80 42 18
Average no. manifestations 2.2 3.5 4.4 4.6 4.6 3.8 3.1 1.9 1.2
Manifestation, no.
Fever 1.0 0.8 0.9 0.9 0.8 0.8 0.6 0.4 0.4
Anorexia 0.1 0.3 0.5 0.4 0.5 0.4 0.4 0.3 0.2
Headache 0.0 0.5 0.5 0.6 0.5 0.4 0.3 0.2 0.3
Arthralgia 0.0 0.2 0.2 0.4 0.5 0.4 0.3 0.2 0.2
Rash 0.4 0.5 0.5 0.5 0.4 0.4 0.3 0.1 0.0
Myalgia 0.0 0.2 0.3 0.5 0.5 0.4 0.3 0.1 0.1
Thirst 0.1 0.2 0.3 0.4 0.4 0.4 0.3 0.2 0.1
Diarrhea 0.3 0.2 0.2 0.3 0.3 0.2 0.2 0.1 0.1
Nausea 0.0 0.1 0.3 0.2 0.2 0.2 0.2 0.1 0.0
Pruritus 0.1 0.3 0.4 0.3 0.3 0.2 0.2 0.1 0.0
Vomiting 0.0 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.0
Retro-orbital pain 0.0 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0
Hemorrhagic manifestations 0.0 0.0 0.0 0.0 0.0 0.0 2.6 0.0 0.0

active surveillance, a finding in line with the lower number need to be given to the potential influence of underlying
of reported symptoms (milder disease) and the possibly disease in treatment for severe dengue virus infection.
greater likelihood of being at home when the active The incidence rates of dengue virus cases occurred
surveillance nurse visited. nonrandomly throughout Kaohsiung City, implying that
The inherent underreporting of dengue virus infections risk factors for dengue virus infection were spatially
by passive surveillance, caused by mild and asymptomatic heterogeneous (Table 4). Hot-spot Lis were detected in
infections, is counterbalanced by the addition of the active different locations during consecutive years (Figure 5),
surveillance component, although the level of impact although some hot spots recurred or were adjacent to other
remains unknown. High retrieval rates for convalescent- hot spots for the epidemics of 2004, 2006, 2008, and 2009.
phase 2 and 3 samples (obtained for 455 [21.8%] of the These hot spots were all shown to overlap with areas of high
2,087 cases) ensure that all reported cases were thoroughly human residential density. Other hot spots and high-value
analyzed by molecular and serologic testing. However, the outliers were detected in areas of low and high population
sensitivity and specificity of the surveillance and laboratory density.
diagnostic systems and the overall cost-effectiveness of the
Table 4. Spatial autocorrelation of dengue incidence rates in
surveillance and control program in Kaohsiung City must Kaohsiung City, Taiwan, 2003–2009
be assessed (36). Year No. cases Moran’s I* p value†
Elderly case-patients in both surveillance systems 2003 62 0.09 <0.001
2004 45 0.08 <0.001
reported fewer symptoms, indicating the urgent need for 2005 99 0.04 <0.001
improved diagnosis and treatment of severe dengue virus 2006 766 0.14 <0.001
infection in this high-risk population. However, improved 2007 153 0.03 <0.001
diagnosis and treatment would require better detection of 2008 326 0.10 <0.001
2009 636 0.09 <0.001
cases that do not fit the currently used criteria for probable *An index of spatial autocorrelation coefficient.
dengue virus infection. In addition, consideration would †p<0.05 indicated that the spatial pattern was nonrandom.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1609
RESEARCH

Figure 5. Local indicator of spatial

association (LISA) cluster maps of
incidence rates for dengue virus
infection during each epidemic period,
Kaohsiung City, Taiwan, 2003–2009.
High-value outlier, high-incidence Li
(smallest administrative unit within
each of 11 districts in Kaohsiung City)
surrounded by low-incidence Lis; not
significant, 0 spatial autocorrelation
presented; Hot spot, high-incidence
Li surrounded by high-incidence Lis.
Hot-spot Lis circled with dashed lines
are those that overlap with clusters
of high residential density; hot-spot
or high-value outlier Lis circled with
solid lines are those that overlap with
clusters of low residential density.

The ambiguous correlation between the incidence of Acknowledgments

dengue virus infections and population density was also We thank the Department of Health, Kaohsiung City
reported by Lin and Wen for the 2002 DENV-2 epidemic Government, for providing all patient and vector data for this
in Kaohsiung City (37), indicating that variations in study.
population density are insufficient for explaining spatial
Ms Lin is a research assistant in the Department of
variations in dengue virus outbreak intensity at the local
Geography at National Taiwan University and in the International
level. To understand the observed variations and to
Health Unit, Department of International Health, Immunology,
predict local occurrence of future outbreaks, it is therefore
and Microbiology at the University of Copenhagen. Her research
necessary to account for additional spatial factors of
interests include spatiotemporal epidemiology of infectious
potential importance.
diseases, especially dengue virus disease.
Our findings on dengue virus outbreaks and high-risk
population groups suggest the need for further research
on demographic parameters, such as age distribution and References
age-dependent behavior at the local level. Variations in
1. Akagi K, Kojima D. Experimental studies for the pathogen of dengue
socioeconomic status, housing standards, and housing fever [in Japanese]. J Formos Med Assoc. 1915;158:1049–78.
density should also be investigated for a potential role 2. Wu YC. Epidemic dengue 2 on Liouchyou Shiang, Pingtung County,
in disease clustering (38–40). In addition, the distinct in 1981 [in Chinese]. Zhonghua Min Guo Wei Sheng Wu Ji Mian Yi
seasonal transmission pattern of dengue virus shown in this Xue Za Zhi. 1986;19:203–11.
3. Taiwan Center for Disease Control. Guidelines for dengue control
study suggests further research into local environmental [in Chinese]. Taipei, Taiwan: the Center; 2009.
factors and control activities that influence vector survival 4. King CC, Wu YC, Chao DY, Lin TH, Lin C, Wang HT, et al. Major
and availability and productivity of vector breeding sites. epidemics of dengue in Taiwan in 1981–2000: related to intensive
Refined spatial analysis combining data for all spatial virus activities in Asia [cited 2011 Nov 30]. http://www.searo.who.
int/EN/Section10/Section332/Section522_2501.htm
factors of identified importance could help identify 5. Harn MR, Chiang YL, Tian KJ, Chang YH, Ko YC. The 1991
imminent hot spots and guide improved prevention and dengue epidemic in Kaohsiung City [in Chinese]. J Formos Med
control efforts. Assoc. 1993;92(Suppl 1):S39–43.

1610 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Dengue Outbreaks in High-Income Area

6. Chen KT, Chou MY, Ma LS. The primary investigation report 26. Chao DY, Lin TH, Hwang KP, Huang JH, Liu CC, King CC. 1998
of dengue outbreak in southern Taiwan [in Chinese]. Taiwan dengue hemorrhagic fever epidemic in Taiwan. Emerg Infect Dis.
Epidemiology Bulletin 1987;93–99. 2004;10:552–4. http://dx.doi.org/10.3201/eid1003.020518
7. Huang JH, Liao TL, Chang SF, Su CL, Chien LJ, Kuo YC, et al. 27. Lee MS, Hwang KP, Chen TC, Lu PL, Chen TP. Clinical
Laboratory-based dengue surveillance in Taiwan, 2005: a molecular characteristics of dengue and dengue hemorrhagic fever in a medical
epidemiologic study. Am J Trop Med Hyg. 2007;77:903–9. center of southern Taiwan during the 2002 epidemic. J Microbiol
8. Shu PY, Su CL, Liao TL, Yang CF, Chang SF, Lin CC, et al. Immunol Infect. 2006;39:121–9.
Molecular characterization of dengue viruses imported into Taiwan 28. Liu CC, Huang KJ, Huang MC, Lin JJ, Wang SM, Liu JJ, et al. High
during 2003–2007: geographic distribution and genotype shift. Am J case-fatality rate of adults with dengue hemorrhagic fever during an
Trop Med Hyg. 2009;80:1039–46. outbreak in non-endemic Taiwan: risk factors for dengue-infected
9. Chen HL, Lin SR, Liu HF, King CC, Hsieh SC, Wang WK. Evolution elders. Am J Infect Dis. 2008;4:10–7. http://dx.doi.org/10.3844/
of dengue virus type 2 during two consecutive outbreaks with an ajidsp.2008.10.17
increase in severity in southern Taiwan in 2001–2002. Am J Trop 29. Guzmán MG, Alvarez M, Rodriguez R, Rosario D, Vázquez S,
Med Hyg. 2008;79:495–505. Valdés L, et al. Fatal dengue hemorrhagic fever in Cuba, 1997.
10. Kan CC, Lee PF, Wen TH, Chao DY, Wu MH, Lin NH, et al. Two Int J Infect Dis. 1999;3:130–5. http://dx.doi.org/10.1016/S1201-
clustering diffusion patterns identified from the 2001–2003 dengue 9712(99)90033-4
epidemic, Kaohsiung, Taiwan. Am J Trop Med Hyg. 2008;79:344– 30. Bravo JR, Guzmán MG, Kouri GP. Why dengue haemorrhagic fever
52. in Cuba? I. Individual risk factors for dengue haemorrhagic fever/
11. Wen TH, Lin NH, Chao DY, Hwang KP, Kan CC, Lin KCM, et dengue shock syndrome (DHF/DSS). Trans R Soc Trop Med Hyg.
al. Spatial-temporal patterns of dengue in areas at risk of dengue 1987;81:816–20. http://dx.doi.org/10.1016/0035-9203(87)90041-1
hemorrhagic fever in Kaohsiung, Taiwan, 2002. Int J Infect Dis. 31. Lye DC, Lee VJ, Sun Y, Leo YS. The benign nature of acute dengue
2010;14:e334–43. http://dx.doi.org/10.1016/j.ijid.2009.06.006 infection in hospitalized older adults in Singapore. Int J Infect Dis.
12. Lin CC, Huang YH, Shu PY, Wu HS, Lin YS, Yeh TM, et al. 2010;14:e410–3. http://dx.doi.org/10.1016/j.ijid.2009.06.026
Characteristic of dengue disease in Taiwan: 2002–2007. Am 32. Ben RJ, Lo CH, Yu HL, Hsueh CW, Feng NH, Chang LC, et al.
J Trop Med Hyg. 2010;82:731–9. http://dx.doi.org/10.4269/ The features of clinical manifestations of dengue fever in Kaohsiung
ajtmh.2010.09-0549 City, Taiwan, 2006. Int J Infect Dis. 2008;12(Suppl 1):e93–4. http://
13. Chen TC, Wang SY, Huang WR, Yen MC. Variation of the East dx.doi.org/10.1016/j.ijid.2008.05.234
Asian summer monsoon rainfall. J Climate. 2004;17:744–62. http:// 33. Hussin N, Jaafar J, Naing NN, Mat HA, Muhamad AH, Mamat
dx.doi.org/10.1175/1520-0442(2004)017<0744:VOTEAS>2.0. MN. A review of dengue fever incidence in Kota Bharu, Kelantan,
CO;2 Malaysia during the years 1998–2003. Southeast Asian J Trop Med
14. Central Weather Bureau. Climate statistics [cited 2011 Nov 11]. Public Health. 2005;36:1179–86.
http://www.cwb.gov.tw/V7e/climate/dailyPrecipitation/dP.htm 34. Guha-Sapir D, Schimmer B. Dengue fever: new paradigms for a
15. Department of Statistics, Ministry of the Interior. The statistics of changing epidemiology. Emerg Themes Epidemiol. 2005;2:1. http://
foreigners in Taiwan 2009 Jan 31 [in Chinese] [cited 2011 Nov 30]. dx.doi.org/10.1186/1742-7622-2-1
http://sowf.moi.gov.tw/stat/week/week9805.doc 35. Shekhar KC, Huat OL. Epidemiology of dengue/dengue hemorrhagic
16. World Health Organization. Dengue haemorrhagic fever: diagnosis, fever in Malaysia—a retrospective epidemiological study. 1973–
treatment, prevention and control. 2nd edition. Geneva: the 1987. Part II: dengue fever (DF). Asia Pac J Public Health. 1992–
Organization; 1997. 3;6:126–33. http://dx.doi.org/10.1177/101053959200600302
17. Shu PY, Chien LJ, Chang SF, Su CL, Kuo YC, Liao TL, et al. 36. Runge-Ranzinger S, Horstick O, Marx M, Kroeger A. What does
Fever screening at airports and imported dengue. Emerg Infect Dis. dengue disease surveillance contribute to predicting and detecting
2005;11:460–2. http://dx.doi.org/10.3201/eid1103.040420 outbreaks and describing trends? Trop Med Int Health. 2008;13:1022–
18. Moran PAP. Notes on continuous stochastic phenomena. Biometrika. 41. http://dx.doi.org/10.1111/j.1365-3156.2008.02112.x
1950;37:17–23. 37. Lin CH, Wen TH. Using geographically weighted regression (GWR)
19. Anselin L. Local indicators of spatial association—LISA. Geogr to explore spatial varying relationships of immature mosquitoes
Anal. 1995;27:93–115. http://dx.doi.org/10.1111/j.1538-4632.1995. and human densities with the incidence of dengue. Int J Environ
tb00338.x Res Public Health. 2011;8:2798–815. http://dx.doi.org/10.3390/
20. Hsieh YH, Chen CWS. Turning points, reproduction number, and ijerph8072798
impact of climatological events for multi-wave dengue outbreaks. 38. Pai HH, Lu YL, Hong YJ, Hsu EL. The differences of dengue vectors
Trop Med Int Health. 2009;14:628–38. http://dx.doi.org/10.1111/ and human behavior between families with and without members
j.1365-3156.2009.02277.x having dengue fever/dengue hemorrhagic fever. Int J Environ Health
21. Wu PC, Lay JG, Guo HR, Lin CY, Lung SC, Su HJ. Higher Res. 2005;15:263–9. http://dx.doi.org/10.1080/09603120500155732
temperature and urbanization affect the spatial patterns of dengue 39. de Mattos Almeida MC, Caiaffa WT, Assunção RM, Proietti FA.
fever transmission in subtropical Taiwan. Sci Total Environ. Spatial vulnerability to dengue in a Brazilian urban area during a
2009;407:2224–33. http://dx.doi.org/10.1016/j.scitotenv.2008.11. 7-year surveillance. J Urban Health. 2007;84:334–45. http://dx.doi.
034 org/10.1007/s11524-006-9154-2
22. Burke DS, Nisalak A, Johnson DE, Scott RM. A prospective-study of 40. Mondini A, Chiaravalloti-Neto F. Spatial correlation of incidence
dengue infections in Bangkok. Am J Trop Med Hyg. 1988;38:172– of dengue with socioeconomic, demographic and environmental
80. variables in a Brazilian city. Sci Total Environ. 2008;393:241–8.
23. World Health Organization. Dengue: guidelines for diagnosis, http://dx.doi.org/10.1016/j.scitotenv.2008.01.010
treatment, prevention and control. Geneva: the Organization; 2009.
24. Harn MR. Clinical study on dengue fever during 1987–1988 Address for correspondence: Chia-Hsien Lin, University of Copenhagen,
epidemic at Kaohsiung City, southern Taiwan [in Chinese].
National Taiwan University, 14F-1, No. 535, Zengzih Rd. Kaohsiung City,
Gaoxiong Yi Xue Ke Xue Za Zhi. 1989;5:58–65.
25. Lai PC, Lee SSJ, Kao CH, Chen YS, Huang CK, Lin WR, et al. Taiwan; email: chlin1983@gmail.com
Characteristics of a dengue hemorrhagic fever outbreak in 2001 in
Kaohsiung. J Microbiol Immunol Infect. 2004;37:266–70.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1611
RESEARCH

Nontuberculous Mycobacteria in
Household Plumbing as Possible
Cause of Chronic Rhinosinusitis
Wellington S. Tichenor,1 Jennifer Thurlow,1 Steven McNulty, Barbara A. Brown-Elliott,
Richard J. Wallace, Jr., and Joseph O. Falkinham III

Symptoms of chronic rhinosinusitis (CRS) often persist persistence of symptoms is the presence of microorganisms
despite treatment. Because nontuberculous mycobacteria that are resistant to typically prescribed antimicrobial
(NTM) are resistant to commonly used antimicrobial drugs drugs, for example, nontuberculous mycobacteria (NTM).
and are found in drinking water that patients may use Recovery of NTM from the sinus cavity has been
for sinus irrigation, we investigated whether some CRS documented in 19 patients, including those with cystic
patients were infected with NTM in New York, New York,
fibrosis (3), HIV infection (4–10), and diabetes (11). NTM
USA, during 2001–2011. Two approaches were chosen: 1)
records of NTM-infected CRS patients were reviewed to
isolation from the sinus cavity has been rarely reported in
identify common features of infection and Mycobacterium immunocompetent, nondiabetic patients who do not have
species; 2) samples from plumbing in households of 8 cystic fibrosis (12–15). One case of infection with NTM
NTM-infected patients were cultured for NTM presence. is documented in a study by Spring and Miller (14). The
In 3 households sampled, M. avium sharing rep-PCR and patient had a 21-year history of rhinosinusitis and exhibited
pulsed field gel electrophoresis fingerprints identified M. left maxillary facial pain, nasal discharge, and congestion.
avium isolates clonally related to the patients’ isolates. We Mycobacterium chelonae, Staphylococcus aureus, and
conclude that patients with treatment-resistant CRS may be Pseudomonas aeruginosa were recovered from sinus cultures.
infected with NTM and should have cultures performed for Successful treatment ultimately required a 3-year course of
NTM so appropriate therapy can be instituted. In addition, multiple intravenous antimicrobial drug combinations and
the results suggest that CRS patients can be infected by
subsequent sinus operations (14). Recently, a new member
NTM in their household plumbing.
of the M. abscessus-chelonae complex, M. franklinii, was
isolated from patients in the northeastern United States who

A subset of patients with chronic rhinosinusitis

(CRS) often experience persistent symptoms,
despite undergoing many medical and surgical modes of
have chronic sinusitis (16).
NTM are environmental opportunistic pathogens found
in natural and human-engineered waters, including drinking
treatment. Current theories regarding the cause of CRS water distribution systems (17) and household plumbing
include immunologic reactions to microorganisms (1,2). (18–20). NTM species can be classified into 2 groups on
Even though they receive various treatments, including the basis of growth rates; rapidly growing mycobacteria
antimicrobial drugs and sinus irrigation, many patients (e.g., M. chelonae and M. abscessus) form colonies in
continue to be symptomatic (2). One possible reason for the <7 days at 37°C, and slowly growing mycobacteria (e.g.,
M. avium and M. intracellulare) take >7 days at 37°C to
Author affiliation: The Center for Allergy, Asthma and Sinusitis, form colonies. Because NTM are resistant to commonly
New York, New York, USA (W.S. Tichenor, J. Thurlow); New York used antimicrobial drugs (21) and are found in drinking
Medical College, Valhalla, New York (W.S. Tichenor); University of water, they might be responsible for antimicrobial drug–
Texas Health Science Center, Tyler, Texas, USA (S. NcNulty, B.A. resistant, chronic rhinosinusitis. We report the isolation,
Brown-Elliott, R.J. Wallace, Jr.); Virginia Polytechnic Institute and identification, and fingerprinting of NTM isolates from
State University, Blacksburg, Virginia, USA (J.O. Falkinham III) patients with CRS and from their household plumbing.
DOI: http://dx.org/10.3201/eid1810.120164
1
These authors contributed equally to this article.

1612 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Mycobacteria in Household Plumbing

Methods Table 1. Characteristics of patients whose sinuses yielded NTM

in study of NTM in household plumbing, New York, New York,
Patient Histories USA, 2001–2011*
Characteristics Value
We reviewed the charts of 33 adult outpatients in Total patients 33 (100)
whom CRS was diagnosed in the medical practice of W.S. Age range, y 25–74
Tichenor, whose endoscopically directed sinus cultures Prior functional endoscopic sinus surgery 30 (91)
Nasal polyps 12 (36)
yielded NTM. The 33 represent ≈1% of patient samples Primary immunodeficiency 10 (30)
collected over a 10-year period. In all patients, CRS had HIV positive 0†
been diagnosed on the basis of a combination of initial Cystic fibrosis carrier state 1 (20)‡
evaluation; appearance of the sinuses by endoscopic Diabetes 0
Repeat culture NTM negative 21 (64)
examination; results of computed tomographic scan; and Repeat culture NTM positive 2 (6)
endoscopically directed cultures for bacteria, fungi, and Repeat culture not performed or lost 10 (30)
NTM. From all patients, bacterial isolates had been cultured Symptoms improved 14 (42)
Symptoms unchanged 6 (18)
at the time of endoscopy. Other persistent microorganism§ 1 (3)
Initial symptoms, NTM identity, surgical history, HIV Refused treatment 3 (9)
status, cystic fibrosis history and carrier status, diabetes Currently treated 9 (27)
*Values are no. (%) patients except as indicated. NTM, nontuberculous
and immune-deficiency status, current nasal irrigations, mycobacteria.
presence of polyps, treatment, repeat culture results, and †No patients were known to be HIV positive; 16 were tested.
‡No patients were known to have cystic fibrosis; 5 were tested.
symptom reduction were assessed (Table 1). Common §Methicillin-resistant Staphylococcus aureus.
patient conditions at the time of nasal endoscopy included
headache, nasal blockage or congestion, thick postnasal or family member to collect hot and cold water samples
drip, and decreased ability to taste or smell. Thirty (91%) (500 mL) and biofilms/sediment from water taps and
of the 33 patients had previously undergone endoscopic showerheads. Biofilm samples were collected by swabbing
sinus surgery; 10 (30%) had histories of primary the inside of taps and showerheads, and swab specimens
immunodeficiency. Twelve (36%) of the 33 patients were placed in tubes containing 2 mL of tap water (from
had evidence of polyps at the time of nasal endoscopy; Blacksburg, VA, USA), sterilized by autoclaving. If in-line
however, no clear association was found between NTM or point-of-use water filters were submitted by the patients,
species and the presence of polyps. Thirty-one (94%) of a 4-cm2 area was swabbed, and the swab was placed in 2
33 patients were using some form of nasal irrigation at the mL of sterile tap water.
time of endoscopy. Of those, 26 were known to have used
tap water to irrigate the sinuses. NTM Isolation, Identification, and Fingerprinting
Patient NTM isolates were identified by various methods,
Patient Sample Collections depending on the laboratory: DNA probe, high-performance
Endoscopically directed samples were taken directly liquid chromatography, gas-liquid chromatography, internal
from the sinuses, middle meatus, or ostiomeatal unit by transcribed spacer region or 16S rDNA sequencing. NTM in
using a flexible catheter with a self-contained Lukens trap water and swab (taps and filters) samples were enumerated
as described (22). Samples (0.5–3 mL) were sent to the and isolated as described (19). Household NTM isolates
microbiology laboratories (Mayo Medical Laboratories, and those from patients were identified by nested PCR of
Rochester, MN, USA; Specialty Laboratories, Valencia, 16S rRNA gene (23) and PCR amplification and analysis
CA, USA; Quest Laboratories, Peterboro, NJ, USA) in of restriction endonuclease digestion fragments of the hsp-
sterile 5-mL containers. 65 gene (24). When the Mycobacterium species of the
patient and household water system isolates were identical,
Household Collections isolates were fingerprinted by rep-PCR (25) and pulsed-
Members of households with occupants who had field gel electrophoresis (PFGE) of AseI and XbaI restriction
NTM-associated CRS volunteered to participate in studies endonuclease digests of genomic DNA (26). To interpret
of their households’ water systems. Informed consent was PFGE in categories of “indistinguishable,” “closely related,”
obtained from each collaborating patient, and the study and “different,” we used previously described criteria for the
was reviewed by the Virginia Tech Institutional Research evaluation of Mycobacterium avium complex (MAC) isolates
Board and granted exempt status. NTM isolates from the (27). With a minimum of 10 interpretable bands, strains
patients were obtained through laboratories that cultured were interpreted as indistinguishable (no band differences),
NTM from endoscopy samples. Containers, swabs, and closely related (1–3 band differences), possibly related (4–6
tubes were sent to each collaborating patient’s household. band differences), and different (>7 band differences). These
Directions were provided for the collaborating patient isolates underwent species confirmation by sequencing

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1613
RESEARCH

of the internal transcribed spacer 1. M. intracellulare and Table 2. NTM isolated from sinus cavity samples of 33 patients in
M. chimaera are indistinguishable without gene/region study of NTM in household plumbing, New York, New York, USA,
sequencing (28). 2001–2011*
NTM species No. (%) patients
Mycobacterium abscessus-chelonae 19 (58)
Results M. chelonae 4 (12)
Review of the charts of the 33 CRS patients showed that M. abscessus 2 (6)
39 NTM isolates belonging to 10 Mycobacterium species M. avium 4 (12)
M. avium complex 2 (6)
were recovered from samples from the ostiomeatal unit or M. immunogenum 1 (3)
paranasal sinuses (Table 2). The patients’ mycobacterial M. asiaticum 1 (3)
isolates were identified by Mayo Clinic, Quest, and M. mucogenicum 1 (3)
M. mageritense 1 (3)
Specialty Laboratories. Two different Mycobacterium M. gordonae 4 (12)
species were isolated from 6 patient samples (Table 2). *NTM, nontuberculous mycobacteria.
Most isolates (25 [64%] of 39) were rapidly growing
mycobacteria, primarily M. abscessus or M. chelonae. One rep-PCR fingerprinting (25). To confirm the relatedness
laboratory that received patient samples did not distinguish between isolates from patient and household plumbing,
M. abscessus from M. chelonae. The predominant slowly PFGE was performed (26) for the same isolates (Figure
growing Mycobacterium species was MAC (6 [15%] of 2). The PFGE band pattern of the isolate from patient
39). M. gordonae was isolated from 4 (12%) of the 33 2 and the pattern from the patient’s household (lanes
patients. Although the organism is normally considered a 10 and 11) appear almost identical (“closely related”).
saprophyte, M. gordonae infection has been reported in The PFGE patterns for 2 isolates from the household
immunodeficient persons (29–31), and thus its isolation of patient 5 were “indistinguishable” and are “closely
should not be dismissed. related” (clonal) to the respective patient isolates and
thereby clonal (Figure 2, panel A). Isolates from patient
NTM Isolates from Households of 8 and the patient’s household plumbing (not shown)
Current CRS Patients gave faint signals by PFGE with repeat testing and both
A total of 80 samples (i.e., 43 water, 31 biofilm, and restriction enzymes. However, the patterns appeared
6 from filters) for NTM isolation were received from the “indistinguishable”(profile not shown). The lack of clear
8 collaborating CRS patients. NTM were isolated from band patterns for the isolates from patient 8 and his or
water, biofilm, or filter samples from at least 1 sample her household plumbing is likely because of the shared
from 5 (63%) of the 8 households sampled and from 35 characteristic of resistance to lysis in the agar plugs (26).
(40%) of the 80 samples (Table 3). The frequency of The absence of a match for patient 1 (not shown) might
NTM recovery from water (44%), biofilm (42%), and be because the person moved throughout the United
filter (50%) samples was not significantly different (p States, and some places where the patient lived were not
= 0.6065, Kruskal-Wallace test). NTM colony counts sampled. Samples of showerheads were collected from 6
varied widely in samples from the different households of the 8 households, and although NTM isolates of the
(Table 4). In 4 households, at least 1 of the samples same species as that of the patient (i.e., M. avium) were
yielded an NTM isolate that was of the same species and recovered from 2 households, none of the showerhead
had the same rep-PCR fingerprint as that of the patient isolates shared the same fingerprint with isolates from
according to published criteria (25) (Figure 1). The band the patient. Notably, the samples from the household
patterns illustrate the large number and wide range of rep- plumbing of the patients with M. gordonae and M.
PCR bands and illustrate the discrimination provided by immunogenum isolates did not yield any NTM.

Table 3. Recovery of NTM from households in study of NTM in household plumbing, New York, New York, USA, 2001–2011*
Patient No. samples No. (%) samples Species found in
household no. Patient isolate collected yielding NTM patient household rep-PCR match PFGE match
1 M. abscessus 9 5 (55) None NA NA
2 M. avium 9 4 (44) Yes Yes Yes
3 M. immunogenum 10 0 NA NA NA
4 M. gordonae 5 2 (40) Yes No –
5 M. avium 10 9 (90) Yes Yes Yes
6 MAC-X† 21 7 (33) Yes Yes No
7 M. gordonae 10 0 NA NA –
8 M. avium 14 8 (57) Yes Yes Yes
*NTM, nontuberculous mycobacteria; NA, not applicable; MAC-X, Mycobacterium avium complex “X” cluster; PFGE, pulsed-field gel electrophoresis.
†MAC-X is a mycobacterium that tests positive by DNA probe analysis for M. avium complex but is negative with specific M. avium and M. Intracellulare
probes and PCR analysis.

1614 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Mycobacteria in Household Plumbing

Table 4. Numbers of NTM in household samples in study of NTM were involved in pathogenesis. No guidance exists for
in household plumbing, New York, New York, USA, 2001–2011* the diagnosis and treatment of NTM sinus infection.
Patient Water Biofilm For pulmonary NTM disease, it is recommended that
household Average Average
no. No. CFU/mL No. CFU/cm
2 multiple cultures be obtained over time (21) to rule out
1 4 5,632 ± 3,372 2 36,000 ± 49,500 transient colonization and avoid sampling deficiencies.
2 5 49 ± 18 6±2 Our experience suggests that multiple cultures may be
3 0 0 necessary to find NTM because endoscopy samples from
4 1 0
5 13 420 ± 1,000 8 23,310 ± 41,700 many patients will be found NTM positive only by 1 of
6 3 17,052 ± 11,200 11 21,100 ± 27,700 2–5 endoscopies. For example, 1 patient had cultures that
7 0 0 yielded M. mageritense, but cultures obtained 1 week later
8 7 27 ± 26 8 513 ± 632
Total 33 2,487 31 13,835 were negative, even in the absence of antibacterial drug
*NTM, nontuberculous mycobacteria. treatment. In addition, smears from 2 patients showed acid-
fast bacilli, but cultures failed to yield any Mycobacterium
Discussion species isolate; yet upon subsequent endoscopy, NTM
Our study confirms the possibility of the involvement
of NTM in sinuses of patients with CRS (3,11,16). CRS
patients who have not responded to medical treatment
should undergo endoscopically directed sinus cultures
for microorganisms, including fungi and NTM and other
bacteria. Endoscopically directed sinus cultures have
been shown to accurately replicate sinus puncture culture
techniques (22). The American Thoracic Society and the
Infectious Diseases Society of America discourage the
use of swabs for sampling because swabs may decrease
the likelihood of recovering NTM (21). Using a suction
device to remove larger volumes of mucus helps increase
the chances of obtaining representative sinus microflora
(22). Spurious recovery of NTM, because of endoscope
contamination, is possible (32), as is the possibility
that glutaraldehyde may not adequately kill NTM (33).
However, in the current study, endoscope contamination is
an unlikely source of NTM because water samples from
the physician’s office did not reveal NTM. In addition, the
patient and household samples were processed in different
laboratories.
Besides establishing NTM as a potential agent of CRS,
our results strongly suggest that in 3 of the 8 CRS patients
studied here, the household plumbing was the source of
infection, on the basis of identity of rep-PCR fingerprints
of patient and household isolates and their clonal
relatedness as determined by PFGE. Clonal variation in
Mycobacterium species isolates is characteristic of isolates
recovered from household plumbing, but because single Figure 1. rep-PCR fingerprint patterns of patient and household
Mycobacterium species isolates are typically recovered isolates, New York, New York, USA, 2001-2011. Lane 1, 100-bp
from patient samples, DNA fingerprint matches are not ladder; lane 2, patient no. 5 Mycobacterium avium isolate AG-P-1;
lane 3, patient no. 5 household filter M. avium isolate AG-F-2–0-2;
always obtained (19,20). A study of persons with NTM lane 4, patient no. 5 household filter M. avium isolate AG-F-2-I-1;
pulmonary disease found that in 7 (41%) of 17 households, lane 5, patient no. 6 M. avium complex “X” cluster isolate GG-P-
patient and household plumbing isolates were identical as 1; lane 6, patient no. 6 household swab M. chimaera isolate GG-
shown by rep-PCR fingerprints (20). Because NTM are Sw-9–1; lane 7, patient no. 8 M. avium isolate GW-P-1; lane 8,
found in household tap water (19,20,34), CRS patients patient no. 8 household water M. avium isolate GW-W-1–1; lane
9, patient no. 8 household swab M avium isolate GW-Sw-7–2;
should avoid sinus irrigation with unsterilized tap water. lane 10, patient no. 2 M. avium isolate BB-P-1; lane 11, patient no.
A major question concerning isolation of NTM from 2 household water M. avium isolate BB-W-4–5; lane 12, 100-bp
the sinus cavities of patients with CRS is whether NTM ladder.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1615
RESEARCH

NTM culture and sensitivity testing take several months

to obtain, patients are typically treated for other possible
infecting microorganisms until the results of the NTM
cultures are reported. Resolution typically occurred only
after an extended course of multiple antimycobacterial
agents given simultaneously. Unfortunately, the
combination of insufficient experience and the absence of
an established treatment protocol for CRS caused by NTM
(21), prevent any meaningful review of treatment regimens
for CRS caused by NTM. Such patients are treated with 2
oral antimycobacterial drugs and urged to irrigate sinuses
with sterile or boiled water or saline, followed by irrigation
with a topical antimycobacterial agent for 3–18 months,
depending on clinical response and, in some cases, on
subsequent positive cultures for NTM
The role of NTM in infectious disease processes is
only starting to be recognized. This work documents that
a proportion of patients with CRS could be infected with
NTM and that sinus samples should be cultured for NTM.
Figure 2. Pulsed-field gel electrophoresis (PFGE) of AseI digest
patterns of patient and household isolates, New York, New York,
In addition, CRS patients should avoid sinus irrigation with
USA, 2001–2011. A) Patient and household isolates. Lane 1, λ tap water because tap water may contain NTM, and it may
ladder; lane 2, patient no. 5 Mycobacterium avium isolate AG-P-1; not be possible to remove it. Sterile saline should be used
lane 3, patient no. 5 household filter M. avium isolate AG-F-2-0-2; instead.
lane 4, patient no. 5 household filter M. avium isolate AG-F-2-I-1
(environmental isolates in lanes 3 and 4 are indistinguishable;
patient isolate in lane 2 considered clonal with 2 environmental
Acknowledgments
isolates [6 bands difference]); with digestion with XbaI, the 3 were We acknowledge the efforts of Myra D. Williams, who
considered closely related.); lane 5, patient no. 6 M. avium complex processed samples and identified, enumerated, and fingerprinted
“X” cluster isolate GG-P-1; lane 6, patient no. 6 household swab M. NTM; and Elena Iakiaeva for performance of the multiplex PCR
chimaera isolate GG-Sw-9–1 (despite overall similarity, isolates in and for interspacer 1 DNA sequencing.
lanes 5 and 6 belong to different species and differ by 10 bands and
are therefore unrelated). B) Additional patient and isolate from the Work in the Mycobacteria/Nocardia Laboratory at the
person’s household. Lane 1, patient no. 2 M. avium isolate BB-P-1; University of Texas Health Science Center was supported by the
lane 2, patient no. 2 household water M. avium isolate BB-W-4–5;
lane 3, λ ladder.
Amon Carter Foundation.
Dr Tichenor is clinical associate professor of medicine at
New York Medical College and maintains a private practice in
New York, New York, focusing on the treatment of patients with
were cultured. Several possible reasons could account for chronic resistant sinusitis. His research interests are nontubercu-
this low yield. First, hydrophobic NTM cells are likely lous mycobacterial infections and primary immune deficiencies as
adhering to the walls of the sinus cavity, and thereby a low they relate to treatment of chronic sinusitis.
number are in fluid removed during endoscopy. Second,
the small volume of mucus removed at the time of culture References
also might reduce the likelihood of recovering NTM (22).
Third, topical anesthetics, typically lidocaine, are used for 1. Bachert C, Zhang N, Patou J, van Zele T, Gevaert P. Role of
anesthesia for endoscopy and might inhibit the growth of staphylococcal superantigens in upper airway disease. Curr Opin
Allergy Clin Immunol. 2008;8:34–8. http://dx.doi.org/10.1097/
many microorganisms, including NTM (35). Although ACI.0b013e3282f4178f
NTM could merely be colonizing the sinuses, several 2. Ponikau JU, Sherris DA, Kephart GM, Kern EB, Congdon DJ,
factors suggest otherwise. The samples that we collected Adolphson CR, et al. Striking deposition of toxic eosinophil major
were primarily mucus, and previously published reports on basic protein in mucus: implications for chronic rhinosinusitis. J
Allergy Clin Immunol. 2005;116:362–9. http://dx.doi.org/10.1016/j.
NTM in sinus samples from immunocompromised CRS jaci.2005.03.049
patients were primarily based on biopsy specimens (3,6,15). 3. Brown SM, DeCelie-Germana JK, Shikowitz MJ, Zahtz GD.
In addition, our patients typically have persistent symptoms Nontuberculous mycobacterial sinusitis in children with cystic
despite treatment with multiple different antimicrobial fibrosis. Int J Ped Otorhinol Extra. 2007;2:9–13. http://dx.doi.
org/10.1016/j.pedex.2006.10.002
drug regimens over several months. Because the results of

1616 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Mycobacteria in Household Plumbing

4. Cibrián F, Quiles I, Anaut P, Gainzaráin J, Vega L, Andia A. Sinusitis 22. Tichenor WS, Adinoff A, Smart B, Hamilos DL. Nasal and sinus
caused by Mycobacterium avium–M. intracellulare in a patient with endoscopy for medical management of resistant rhinosinusitis,
HIV infection. Enferm Infecc Microbiol Clin. 1996;14:401–2. including postsurgical patients. J Allergy Clin Immunol.
5. Li C, Szuba M, Schuman P, Crane L, Vazquez JA. Mycobacterium 2008;121:917–27. http://dx.doi.org/10.1016/j.jaci.2007.08.065
kansasii sinusitis in a patient with AIDS. Clin Infect Dis. 23. Wilton S, Cousins D. Detection and identification of multiple
1994;19:792–3. http://dx.doi.org/10.1093/clinids/19.4.792 mycobacterial pathogens by DNA amplification in a single tube.
6. Naguib MT, Byers JM, Slater LN. Paranasal sinus infection due to PCR Methods Appl. 1992;1:269–73.
atypical mycobacteria in two patients with AIDS. Clin Infect Dis. 24. Telenti A, Marchesi F, Balz M, Bally F, Böttger EC, Bodmer T. Rapid
1994;19:789–91. http://dx.doi.org/10.1093/clinids/19.4.789 identification of mycobacteria to the species level by polymerase
7. Sussman SJ. Sinusitis caused by Mycobacterium avium-intracellulare chain reaction and restriction enzyme analysis. J Clin Microbiol.
in a patient with human immunodeficiency virus. Pediatr Infect Dis 1993;31:175–8.
J. 1995;14:726–7. http://dx.doi.org/10.1097/00006454-199508000- 25. Cangelosi GA, Freeman R, Lewis KN, Livingston-Rosanoff
00024 D, Shah KS, Milan SJ, et al. Evaluation of high-throughput
8. Tsi L, Gálvez A, Broto J, García Restoy E, Gual J. Sinusitis in HIV repetitive sequence–based PCR system for DNA fingerprinting
infection. Acta Otorrinolaringol Esp. 1994;45:301–2. of Mycobacterium tuberculosis and Mycobacterium avium
9. Upadhyay S, Marks S, Arden R, Crane L, Cohn A. Bacteriology complex strains. J Clin Microbiol. 2004;42:2685–93. http://dx.doi.
of sinusitis in human immunodeficiency virus–positive patients: org/10.1128/JCM.42.6.2685-2693.2004
implications for management. Laryngoscope. 1995;105:1058–60. 26. Mazurek GH, Hartman S, Zhang Y, Brown BA, Hector JSR,
http://dx.doi.org/10.1288/00005537-199510000-00009 Murphy D, et al. Large DNA restriction fragment polymorphism in
10. Zurlo JJ, Feuerstein IM, Lebovics R, Lane HC. Sinusitis in the Mycobacterium avium–M. intracellulare complex: a potential
HIV-1 infection. Am J Med. 1992;93:157–62. http://dx.doi. epidemiologic tool. J Clin Microbiol. 1993;31:390–4.
org/10.1016/0002-9343(92)90045-D 27. Wallace RJ, Zhang Y, Brown-Elliott BA, Yakrus MA, Wilson
11. Ferguson BJ, Kapadia SB, Carrau RL. Mycobacterium avium RW, Mann L, et al. Repeat positive cultures in Mycobacterium
complex infection of the paranasal sinuses. Otolaryngol Head intracellulare lung disease after macrolide therapy represent new
Neck Surg. 1997;117:S160–2. http://dx.doi.org/10.1016/S0194- infections in patients with nodular bronchiectasis. J Infect Dis.
5998(97)70089-5 2002;186:266–73. http://dx.doi.org/10.1086/341207
12. Eron LJ, Huckins C, Park CH, Poretz DM, Gelman HK, Ball MF. 28. Tortoli E, Rindi L, Garcia MJ, Chiaradonna P, Dei R, Garzelli C, et
Mycobacterium chelonei infects the maxillary sinus: a rare case. Va al. Proposal to elevate the genetic variant MAC-A, included in the
Med. 1981;108:335–8. Mycobacterium avium complex, to species rank as Mycobacterium
13. Mra Z, Roach J, Brook A. Infectious and neoplastic diseases of the chimaera sp. nov. Int J Syst Evol Microbiol. 2004;54:1277–85.
sphenoid sinus—a report of 10 cases. Rhinology. 2002;40:34–40. http://dx.doi.org/10.1099/ijs.0.02777-0
14. Spring PM, Miller RH. Initial report of primary sinusitis caused 29. Weinberger M, Berg SL, Feuerstein IM, Pizzo PA, Witebsky FG.
by an atypical pathogen (Mycobacterium chelonae) in an Disseminated infection with Mycobacterium gordonae: report
immunocompetent adult. Ear Nose Throat J. 1999;78:358–9, 362–4. of a case and critical review of the literature. Clin Infect Dis.
15. Weiss RL, Zahtz GD, Isenberg H. Nontuberculous mycobacterial 1992;14:1229–39. http://dx.doi.org/10.1093/clinids/14.6.1229
infection of the frontal sinus in a child. Otolaryngol Head Neck Surg. 30. Maslo C, Hadacek B, Maresca A, Vallee E, Coulaud JP. Infections
1997;116:110–2. http://dx.doi.org/10.1016/S0194-5998(97)70360-7 à Mycobacterium gordonae au cours de l’infection par le virus de
16. Simmon KE, Brown-Elliott BA, Ridge PG, Durtschi JD, Mann l’immunodéficience humaine. Presse Med. 1995;24:1157–60.
LB, Slechta ES, et al. Mycobacterium chelonae-abscessus complex 31. Eckburg PB, Buadu EO, Stark P, Sarinas PSA, Chitkara RK,
associated with sinopulmonary disease, northeastern USA. Kuschner WG. Clinical and chest radiographic findings among
Emerg Infect Dis. 2011;17:1692–700. http://dx.doi.org/10.3201/ persons with sputum culture positive for Mycobacterium gordonae.
eid1709.101667 Chest. 2000;117:96–102. http://dx.doi.org/10.1378/chest.117.1.96
17. Falkinham JO III. Surrounded by mycobacteria: nontuberculous 32. Wallace RJ Jr, Brown B, Griffith D. Nosocomial outbreaks/pseudo-
mycobacteria in the human environment. J Appl Microbiol. outbreaks caused by nontuberculous mycobacteria. Annu Rev
2009;107:356–67. http://dx.doi.org/10.1111/j.1365-2672.2009. Microbiol. 1998;52:453–90. http://dx.doi.org/10.1146/annurev.
04161.x micro.52.1.453
18. Nishiuchi Y, Maekura R, Kitada S, Tamaru A, Taguri T, Kira 33. Griffiths PA, Babb JR, Bradley CR, Fraise AP. Glutaraldehyde-
Y, et al. The recovery of Mycobacterium avium-intracellulare resistant Mycobacterium chelonae from endoscope washer
complex (MAC) from the residential bathrooms of patients with disinfectors. J Appl Microbiol. 1997;82:519–26. http://dx.doi.
pulmonary MAC. Clin Infect Dis. 2007;45:347–51. http://dx.doi. org/10.1046/j.1365-2672.1997.00171.x
org/10.1086/519383 34. Feazel LM, Baumgartner LK, Peterson KL, Frank DN, Harris JK,
19. Falkinham JO III, Iseman MD, De Haas P, van Soolingen D. Pace NR. Opportunistic pathogens enriched in showerhead biofilms.
Mycobacterium avium in a shower linked to pulmonary disease. J Proc Natl Acad Sci U S A. 2009;106:16393–9. http://dx.doi.
Water Health. 2008;6:209–13. org/10.1073/pnas.0908446106
20. Falkinham JO III. Nontuberculous mycobacteria from household 35. Schmidt RM, Rosenkranz HS. Antimicrobial activity of local
plumbing of patients with nontuberculous mycobacterial disease. anesthetics: lidocaine and procaine. J Infect Dis. 1970;121:597–607.
Emerg Infect Dis. 2011;17:419–24. http://dx.doi.org/10.1093/infdis/121.6.597
21. Griffith DE, Aksamit T, Brown-Elliott BA, Catanzaro A, Daley
C, Gordin F, et al. An official ATS/IDSA statement: diagnosis, Address for correspondence: Wellington S. Tichenor, 642 Park Ave, New
treatment, and prevention of nontuberculous mycobacterial diseases.
York, NY, 10065, USA; email: wtichenor@sinuses.com
Am J Respir Crit Care Med. 2007;175:367–416. http://dx.doi.
org/10.1164/rccm.200604-571ST

Search past issues of EID at www.cdc.gov/eid

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1617
RESEARCH

Autochthonous and Dormant

Cryptococcus gattii Infections
in Europe
Ferry Hagen, M. Francisca Colom, Daniëlle Swinne, Kathrin Tintelnot, Roberta Iatta,
Maria Teresa Montagna, Josep M. Torres-Rodriguez, Massimo Cogliati, Aristea Velegraki,
Arjan Burggraaf, Alwin Kamermans, Johanna M. Sweere, Jacques F. Meis, Corné H.W. Klaassen,
and Teun Boekhout

Until recently, Cryptococcus gattii infections occurred immunocompromised persons, and C. gattii mainly infects
mainly in tropical and subtropical climate zones. However, apparently immunocompetent persons. C. neoformans is
during the past decade, C. gattii infections in humans and found globally, and C. gattii has been mostly limited to
animals in Europe have increased. To determine whether the tropical and subtropical areas in Central Africa, northern
infections in Europe were acquired from an autochthonous Australia, and Central and South America (1). However,
source or associated with travel, we used multilocus
this distribution pattern changed after an unprecedented
sequence typing to compare 100 isolates from Europe (57
from 40 human patients, 22 from the environment, and 21
outbreak of C. gattii emerged in the temperate climate of
from animals) with 191 isolates from around the world. Of British Columbia, Canada, and expanded to the Pacific
the 57 human patient isolates, 47 (83%) were obtained since Northwest region of Canada and the United States (1,2).
1995. Among the 40 patients, 24 (60%) probably acquired Epidemiologic studies have shown that C. gattii occurs
the C. gattii infection outside Europe; the remaining 16 in areas other than tropical or subtropical zones, such as
(40%) probably acquired the infection within Europe. Human in Mediterranean Europe, northern Europe, and western
patient isolates from Mediterranean Europe clustered into a Australia (3–5).
distinct genotype with animal and environmental isolates. For the purpose of studying the epidemiology of C.
These results indicate that reactivation of dormant C. gattii gattii, a broad variety of molecular biological techniques
infections can occur many years after the infectious agent have been developed, including PCR fingerprinting,
was acquired elsewhere.
restriction fragment length polymorphism analysis of
the PLB1 and URA5 loci, amplified fragment length

D uring the past decade, the basidiomycetous yeast

Cryptococcus gattii has stepped out of the shadows of
its sibling C. neoformans. The latter species mainly infects
polymorphism (AFLP) fingerprint analysis, and several
multilocus sequence typing (MLST) approaches (6–9).
These laboratory investigations have shown that C. gattii
can be divided into 5 distinct genotypes: AFLP4/VGI,
Author affiliations: Royal Netherlands Academy of Arts and
AFLP5/VGIII, AFLP6/VGII, AFLP7/VGIV, and AFLP10/
Sciences Fungal Biodiversity Centre, Utrecht, the Netherlands (F.
VGIV (8,9). Serotype B strains occur in genotypes AFLP4/
Hagen, A. Burggraaf, A. Kamermans, J.M. Sweere, T. Boekhout);
VGI, AFLP6/VGII, and AFLP10/VGIV; serotype C strains
University Medical Center Utrecht, Utrecht (F. Hagen, T Boekhout);
are restricted to genotypes AFLP5/VGIII and AFLP7/
Universidad Miguel Hernàndez, Alicante, Spain (M.F. Colom);
VGIV (8).
Institute of Public Health, Brussels, Belgium (D. Swinne); Robert
Recently, a consensus MLST scheme was proposed for
Koch-Institut, Berlin, Germany (K. Tintelnot); Università degli Studi
epidemiologic investigations of C. gattii and C. neoformans,
di Bari, Bari, Italy (R. Iatta, M.T. Montagna); Autonomous University
specifically, the nuclear loci CAP59, GPD1, IGS1, LAC1,
of Barcelona, Barcelona, Spain (J.M. Torres-Rodriguez); Università
PLB1, SOD1, and URA5 (9). So far, this consensus MLST
degli Studi di Milano, Milan, Italy (M. Cogliati); University of Athens,
scheme has been used to study the population structure of
Athens, Greece (A. Velegraki); University College Utrecht, Utrecht
C. neoformans strains from Thailand and C. gattii strains
(J.M. Sweere); Radboud University Nijmegen Medical Centre,
from Australia (3,10).
Nijmegen, the Netherlands (J.F. Meis); and Canisius Wilhelmina
We investigated the occurrence of C. gattii in Europe,
Hospital, Nijmegen (J.F. Meis, C.H.W. Klaassen)
focusing on whether this pathogen is emerging and, if so,
DOI: http://dx.doi.org/10.3201/eid1810.120068 how to explain this emergence pattern. Furthermore, we

1618 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Cryptococcus gattii in Europe

explored whether the infections originated from Europe of both primers (Biolegio, Nijmegen, the Netherlands)
or were introduced from other continents. To achieve (online Technical Appendix Table 2), and ≈100 ng of
these goals, members of the European Confederation of genomic DNA. PCRs were conducted with an initial
Medical Mycology were asked to send recently obtained denaturation step at 94°C for 5 min, followed by 35
human patient isolates of the species for detailed AFLP cycles of denaturation at 94°C for 30s, annealing for 30s
and MLST analyses. Thus, the genetic diversity of the (see Technical Appendix Table 2 for optimal annealing
yeast was used to trace its geographic origin to identify temperatures), extension at 72°C for 1 min, followed by
where the infections were acquired. AFLP genotyping 72°C for 5 min and a final dwell at 21°C.
results were combined with published C. gattii MLST Sequencing reactions were conducted with the BigDye
results from Byrnes et al. (11) and Fraser et al. (12), which version 3.1 chemistry kit (Applied Biosystems, Foster City,
were extended to match the Cryptococcus consensus CA, USA) as described (13). For all amplification products
MLST scheme (9). Our study produced the following 5 except CAP59, the initial amplification primers were used
conclusions: all hitherto known genotypes of C. gattii for sequencing reactions. For CAP59, the newly designed
are emerging in Europe; genotype AFLP4/VGI isolates forward primer CAP59L-Fwd and the original reverse
predominate; a C. gattii cluster, which is endemic to primer JOHE15438 (12) were used.
Mediterranean Europe and genetically distinct from the
other populations, exists; several human infections are Sequence Alignment and Phylogenetic and
caused by travel-related acquisition of C. gattii outside Recombination Analyses
Europe; and autochthonous cases occur in Europe. Consensus sequences were assembled and checked
for ambiguities by using SeqMan version 8.0.2
Materials and Methods (DNASTAR, Madison, WI, USA). Sequence alignments
were generated with MEGA version 5 (14) by using the
Strains and Media standard settings and manual correction. The genome
We compared a collection of 107 isolates collected sequence databases of reference strains H99 (culture
from Europe with 194 isolates collected globally (Table 1; collection no. CBS8710; C. neoformans variety grubii;
online Technical Appendix Table 1, wwwnc.cdc.gov/EID/ AFLP1/VNI; Broad Institute [www.broadinstitute.
pdfs/12-0068-Techapp.pdf). All isolates were checked for org/annotation/genome/cryptococcus_neoformans/])
purity and cultivated on malt extract agar medium (Oxoid, and JEC21 (culture collection no. CBS10513; C.
Basingstoke, UK). Cultures were incubated for 2 days at neoformans variety neoformans; AFLP2/VNIV; Stanford
30°C. A working collection was made by growing C. gattii Genome Technology Center [www-sequence.stanford.
strains on malt extract agar slants for 2 days at 30°C, after edu/group/C.neoformans/]) were used to extract the
which the strains were stored at 4°C. Strains were stored corresponding sequences for all 10 investigated nuclear
long term at −80°C by using the Microbank system (Pro- loci to serve as an outgroup. The best fitting nucleotide
Lab Diagnostics, Richmond Hill, Ontario, Canada). substitution model was determined by using MrModeltest
version 2 (15) and was conducted for the complete C.
Amplification and Sequencing of MLST Loci gattii 10-loci MLST, for the accepted consensus MLST
Genomic DNA extraction, AFLP genotyping, and scheme (CAP59, GPD1, IGS1, LAC1, PLB1, SOD1,
mating-type determination by amplification of either the and URA5) (9), and a previously launched C. gattii
STE12a or STE12α locus were performed as described (8). MLST scheme (CAP10, GPD1, IGS1, LAC1, MPD1,
The 7 nuclear consensus MLST loci (CAP59, GPD1, IGS1, PLB1, and TEF1) (12). As a result, the HKY G+I model
LAC1, PLB1, SOD1, and URA5) were amplified by using the (Hasegawa-Kishino-Yano plus gamma distributed with
preferred primer combinations (9). To compare the current invariant sites) was the best model to use for analyzing the
set of C. gattii strains with those from a published C. gattii phylogeny of the C. gattii isolates for all 3 datasets. The
population biology study, we included the 3 nuclear loci evolutionary history was inferred by using the maximum-
that are not part of the consensus MLST scheme (CAP10, likelihood method in MEGA version 5 (14). A bootstrap
MPD1, and TEF1α) (12). We also included isolates from consensus tree was inferred from 1,000 replicates to show
a global study by Byrnes et al. (11) and Fraser et al. (12) the relevant lineages obtained in this analysis.
and subjected them to amplification and sequencing of the We calculated the haplotype diversity (HR), equal to
CAP59, SOD1, and URA5 loci. the Simpson diversity index (D), by using the Microsoft
Amplifications were conducted in a 25-μL PCR Excel (Microsoft, Redmond, WA, USA) add-in called
mixture containing 37.5 mmol/L MgCl2 (Bioline, London, Haplotype Analysis (16). For this purpose, sequences were
UK), 1× PCR buffer (Bioline), 1.9 mmol/L dNTPs collapsed into sequence type numbers (online Technical
(Bioline), 0.5 U Taq DNA polymerase (Bioline), 5 pmol Appendix Table 1).

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1619
RESEARCH

Results Phylogeographic Origin

Sequence type diversity was calculated for each of
AFLP Genotypes and Geographic Distribution the MLST loci, all 10 loci, and the combined datasets
The 301 C. gattii isolates collected from Europe and other according to Fraser et al. (12) and Meyer et al. (9) (online
areas around the world could be divided into the following Technical Appendix Table 3). The CAP10 locus showed
genotypes: 146 AFLP4/VGI (50.2%; 72 from Europe), 22 the lowest overall diversity (nST = 19; DST = 0.765) and
AFLP5/VGIII (7.6%; 1 from Europe), 108 AFLP6/VGII the IGS1 locus showed the highest diversity (nST = 52;
(37.1%; 23 from Europe), 13 AFLP7/VGIV (4.5%; 2 from DST = 0.930). The 10-loci MLST dataset showed the highest
Europe), and 2 AFLP10/VGIV (0.7%; both from Europe). diversity (nST = 150; DST = 0.975), followed by the MLST
From 10 isolates (7 AFLP8 [5 from Europe] and 3 AFLP9 scheme of Meyer et al. (DST = 0.971) (9) and Fraser et al.
[2 from Europe]), genotypes represented interspecies C. (DST of 0.959) (12). The latter MLST scheme differentiated
neoformans × C. gattii hybrids. These 10 isolates were more sequence types than the consensus MLST scheme
excluded from further analysis because amplified fragments (nST = 136 vs. 127).
of hybrid isolates will result in mixtures of different alleles. Maximum-likelihood analysis of the MLST data
The 57 human patient isolates from Europe were obtained showed that the C. gattii isolates clustered in 5 monophyletic
from 40 patients (online Technical Appendix Table 1). The clusters, which were highly supported and agreed with
genotypic diversity of the remaining 291 C. gattii isolates the AFLP genotypes (Figure; online Technical Appendix
(100 from Europe, 191 from other areas) with a haploid Figure). When each genotypic cluster was separately
genotype is shown in the Table. analyzed, support values of the branches were low, <75

Table. Distribution of Cryptococcus gattii strains*

Genotype
Source of isolation AFLP4/VGI AFLP5/VGIII AFLP6/VGII AFLP7/VGIV AFLP10/VGIV Total† Total (%)
All C. gattii isolates 146 22 108 13 2 0 291 (100)
Human 84 16 68 12 2 0 182 (62.5)‡
Environment 37 5 17 0 0 0 59 (20.3)‡
Animal 24 0 23 1 0 0 48 (16.5)‡
Unknown 1 1 0 0 0 0 2 (0.7)‡
Africa
Human 18 0 3 8 0 29 36 (12.7)§
Environment 6 0 0 0 0 6
Animal 0 0 0 1 0 1
Asia (clinical) 19 0 5 2 0 26 26 (8.9)§
Australia
Human 2 1 8 0 0 11 18 (6.2)§
Environment 5 0 1 0 0 6
Animal 0 0 1 0 0 1
Europe¶
Human 29# 1** 23†† 2‡‡ 2§§ 57 100 (34.4)§
Environment 22 0 0 0 0 22
Animal 21 0 0 0 0 21
North America¶
Human 3 8 19 0 0 30 69 (23.7)§
Environment 4 3 10 0 0 17
Animal 2 0 20 0 0 22
South America¶
Human 12 3 10 0 0 25 36 (12.4)§
Environment 0 2 6 0 0 8
Animal 1 0 2 0 0 3
Unknown
Human 1 3 0 0 0 4 6 (2.1)§
Unknown 1 1 0 0 0 2
*The number of C. gattii isolates is provided for each geographic area and further subdivided per genotype and source of isolation. AFLP, amplified
fragment length polymorphism; human, human clinical patient.
†Total number of isolates per geographic area.
‡The percentage of isolates is given for the source of isolation subset.
§The percentage of isolates is given as a percentage of all isolates.
¶Ten interspecies hybrid C. gattii × C. neoformans isolates were excluded from the set of isolates from Europe (n = 7), North America (n = 1), and South
America (n = 2).
#Twelve patients with an autochthonous acquired infection (16 isolates) and 11 patients with an infection acquired outside Europe (13 isolates).
**Infection acquired outside Europe.
††Four patients with an autochthonous infection (10 isolates) and 10 patients with an infection acquired outside Europe (13 isolates).
‡‡Two phenotypically different isolates from an emigrant from Zambia.
§§Two phenotypically different isolates from an emigrant from Mexico.

1620 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Cryptococcus gattii in Europe

(online Technical Appendix Figure). For genotype AFLP4/ the genotype that caused the Vancouver Island outbreak
VGI isolates, bootstrap support values for nearly all (Figure 1; online Technical Appendix Figure) (17–19). The
branches were low; for the second largest group formed outbreak-related sets of AFLP6A/VGIIa and AFLP6C/
by AFLP6/VGII isolates, branches were better supported. VGIIc isolates from the Vancouver Island and Pacific
Phylogenetic analysis demonstrated that several Northwest outbreaks, respectively, are within outbreak-
human patient genotype AFLP4/VGI isolates from specific clusters (Figure, online Technical Appendix
Europe had an autochthonous origin because they formed Figure). A similar finding was observed for a set of 5 human
a separate cluster with environmental and animal isolates patient C. gattii AFLP6B/VGIIb isolates (IP1996/1120–1
from the same area. A set of isolates from human patients and −2, IP1999/901–1 and −2, and IP2000/87) in France,
on the Iberian Peninsula (CCA232, CCA242L, CCA242T,
CCA311, CCA312, CL6148) were found to be genetically
indistinguishable from isolates obtained from animals
or the environment in the same area (indicated in online
Technical Appendix Figure as the European Mediterranean
cluster). The human patient isolates from Europe with
genotype AFLP4/VGI, which were probably acquired
within Europe because patients did not have histories of
travel outside Europe, are IP2005/215 and IP2006/958
(France), RKI85/888 (Germany), 5UM and 75UM (Italy),
RKI97/482 (Portugal), and CBS2502 (the Netherlands).
Some of these human patient isolates were closely related
(e.g., IP2005/215 and RKI97/482) or even genetically
indistinguishable (e.g., CBS2502 and RKI85/888).
A large proportion of the human patient and
environmental isolates of C. gattii AFLP4/VGI from Italy
and Spain formed a novel autochthonous Mediterranean
MLST cluster that was genetically homogeneous,
irrespective of their origin or mating type (Figure, online
Technical Appendix Figure). C. gattii AFLP4/VGI was
involved with numerous small outbreaks among goats in
Spain, and the isolates were genetically indistinguishable
from recently obtained human patient, animal, and
environmental isolates from different provinces in Spain as
well as from AFLP4/VGI mating-type a isolates from Italy
(online Technical Appendix Figure).
Several C. gattii genotype AFLP6/VGII infections were
found to have originated in Europe (e.g., the IP1998/1037–
1 and −2, IP2003/125, and CCA242 isolates), and all fell
within a cluster that could not be linked to an environmental
Figure. Maximum-likelihood phylogenetic analysis based on 10-
source. The same holds true for the human patient isolates loci multilocus sequence type data of Cryptococcus gattii isolates
from Greece (AV54S, –W, and IUM01–4731), which came (condensed). Phylogenetic relatedness of 150 STs representing the
from the same patient who had no history of travel outside 291 C. gattii isolates, calculated by using the maximum-likelihood
Greece. algorithm and rooted by using the 2 C. neoformans reference
In the phylogenetic analysis, isolates from human strains CBS8710 (genotype AFLP1/VNI) and CBS10513 (genotype
AFLP2/VNIV). Closely related sequence types were collapsed into
patients in Europe were also observed next to those 1 branch shown by multiple sequence type numbers. Boldface
originating from other geographic areas. The most striking indicates sequence types that are within a shaded area belong
example was a set of 4 human patient isolates from citizens to a specified C. gattii cluster; B, M, PNW, and VIO represent
of Denmark, the Netherlands, Germany, and Switzerland, clusters from Brazil; Mediterranean Europe; the US Pacific
in whom cryptococcosis developed after they had visited Northwest outbreak; and the Vancouver Island, British Columbia,
Canada, outbreak, respectively. AFLP, amplified fragment length
Vancouver Island, British Columbia, Canada. The isolates polymorphism; ST, sequence type. Scale bar indicates number
obtained from these tourists (CBS10485, RKI06/496, of substitutions per site. See online Technical Appendix Figure
RKI01/774, and CBS10866, respectively) had MLST (wwwnc.cdc.gov/EID/pdfs/12-0068-Techapp.pdf) for a detailed
profiles identical to that of C. gattii AFLP6A/VGIIa, phylogenetic analysis.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1621
RESEARCH

which had been obtained from patients who had emigrated Discussion
from Africa to France and which were genetically In Europe, C. gattii has been reported as a rare cause of
indistinguishable from an isolate (IP2001/935–1) from a apparently autochthonous cryptococcal infections (1). The
resident of Senegal (online Technical Appendix Figure). earliest documented case that turned out to be caused by
This set of 5 isolates from France and 1 from Senegal were C. gattii in Europe was made by Curtis in 1896 (20). Until
closely related to isolates from the Vancouver Island and the 1980s, cryptococcosis was rarely observed in Europe
Pacific Northwest outbreaks; however, it seems unlikely and C. gattii infections were especially rare; only 2 cases
that the patient from Senegal had traveled to these outbreak have retrospectively been found (20,22). In the 1980s,
areas, and none of the patients in France reported having infections were reported for 2 immunocompetent citizens
traveled to Canada or the United States. of Germany, who had never traveled abroad (23,24).
A set of C. gattii AFLP4/VGI mating-type α isolates Subsequent case reports described C. gattii infections that
from Portugal (IP1997/18) and Belgium (IHEM19725B were imported or acquired in Europe (25,26). Since 1995,
and –S) was found to be indistinguishable from human the number of reports of C. gattii infections increased and
patient isolates from the Democratic Republic of the describe C. gattii infections in humans and animals from
Congo and Rwanda (B3939 and CBS6289, which were Greece (27), Italy (28,29), and Spain (5,30–33). In the
both mating-type a, and IHEM10602S, IHEM10769S, and current study, 8 (20%) cases were observed until 1995, and
IHEM10769W, which were mating-type α) (Figure; online 32 (80%) cases in humans have been observed since 1995.
Technical Appendix Figure). Both phenotypically different We excluded cases for which an isolate was not available
isolates of IHEM19725 were obtained from an HIV-infected for confirmation. This exclusion is especially relevant
patient from Rwanda who had emigrated to Belgium. The in that we re-identified a reported C. gattii infection as
CBS1622 isolate from Europe, obtained from a patient with actually being caused by C. neoformans (34). These data
the oldest documented case of a C. gattii infection (20), suggest that during the past 2 decades, C. gattii has been
was found to be genetically indistinguishable from a set of emerging in Europe.
isolates from North America. In the current study, 16 (40%) of the human patient
The single C. gattii AFLP5/VGIII isolate from a isolates from Europe were found to have an autochthonous
24-year-old immunocompetent patient from Germany origin in Europe that either could be linked to environmental
(isolate RKI97/310) (21) clustered with human patient isolates from the Mediterranean area or that came from
and environmental isolates from Mexico (online Technical patients who had never traveled outside their resident
Appendix Figure). The 2 C. gattii AFLP7/VGIV isolates country. A total of 24 (60%) isolates could be linked to
CBS7952D and CBS7952S, obtained from an HIV-infected C. gattii–endemic regions in Brazil, the United States,
patient in Sweden, were genetically indistinguishable from Africa, and the Vancouver Island outbreak region. These
each other and had a unique MLST profile that clustered observations demonstrate that C. gattii infections can be
with human patient isolates from Africa. According to the imported subclinically and can cause infections after being
patient’s history, years before the onset of cryptococcal dormant for many years.
infection, she had emigrated from Zambia to Sweden Nearly all isolates from Mediterranean Europe
(online Technical Appendix Figure). Another example belonged to genotype AFLP4/VGI, and MLST analysis
of a reactivated dormant C. gattii infection is that of the showed that these isolates form a separate cluster within
human patient isolate from Italy, IUM92–6682 (AFLP4/ this genotype (Figure; online Technical Appendix Figure)
VGI), obtained from an immunocompetent immigrant from (5). C. gattii genotype AFLP4/VGI has also recently been
Brazil, which had an identical MLST profile to 6 human reported from the environment in the Netherlands, but
patient mating-type α isolates (IHEM14934, IHEM14956, these isolates were not similar to any of the AFLP4/VGI
IHEM14965, IHEM14968, IHEM14976, and IHEM14984) isolates from Mediterranean Europe or to isolate CBS2502,
from Brazil (Figure, online Technical Appendix Figure). which was isolated in 1957 from a pregnant citizen of the
Infection was acquired outside Europe for 24 of these Netherlands, who had never traveled abroad but who died
patients (31 isolates) and within Europe for 16 patients of cryptococcosis (4). Isolate CBS2502 is genotypically
(26 isolates) (Table; online Technical Appendix Figure). identical to isolate RKI85/888, which was isolated from a
Among these 57 human patient isolates, most (47 [82.5%]) previously healthy citizen of Germany, who also had never
were obtained since 1995, the remaining 10 (17.5%) were traveled outside Germany. These 2 cases strongly suggest
isolated during 1895–1994. One of these isolates originated that different genotypes of C. gattii AFLP4/VGI occur in
from 1985, another 2 isolates (CBS1622 and CBS2502), the environment of northwestern Europe because isolates
from 1895 and 1957, were retrospectively found to from patients were genetically different from the recently
represent C. gattii isolates from Europe (20,22). reported isolates from the environment of the Netherlands
(this study; 4).

1622 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Cryptococcus gattii in Europe

In conclusion, C. gattii is emerging in Europe, and 3. Carriconde F, Gilgado F, Arthur I, Ellis D, Malik R, van de
the isolates from Europe can be divided into 5 genotypic Wiele N, et al. Clonality and α-a recombination in the Australian
Cryptococcus gattii VGII population—an emerging outbreak in
clusters. Most C. gattii infections in Europe are probably Australia. PLoS One. 2011;6:e16936. http://dx.doi.org/10.1371/
autochthonous, and several infections are proven to journal.pone.0016936
have been acquired outside the European continent, e.g., 4. Chowdhary A, Randhawa HS, Boekhout T, Hagen F, Klaassen
during visits to C. gattii–endemic regions, such as the CHW, Meis JF. Temperate climate niche for Cryptococcus gattii in
northern Europe. Emerg Infect Dis. 2012;18:172–4. http://dx.doi.
Vancouver Island outbreak area or before migration to org/10.3201/eid1801.111190
Europe from C. gattii–endemic regions in Africa and South 5. Colom MF, Hagen F, Gonzalez A, Mellado A, Morera N, Linares C,
America. Reactivation of dormant C. gattii infections et al. Ceratonia siliqua (carob) trees as natural habitat and source of
and of infections acquired outside Europe after immune infection by Cryptococcus gattii in the Mediterranean environment.
Med Mycol. 2012;50:67–73.
suppression occurs more often than previously assumed. 6. Boekhout T, Theelen B, Diaz M, Fell JW, Hop WC, Abeln EC, et al.
This finding might suggest that C. gattii infections caused Hybrid genotypes in the pathogenic yeast Cryptococcus neoformans.
by certain genotypes are associated with altered immune Microbiology. 2001;147:891–907.
status of the human host. Thus, C. gattii is probably a more 7. Bovers M, Hagen F, Boekhout T. Diversity of the Cryptococcus
neoformans–Cryptococcus gattii species complex. Rev Iberoam
opportunistic pathogen, as has recently been hypothesized, Micol. 2008;25:S4–12. http://dx.doi.org/10.1016/S1130-1406(08)
than a strictly primary pathogen (18,35). C. gattii isolates 70019-6
with genotypes AFLP5/VGIII and AFLP7/VGIV were 8. Hagen F, Illnait-Zaragozi MT, Bartlett KH, Swinne D, Geertsen E,
rarely found in Europe and were all acquired outside the Klaassen CH, et al. In vitro antifungal susceptibilities and amplified
fragment length polymorphism genotyping of a worldwide collection
European continent. However, these genotypes have of 350 clinical, veterinary, and environmental Cryptococcus gattii
frequently been isolated from HIV-infected persons and isolates. Antimicrob Agents Chemother. 2010;54:5139–45. http://
other immunocompromised patients in Africa and the dx.doi.org/10.1128/AAC.00746-10
American continents (1,7,36). A connection seems to exist 9. Meyer W, Aanensen DM, Boekhout T, Cogliati M, Diaz MR,
Esposto MC, et al. Consensus multi-locus sequence typing scheme
between these C. gattii genotypes and the host’s immune for Cryptococcus neoformans and Cryptococcus gattii. Med Mycol.
status. Further epidemiologic and immunologic research is 2009;47:561–70. http://dx.doi.org/10.1080/13693780902953886
needed to unravel this apparent correlation. 10. Simwami SP, Khayhan K, Henk DA, Aanensen DM, Boekhout T,
Hagen F, et al. Low diversity Cryptococcus neoformans variety
grubii multilocus sequence types from Thailand are consistent with
Acknowledgments an ancestral African origin. PLoS Pathog. 2011;7:e1001343. http://
We thank Karen Bartlett, Alf Botha, Eddy Byrnes III, Manuel dx.doi.org/10.1371/journal.ppat.1001343
Cuenca-Estrella, Françoise Dromer, Martine Gari-Toussaint, 11. Byrnes EJ III, Li W, Lewit Y, Ma H, Voelz K, Ren P, et al. Emergence
Jesús Guinea, Joseph Heitman, Connie Jarstrand, Margareth Ip, and pathogenicity of highly virulent Cryptococcus gattii genotypes
in the northwest United States. PLoS Pathog. 2010;6:e1000850.
Greetje Kampinga, Maarten Scholing, and Marianna Viviani for http://dx.doi.org/10.1371/journal.ppat.1000850
providing Cryptococcus spp. isolates. 12. Fraser JA, Giles SS, Wenink EC, Geunes-Boyer SG, Wright JR,
Diezmann S, et al. Same-sex mating and the origin of the Vancouver
This work was conducted at the CBS-KNAW Fungal Island Cryptococcus gattii outbreak. Nature. 2005;437:1360–4.
Biodiversity Centre in Utrecht, the Netherlands. The data are http://dx.doi.org/10.1038/nature04220
available online (www.cbs.knaw.nl/publications/online.aspx). 13. Bovers M, Hagen F, Kuramae EE, Boekhout T. Six monophyletic
lineages identified within Cryptococcus neoformans and
Research performed by F.H. was financially supported by the Odo
Cryptococcus gattii by multi-locus sequence typing. Fungal Genet
van Vloten Foundation, the Netherlands. Biol. 2008;45:400–21. http://dx.doi.org/10.1016/j.fgb.2007.12.004
14. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar
Dr Hagen is a medical molecular microbiologist trainee at S. MEGA5: Molecular Evolutionary Genetics Analysis using
the Canisius-Wilhelmina Hospital in Nijmegen, the Netherlands. maximum likelihood, evolutionary distance, and maximum
His research interests include medical mycology, especially the parsimony methods. Mol Biol Evol. 2011;28:2731–9. http://dx.doi.
development of novel molecular approaches to understand the org/10.1093/molbev/msr121
15. Nylander JAA. MrModeltest v.2. Uppsala: Evolutionary Biology
epidemiology of fungal pathogens. Centre, Uppsala University: 2004 [cited 2012 Aug 9]. http://www.
abc.se/~nylander/
16. Eliades N-G, Eliades DG. Haplotype Analysis: software for analysis
References of haplotype data. Forest Goettingen (Germany): Genetics and
Forest Tree Breeding, Georg-August University Goettingen,. 2009
1. Springer DJ, Chaturvedi V. Projecting global occurrence of [cited 2012 Aug 9]. http://www.uni-goettingen.de/en/134935.html
Cryptococcus gattii. Emerg Infect Dis. 2010;16:14–20. http:// 17. Georgi A, Schneemann M, Tintelnot K, Calligaris-Maibach RC,
dx.doi.org/10.3201/eid1601.090369 Meyer S, Weber R, et al. Cryptococcus gattii meningoencephalitis
2. Kidd SE, Hagen F, Tscharke RL, Huynh M, Bartlett KH, Fyfe M, et in an immunocompetent person 13 months after exposure. Infection.
al. A rare genotype of Cryptococcus gattii caused the cryptococcosis 2009;37:370–3. http://dx.doi.org/10.1007/s15010-008-8211-z
outbreak on Vancouver Island (British Columbia, Canada). Proc Natl
Acad Sci U S A. 2004;101:17258–63. http://dx.doi.org/10.1073/
pnas.0402981101

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1623
RESEARCH

18. Hagen F, van Assen S, Luijckx GJ, Boekhout T, Kampinga GA. 29. Montagna MT, Viviani MA, Pulito A, Aralla C, Tortorano AM,
Activated dormant Cryptococcus gattii infection in a Dutch Fiore L, et al. Cryptococcus neoformans var. gattii in Italy. Note
tourist who visited Vancouver Island (Canada): a molecular II. Environmental investigation related to an autochthonous clinical
epidemiological approach. Med Mycol. 2010;48:528–31. http:// case in Apulia. J Mycol Med. 1997;7:93–6.
dx.doi.org/10.3109/13693780903300319 30. Baró T, Torres-Rodríguez JM, De Mendoza MH, Morera Y, Alía
19. Lindberg J, Hagen F, Laursen A, Stenderup J, Boekhout T. C. First identification of autochthonous Cryptococcus neoformans
Cryptococcus gattii risk for tourists visiting Vancouver Island, var. gattii isloated from goats with predominantly severe pulmonary
Canada. Emerg Infect Dis. 2007;13:178–9. http://dx.doi. disease in Spain. J Clin Microbiol. 1998;36:458–61.
org/10.3201/eid1301.060945 31. Colom MF, Frasés S, Ferrer C, Jover A, Andreu M, Reus S, et al.
20. Curtis F. Contribution a l’étude de la saccharomycose humaine. Ann First case of human cryptococcosis due to Cryptococcus neoformans
Inst Pasteur (Paris). 1896;10:449–68. var. gattii in Spain. J Clin Microbiol. 2005;43:3548–50. http://
21. Grosse P, Tintelnot K, Söllner O, Schmitz B. Encephalomyelitis dx.doi.org/10.1128/JCM.43.7.3548-3550.2005
due to Cryptococcus neoformans var. gattii presenting as spinal 32. Guinea J, Hagen F, Peláez T, Boekhout T, Tahoune H, Torres-
tumour: case report and review of the literature. J Neurol Neurosurg Narbona M, et al. Antifungal susceptibility, serotyping, and
Psychiatry. 2001;70:113–6. http://dx.doi.org/10.1136/jnnp.70.1.113 genotyping of clinical Cryptococcus neoformans isolates collected
22. Janssens J, Beetstra A. Torulosis in a pregnant female, with during 18 years in a single institution in Madrid, Spain. Med Mycol.
localization in the lungs [in Dutch]. Ned Tijdschr Geneeskd. 2010;48:942–8. http://dx.doi.org/10.3109/13693781003690067
1957;101:824–6. 33. Solla I, Morano LE, Vasallo F, Cuenca-Estrella M. Cryptococcus
23. Kohl KH, Hof H, Schrettenbrunner A, Seeliger HP, Kwon-Chung gattii meningitis: observation in a Spanish patient [in Spanish].
KJ. Cryptococcus neoformans var. gattii in Europe. Lancet. Enferm Infecc Microbiol Clin. 2008;26:395–6. http://dx.doi.
1985;325:1515. http://dx.doi.org/10.1016/S0140-6736(85)92300-1 org/10.1157/13123846
24. Schaberg T, Mai J, Thalmann U, Seibold M, Staib F. Cryptococcoma 34. Bodasing N, Seaton RA, Shankland GS, Kennedy D. Cryptococcus
of the lung—a contribution to diagnosis and therapy [in German]. neoformans var. gattii meningitis in an HIV-positive patient: first
Internist (Berl). 1988;29:510–5. observation in the United Kingdom. J Infect. 2004;49:253–5. http://
25. Dromer F, Ronin O, Dupont B. Isolation of Cryptococcus neoformans dx.doi.org/10.1016/j.jinf.2003.06.001
var. gattii from an Asian patient in France: evidence for dormant 35. MacDougall L, Fyfe M, Romney M, Starr M, Galanis E. Risk factors
infection in healthy subjects. J Med Vet Mycol. 1992;30:395–7. for Cryptococcus gattii infection, British Columbia, Canada. Emerg
http://dx.doi.org/10.1080/02681219280000511 Infect Dis. 2011;17:193–9.
26. Dromer F, Mathoulin S, Dupont B, Laporte A. Epidemiology of 36. Litvintseva AP, Thakur R, Reller LB, Mitchell TG. Prevalence of
cryptococcosis in France: a 9-year survey (1985–1993). French clinical isolates of Cryptococcus gattii serotype C among patients
Cryptococcosis Study Group. Clin Infect Dis. 1996;23:82–90. http:// with AIDS in sub-Saharan Africa. J Infect Dis. 2005;192:888–92.
dx.doi.org/10.1093/clinids/23.1.82 http://dx.doi.org/10.1086/432486
27. Velegraki A, Kiosses VG, Pitsouni H, Toukas D, Daniilidis VD,
Legakis NJ. First report of Cryptococcus neoformans var. gattii Address for correspondence: Ferry Hagen, Canisius-Wilhelmina Hospital,
serotype B from Greece. Med Mycol. 2001;39:419–22.
Department of Medical Microbiology and Infectious Diseases, Weg door
28. Iatta R, Hagen F, Fico C, Lopatriello N, Boekhout T, Montagna MT.
Cryptococcus gattii infection in an immunocompetent patient from Jonkerbos 100, NL-6532 SZ Nijmegen, the Netherlands; email: f.hagen@
southern Italy. Mycopathologia. 2012 ;174:87–92. gmail.com

1624 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Echinococcus more abundant than foxes. E. multilocularis was reported
in 7 (4.1%) of 171 coyotes in the northcentral United States
multilocularis in in the late 1960s (3), and subsequently prevalences ranging
from 19.0% to 35.0% have been reported in coyotes in the
Urban Coyotes, central United States (4).

Alberta, Canada
In Canada, E. multilocularis was detected in 10
(23.0%) of 43 coyotes in Riding Mountain National
Park, Manitoba (5). In Alberta, 1 case was recorded from
Stefano Catalano,1 Manigandan Lejeune,1 the aspen parkland in 1973 (5) but it was not found in
Stefano Liccioli, Guilherme G. Verocai, coyotes from forested regions and southern prairies (6,7).
Karen M. Gesy, Emily J. Jenkins, Susan J. Kutz, Nonetheless, E. multilocularis is generally considered
Carmen Fuentealba, Padraig J. Duignan, enzootic to central and southern Alberta on the basis of
and Alessandro Massolo its prevalence in rodent intermediate hosts. During the
1970s, sixty-three (22.3%) of 283 deer mice (Peromyscus
Echinococcus multilocularis is a zoonotic parasite maniculatus) trapped in periurban areas of Edmonton were
in wild canids. We determined its frequency in urban positive for alveolar hydatid cysts (8), and E. multilocularis
coyotes (Canis latrans) in Alberta, Canada. We detected
was also detected in 2 deer mice collected <1.8 km from
E. multilocularis in 23 of 91 coyotes in this region. This
parasite is a public health concern throughout the Northern
Lethbridge in southern Alberta (9).
Hemisphere, partly because of increased urbanization of Because mice and voles (family Cricetidae, including
wild canids. Peromyscus spp.) have been reported as main prey (70.1%)
of coyotes in Calgary (10), and coyotes are common in
urban areas of Calgary and Edmonton, we suspected a role
chinococcus multilocularis is the causative agent of
E alveolar echinococcosis in humans. This disease is
a serious problem because it requires costly long-term
for this carnivore in the maintenance of the wild animal
cycle of E. multilocularis in such urban settings. Thus, we
aimed to ascertain the frequency of E. multilocularis in
therapy, has high case-fatality rate, and is increasing coyotes from metropolitan areas in Alberta, Canada.
in incidence in Europe (1). This parasitic cestode has a
predominantly wild animal cycle involving foxes (Vulpes The Study
spp.) and other wild canids, including coyotes (Canis Ninety-one hunted or road-killed coyotes were
latrans), as definitive hosts. However, it can also establish collected during October 2009–July 2011. Most (n
an anthropogenic life cycle in which dogs and cats are the = 83) of the carcasses were from the Calgary census
final hosts. Rodents are the primary intermediate hosts in metropolitan area (CMA) (Figure 1). The remainder (n =
which the alveolar/multivesicular hydatid cysts grow and 8) were opportunistically collected from the Edmonton
are often fatal. Humans are aberrant intermediate hosts for CMA. Of those from the Calgary CMA, the exact location
E. multilocularis (2). of collection was known for 60 animals: 27 were from
In North America, E. multilocularis was believed to Calgary and 33 were from the rural fringe, including 2 near
be restricted to the northern tundra zone of Alaska, USA, Strathmore. Of the carcasses from the Edmonton CMA, 7
and Canada until it was reported in red foxes (Vulpes were from Edmonton and 1 was from a periurban site. Sex
vulpes) from North Dakota, USA (3). This parasite has and age of 90 of the coyotes were recorded.
now been reported in the southern half of 3 provinces in Before necropsy, all carcasses were stored at −20°C.
Canada (Manitoba, Saskatchewan, and Alberta) and in 13 Gastrointestinal tracts collected at necropsy were refrozen
contiguous states in the United States (1). at −80°C for 3–5 days to inactivate Echinococcus spp.
Foxes are the traditional definitive hosts for E. eggs. Once thawed and dissected, intestinal contents were
multilocularis worldwide. However, in North America, washed, cleared of debris, and passed through a sieve (500-
coyotes may be prominent hosts, particularly when they are μm pores), and the material in the sieve was examined for
Echinococcus spp.
Author affiliations: University of Calgary, Calgary, Alberta, Canada
Adult tapeworms were counted and identified as E.
(S. Catalano, M. Lejeune, S. Liccioli, G.G. Verocai, S.J. Kutz, P.J.
multilocularis on the basis of morphologic features (Figure
Duignan, A. Massolo); University of Saskatchewan, Saskatoon,
2). To confirm morphologic identification, PCR was
Saskatchewan, Canada (K.M. Gesy, E.J. Jenkins); and Ross
performed by using species-specific primers (11). Briefly,
University, Basseterre, Saint Kitts, Saint Kitts, and Nevis (C.
a representative adult worm from each positive animal was
Fuentealba)

DOI: http://dx.doi.org/10.3201/eid1810.120119 1
These authors contributed equally to this article.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1625
DISPATCHES

Figure 1. Calgary and Edmonton,

Alberta, Canada, census metropolitan
areas in which 91 coyote carcasses
were collected during 2009–2011 and
tested for Echinococcus multilocularis.
Reference maps (2006) were obtained
from the Geography Division, Statistics
Canada (www12.statcan.gc.ca/census-
recensement/2006/geo/index-eng.cfm).
Urban core areas and surrounding rural
fringes are indicated. For Edmonton, 5
(62.5%) of 8 carcasses were positive.
For Calgary, 18 (20.5%) of 83 carcasses
were positive: 9 (27.3%) of 33 from the
rural fringe, 4 (14.8%) of 27 from the
urban area, and 5 (21.7%) of 23 whose
locations of collection were not accurate
enough to be classified as urban or from
the rural fringe.

lysed in 50 μL of DNA extraction buffer (500 mmol/L Conclusions

KCl, 100 mmol/L Tris-HCl, pH 8.3, 15 mmol/L MgCl2, We demonstrated that E. multilocularis is common in
10 mol/L dithiothreitol, and 4.5% Tween 20) containing 2 coyotes of metropolitan areas of Calgary and Edmonton,
μL of proteinase K. This lysate was further diluted (1:20 Alberta, Canada, including their urban cores. This finding
in double-distilled water), and 2 μL was used for PCR. might indicate an emerging phenomenon similar to that
Amplicons of an expected 395 bp confirmed infection with observed in Europe with infiltration of urban centers by E.
E. multilocularis. multilocularis caused by an increase in red foxes in cities
E. multilocularis was identified in 23 (25.3%) of 91 such as Copenhagen, Geneva, and Zurich (2). In Alberta,
coyotes by using morphologic and molecular identification.
Among positive animals, 18 (20.5%) of 83 were from the
Calgary CMA and 5 (62.5%) of 8 were from the Edmonton
CMA. In the Calgary CMA, 4 (14.8%) of 27 positive
animals were found in the city and 9 (27.3%) of 33 were
found in the rural fringe (Figure 1). Five (21.73%) of 23
coyotes for which the location was not recorded were also
positive.
E. multilocularis intensity (number of cestodes per
host) ranged from 1 to 1,400 (median 20.5). The frequency
of infection was significantly higher in male coyotes (n
= 44, 34.19%) than in female coyotes (n = 46, 15.2%; Figure 2. Differential interference contrast micrograph of a
χ2 4.337, df 1, Pexact = 0.05) (Table). No difference was representative Echinococcus multilocularis isolate from a coyote
carcass in Alberta, Canada, October 2009–July 2011. The parasite
detected between 43 juvenile coyotes and 47 adult coyotes
was 2,059.72 μm long, as measured by using an Olympus BX53
(Table). microscope and software (http://microscope.olympus-global.com/
en/ga/product/bx53/sf04.cfm). Scale bar = 200 μm.

1626 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 E. multilocularis in Coyotes, Alberta, Canada

Table. Echinococcus multilocularis in coyotes carcasses collected in Calgary (n = 83) and Edmonton (n = 8) census metropolitan
areas, Alberta, Canada, October 2009–July 2011*
No. (%) positive or
2
Characteristic Total median (range) No. negative IQ distance Ȥ (z) df pexact value†
Sex‡
M 44 15 (34.1) 29 NA
F 46 7 (15.2) 39 NA 4.337 1 0.05
Parasite intensity
M NA 9 (1–1,400) NA 83
F NA 59 (9–822) NA 137 1.406) 0.19
Age‡
Juvenile 43 14 (33.3) 29 NA
Adult 47 8 (17.0) 39 NA 1.661 1 0.226
Parasite intensity
Juvenile NA 9 (1–151) NA 71
Adult NA 32 (1–1,400) NA 520 0.737) 0.518
*Values in boldface indicate a significant difference. Higher prevalence in male coyotes suggests a role for sex in parasite dispersion. Frequencies of
cestodes in males vs. females and juveniles vs. adults were analyzed by using Ȥ2 test. Parasite intensity (no. parasites per host) among sex and age
classes was compared by using Mann-Whitney test for independent samples. IQ, interquartile distance; NA, not applicable.
†Probability of distribution was estimated by using the permutation approach (pexact).
‡Sex and age of 1 coyote were not recorded.

this phenomenon may be caused by coyotes occupying target veterinarians and dog owners because domestic dogs
the urban landscape or by city sprawl invading the natural probably represent the main infection route for humans
habitats of coyotes. in North America (2,12). Genetic characterization of E.
Our data suggest that E. multilocularis is enzootic in multilocularis and spatially explicit transmission models
coyotes in Alberta and that perpetuation of the wild animal should also be developed to better assess risks of this
cycle of E. multilocularis within cities and surroundings emerging zoonosis in North America and worldwide.
and potential infection of domestic dogs may pose a
zoonotic risk, as documented on Saint Lawrence Island, Acknowledgments
Alaska, and in China (2,12). With a considerable increase in We thank Bill Bruce, officers of the Animal and Bylaw
domestic dog population of Calgary (32.1% increase since Services of the city of Calgary, officers of the Alberta Fish and
2001, a total of 122,325 dogs in 2010; Animal and Bylaw Wildlife and the Alberta Provinicial Parks, the Municipality
Services Survey 2010, www.calgary.ca/CSPS/ABS/Pages/ District of Rocky View, City Roads of the City of Calgary,
home.aspx) and substantial human population growth Colleen St. Claire and collaborators, and Mark Edwards for
(32.9% increase in Calgary since 1999; Statistics Canada, carcass collection; and Margo Pybus, Susan Cork, and William
2009, www.statcan.gc.ca/start-debut-eng.html), awareness Samuel for helpful suggestions.
is needed of potential transmission risks associated with
This study and A.M. were supported by the Animal and
changing city landscapes and E. multilocularis in the urban
Bylaw Services of the city of Calgary and the Faculty of Veterinary
environment.
Services of the University of Calgary. S.J.K was supported by
In Canada, only 1 autochthonous human case of
Natural Sciences and Engineering Research Council of Canada.
alveolar echinococcosis has been reported in Manitoba (13).
However, imported cases have been described. In Alberta, Mr Catalano is enrolled in the MSc graduate program at the
there are no known reports of alveolar echinococcosis. This Faculty of Veterinary Medicine, University of Calgary, Alberta,
finding may be caused by the long incubation time required Canada. His research interests include wildlife diseases and the
for clinical manifestation in humans (12) or a strain of E. ecology of parasites in wild animal communities.
multilocularis with a low zoonotic potential. Although
there is little evidence of human risk from the strain of E.
References
multilocularis in central North America (14), a human case
caused by this strain has been confirmed (15). 1. Moro P, Schantz PM. Echinococcosis: a review. Int J Infect Dis.
Our finding of E. multilocularis in coyotes in urban 2009;13:125–33. http://dx.doi.org/10.1016/j.ijid.2008.03.037
regions in Alberta suggests that surveillance for this parasite 2. Eckert J, Deplazes P. Biological, epidemiological, and clinical
aspects of echinococcosis, a zoonosis of increasing concern.
should be increased in North America. Although removal
Clin Microbiol Rev. 2004;17:107–35. http://dx.doi.org/10.1128/
of this parasite from domestic dogs and cats is effective, CMR.17.1.107-135.2004
eradication from free-ranging definitive hosts may be 3. Leiby PD, Carney WP, Woods CE. Studies on sylvatic
unfeasible (2,12). Interventions other than improving echinococcosis. III. Host occurrence and geographic distribution
of Echinococcus multilocularis in the north central United States. J
public awareness about prevention and transmission risk are
Parasitol. 1970;56:1141–50. http://dx.doi.org/10.2307/3277560
probably unnecessary, and public health messages should

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1627
DISPATCHES

4. Kazacos KR. Cystic and alveolar hydatid disease caused by 11. Trachsel D, Deplazes P, Mathis A. Identification of taeniid eggs in
Echinococcus species in the contiguous United States. Compendium the faeces from carnivores based on multiplex PCR using targets in
on Continuing Education for the Practising Veterinarian. 2003;25 mitochondrial DNA. Parasitology. 2007;134:911–20. http://dx.doi.
(Suppl. 7A):16–20. org/10.1017/S0031182007002235
5. Samuel WM, Ramalingam S, Carbyn L. Helminths of coyotes 12. Deplazes P, Knapen FV, Schweiger A, Overgaauw PA. Role of pet
(Canis latrans Say), wolves (Canis lupus L.) and red foxes (Vulpes dogs and cats in the transmission of helminthic zoonoses in Europe,
vulpes L.) of southwestern Manitoba. Can J Zool. 1978;56:2614–7. with a focus on echinococcosis and toxocarosis. Vet Parasitol.
http://dx.doi.org/10.1139/z78-351 2011;182:41–53. http://dx.doi.org/10.1016/j.vetpar.2011.07.014
6. Holmes J, Podesta R. The helminths of wolves and coyotes from 13. James E, Boyd W. Echinococcus alveolaris: (with the report of a
the forested regions of Alberta. Canadian Journal of Zoology. case). Can Med Assoc J. 1937;36:354–6.
1968;46:1193–203. http://dx.doi.org/10.1139/z68-169 14. Hildreth MB, Sriram S, Gottstein B, Wilson M, Schantz PM. Failure
7. Thompson RC, Colwell D, Shury T, Appelbee A, Read C, Njiru Z, to identify alveolar echinococcosis in trappers from South Dakota
et al. The molecular epidemiology of Cryptosporidium and Giardia in spite of high prevalence of Echinococcus multilocularis in wild
infections in coyotes from Alberta, Canada, and observations on canids. J Parasitol. 2000;86:75–7.
some cohabiting parasites. Vet Parasitol. 2009;159:167–70. http:// 15. Yamasaki H, Nakao M, Nakaya K, Schantz PM, Ito A. Genetic
dx.doi.org/10.1016/j.vetpar.2008.10.003 analysis of Echinococcus multilocularis originating from a patient
8. Holmes JC, Mahrt JL, Samuel WM. The occurrence of Echinococcus with alveolar echinococcosis occurring in Minnesota in 1977. Am J
multilocularis Leuckart, 1863 in Alberta. Can J Zool. 1971;49:575– Trop Med Hyg. 2008;79:245–7.
6. http://dx.doi.org/10.1139/z71-090
9. Chalmers GA, Barrett MW. Echinococcus multilocularis Leuckart, Address for correspondence: Alessandro Massolo, Department of
1863 in rodents in southern Alberta. Can J Zool. 1974;52:1091.
Ecosystem and Public Health, University of Calgary, 3280 Hospital Dr
http://dx.doi.org/10.1139/z74-145
10. Lukasik VM, Alexander SM. Coyote diet and conflict in urban parks NW, Calgary AB T2N 4Z6, Canada; email: amassolo@ucalgary.ca
in Calgary, Alberta. Presented at: Canadian Parks for Tomorrow,
40th Anniversary Conference, May 8–11, 2008; University of
Calgary, Calgary, Alberta, Canada.

1628 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Orthobunyavirus activity in the Yucatan Peninsula (1–4), we investigated
whether orthobunyaviruses commonly infect humans in
Antibodies in this region.

Humans, Yucatan The Study

Peninsula, Mexico
Serum samples were obtained from 823 febrile
patients at the Secretaria de Salud de Yucatán and other
health institutions in Merida during January–October
Bradley J. Blitvich, Rungrat Saiyasombat, 2007. The patients resided in all 3 states of the Yucatan
Lourdes G. Talavera-Aguilar, Peninsula of Mexico: Yucatan (n = 809), Quintana Roo (n
Julian E. Garcia-Rejon, Jose A. Farfan-Ale, = 8) and Campeche (n = 6). The study was approved by the
Carlos Machain-Williams, Institutional Biosafety Committees at Iowa State University
and Maria A. Loroño-Pino (Ames, IA, USA) and the Universidad Autónoma de
Yucatán (Mérida, Mexico).
We performed a serologic investigation to determine All serum samples were examined at a dilution of 1:20
whether orthobunyaviruses commonly infect humans in by plaque reduction neutralization test (PRNT) by using
the Yucatan Peninsula of Mexico. Orthobunyavirus-specific
CVV (strain CVV-478), and PRNTs were performed as
antibodies were detected by plaque reduction neutralization
test in 146 (18%) of 823 persons tested. Further studies
described (7). A subset of serum samples with antibodies
are needed to determine health risks for humans from this that neutralized CVV were titrated and further analyzed by
potentially deadly group of viruses. PRNT by using CVV, CHLV (strain CHLV-Mex07), KRIV
(strain KRIV-Mex07), SOURV (strain NJO-94f), Maguari
virus (strain BeAr7272), and Wyeomyia virus (strain

W e previously reported the isolation of Cache Valley

virus (CVV), Kairi virus (KRIV), Cholul virus
(CHLV), and South River virus (SOURV) from mosquitoes
prototype). All of these viruses belong to the Bunyamwera
(BUN) serogroup except SOURV, which belongs to the
California (CAL) serogroup.
in the Yucatan Peninsula of Mexico (1–3). Antibodies to Titers were expressed as the reciprocal of highest
CVV, CHLV, and SOURV were also detected in livestock serum dilutions yielding >90% reduction in the number
in this region (4). These viruses belong to the genus of plaques (PRNT90). For etiologic diagnosis, the PRNT90
Orthobunyavirus (5). All viruses in this genus possess a antibody titer for each virus was required to be >4-fold
tripartite, single-stranded, negative-sense RNA genome. greater than that to the other viruses tested.
CVV is a recognized human pathogen (5) that has Antibodies that neutralized CVV were detected in 146
been linked to severe encephalitis and multiorgan failure. (18%) of 823 study participants. The mean ages of patients
KRIV has not been implicated as a cause of human with and without antibodies that neutralized CVV were 32.0
disease, although antibodies to this virus have been and 22.3 years, respectively. Logistic regression analysis
detected in humans in Argentina (6). Recent data suggest showed that the risk for infection increased significantly
that CHLV is a reassortant that acquired its small RNA with age (p = 0.0001).
segment from CVV and medium and large RNA segments Serum samples from 50 seropositive patients were
from Potosi virus (POTV) (1). No clear evidence exists titrated and analyzed by comparative PRNT to identify
for human susceptibility to infection with CHLV or the orthobunyaviruses responsible for these infections.
SOURV. However, diagnostic laboratories rarely test Six persons were seropositive for CVV, 5 for CHLV, and
for orthobunyavirus infection; therefore, the true disease 1 for SOURV or a SOURV-like virus; 38 had antibodies
incidence and seroprevalence of these viruses remains to an undetermined orthobunyavirus (Table). Because
to be determined. Because orthobunyaviruses comprise SOURV was the only CAL serogroup virus used in this
a neglected but potentially deadly group of viruses and study, and another CAL serogroup virus may have been
recent studies have provided evidence of orthobunyavirus responsible for the infection, the person who had a SOURV
PRNT90 titer >4-fold than that to the other viruses tested
Author affiliations: Iowa State University College of Veterinary received a conservative PRNT diagnosis of seropositive for
Medicine, Ames, Iowa, USA (B.J. Blitvich, R Saiyasombat); and SOURV or a SOURV-like virus. Because interserogroup
Universidad Autónoma de Yucatán Centro de Investigaciones crossreactivity of neutralizing antibodies to viruses in
Regionales Dr. Hideyo Noguchi, Mérida, México (L.G. Talavera- the BUN and CAL serogroups has not been seen, the 17
Aguilar, J.E. Garcia-Rejon, J.A. Farfan-Ale, C. Machain-Williams, persons with antibodies that neutralized SOURV and >1 of
M.A. Loroño-Pino) the BUN serogroup viruses might have been exposed to >1
DOI: http://dx.doi.org/10.3201/eid1810.120492 viruses from each serogroup.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1629
DISPATCHES

CHLV and POTV share the same medium RNA Peninsula, whereas no direct evidence has been found for
segment, so antibodies for these viruses cannot be POTV in this region.
differentiated by PRNT. Furthermore, antibodies to CHLV As already noted, serum samples from 38 (76%) of
and POTV cannot be differentiated by complement fixation the study participants analyzed by comparative PRNT
test (4). Thus, we cannot dismiss POTV as a possible had antibodies to an undetermined orthobunyavirus.
cause of infection in some or all of the study participants Most of these persons had low PRNT90 titers; the highest
who were seropositive for CHLV. However, it appears PRNT90 titer for 29 of these persons did not exceed 40.
more likely that these persons had been infected with Because neutralizing antibody levels decline over time,
CHLV because this virus has been isolated in the Yucatan these findings may indicate that many of these infections

Table. Endpoint titers of serum samples collected from persons in Mexico and analyzed by using comparative PRNT*
Patient Demographic characteristics PRNT90 titers
ID no. Residence Age, y/sex CVV CHLV KRIV SOURV MAGV WYOV Diagnosis
28 Yucatan 18/M 80 20 – – 40 – UND
34 Yucatan 39/F 80 20 – – 20 – CVV
52 Quintana Roo 39/F 160 – – 40 80 – UND
54 Yucatan 32/M 40 20 – – 40 – UND
62 Yucatan 44/M 40 – – 20 20 – UND
72 Yucatan 23/M 80 – – – 40 – UND
81 Yucatan 60/F 20 20 – 20 – – UND
92 Yucatan 13/M 160 20 – 20 80 – UND
93 Yucatan 42/F 40 20 – – 20 – UND
113 Yucatan 24/F 20 – – – – – UND
114 Yucatan 29/F 40 – – – 20 – UND
120 Yucatan 60/F 20 – – 320 – – SOURV or
SOURV-like virus
159 Yucatan 54/M 20 – – – 20 – UND
161 Yucatan 53/M 20 20 – – – – UND
163 Yucatan 27/F 160 80 40 – 40 – UND
167 Yucatan 16/M 160 – – – 40 – CVV
183 Yucatan 69/F 20 – – – 20 – UND
184 Yucatan 34/M 160 40 – – 80 – UND
185 Yucatan 25/F 20 – – – – – UND
192 Campeche 54/F 80 40 – – 20 – UND
193 Yucatan 16/F 80 – – – 20 – CVV
194 Yucatan 69/F 20 – – 20 – – UND
200 Yucatan 3/F 40 – – 40 40 – UND
205 Yucatan 53/M 40 – – 20 20 – UND
208 Yucatan 57/F 20 160 – – – – CHLV
210 Yucatan 42/M 20 – – – 20 – UND
224 Yucatan 34/M 20 20 – 20 – – UND
234 Yucatan 39/F 20 80 – – 20 – CHLV
236 Yucatan 74/F 20 – – 20 – – UND
386 Yucatan 14/F 20 – – 20 – – UND
388 Yucatan 60/M 160 1,280 160 – 40 – CHLV
389 Yucatan 5/M 20 – – 20 – – UND
390 Yucatan 33/M 40 20 40 20 40 – UND
392 Yucatan 22/M 20 – – – – – UND
393 Yucatan 29/M 40 20 – – 40 – UND
396 Yucatan 34/F 80 – – – 20 – CVV
397 Yucatan 32/M 80 – – – 40 – UND
399 Yucatan 27/M 320 1,280 160 – 320 – CHLV
400 Yucatan 37/F 20 – – – – – UND
401 Yucatan 30/F 160 – – – 40 – CVV
402 Yucatan 18/M 20 – – – 20 – UND
403 Yucatan 50/F 20 20 – – – – UND
407 Yucatan 27/M 80 20 – – 40 – UND
408 Yucatan 40/F 20 – – – – – UND
412 Yucatan 60/M 20 320 80 40 40 – CHLV
415 Yucatan 17/F 20 – – 20 – – UND
420 Yucatan 16/F 160 – – – 20 – CVV
429 Yucatan 37/F 20 – – – – – UND
442 Yucatan 10/F 20 40 – 20 – – UND
455 Yucatan 30/F 20 – – 20 20 – UND
*PRNT, plaque reduction neutralization test; CVV, Cache Valley virus; CHLV, Cholul virus; KRIV, Kairi virus; SOURV, South River virus; MAGV, Maguari
virus, WYOV, Wyeomyia virus; –, titer <20; UND, undetermined orthobunyavirus.

1630 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Orthobunyavirus Antibodies in Humans, Mexico

occurred years ago, and the trace amounts of neutralizing orthobunyaviruses represent an unrecognized cause of
antibodies that remained were insufficient to yield a >4- illness in humans and vertebrate animals in Mexico.
fold difference between the titers of the virus responsible
for the infection and the other viruses used in the PRNTs. Acknowledgments
Another explanation is that some of these persons had We thank Amanda J. Panella for providing labile serum
been infected with an orthobunyavirus not included in the factor; Robert B. Tesh for providing isolates of SOURV, Maguari
PRNTs, although all orthobunyaviruses known to occur in virus, and Wyeomyia virus; and Nubia Rivero-Cárdenas for
the Yucatan Peninsula were represented. technical assistance with the serum sample collections.
This study was supported by the Iowa State University Plant
Conclusions
Sciences Institute Virus–Insect Interactions Initiative.
We found 18% of the 823 Yucatan residents
participating in our study had evidence of orthobunyavirus Dr Blitvich is an associate professor in the College of
exposure. This number is presumably an underestimate; Veterinary Medicine at Iowa State University, Ames, Iowa.
additional seropositive persons might have been identified His research interests are the mechanisms of vector and host
if the initial PRNTs had not been restricted to CVV. In interactions in arbovirus transmission cycles.
particular, additional seropositive persons likely would
have been detected if SOURV was used in the initial
References
PRNTs, because a screening algorithm that includes only
a BUN serogroup virus would likely miss many CAL 1. Blitvich BJ, Saiyasombat R, Dorman KD, Garcia-Rejon JE, Farfan-
serogroup virus infections. Nevertheless, we provide Ale JA, Lorono-Pino MA. Sequence and phylogenetic data indicate
evidence that orthobunyaviruses commonly infect humans that an orthobunyavirus recently detected in the Yucatan Peninsula
of Mexico is a novel reassortant of Potosi and Cache Valley viruses.
in the Yucatan Peninsula.
Arch Virol. 2012;157:1199–204. http://dx.doi.org/10.1007/s00705-
Previous serosurveys have provided information on 012-1279-x
the seroprevalence of orthobunyaviruses in humans in the 2. Farfan-Ale JA, Lorono-Pino MA, Garcia-Rejon JE, Hovav E,
United States. For example, antibodies that neutralized Powers AM, Lin M, et al. Detection of RNA from a novel West
Nile–like virus and high prevalence of an insect-specific flavivirus
CVV were detected in 42/356 (12%) residents in Maryland
in mosquitoes in the Yucatan Peninsula of Mexico. Am J Trop Med
and Virginia in 1961–1963 (8). Antibodies that neutralized Hyg. 2009;80:85–95.
Maguari virus or Tensaw virus were detected in 71/≈300 3. Farfan-Ale JA, Lorono-Pino MA, Garcia-Rejon JE, Soto V, Lin M,
humans in Florida in the 1980s (9); as observed in our Staley M, et al. Detection of flaviviruses and orthobunyaviruses
in mosquitoes in the Yucatan Peninsula of Mexico in 2008. Vector
study, the highest PRNT titers for most of the seropositive
Borne Zoonotic Dis. 2010;10:777–83. http://dx.doi.org/10.1089/
persons in that study did not exceed 40. vbz.2009.0196
All persons in our study cohort initially sought care 4. Blitvich BJ, Saiyasombat R, Travassos da Rosa A, Tesh RB, Calisher
for unspecified fever; however, we could not determine CH, Garcia-Rejon JE, et al. Orthobunyaviruses are a common cause
of infection of livestock in the Yucatan Peninsula of Mexico. Am J
whether any of these febrile illnesses were a direct
Trop Med Hyg. In press.
consequence of orthobunyavirus infection. The detection 5. Schmaljohn CS, Nichol ST. Bunyaviridae. In: Knipe DM, Howley
of acute orthobunyavirus infections is limited because no PM, editors. Fields virology, 5th ed. Philadelphia: Lippincott
IgM-capture ELISA for orthobunyavirus diagnosis exists. Williams & Wilkins; 2007.
6. Tauro LB, Almeida FL, Contigiani MS. First detection of human
PRNTs can be used to identify recent orthobunyavirus
infection by Cache Valley and Kairi viruses (Orthobunyavirus) in
infections when paired acute and convalescent serum Argentina. Trans R Soc Trop Med Hyg. 2009;103:197–9. http://
samples are available, but for our study, only single dx.doi.org/10.1016/j.trstmh.2008.09.004
serum samples were available from each participant. 7. Rodríguez ML, Rodriguez DR, Blitvich BJ, Lopez MA, Fernandez-
Salas I, Jimenez JR, et al. Serologic surveillance for West Nile
Orthobunyavirus viremias in humans are transient and of
virus and other flaviviruses in febrile patients, encephalitic patients,
low magnitude, which makes reverse transcription PCR and asymptomatic blood donors in northern Mexico. Vector
ineffective for the detection of orthobunyavirus RNA in Borne Zoonotic Dis. 2010;10:151–7. http://dx.doi.org/10.1089/
serum samples. However, a duplex reverse transcription vbz.2008.0203
8. Buescher EL, Byrne RJ, Clarke GC, Gould DJ, Russell PK, Scheider
PCR was recently developed for the detection of CVV
FG, et al. Cache Valley virus in the Del Mar Va Peninsula. I.
RNA in human cerebrospinal fluid (10). Virologic and serologic evidence of infection. Am J Trop Med Hyg.
In conclusion, we provide evidence that 1970;19:493–502.
orthobunyaviruses commonly infect humans in the 9. Calisher CH, Lazuick JS, Lieb S, Monath TP, Castro KG. Human
infections with Tensaw virus in south Florida: evidence that Tensaw
Yucatan Peninsula. These viruses are also a common
virus subtypes stimulate the production of antibodies reactive with
cause of infection of livestock in this region (4). Our closely related Bunyamwera serogroup viruses. Am J Trop Med
findings underscore the need to determine whether Hyg. 1988;39:117–22.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1631
DISPATCHES

10. Wang HS, Nattanmai S, Kramer LD, Bernard KA. Tavakoli NP. A Address for correspondence: Bradley J. Blitvich, 2116 Veterinary
duplex real-time reverse transcriptase polymerase chain reaction Medicine, Iowa State University, Ames, IA 50011, USA; email: blitvich@
assay for the detection of California serogroup and Cache Valley
iastate.edu
viruses. Diagn Microbiol Infect Dis. 2009;65:150–7. http://dx.doi.
org/10.1016/j.diagmicrobio.2009.07.001

1632 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Tetanus as Cause season (November and December). Thus, the observed
number of presumed tetanus cases during the 2008 breeding
of Mass Die-off of season (9/60) was 8.4× greater than the number during the
2007 breeding season (1/56).
Captive Japanese The soil in the monkeys’ enclosure was clay-like

Macaques, Japan,
and without vegetation. In 2008 before the increase in
deaths, there were no changes in maintenance procedures,

2008
such as feeding, at the facility and no evident pathogenic
contamination of the monkeys’ food or environment.
We performed necropsies on 3 monkeys (animal nos.
Tomomi Nakano, Shin-ichi Nakamura, 1, 2, and 3) that died 5, 2, and 3 days, respectively, after
Akihiko Yamamoto, Motohide Takahashi, the onset of symptoms. At death, all showed a specific
and Yumi Une posture: the jaw was elevated, the back straightened, and
the tail tightly stretched; the forelimbs were crossed in front
In 2008 in Japan, 15/60 captive Japanese macaques
of the body with the wrists bent; and the hind limbs were
died. Clostridium tetani was isolated from 1 monkey, and
11 had tetanus-specific symptoms. We conclude the
extended backward (Figure 2). Rigidity was abnormally
outbreak resulted from severe environmental C. tetani severe and did not remit after death; at necropsy, the mouth
contamination. Similar outbreaks could be prevented by was difficult to open. Congestion of the visceral organs and
vaccinating all monkeys, disinfecting housing areas/play pulmonary edema were noted, but there were no findings
equipment, replacing highly C. tetani–contaminated soil, to suggest poisoning, such as foreign bodies in the stomach
and conducting epidemiologic surveys. or erosive changes in the gastrointestinal tract. No wound
that might have led to infection was found in monkeys 1 or
2, but a lesion with purulent incrustation was present on a
T etanus is a wound infection caused by a potent neuro-
toxin produced by Clostridium tetani. The bacterium
is difficult to isolate, and no pathologically characteristic
toe tip on the right hind limb of monkey 3. C. tetani was
isolated from this lesion, and the tetanus toxin gene was
detected by PCR. A mouse toxicity test confirmed tetanus
lesion is present during infection; thus, tetanus diagnosis
toxin activity.
is based on tetanus-specific clinical symptoms (1–4).
We obtained samples from the soil in monkey
Tetanus is a highly lethal zoonosis, and cases usually occur
enclosures, from wooden playground equipment, and
sporadically. Outbreaks among humans have occurred
from the soil surrounding the enclosures and tested
only after earthquakes and tsunamis (4). We report on an
them for C. tetani; 67%, 75%, and 53% of the samples,
outbreak of tetanus in 2008 among a captive colony of
respectively, were positive for C. tetani, indicating marked
Japanese macaques (Macaca fuscata) in Japan.

The Study
In 2008, deaths suddenly increased among Japanese
macaques housed in a facility in the Kantou area of Japan.
At that time, the facility, which had been in service for
>40 years, housed ≈60 macaques, 15 (25%) of which died.
This mortality rate was much higher than that during 2006
(10.9%, 7/64 monkeys), 2007 (7.1%, 4/56), 2009 (13.8%,
9/65), 2010 (5.2%, 3/58), and 2011 (5.7%, 4/70) (Figure
1). A total of 42 monkeys died during 2006–2011, and
investigations at the time of death showed that 14 of the
monkeys had tetanus-specific symptoms: 1 of 4 that died
in 2007, 11 of 15 that died in 2008, and 2 of 9 that died in
2009). Nine of the 11 monkeys that died with characteristic
symptoms of tetanus in 2008 died during the breeding Figure 1. Number of deaths during 2006–2009 among macaques
(Macaca fuscata) housed in an animal facility in the Kantou area
of Japan. Grey boxes, monkeys with tetanus-specific clinical
Author affiliations: Azabu University, Kanagawa, Japan (T. Nakano, symptoms; white boxes, monkeys without tetanus-specific clinical
S. Nakamura, Y. Une); and National Institute for Infectious symptoms. 1, January–March; 2, April–June; 3, July–September; 4,
Diseases, Tokyo, Japan (A. Yamamoto, M. Takahashi) October–December; n, total number of monkeys. *Juvenile animal;
†Accident at time monkeys captured for vaccination (death due to
DOI: http://dx.doi.org/10.3201/eid1810.120503 hyperthermia).

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1633
DISPATCHES

contamination. C. tetani was not isolated from the monkeys’ Puerto Rico, showed a high incidence of tetanus among
food or from soil sampled >1 km from the facility. We the monkeys during the breeding seasons, but the report
performed pulsed-field gel electrophoresis on isolates from did not clarify the cause (6).
the soil at the facility and from monkey number 3, and the In facilities maintaining animals, the soil is often
results were identical, showing >90% homology. contaminated with C. tetani at a relatively high rate (1,2).
On 3 occasions (October 27 and December 17, 2009, In the facility in Japan, C. tetani was isolated at a high rate
and December 8, 2010), macaques housed in the facility from soil and from play structures. The genotype of these
(total 65) were intramuscularly administered 0.5 mL of isolates was consistent with that for an isolate obtained
tetanus toxoid (Nisseiken Co., Ltd., Tokyo, Japan). In 1 from a monkey housed at the facility, suggesting that the
monkey, the prevaccination serum level of tetanus toxoid soil was the source of the infection.
antibody was higher than the level for tetanus prevention The facility has >40 years’ experience raising monkeys,
(0.1 IU/mL). At 51 days after the first vaccination, 83.3% and the cause of the sudden outbreak in 2008 is unclear.
(51/61) of the animals were antibody-positive, and 1 year The outbreak was concentrated during the breeding season,
after the second vaccination, 100% were antibody-positive. suggesting that injuries sustained through fighting during
Since then, no tetanus symptoms have occurred in any of the mating season in an environment with severe C. tetani
the monkeys. Caretakers for monkeys at the facility were contamination may have led to the outbreak. C. tetani is
examined at a community medical office and inoculated present in the intestinal contents of various animal species
with tetanus toxoid. (1,3). Thus, bacteria in the feces of infected monkeys
may have added to the level of indigenous C. tetani
Conclusions contamination in the soil.
On the basis of these findings, we diagnosed the In Japan, tetanus is still reported in >100 persons each
disease as tetanus, and we concluded that it was an year: 115 cases were reported in 2005, 117 in 2006, 89 in
unprecedented, large-scale outbreak. Many animal 2007, 124 in 2008, 113 in 2009,and 106 in 2010) (8). It is a
exhibition facilities in Japan maintain Japanese macaques, highly lethal zoonosis and a disease of concern with regard
and tetanus has been reported in captive macaques in other to public and animal health. After tetanus was diagnosed in
countries (5–7). Results of a 5-year study (July 1, 1976– the monkeys, we immediately administered tetanus vaccine
June 30, 1981) among the free-ranging rhesus monkey to monkey caretakers at the facility and thoroughly enforced
(Macaca mulatta) colony on the island of Cayo Santiago, hygiene practices. To prevent tetanus infection in animals
and animal caretakers in such facilities and in visitors, we
recommend that newborn monkeys be vaccinated, housing
areas and play equipment be disinfected, soil highly
contaminated with C. tetani be replaced, and epidemiologic
surveys be conducted.

This work was supported in part by a grant-in-aid from the

Ministry of Health, Labor and Welfare, Japan.
Miss Nakano, a graduate of Azabu University, Kanagawa,
Japan, works as a clinical veterinarian; this paper is her graduation
thesis. Her research interest is in infectious diseases of monkeys.

References

1. Becker DG, Lineaweaver WC, Edlich R, Hill LG, Mahler CA, Cox
MJ, et al. Management and prevention of tetanus. J Long Term
Eff Med Implants. 2003;13:139–54. http://dx.doi.org/10.1615/
JLongTermEffMedImplants.v13.i3.20
2. Brook I. Current concepts in the management of Clostridium tetani
infection. Expert Rev Anti Infect Ther. 2008;6:327–36. http://dx.doi.
org/10.1586/14787210.6.3.327
3. Novak RT, Thomas CG. Tetanus. In: CDC health information for
Figure 2. A) Opisthotonos as a tetanus-specific clinical symptom international travel 2012: the yellow book [cited 2012 Apr 27]. http://
in a 1-year-old male Japanese macaque (Macaca fuscata). B) wwwnc.cdc.gov/travel/yellowbook/2012/chapter-3-infectious-
Opisthotonos with severe rigid posture in an adult male Japanese diseases-related-to-travel/tetanus.htm
macaque.

1634 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Tetanus in Captive Japanese Macaques

4. World Health Organization. Current recommendations for treatment 7. Springer DA, Phillippi-Falkenstein K, Smith G. Retrospective
of tetanus during humanitarian emergencies [cited 2012 Jan 15]. analysis of wound characteristics and tetanus development in
http://whqlibdoc.who.int/hq/2010/WHO_HSE_GAR_DCE_ captive macaques. J Zoo Wildl Med. 2009;40:95–102. http://dx.doi.
2010.2_eng.pdf org/10.1638/2008-0055.1
5. Digiacomo RF, Missakian EA. Tetanus in a free-ranging colony of 8. Yamamoto A, Takahashi M. Clostridium tetani [in Japanese]. Nihon
Macaca mulatta: a clinical and epizootiologic study. Lab Anim Sci. Rinsho. 2010;68(Suppl 6):220–3.
1972;22:378–83.
6. Rawlins GR, Kessler MJ. A five-year study of tetanus in the Address for correspondence: Yumi Une, Laboratory of Veterinary
Cayo Santiago rhesus monkey colony: behavioral description
Pathology, School of Veterinary Medicine, Azabu University, Fuchinobe
and epizootiology. Am J Primatol. 1982;3:23–39. http://dx.doi.
org/10.1002/ajp.1350030103 1-17-71, Chuo-ku, Sagamihara, Kanagawa 252-5201, Japan; email: une@
azabu-u.ac.jp

etymologia
Tetanus
[tet′ə-nəs]
From the Greek tetanos (“tension,” from teinein, “to stretch”), an often fatal infectious disease caused by the
anaerobic bacillus Clostridium tetani. Tetanus was well known to the ancients; Greek physician Aretaeus wrote
in the first century CE, “Tetanus in all its varieties, is a spasm of an exceedingly painful nature, very swift to prove
fatal, but neither easy to be removed.” Active immunization with tetanus toxoid was described in 1890, but cases
continue to be reported (275 in the United States from 2001 through 2010), almost exclusively in persons who were
never vaccinated or had not received a booster immunization in the previous 10 years. In developing countries,
neonatal tetanus—when infants are infected through nonsterile delivery—is a major contributor to infant mortality.
Worldwide, an estimated 59,000 infants died of neonatal tetanus in 2008.

Sources

1. Centers for Disease Control and Prevention. Tetanus. In: Epidemiology and prevention of vaccine-preventable diseases. Atlanta: The
Centers; 2012. p. 291–300.
2. Dorland’s Illustrated Medical Dictionary. 32nd ed. Philadelphia: Elsevier Saunders; 2012.
3. Pearce JM. Notes on tetanus (lockjaw). J Neurol Neurosurg Psychiatry. 1996;60:332. http://dx.doi.org/10.1136/jnnp.60.3.332
4. Reddy P, Bleck TP. Clostridium tetani (tetanus). In: Mandell GL, Bennett JE, Dolin R, editors. Principles and practices of infectious dis-
eases. 7th ed. Philadelphia: Churchill Livingstone; 2010. p. 3091–6.
5. World Health Organization. Neonatal tetanus. October 4, 2011 [cited 2012 Aug 27]. http://www.who.int/immunization_monitoring/
diseases/neonatal_tetanus/en/index.html

Address for correspondence: Ronnie Henry, Centers for Disease Control and Prevention, 1600 Clifton Rd NE, Mailstop E03, Atlanta, GA 30333,
USA; email: boq3@cdc.gov

DOI: http://dx.doi.org/10.3201/eid1810.ET1810

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1635
DISPATCHES

Human Infection
with Candidatus
Neoehrlichia
mikurensis, China
Hao Li,1 Jia-Fu Jiang,1 Wei Liu,1
Yuan-Chun Zheng, Qiu-Bo Huo, Kun Tang,
Shuang-Yan Zuo, Kun Liu, Bao-Gui Jiang,
Hong Yang, and Wu-Chun Cao

To identify Candidatus Neoehrlichia mikurensis infection

in northeastern China, we tested blood samples from 622
febrile patients. We identified in 7 infected patients and
natural foci for this bacterium. Field surveys showed that
1.6% of ticks and 3.8% of rodents collected from residences
of patients were also infected.

andidatus Neoehrlichia mikurensis was detected in

C 1999 in Ixodes ricinus ticks in the Netherlands and
referred to as an Ehrlichia spp.–like agent (1). It was then
Figure 1. Location of Mudanjiang, Heilongjiang Pro-vince, China,
where Candidatus Neoehrlichia mikurensis was detected.
classified as a new member of family Anaplasmataceae on
the basis of ultrastructure and phylogenetic analysis (2).
The agent was detected in ticks and small wild mammals
in Europe and Asia (1–6) and has recently been reported to were collected and treated with EDTA. DNA as extracted
infect humans, especially immunocompromised patients in by using the QIAmp DNA Blood Mini Kit (QIAGEN,
Europe (7–10). However, no cases of infection have been Germantown, MD, USA).
identified outside Europe. Moreover, the agent has not For a broad-range assay, a nested PCR specific for
yet been isolated in pure culture, and its antigens are not the 16S rRNA gene (rrs) was used to detect organisms
available. in the family Anaplasmataceae. For positive samples, 2
To investigate human infections with tick-borne agents heminested PCRs were used to amplify the entire rrs gene.
in China, we initiated a surveillance study at Mudanjiang For further confirmation, a nested PCR specific for the
Forestry Central Hospital (Mudanjiang, China). This 60-kDa heat shock protein gene (groEL) was performed.
hospital is one of the largest hospitals treating patients with Detailed cycling conditions for all amplifications are
tick-borne infectious diseases in northeastern China, where described in the Technical Appendix (wwwnc.cdc.gov/
various tick-borne agents have been detected in ticks and EID/pdfs/12-0594-Techapp.pdf).
animal hosts (11–15). Seven patients were found to be infected with
Candidatus N. mikurensis by amplifications of the rrs and
The Study groEL genes. Amplified rrs gene (1,501 bp) and partial
During May 2–July 30, 2010, a total of 622 febrile groEL gene (1,230 bp) sequences from these patients were
patients, who had histories of recent tick bites and sought identical. These sequences were also identical to genes of
treatment at Mudanjiang Forestry Central Hospital (Figure Candidatus N. mikurensis detected in ticks and rodents in
1) were screened for the infections of tick-borne agents. the Asian region of Russia (5).
When patients were admitted, peripheral blood samples Serum samples were collected from patients during the
acute (2–12 days after onset of illness) or convalescent (34–
Author affiliations: Beijing Institute of Microbiology and Epi- 42 days after onset of illness) phases of illness. All samples
demiology, Beijing, People’s Republic of China (H. Li, J.-F. Jiang, were negative for IgG against Anaplasma phagocytophilum,
W. Liu, K. Tang, S.-Y. Zuo, K. Liu, B.-G. Jiang, H. Yang, W.-C. Cao); Ehrlichia chaffeensis, Borrelia burgdorferi, Rickettsia
and Mudanjiang Forestry Central Hospital, Mudanjiang, People’s heilongjiangensis, and tick-borne encephalitis virus when
Republic of China (Y.-C. Zheng, Q.-B. Huo) tested by indirect immunofluorescence assay.
DOI: http://dx.doi.org/10.3201/eid1810.120594 1
These authors contributed equally to this article.

1636 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Human Infection with Candidatus N. mikurensis

All 7 patients were farmers residing in the villages in Table. Prevalence of Candidatus Neoehrlichia mikurensis in ticks
Mudanjiang. Their median age was 41 years (range 29–67 and rodents, Mudanjiang, China
years) and 5 were men. None had been vaccinated against Species No. positive/no. tested (%)
tick-borne encephalitis. The patients had onset of illness Tick
Ixodes persulcatus 6/316 (1.9)
during May 20–July 13, 2010. The median time from the
Haemaphysalis concinna 2/187 (0.8)
tick bite to the onset of illness and from the onset of illness Dermacentor silvarum 0/13 (0)
to the physician visit was 8 days (range 2–35 days) and 7 Total 8/516 (1.6)
days (range 1–12 days), respectively. Rodent
All patients were otherwise healthy, and none had a Clethrionomys rufocanus 5/109 (4.6)
history of underlying immunocompromised conditions. Rattus norvegicus 2/35 (5.7)
Apodemus peninsulae 0/30 (0)
Fever, headache, and malaise were reported for all 7 patients.
Apodemus agrarius 0/25 (0)
Other major manifestations included nausea (5/7), vomiting Mus musculus 0/9 (0)
(5/7), myalgia (4/7), and stiff neck (4/7). Less common Tamias sibiricus 1/3 (33.3)
symptoms were arthralgias (2/7), cough (2/7), diarrhea (1/7), Total 8/211 (3.8)
confusion (1/7), and erythema (1/7). Skin erythema (multiple
and oval) was seen on the neck of 1 patient. Conclusions
Laboratory test results showed leukopenia in 1 We have detected Candidatus N. mikurensis DNA in
patient, leukocytosis in 1 patient, thrombocytopenia in 2 blood samples from 7 patients collected during the period
patients, and anemia in 2 patients. Serum levels of alanine of acute illness, which suggests that this bacterium was the
aminotransferase and aspartate aminotransferase were etiologic agent of the infections. Our findings demonstrated
within reference ranges for all patients. Wright–Giemsa human infections with Candidatus N. mikurensis in
stained peripheral blood smears did not show morulae or China. The rrs and groEL gene nucleotide sequences of
other blood parasites. this Candidatus N. mikurensis variant were identical to
To identify local natural foci, we performed a field those obtained from ticks and rodents in the Asian region
investigation on infections of Candidatus N. mikurensis in of Russia, which have not been reported to cause human
ticks and rodents from areas of residences of the patients. infection.
During May–July 2010, a total of 516 host-seeking ticks, Unlike reported cases in elderly or immunocompromised
including 316 I. persulcatus, 187 Haemaphysalis concinna, patients in whom disease developed (7–10), all 7 patients
and 13 Dermacentor silvarum, were collected on vegetation in our study had relatively mild disease. Major clinical
and individually examined. Candidatus N. mikurensis manifestations and laboratory findings of the cases in our
DNA was detected in 6 (1.9%) I. persulcatus and 2 (0.8%) report, such as leukocytosis, were not similar to those of
H. concinna ticks, but no DNA was detected in D. silvarum previously reported cases. It is noteworthy that the patients
ticks (Table). reported in this study were previously healthy. Thus, their
A total of 211 rodents of various species were captured clinical manifestations might be typical of Candidatus N.
by using snap traps. After rodent species was identified, mikurensis infection in an otherwise healthy population.
spleen specimens were collected for DNA extraction and However, the number of cases in our study was limited, and
PCR. Eight rodents of 3 species, 5 (4.6%) Clethrionomys clinical data were not inclusive. Clinical characteristics of
rufocanus, 2 (5.7%) Rattus norvegicus, and 1 (33.3%) Candidatus N. mikurensis infection should include detailed
Tamias sibiricus, were positive for Candidatus N. descriptions of additional cases.
mikurensis (Table). Our finding of a Candidatus N. mikurensis variant in
Nucleotide sequences of rrs and groEL genes of 8 ticks 1.6% of ticks and 3.8% of rodents tested suggested natural
and 8 rodents were identical to each other and to sequences foci of the bacterium in Mudanjiang. Therefore, clinical
obtained from the 7 patients. Phylogenetic analysis of rrs diagnosis of Candidatus N. mikurensis infection should be
genes showed that nucleotide sequences identified were considered in patients who have been exposed to areas with
identical to those of Candidatus N. mikurensis from Japan high rates of tick activity. It is noteworthy that Candidatus
and the Asian region of Russia but different from sequences N. mikurensis was originally detected in R. norvegicus
from Europe (99.6%–99.8% similarity) (Figure 2, panel from Guangzhou Province in southeastern China (4),
A). Similar phylogenetic relationships were observed in thereby indicating the potential threat to humans in areas
a neighbor-joining tree based on groEL gene nucleotide other than northeastern China.
sequences. In comparison with sequences from humans In summary, we identified Candidatus N. mikurensis
and ticks in Europe, the groEL gene sequences identified as an emerging human pathogen in China. Further studies
in the study showed 97.6%–98.4% similarity (Figure 2, should be conducted to isolate this bacterium and investigate
panel B). its epidemiologic, genetic, and pathogenic features. To

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1637
DISPATCHES

Mr Li is a medical student in the State Key Laboratory of

Pathogen and Biosecurity, Beijing Institute of Microbiology and
Epidemiology. His research interests include the microbiology,
epidemiology, and ecology of vector-borne diseases.

References

1. Schouls LM, Van De Pol I, Rijpkema SG, Schot CS. Detection and
identification of Ehrlichia, Borrelia burgdorferi sensu lato, and
Bartonella species in Dutch Ixodes ricinus ticks. J Clin Microbiol.
1999;37:2215–22.
2. Kawahara M, Rikihisa Y, Isogai E, Takahashi M, Misumi H, Suto
C, et al. Ultrastructure and phylogenetic analysis of ‘Candidatus
Neoehrlichia mikurensis’ in the family Anaplasmataceae, isolated
from wild rats and found in Ixodes ovatus ticks. Int J Syst Evol
Microbiol. 2004;54:1837–43. http://dx.doi.org/10.1099/ijs.0.63260-
0
3. Andersson M, Raberg L. Wild rodents and novel human pathogen
Candidatus Neoehrlichia mikurensis, southern Sweden. Emerg Infect
Dis. 2011;17:1716–8. http://dx.doi.org/10.3201/eid1709.101058
4. Pan H, Liu S, Ma Y, Tong S, Sun Y. Ehrlichia-like organism gene
found in small mammals in the suburban district of Guangzhou
of China. Ann N Y Acad Sci. 2003;990:107–11. http://dx.doi.
org/10.1111/j.1749-6632.2003.tb07346.x
5. Rar VA, Livanova NN, Panov VV, Doroschenko EK, Pukhovskaya
NM, Vysochina NP, et al. Genetic diversity of Anaplasma and
Ehrlichia in the Asian part of Russia. Ticks Tick Borne Dis.
2010;1:57–65. http://dx.doi.org/10.1016/j.ttbdis.2010.01.002
6. Richter D, Matuschka FR. “Candidatus Neoehrlichia mikurensis,”
Anaplasma phagocytophilum, and Lyme disease spirochetes in
questing European vector ticks and in feeding ticks removed from
people. J Clin Microbiol. 2012;50:943–7. http://dx.doi.org/10.1128/
JCM.05802-11
7. von Loewenich FD, Geissdorfer W, Disque C, Matten J, Schett G,
Figure 2. A) Neighbor-joining trees based on the 16S rRNA gene (rrs)
Sakka SG, et al. Detection of “Candidatus Neoehrlichia mikurensis”
and B) the 60-kDa heat shock protein gene (groEL) of Candidatus
in two patients with severe febrile illnesses: evidence for a European
Neoehrlichia mikurensis, China, generated by using Molecular sequence variant. J Clin Microbiol. 2010;48:2630–5. http://dx.doi.
Evolutionary Genetics Analysis software version 4.0, (www. org/10.1128/JCM.00588-10
megasoftware.net/) the maximum composite-likelihood method, 8. Fehr JS, Bloemberg GV, Ritter C, Hombach M, Luscher TF, Weber R,
and bootstrap analysis of 1,000 replicates. Asterisks indicate et al. Septicemia caused by tick-borne bacterial pathogen Candidatus
nucleotide sequences of Candidatus N. mikurensis determined in Neoehrlichia mikurensis. Emerg Infect Dis. 2010;16:1127–9. http://
this study. Numbers on branches indicate percentage of replicates dx.doi.org/10.3201/eid1607.091907
that reproduced the topology for each clade. Scale bars indicate 9. Pekova S, Vydra J, Kabickova H, Frankova S, Haugvicova R, Mazal
estimated evolutionary distance. A total of 1,303 positions for rrs O, et al. Candidatus Neoehrlichia mikurensis infection identified in
and 953 positions for groEL were analyzed. Sources of Candidatus 2 hematooncologic patients: benefit of molecular techniques for rare
N. mikurensis sequences are shown between species names and pathogen detection. Diagn Microbiol Infect Dis. 2011;69:266–70.
GenBank accession numbers. http://dx.doi.org/10.1016/j.diagmicrobio.2010.10.004
10. Welinder-Olsson C, Kjellin E, Vaht K, Jacobsson S, Wenneras C.
First case of human “Candidatus Neoehrlichia mikurensis” infection
in a febrile patient with chronic lymphocytic leukemia. J Clin
Microbiol. 2010;48:1956–9. http://dx.doi.org/10.1128/JCM.02423-
09
11. Cao WC, Zhao QM, Zhang PH, Dumler JS, Zhang XT, Fang LQ,
guide diagnostic testing and treatment, physicians should
et al. Granulocytic Ehrlichiae in Ixodes persulcatus ticks from an
be aware that human Candidatus N. mikurensis infections area in China where Lyme disease is endemic. J Clin Microbiol.
are in Heilongjiang Province and that PCR can be used as 2000;38:4208–10.
a diagnostic technique for identifying suspected infections. 12. Cao WC, Zhao QM, Zhang PH, Yang H, Wu XM, Wen BH, et al.
Prevalence of Anaplasma phagocytophila and Borrelia burgdorferi
in Ixodes persulcatus ticks from northeastern China. Am J Trop Med
This study was supported by the Natural Science Foundation Hyg. 2003;68:547–50.
13. Cao WC, Zhan L, De Vlas SJ, Wen BH, Yang H, Richardus JH,
of China (81130086, 81072250, and 81172729), the Chinese et al. Molecular detection of spotted fever group Rickettsia in
Basic Research Project (2010CB530201), and the Special Fund Dermacentor silvarum from a forest area of northeastern China.
for Health Research in the Public Interest (201202019). J Med Entomol. 2008;45:741–4. http://dx.doi.org/10.1603/0022-
2585(2008)45[741:MDOSFG]2.0.CO;2

1638 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Human Infection with Candidatus N. mikurensis

14. Zhan L, Cao WC, Chu CY, Jiang BG, Zhang F, Liu W, et al. Tick- Address for correspondence: Wu-Chun Cao, State Key Laboratory
borne agents in rodents, China, 2004–2006. Emerg Infect Dis. of Pathogen and Biosecurity, Beijing Institute of Microbiology and
2009;15:1904–8. http://dx.doi.org/10.3201/eid1512.081141
Epidemiology, 20 Dong-Da St, Fengtai District, Beijing 100071, People’s
15. Zhan L, Cao WC, Jiang JF, Zhang XA, Liu YX, Wu XM, et al.
Anaplasma phagocytophilum from rodents and sheep, China. Republic of China: email: caowc@nic.bmi.ac.cn
Emerg Infect Dis. 2010;16:764–8. http://dx.doi.org/10.3201/
eid1605.091293

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1639
DISPATCHES

Anthroponotic 1985. Visitors are allowed to bring or buy food to feed the
animals, watch them from a short distance, or play with
Enteric Parasites them (Figure, panel A).
We collected fecal droppings at 3 locations with
in Monkeys in different animal densities. A total of 187 specimens

Public Park, China

were collected from the Macaque Garden, where animal
density was the highest, at ≈400 animals in a small open
area between 2 mountains. Another 74 specimens were
Jianbin Ye, Lihua Xiao, Jingbo Ma, Meijin Guo, collected from the Tanquan Spring area, where animal
Lili Liu, and Yaoyu Feng density was the lowest. The remaining 150 specimens were
collected from the Hongfu Temple, which had moderate
Cryptosporidium spp., Giardia duodenalis, and animal density. Twenty-three 100-mL grab samples of
Enterocytozoon bieneusi were detected in 45, 35, and high-turbidity water were collected at various points of a
116 of 411 free-range rhesus monkeys, respectively, in a
small lake near the Macaque Garden and Tanquan Spring,
popular public park in the People’s Republic of China. Most
genotypes and subtypes detected were anthroponotic,
where the rhesus monkeys frequently bathed (Figure).
indicating these animals might be reservoirs for human We detected Cryptosporidium spp., G. duodenalis sub-
cryptosporidiosis, giardiasis, and microsporidiosis. types, and E. bieneusi genotypes in the fecal specimens and
differentiated them by using PCR and sequence analysis of
the small subunit rRNA gene (5), triosephosphate isomerase

C ryptosporidiosis, giardiasis, and microsporidiosis

are enteric diseases in humans and are mainly
caused by Cryptosporidium spp., Giardia duodenalis,
and Enterocytozoon bieneusi, respectively (1–3). These
protozoan parasites are also commonly found in animals
and are considered zoonotic. However, the role of
nonhuman primates in the transmission of the diseases
remains unclear because few studies have been conducted
on the genetic characteristics of the parasites in these
animals. In a recent study in Kenya, 5 (2.0%) and 13
(5.5%) of 235 captive baboons had human-pathogenic
C. hominis subtypes and E. bieneusi genotypes,
respectively. This finding implies that nonhuman primates
might be reservoirs for human cryptosporidiosis and
microsporidiosis (4). We determined the genotypes and
subtypes of Cryptosporidium spp., G. duodenalis, and E.
bieneusi in free-range rhesus monkeys in a popular public
park to assess the potential for transmission of these
parasites from rhesus monkeys to humans.

The Study
In November 2010, we collected 411 fecal specimens
from rhesus monkeys (Macaca mulatta) in Qianling Park,
Guiyang, People’s Republic of China (www.qlpark.cn).
The park, a major tourist attraction of the city, is visited
by 10,000–70,000 persons each day. It has the highest
number (≈700) of domesticated free-range monkeys in
China, which originated from a troop of 20 animals in

Author affiliations: East China University of Science and Technology,

Shanghai, People’s Republic of China (J. Ye, J. Ma, M. Guo, L. Liu,
Y. Feng); and Centers for Disease Control and Prevention, Atlanta, Figure. Potential zoonotic and waterborne pathways of parasites in
Georgia, USA (J. Ye, L. Xiao) Qianling Park, Guiyang, China. A) Close contact of rhesus monkeys
with humans. B) Potential contamination of recreational water with
DOI: http://dx.doi.org/10.3201/eid1810.120653 pathogens from rhesus monkeys.

1640 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Anthroponotic Parasites in Monkeys, China

gene (6), and ribosomal internal transcribed spacer (4), successfully subtyped, 7 subtypes in 5 subtype families
respectively. We similarly analyzed water samples after were identified: 4 families (Ia, Id, Ie, and If) of C. hominis
concentrating pathogens by centrifugation at 3,000 × g for and 1 family (IIc) of C. parvum (Table). The most common
15 min. We subtyped C. hominis and C. parvum by using subtypes were IaA13R8 (8 specimens), IdA20 (13),
sequence analysis of the 60-kDa glycoprotein gene (5). We IeA11G3T3 (13), and IIcA5G3a (5).
analyzed each specimen at least 2× by using PCR, with G. duodenalis was identified in 35 (8.5%) of the 411
the inclusion of positive and negative controls in each run. fecal specimens. The rates at Macaque Garden (10.2%;
We used the χ2 test to compare differences in rates of each p = 0.016) and Hongfu Temple (10.0%; p = 0.018) were
parasite. significantly higher than at Tanquan Spring (1.4%).
We detected Cryptosporidium spp. in 45 (10.9%) of All positive specimens except for 1 were successfully
the 411 fecal specimens, belonging to 3 species: C. hominis genotyped and subtyped and belonged to assemblages A
(39 specimens), C. parvum (5), and C. felis (1). The rate (10) and B (24). All assemblage A isolates belonged to
at Macaque Garden (16.6%) was significantly higher subtype A2 (10 specimens). In assemblage B, 7 subtypes
than at Hongfu Temple (6.0%; p = 0.003) and Tanquan were identified: 1 known subtype in 11 specimens and 6
Spring (6.8%; p = 0.038) (Table). Among 44 specimens new subtypes at low frequencies (Table).

Table. Anthroponotic enteric parasites in free-range rhesus monkeys (Macaca mulatta) and water samples in Qianling Park, Guiyang,
China*
Fecal specimens positive for organism
Species, genotype, or GenBank Macaque Garden, Hongfu Temple, n Tanquan Spring, n Water samples positive
subtype accession no. n = 187 = 150 = 74 for organism, n = 23
Cryptosporidium
C. hominis
IaA13R7 EU095261 2 0 0 0
IaA13R8 JX000568† 6 2 0 2
IaA14R7 JX000569† 1 0 1 0
IdA20 EU095265 10 3 0 0
IeA11G3T3 DQ665689 8 3 2 7
IfA16G2 JX000570† 1 0 0 0
Unknown‡ NA 0 0 0 2
C. parvum
IIcA5G3a AY738195 2 1 2 0
C. felis 1 0 0 0
Subtotal (mean %; 95% CI)§ 31/187 9/150 5/74 11/23
(16.6; 10.7–22.4) (6.0; 2.1–9.9) (6.8; 0.8–12.7) (47.8; 19.6–76.1)
Giardia duodenalis
Assemblage A2 U57897 10 0 0 0
Assemblage B
B1 AY368164 3 7 1 8
B2d JX000562† 0 2 0 0
B3d JX000563† 0 1 0 0
B4d JX000564† 2 0 0 0
B5d JX000565† 2 3 0 3
B6d JX000566† 2 0 0 0
B7d JX000567† 0 1 0 0
Unknown‡ NA 0 1 0 1
Subtotal (mean %; 95% CI)§ 19/187 15/150 1/74 12/23
(10.2; 5.6–14.7) (10.0; 4.9–15.1) (1.4; 0–4.0) (52.2; 22.7–81.7)
Enterocytozoon bieneusi
Peru11 AY371286 46 21 2 4
WL15 AY237223 15 9 4 5
EbpC AY371279 4 0 0 1
Type IV AY371277 6 0 0 2
LW1d JX000571† 0 0 0 1
Macaque1¶ JX000572† 0 0 1 0
Macaque2¶ JX00057† 1 0 0 0
Unknown‡ NA 4 2 1 0
Subtotal (mean %; 95% CI)§ 76/187 32/150 8/74 13/23
(40.6; 31.5–49.8) (21.3; 13.9–28.7) (10.8; 3.3–18.3) (56.5; 25.8–87.2)
*Values are number of positive samples unless otherwise indicated. NA, not applicable.
†From this study.
‡PCR positive but sequence unavailable.
§No. positive /no. tested (% positive; 95% CI).
¶New subtypes or genotypes identified during this study.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1641
DISPATCHES

E. bieneusi was identified in 116 (28.2%) of the 411 rhesus monkeys, as supported by the higher occurrence of
fecal specimens. The occurrence rate at Macaque Garden Cryptosporidium spp., G. duodenalis, and E. bieneusi at
(40.6%) was significantly higher than at Hongfu Temple places with higher animal density.
(21.3%; p = 0.00016) and Tanquan Spring (10.8%; p Our results indicate that rhesus monkeys in close
= 0.000033). Among the 109 specimens successfully contact with humans are commonly infected with human-
sequenced, 6 genotypes were identified: 4 known genotypes pathogenic C. hominis, C. parvum, and G. duodenalis
(Peru11 [69 specimens], WL15 [28], EbpC [4], and Type subtypes and E. bieneusi genotypes. Therefore, they can
IV [6]) and 2 new genotypes at low frequencies (Table). serve as reservoirs of human cryptosporidiosis, giardiasis,
Cryptosporidium spp., G. duodenalis, and E. bieneusi and microsporidiosis. Zoonotic transmission of infection
were detected in 11 (47.8%), 12 (52.2%), and 13 (56.5%), from these monkeys can occur directly by close contact
respectively, of the 23 water samples collected from the of monkeys and humans (Figure, panel A), or indirectly
lake where the animals bathed. Fewer C. hominis and through contamination of drinking water or recreational
G. duodenalis subtypes and E. bieneusi genotypes were water (Figure, panel B). Efforts should be made to educate
detected in water samples than in fecal specimens. Most the public about the potential risk for zoonotic transmission
of the common C. hominis (IaA13R8 and IeA11G3T3) and of enteric pathogens from rhesus monkeys and to minimize
G. duodenalis (B1 and B5) subtypes and all common E. contamination of drinking and recreational water by
bieneusi genotypes (Peru11, W15, EbpC, and Type IV) in parasites of rhesus monkey origin.
animals were found in water samples (Table). We deposited
unique nucleotide sequences obtained in GenBank under
This work was supported in part by the National Natural
accession nos. JX000562–JX000573.
Science Foundation of China (31110103901, 31229005,
41001316), the National Basic Research Program of China
Conclusions
(973 Project) (2011CB200903), Fundamental Research Funds
All C. hominis and C. parvum subtypes found in this
for the Central Universities, and Open Funding Projects of the
study are well-known parasites of humans and have rarely
State Key Laboratory of Veterinary Etiological Biology at the
been found in animals. The C. hominis subtype families
Lanzhou Veterinary Research Institute and State Key Laboratory
Ia, Id, Ie, and If had been reported in humans and urban
of Bioreactor Engineering at East China University of Science
wastewater in China (5,7–9). Although it has not been
and Technology.
found in humans in China, the C. parvum IIc subtype
family identified in rhesus monkeys in this study is a well- Mr Ye is a doctoral candidate at the East China University
known anthroponotic parasite in developing countries (1). of Science and Technology. His research interest focuses on
The G. duodenalis subtypes found in Qianling Park in the molecular epidemiology of cryptosporidia, giardia, and
are also major pathogens in humans. The subtype A2 of microsporidia.
assemblage A is a common pathogen in humans in most
areas studied and is less frequently found in animals than References
the A1 subtype (2). The dominant B1 subtype found in
1. Xiao L. Molecular epidemiology of cryptosporidiosis: an update.
Qianling Park is also identical to an assemblage B subtype Exp Parasitol. 2010;124:80–9. http://dx.doi.org/10.1016/j.exppara.
(GenBank accession no. GU564280) previously identified 2009.03.018
in humans in China (9). 2. Feng Y, Xiao L. Zoonotic potential and molecular epidemiology of
Giardia species and giardiasis. Clin Microbiol Rev. 2011;24:110–
Most E. bieneusi genotypes identified in this study
40. http://dx.doi.org/10.1128/CMR.00033-10
had also been reported in humans. Among the dominant E. 3. Santín M, Fayer R. Microsporidiosis: Enterocytozoon bieneusi in
bieneusi genotypes, Peru11 had been seen only in humans domesticated and wild animals. Res Vet Sci. 2011;90:363–71. http://
and baboons (3,4). Although genotypes IV, EbpC, and dx.doi.org/10.1016/j.rvsc.2010.07.014
4. Li W, Kiulia NM, Mwenda JM, Nyachieo A, Taylor MB, Zhang X,
WL15 have been reported in animals, they are common
et al. Cyclospora papionis, Cryptosporidium hominis, and human-
parasites of humans in many areas (3). pathogenic Enterocytozoon bieneusi in captive baboons in Kenya.
The origin of Cryptosporidium spp., G. duodenalis, J Clin Microbiol. 2011;49:4326–9. http://dx.doi.org/10.1128/JCM.
and E. bieneusi parasites in the rhesus monkey population 05051-11
5. Feng Y, Li N, Duan L, Xiao L. Cryptosporidium genotype and
is not clear. Because these parasites are common human
subtype distribution in raw wastewater in Shanghai, China: evidence
pathogens, they could have been introduced by humans. for possible unique Cryptosporidium hominis transmission. J Clin
However, rhesus monkeys can be natural hosts of these Microbiol. 2009;47:153–7. http://dx.doi.org/10.1128/JCM.01777-
organisms, as supported by recent identification of some 08
6. Sulaiman IM, Fayer R, Bern C, Gilman RH, Trout JM, Schantz PM,
of these organisms in newly captive baboons from rural
et al. Triosephosphate isomerase gene characterization and potential
and forested areas (4). Regardless of the initial origin of zoonotic transmission of Giardia duodenalis. Emerg Infect Dis.
the parasites, they can be transmitted efficiently among 2003;9:1444–52. http://dx.doi.org/10.3201/eid0911.030084

1642 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Anthroponotic Parasites in Monkeys, China

7. Feng Y, Wang L, Duan L, Gomez-Puerta LA, Zhang L, Zhao X, et Address for correspondence: Yaoyu Feng, State Key Laboratory of
al. Extended outbreak of cryptosporidiosis in a pediatric hospital, Bioreactor Engineering, School of Resources and Environmental
China. Emerg Infect Dis. 2012;18:312–4. http://dx.doi.org/10.3201/
Engineering, East China University of Science and Technology, Shanghai
eid1802.110666
8. Peng MM, Matos O, Gatei W, Das P, Stantic-Pavlinic M, Bern C, 200237, People’s Republic of China; email: yyfeng@ecust.edu.cn
et al. A comparison of Cryptosporidium subgenotypes from several
geographic regions. J Eukaryot Microbiol. 2001;48(Suppl):28S–
31S. http://dx.doi.org/10.1111/j.1550-7408.2001.tb00442.x The opinions expressed by authors contributing to this
9. Wang R, Zhang X, Zhu H, Zhang L, Feng Y, Jian F, et al. Genetic journal do not necessarily reflect the opinions of the Centers for
characterizations of Cryptosporidium spp. and Giardia duodenalis Disease Control and Prevention or the institutions with which
in humans in Henan, China. Exp Parasitol. 2011;127:42–5. http:// the authors are affiliated.
dx.doi.org/10.1016/j.exppara.2010.06.034

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1643
DISPATCHES

Schmallenberg depth phylogenetic analysis difficult. Therefore, a detailed

taxonomic classification of SBV could not be made initially
Virus as Possible when the virus emerged.
The first report of SBV showed highest similarities
Ancestor of of M- and L-segment sequences to partial Aino virus and

Shamonda Virus
AKAV sequences, whereas the N gene was most closely
related to Shamonda virus (SHAV) (1). Additionally,
results of recent investigations on complete N and M genes
Katja V. Goller,1 Dirk Höper,1 Horst Schirrmeier, and partial L genes of SHAV, Douglas virus (DOUV), and
Thomas C. Mettenleiter, and Martin Beer Sathuperi virus (SATV) suggested that SBV is a reassortant
consisting of the M segment from SATV and the S and
Schmallenberg virus (SBV), an orthobunyavirus of the L segments from SHAV (9). Conversely, in 2001, SHAV
Simbu serogroup, recently emerged in Europe and has been was described as a reassortant virus comprising the S and L
suggested to be a Shamonda/Sathuperi virus reassortant.
segments of SATV and the M segment from the unclassified
Results of full-genome and serologic investigations indicate
that SBV belongs to the species Sathuperi virus and is a
Yaba-7 virus (8). To clarify the phylogenetic relationships
possible ancestor of the reassortant Shamonda virus. and classification of SBV within the Simbu serogroup,
we conducted genetic and serologic investigations of its
relationship to 9 other Simbu serogroup viruses.

A novel virus, Schmallenberg virus (SBV), was dis-

covered in Europe in October 2011, and since then,
cases of SBV infection have been reported in sheep, cattle,
The Study
To enable comparative sequence analysis and
and goats in several European countries (1–4). Preliminary phylogenetic investigations, we determined almost full-
phylogenetic analyses revealed that SBV is a member of length S-, M-, and L-segment sequences for 9 Simbu
the genus Orthobunyavirus within the family Bunyaviridae serogroup viruses belonging to 5 species (Table 1): SHAV,
and is related to Simbu serogroup viruses (1). Similar to Peaton virus, and Sango virus, species Shamonda virus;
Akabane virus (AKAV), another Simbu serogroup virus, DOUV and SATV, species Sathuperi virus; Aino virus
SBV can cause fatal congenital defects by infection and Shuni virus, species Shuni virus; Sabo virus, species
of fetuses during a susceptible stage in pregnancy (2). Akabane virus; and Simbu virus, species Simbu virus.
Vaccines for SBV are not available. Thus, SBV poses a Sample preparation and sequencing were done by using the
serious threat to naive populations of ruminant livestock Genome Sequencer FLX (Roche, Mannheim, Germany)
in Europe. as described (10). Sequence data obtained in this study are
Orthobunyaviruses are arthropod-borne viruses with archived in the International Nucleotide Sequence Database
a negative-stranded tripartite RNA genome comprising Collaboration databases (www.insdc.org; accession nos.
large (L), medium (M), and small (S) segments. Genetic HE795087–HE795110 and HE800141–HE800143). In
reassortment occurs naturally among these viruses, which addition to the newly determined sequences, published full-
results in the emergence of new virus strains that have genome sequences of AKAV and Oropouche virus (OROV)
altered biologic properties (5). The L segment encodes the from the National Center for Biotechnology Information
RNA-dependent RNA polymerase; antigenic determinants reference genome database (www.ncbi.nlm.nih.gov/
are the M-encoded viral surface glycoproteins Gn and Gc, sites/genome) were used for sequence comparisons and
which are responsible for viral attachment, cell fusion, the reconstruction of phylogenetic relationships. Coding
hemagglutination, and the induction of neutralizing sequences of each genome segment were aligned by using
antibodies, and the S-encoded nucleocapsid protein N, which ClustalW (www.clustal.org) for codons, and phylogenetic
plays a role in complement fixation (6). In the pregenomics analyses were performed by using maximum-likelihood
era, orthobunyavirus relationships were determined solely methods in MEGA5 (11). For the N and L gene analysis
by serologic cross-reactivity analyses (7), but since DNA Tamura-Nei parameter, the M gene analysis Tamura
sequencing became available, phylogenetic relationships 3-parameter was used. The robustness of the trees was
have additionally been assessed by comparison of partial tested by bootstrap analysis by using 1,000 replications.
genome sequences (8,9). However, published full-length Sequence identities were calculated by using BioEdit
genome sequence information is sparse, which makes in- version 7.0.9.0 (12).
SBV N gene nucleotide sequence identities to other
Author affiliation: Friedrich-Loeffler-Institut, Greifswald-Insel Riems,
viruses ranged from 69.8% (OROV; 69.9% aa identity)
Germany

DOI: http://dx.doi.org/10.3201/eid1810.120835 1
These authors contributed equally to this article.

1644 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Schmallenberg Virus and Shamonda Virus

Table 1. Viruses, isolates, and sequence lengths used in Thus, phylogeny and cross-reactivity identified SHAV, but
classification of Schmallenberg virus within the Simbu serogroup not SBV, as a reassortant within the Simbu serogroup.
Sequence length, nt
Virus Isolate S M L
Aino 38K 834 4,335 6,966
Conclusions
Douglas 93–6 813 4,365 6,857 Although our results do not support the suggestions
Peaton CSIRO 110 851 4,324 6,829 of Yanase et al. (9), they are fully consistent with the
Sabo IB AN 9398 894 4,307 6,857 conclusions of Saeed et al. (8). On the basis of our results
Sango An 5077 838 4,314 6,828
Sathuperi NA 843 4,330 6,861 and those of Saeed et al. (8), we suggest that SHAV should
Schmallenberg BH80/11–4 830 4,415 6,864 be reclassified into the species Sathuperi virus and that the
Shamonda Ib An 5550 927 4,314 6,863 species Shamonda virus should be renamed Peaton virus
Shuni Ib An 10107 850 4,326 6,880
Simbu SA Ar 53 860 4,417 6,895 or Sango virus.
*S, small segment; M, medium segment; L, large segment; NA, not In addition to showing that SBV belongs to the
available. species Sathuperi virus, our results show that the virus
is most likely not a reassortant and is likely to be one of
to 97.7% (SHAV; 100% aa) (Table 2, Appendix, www. the ancestors of SHAV, whereas SHAV is a reassortant
cdcnc.gov/EID/article/18/9/12-0835-T2.htm). The L gene comprising the SBV S and L genomic segments and the
sequence of SBV had the lowest identity to OROV (60.4% M segment from an unclassified virus. These detailed
nt; 57.5% aa) and highest identity to SHAV (92.9% nt; insights into the phylogeny of SBV could be the basis for
98.4% aa); the SBV M gene showed the highest sequence
identity to SATV (82.1% nt; 90.1% aa), whereas identity
of SBV and SHAV M gene was low (48.2% nt; 36.5%
aa). In general, identity of the SHAV M gene to the other
Simbu serogroup viruses was low, from 45.6% nt (33.4%
aa; OROV) to 55.0% nt (47.9% aa; Sango virus), which
indicates that its M segment belongs to another virus,
as previously suggested (8). On the other hand, the high
sequence identity of all SBV genes to SATV and DOUV
indicates that SBV belongs to the species Sathuperi virus.
A phylogenetic tree derived from the M gene (Figure,
panel A) demonstrates that SBV and DOUV cluster closely
with SATV, whereas SHAV was placed distantly from
all other viruses. In contrast, phylogenetic trees of the L
(Figure, panel B) and N (Figure, panel C) genes showed
a close relationship between SBV and SHAV. Results for
all genome segments show that SBV should be classified
within the species Sathuperi virus and is likely to be the
ancestor of SHAV, which is in contrast a reassortant virus
comprising the S and L segments from SBV and the M
segment from another virus, as proposed previously (8).
To further clarify the placement of SBV within the
Simbu serogroup, we determined the ability of SBV
antibodies to neutralize other Simbu serogroup viruses. We
performed titer reduction assays of the 9 other viruses with
anti-SBV serum; the viruses were propagated in BHK-21 Figure. Maximum-likelihood trees showing phylogenetic relation-
cells clone CT (L164, Collection of Cell Lines in Veterinary ships of Simbu serogroup viruses for the M (A), L (B), and S coding
Medicine, Friedrich-Loeffler-Institute, Riems, Germany). regions (C). The bar plot in panel A indicates the percentage of
titer reduction of each virus by anti–Schmallenberg virus serum.
Neutralizing potencies of SBV antibodies against Simbu GenBank accession numbers are shown. Numbers at nodes
serogroup virus strains were tested by titrating virus in the indicate percentage of 1,000 bootstrap replicates (values <50 are
presence of anti-SBV serum (neutralizing titer 32); results not shown). Scale bars indicate nucleotide substitutions per site.
were expressed as percentage titer reduction in relation to a DOUV, Douglas virus; SATV, Sathuperi virus; SBV, Schmallenberg
parallel test without antiserum. Although DOUV and SATV virus; SHUV, Shuni virus; AINOV, Aino virus; SIMV, Simbu virus;
PEAV, Peaton virus; SANV, Sango virus; AKAV, Akabane virus;
were well neutralized, SHAV titers were not reduced at all SABOV, Sabo virus; SHAV, Shamonda virus; OROV, Oropouche
(Figure, panel A), which supports the M gene phylogeny. virus. ND, not determined.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1645
DISPATCHES

the development of efficient, cross-protective vaccines. 4. Friedrich-Loeffler-Institut. FLI: Schmallenberg virus [cited 2012
Our results also highlight the importance of full-genome May 31]. http://www.fli.bund.de/en/startseite/current-news/animal-
disease-situation/new-orthobunyavirus-detected-in-cattle-in-
analyses to identify potential genetic reassortments and germany.html
to investigate the evolutionary history of viruses with 5. Bowen MD, Trappier SG, Sanchez AJ, Meyer RF, Goldsmith
segmented genomes. CS, Zaki SR, et al. A reassortant bunyavirus isolated from
acute hemorrhagic fever cases in Kenya and Somalia. Virology.
2001;291:185–90. http://dx.doi.org/10.1006/viro.2001.1201
Acknowledgments 6. King AMQ, Adams MJ, Carstens EB, Lefkowitz EJ, editors. Virus
We thank Robert Tesh and his colleagues at the University taxonomy: classification and nomenclature of viruses. Ninth report
of Texas Medical Branch for providing the strains of the Simbu of the International Committee on Taxonomy of Viruses. San Diego:
Elsevier Academic Press; 2011.
serogroup viruses. We are indebted to Moctezuma Reimann and 7. Kinney RM, Calisher CH. Antigenic relationships among
Bianka Hillmann for excellent technical assistance. Simbu serogroup (Bunyaviridae) viruses. Am J Trop Med Hyg.
1981;30:1307–18.
This work was funded by the European Union FP7 projects 8. Saeed MF, Li L, Wang H, Weaver SC, Barrett AD. Phylogeny
EMPERIE (no. 223498) and RAPIDIA-Field (no. 289364) and the of the Simbu serogroup of the genus Bunyavirus. J Gen Virol.
Federal Ministry of Food, Agriculture and Consumer Protection, 2001;82:2173–81.
9. Yanase T, Kato T, Aizawa M, Shuto Y, Shirafuji H, Yamakawa M, et
Germany.
al. Genetic reassortment between Sathuperi and Shamonda viruses
Dr Goller is a molecular biologist and postdoctoral scientist of the genus Orthobunyavirus in nature: implications for their
genetic relationship to Schmallenberg virus. Arch Virol. 2012. Epub
at the Friedrich-Loeffler-Institute, Institute of Diagnostic Viro- ahead of print. http://dx.doi.org/10.1007/s00705-012-1341-8
logy, Greifswald-Insel Riems, Germany. Her research interests 10. Becker N, Jöst H, Ziegler U, Eiden M, Höper D, Emmerich P, et
are emerging animal viruses, molecular diagnostics, and al. Epizootic emergence of Usutu virus in wild and captive birds in
epidemiology, as well as phylogenetic analyses. Germany. PLoS One. 2012;7:e32604-e.
11. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S.
MEGA5: molecular evolutionary genetics analysis using maximum
likelihood, evolutionary distance, and maximum parsimony methods.
References
Mol Biol Evol. 2011;28:2731–9. http://dx.doi.org/10.1093/molbev/
msr121
1. Hoffmann B, Scheuch M, Höper D, Jungblut R, Holsteg M,
12. Hall TA. BioEdit: a user-friendly biological sequence alignment
Schirrmeier H, et al. Novel orthobunyavirus in cattle, Europe,
editor and analysis program for Windows 95/98/NT. Nucleic Acids
2011. Emerg Infect Dis. 2012;18:469–72. http://dx.doi.org/10.3201/
Symposium Series. 1999;41:95–8.
eid1803.111905
2. Garigliany M-M, Hoffmann B, Dive M, Sartelet A, Bayrou C,
Cassart D, et al. Schmallenberg virus in calf born at term with Address for correspondence: Martin Beer, Institute of Diagnostic
porencephaly, Belgium. Emerg Infect Dis. 2012;18:1005–6. http:// Virology, Friedrich-Loeffler-Institut, Südufer 10, 17493 Greifswald-Insel
dx.doi.org/10.3201/eid1806.120104 Riems, Germany; email: martin.beer@fli.bund.de
3. Elbers ARW, Loeffen WLA, Quak S, de Boer-Luijtze E, van der
Spek AN, Bouwstra R, et al. Seroprevalence of Schmallenberg virus
antibodies among dairy cattle, the Netherlands, winter 2011–2012. Use of trade names is for identification only and does not
Emerg Infect Dis. 2012;18:1065–71. http://dx.doi.org/10.3201/ imply endorsement by the Public Health Service or by the US
eid1807.120323 Department of Health and Human Services.

1646 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Monkey Bites Combined Joint Task Force–1 in eastern Afghanistan. We
evaluated these records to identify and describe monkey
among US Military bites and high-risk exposures among US military mem-
bers serving in eastern Afghanistan during September–De-
Members, cember 2011. For this study, eastern Afghanistan refers to

Afghanistan, 2011
North Atlantic Treaty Organization Regional Command
East, which covers ≈43,000 square miles (110,000 km2).
The US military population in eastern Afghanistan during
Luke E. Mease1 and Katheryn A. Baker2 the study period was ≈23,500 persons. Case information
obtained included patient age, sex, rank, branch of military
Bites from Macaca mulatta monkeys, native to Afghani- service, animal exposures, and treatment details.
stan, can cause serious infections. To determine risk for US We evaluated the cases for the 5 parameters that com-
military members in Afghanistan, we reviewed records for
prise appropriate initial treatment according to the litera-
September–December 2011. Among 126 animal bites and
exposures, 10 were monkey bites. Command emphasis is
ture. The parameters are wound care (appropriate cleans-
vital for preventing monkey bites; provider training and bite ing of the wound) (7), antiviral medications for B-virus
reporting promote postexposure treatment. (valacyclovir) (8), antimicrobial drugs for oral bacteria
(amoxicillin/clavulanic acid or clindamycin plus sulfa-
methoxazole/trimethoprim) (3), verification of up-to-date

M ilitary members deployed to Afghanistan face many

risks; among these are bites from Macaca mulatta
monkeys and possible subsequent infections. In August
tetanus vaccination status or vaccine administration in ac-
cordance with Advisory Committee on Immunization Prac-
tices guidelines (9), and rabies postexposure prophylaxis
2011, a 24-year-old US Army soldier died of a rabies in- (PEP). US military policy advised that rabies PEP should
fection contracted while in eastern Afghanistan. This trag- adhere to World Health Organization guidelines (10),
edy highlights the threat that animal bites pose to deployed which recommend giving human rabies immunoglobulin
military members. plus 5 doses of rabies vaccine. In accordance with the same
During 2001–2010, a total of 643 animal bites among policy, adherence to Advisory Committee on Immuniza-
deployed US military members were reported (1). Dogs tion Practices guidelines for rabies PEP with human rabies
were implicated in 50% of these bites, but several other immunoglobulin plus 4 doses of rabies vaccine was also
animals pose risk as well. Prominent among these is the acceptable (11).
nonhuman primate M. mulatta (rhesus macaque), native When appropriate initial treatment was not adminis-
to and commonly kept as a pet in Afghanistan (2) (Fig- tered, subsequent follow-up was conducted to ensure that
ure). Risks from M. mulatta monkey bites include physical patients received required treatment. Appropriate treatment
trauma and/or infection with B-virus (Macacine herpes- was accomplished by contacting and coordinating with the
virus 1), oral bacteria (including Clostridium tetani), and responsible provider, the patients, and their commanders.
rabies virus. Although not well characterized in Afghani- During the study period, we identified 126 cases of
stan, the risk for exposure to M. mulatta monkeys has been animal bites or serious exposures (involving animal neu-
described (3) for researchers (4), tourism workers (5), and ral tissue or saliva affecting the mucosal surfaces or open
US pet owners (6). We examined this risk for US military wounds of the patient). Among these cases, 10 were cases
members deployed to eastern Afghanistan. The work pre- of monkey bites.
sented herein was reviewed and deemed exempt from in- Among the 10 military members who had been bitten
ternal review board oversight by the Joint Combat Casualty by monkeys, age range was 22–44 years (Table); most (7)
Research Team, the human subjects review board respon- were <30 years of age, and 8 were male. All were junior
sible for oversight of human subjects research affecting US enlisted or noncommissioned officers; 8 were members of
military members in Afghanistan. the Army, and 2 were members of the Air Force (Table).
In terms of treatment, 6 received appropriate wound
The Study care and washing, 5 received appropriate B-virus pro-
Information about all reported animal bites and ex- phylaxis, and 8 received appropriate antimicrobial drugs
posures affecting US military and coalition personnel (Table). In terms of prophylaxis, only 4 were evaluated for
is collected by preventive medicine officers assigned to
1
Current affiliation: Army Health Clinic, Dugway Proving Ground,
Author affiliation: US Army Combined Joint Task Force–1, Bagram
Utah, USA.
Air Field, Afghanistan 2
Current affiliation: General Leonard Wood Army Community
DOI: http://dx.doi.org/10.3201/eid1810.120419 Hospital, Fort Leonard Wood, Missouri, USA.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1647
DISPATCHES

civilians, and 1 belonged to US military members. For the

other 2, no ownership data were available; they could have
been wild or pets. One monkey was euthanized and sent to
US Army Veterinary Laboratory Europe for testing; brain
samples were negative for rabies and B-virus.

Conclusions
Our identification of 126 reported bites or exposures
over just 4 months suggests that the 643 animal bites re-
ported for all deployed US military members for the past
decade greatly underestimate the true number of animal
bites in this population. The number of bites and exposures
identified in this study might represent more accurate re-
porting because of increased attention to animal bites after
the US soldier died in August 2011. It is possible that be-
fore that time, only more severe bites and exposures were
reported but that after that time, more lower-risk exposures
might have been reported.
The risk for monkey bites in other populations has
been described. The 10 monkey bites reported in this study
demonstrate that US and coalition military members in Af-
ghanistan are also at risk for the trauma and the B-virus,
bacterial, tetanus, and rabies infections that can result from
monkey bites and exposures. The demographics of the pop-
ulation bitten (Army, age <30 years, and male) is represen-
tative of the underlying population at risk.
Figure. Pet monkey (Macaca mulatta), Afghanistan, 2011. Most monkey-bite patients received appropriate care.
Photograph courtesy of Ronald Havard. This care is laudable, considering the recognized difficul-
ties in treating monkey bites (12). Some patients, however,
did not receive appropriate medical treatment initially. Be-
cause treatment of monkey bites is not a standard part of
US medical education, inadequate treatment could reflect
tetanus status, and 8 received appropriate rabies PEP. Be- insufficient training and lack of familiarity among US-
yond the initial trauma and follow-up visits for rabies PEP, trained health care providers. It is imperative that before
no visits for any illness possibly associated with the bite or providers are deployed to Afghanistan, they receive proper
exposure were recorded. instruction on the care of animal bites and exposures. Ap-
All cases involved different monkeys, 8 of which were propriate reporting of any animal bite to military preventive
kept as pets. Of these 8 pet monkeys, 4 belonged to Afghan medicine personnel is crucial because it permits oversight
National Security Forces (ANSF), 3 belonged to Afghan of care and timely correction of deficiencies.

Table. Characteristics of US military members bitten by monkeys, eastern Afghanistan, September–December, 2011*

Treatment received
Patient Military Wound Antimicrobial Tetanus Rabies vaccine, Monkey
no. Age, y/sex branch care Valacyclovir drug vaccine HRIG ownership
1 39/M Army – + + + + ANSF
2 27/M Army + + + + + CIV†
3 22/M Army – + + – + CIV
4 44/F Army + – + – – CIV
5 31/M Army + – + + + ANSF
6 26/M Air Force + – – – – US military
7 26/M Army – + – – + ANSF
8 27/M Army + – + + + ANSF
9 22/M Army – – + – + Unknown
10 25/F Air Force + + + – + Unknown
*HRIG, human rabies immunoglobulin; –, not administered; +, administered; ANSF, Afghan National Security Forces; CIV, Afghan civilian.
†Monkey euthanized. Brain, tested at US Army Veterinary Laboratory Europe, was negative for rabies and B-virus.

1648 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Monkey Bites among US Military Members

Most (7/10) monkeys involved were pets owned by References

ANSF or Afghan civilians. As the mission in Afghanistan
1. Armed Forces Health Surveillance Center. Animal bites, active
shifts from combat to ANSF mentoring and reconstruction, and reserve components, U.S. Armed Forces, 2001–2010. MSMR.
US and coalition troops will come into increasingly close 2011;18:12–5.
contact with ANSF and Afghan civilians. Accordingly, 2. Stevens K, Dehgan A, Karlstetter M, Rawan F, Tawhid MI, Ostrows-
the likelihood of deployed US military members being ex- ki S, et al. Large mammals surviving conflict in the eastern forests
of Afghanistan. Oryx. 2011;45:265–71. http://dx.doi.org/10.1017/
posed to monkeys in Afghanistan will probably increase. S0030605310000517
However, although risk for contact with monkeys might 3. Abrahamian FM, Goldstein EJ. Microbiology of animal bite wound
increase, an increase in bites is not inevitable. Explicit or- infections. Clin Microbiol Rev. 2011;24:231–46. http://dx.doi.
ders prohibit deployed US military members from adopt- org/10.1128/CMR.00041-10
4. Estep RD, Messaoudi I, Wong SW. Simian herpesviruses and their
ing local mascots and from interacting with animals or pets risk to humans. Vaccine. 2010;28(Suppl 2):B78–84. http://dx.doi.
owned by ANSF or Afghan civilians. To mitigate the risk org/10.1016/j.vaccine.2009.11.026
for animal bites, it is crucial that commanders enforce these 5. Engel GA, Jones-Engel L, Schillaci MA, Suaryana KG, Putra A,
regulations (13). Fuentes A, et al. Human exposure to herpesvirus B–seropositive
macaques, Bali, Indonesia. Emerg Infect Dis. 2002;8:789–95. http://
The risk of being bitten by a monkey could increase dx.doi.org/10.3201/eid0808.010467
as US forces work more closely with ANSF and Afghan 6. Ostrowski SR, Leslie MJ, Parrott T, Abelt S, Piercy PE. B-virus
civilians. Bites could be prevented by appropriate emphasis from pet macaque monkeys: an emerging threat in the United
from command and enforcement of existing policies pro- States? Emerg Infect Dis. 1998;4:117–21. http://dx.doi.org/10.3201/
eid0401.980117
hibiting pet adoption and animal contact. Treatment of pa- 7. Centers for Disease Control and Prevention. CDC health informa-
tients who are bitten could be improved by further training tion for international travel 2012. The yellow book. Gary W. Bru-
of military health care providers on appropriate treatment nette, editor. New York: Oxford University Press; 2012.
for animal bites, including monkey bites. 8. Cohen JI, Davenport DS, Stewart JA, Deitchman S, Hilliard JK,
Chapman LE. Recommendations for prevention of and therapy for
exposure to B virus (Cercopithecine herpesvirus 1). Clin Infect Dis.
Acknowledgments 2002;35:1191–203. http://dx.doi.org/10.1086/344754
We thank David Broussard, Ronald Havard, Jason Lennen, 9. Centers for Disease Control and Prevention Updated recommenda-
tions for use of tetanus toxoid, reduced diphtheria toxoid and acel-
and James Reynolds for their review and critique of this study and
lular pertussis (Tdap) vaccine from the Advisory Committee on
report. We thank James Geracci for his review of this report and Immunization Practices, 2010. MMWR Morb Mortal Wkly Rep.
for his leadership. 2011;60:13–5.
10. World Health Organization. Rabies vaccines: WHO position pa-
The work reported here was performed by L.E.M. and per—recommendations. Vaccine. 2010;28:7140–2. http://dx.doi.
K.A.B. as part of their duties in the US Army, Combined Joint org/10.1016/j.vaccine.2010.08.082
Task Force-1, Afghanistan. They have no financial or nonfinan- 11. Centers for Disease Control and Prevention Use of a reduced (4-
dose) vaccine schedule for postexposure prophylaxis to prevent hu-
cial competing interests. man rabies—recommendations of the Advisory Committee on Im-
munization Practices. Ann Emerg Med. 2010;56:64–7. http://dx.doi.
Dr Mease was the Division Preventive Medicine Officer for
org/10.1016/j.annemergmed.2010.05.020
Combined Joint Task Force–1, Afghanistan and is the officer in 12. Tregle RW Jr, Loe CL, Earhart RH III, d’Autremont SB. Cerco-
charge at the Army Health Clinic in Utah. His research interests pithecine herpesvirus 1 risk in a child bitten by a bonnet macaque
include vector-borne diseases and dermatologic disorders in mili- monkey. J Emerg Med. 2011;41:e89–90. http://dx.doi.org/10.1016/j.
jemermed.2010.02.011
tary and occupational settings.
13. Chretien JP. Protecting service members in war–non-battle morbid-
Ms Baker was the Division Army Public Health Nurse for ity and command responsibility. N Engl J Med. 2012;366:677–9.
http://dx.doi.org/10.1056/NEJMp1112981
the Combined Joint Task Force-1, Afghanistan, and is a staff
public health nurse at the Community Health Resource Center,
Address for correspondence: Luke E. Mease, Army Health Clinic, 5116
General Leonard Wood Army Community Hospital in Missouri.
Kister Ave, Dugway, UT 84022, USA; email: luke.mease@us.army.mil
Her research interests include human papillomavirus, sexually
transmitted diseases, and latent tuberculosis infection in military
settings.

Search past issues of EID at www.cdc.gov/eid

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1649
DISPATCHES

Human Parvovirus 4 of PARV4 have also been suggested in South Africa (6),
Taiwan (8), India, (9), China (10), and Thailand (11).
in Nasal and Fecal PARV4 has been classified into 3 genotypes.
Genotypes 1 and 2 are found in North America, Europe, and
Specimens from Asia (1–3,9–11), and genotype 3 is found in in sub-Saharan

Children, Ghana
Africa (7,12). To investigate whether PARV4 is found in
the respiratory or intestinal tract, we analyzed previously
collected specimens from 1,904 children in Ghana.
Jan Felix Drexler, Ulrike Reber, Doreen Muth,
Petra Herzog, Augustina Annan, Fabian Ebach, The Study
Nimarko Sarpong, Samuel Acquah, Ethical approval for this study was provided by the
Julia Adlkofer, Yaw Adu-Sarkodie, Committee on Human Research Publication and Ethics,
Marcus Panning, Egbert Tannich, Jürgen May, Kwame Nkrumah University of Science and Technology,
Christian Drosten, and Anna Maria Eis-Hübinger Kumasi, Ghana. Informed consent was obtained from
parents or guardians of all children.
Nonparenteral transmission might contribute to human A total of 1,904 anonymous nasal and fecal
parvovirus 4 (PARV4) infections in sub-Saharan Africa. specimens were obtained during a study on molecular
PARV4 DNA was detected in 8 (0.83%) of 961 nasal samples
diagnostics for respiratory and enteric tract infections
and 5 (0.53%) of 943 fecal samples from 1,904 children in
Ghana. Virus concentrations ≈6–7 log10 copies/mL suggest
in symptomatic children <15 years of age at the
respiratory or fecal–oral modes of PARV4 transmission. Presbyterian Hospital in Agogo, Ghana. Nasal swab
specimens were obtained from children with upper
or lower respiratory tract symptoms. Fecal samples

H uman parvovirus 4 (PARV4; human partetravirus) is

a single-stranded DNA virus discovered in 2005 (1).
PARV4 has been detected in persons at risk for parenteral
were obtained from 504 children with gastrointestinal
symptoms (53.4% of sampled children; 294 [58.3%] of
symptomatic children with vomiting, 190 [37.7%] with
infections, suggesting blood-borne transmission (2,3) diarrhea, and 144 [28.6%] with acute malnutrition; 9
although other transmission routes have not been ruled out. [1.8%] with incomplete clinical data) and 439 (46.6%)
Studies in northern Europe demonstrated a high prevalence children without gastrointestinal symptoms.
of antibodies against PARV4 in injection drug users, A total of 961 nasal swabs were obtained during
persons co-infected with HIV and hepatitis C virus, and February–November 2008 from 520 boys and 441 girls
persons with hemophilia who were exposed to nonvirally (median age 19 months, range 0–162 months, interquartile
inactivated clotting factors; however, antibodies were not range 8–38 months). Nasal swabs were placed in 1.5 mL
detected in the general population (4,5). of RNAlater (QIAGEN, Hilden, Germany). A total of 943
In contrast, PARV4 seroprevalence was 25%–37% in fecal samples were obtained during May–October 2009
adults in the Democratic Republic of Congo, Cameroon, from 500 boys and 443 girls (median age 36 months, range
and Burkina Faso who were not infected with HIV and 0–165 months, interquartile range 17–78 months). Fecal
hepatitis C virus. (6). PARV4 DNA was detected in blood samples were prepared as 10% suspensions in phosphate-
of 8.6% of children 15 or 24 months of age in Ghana (7). buffered saline. No paired nasal and fecal specimens were
There was no history of exposure to multiple-use needles available from individual patients.
or blood transfusion in any of these children. These data Viral DNA was purified from 140 μL of nasal swab
suggested alternative modes of PARV4 transmission in suspension or 200 μL of fecal suspension by using
countries in Africa. Nonparenteral modes of transmission QIAamp Viral RNA and DNA Stool Mini Kits (QIAGEN),
respectively. Two real-time PCRs were performed. One
Author affiliations: University of Bonn Medical Centre, Bonn, primer/probe set was designed to detect PARV4 genotypes
Germany (J.F. Drexler, U. Reber, D. Muth, A. Annan, F. Ebach, C. 1 or 2 viruses (13), and a second primer set was designed
Drosten, A.M. Eis-Hübinger); Bernhard Nocht Institute for Tropical to detect PARV4 genotype 3 viruses (7). The sensitivity of
Medicine, Hamburg, Germany (P. Herzog, J. Adlkofer, E. Tannich, both protocols was 1–2 genome copies/reaction. Absolute
J. May); Kumasi Centre for Collaborative Research in Tropical quantification of PARV4 genome copy numbers relied on
Medicine, Kumasi, Ghana (A. Annan, N. Sarpong, S. Acquah); photometrically quantified genotype 3 plasmid standards,
Kwame Nkrumah University of Science and Technology, Kumasi (Y. as described (7).
Adu-Sarkodie); and Freiburg University Medical Center, Freiburg, To exclude bias from DNA purification methods,
Germany (M. Panning) PARV4-negative nasal and fecal specimens were
DOI: http://dx.doi.org/10.3201/eid1810.111373 spiked with quantified plasmid standards. Subsequent

1650 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 PARV4 in Nasal and Fecal Specimens from Children

Table. Nucleotide sequence divergence of parvovirus 4 strains from nasal swab and fecal samples from children, Ghana, from
genotype 1, 2, and 3 prototype strains*
Nucleotide position according to Nucleotide sequence divergence from parvovirus 4 reference strains, %
Specimen type GenBank accession no. Genotype 1 (GenBank Genotype 2 BR10627–5 Genotype 3 NG-OR
and no. EU874248 AY622943) (GenBank DQ873390) (GenBank EU874248)
Nasal swab
N1 1700–4660 6.56 7.39 0.92
N2 299–4660 7.51 8.07 0.88
N3 50–4660 7.37 8.38† 0.83
N4 1962–2056‡ 9.16 6.73 2.14
N4 2117–3413 4.97 5.31 0.93
N5 1962–2056 9.16 6.73 2.14
N5 2117–4183 5.50 6.34 0.98
N6 299–4660 7.51 8.10 0.90
N7 1962–2056 9.16 6.73 2.14
N7 2431–2914 6.24 7.01 1.25
N7 3068–3246 4.61 5.19 1.12
N8 624–3246 7.36 7.84 0.84
Feces
F1 1700–4183 6.20 6.82 0.89
F2 1700–4460 6.56 7.39 0.92
F3 1700–3716 6.08 6.52 0.85
F4 1700–4183 6.02 6.78 0.89
F5 1700–4183 6.93 6.73 1.04
*Pairwise nucleotide divergence was calculated by using the DNA distance matrix in BioEdit (www.mbio.ncsu.edu/BioEdit/bioedit.html).
†Because the homologs of the first 92 nt of strain N3 are not given in the prototype strain BR10627–5, calculation of divergence started at N3 nt position
93.
‡Nucleotide sequence of the PCR product (primer sequences trimmed) was amplified by using screening PCR designed for detection of PARV4 genotype
3 as described (7).

quantification was equivalent between techniques and

specimens, and differences between specimen types in
several experiments were <0.5 log10 copies/mL. Standard
procedures were used to prevent PCR contamination.
Determination of PARV4 genotypes was conducted by
nucleotide sequencing of several genomic target regions
(Table).
Eight (0.83%) of 961 nasal swabs and 5 (0.53%) of 943
fecal samples tested were positive for PARV4 DNA. Virus
concentrations ranged from 1.3 × 103 to 1.8 × 107 copies/
mL (median 1.0 × 104 copies/mL) in nasal swab suspensions
and from 2.3 × 103 to 4.6 × 106 copies/mL (median
6.8 × 104 copies/mL) in fecal suspensions (Figure 1).
The difference in virus concentrations between the 2 groups
was not significant (p = 0.056, by Mann-Whitney U test).
Nucleotide sequencing of amplicons generated by
screening PCRs and sequencing of additional genomic
regions classified all viruses as PARV4 genotype 3 (Table)
(GenBank accession numbers JN183920–JN183932). This
result was confirmed by phylogenetic analysis of a 483-
nt fragment of the capsid-encoding open reading frame 2
(Figure 2).
Ages of the 8 children with PARV4-positive nasal Figure 1. Parvovirus 4 DNA loads in virus-positive nasal and fecal
swab specimens ranged from 9 to 58 months (median 32 specimens from children, Ghana. Virus concentrations are given
months). Ages of the 5 children with PARV4-positive on a log scale on the y-axis. Each dot represents 1 specimen.
Horizontal lines represent median values for each sample type.
fecal samples were 1, 36, 43, 57, and 124 months. Nasal
For calculation of statistical significance of the difference in viral
swab specimens with the highest viral loads were from a quantities between sample types, the Mann-Whitney U test was
9-month-old boy and a 29-month-old girl. Fecal samples used. Virus quantities in nasal swabs and feces are given for
with the highest viral loads were from 2 boys 43 and 57 sample suspensions (nasal swabs in 1.5 mL of stabilizing reagent
months of age. and feces in a 10% suspension in phosphate-buffered saline).

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1651
DISPATCHES

drug users to the general population. Likewise, the higher

prevalence of PARV4 antibodies in HIV-infected blood
donors in South Africa compared with uninfected donors (6)
appears incompatible with PARV4 transmission primarily by
the respiratory route. Therefore, our results do not contradict
those of a study conducted in Scotland, which showed no
PARV4 in respiratory specimens (15).
Because of the small number of children with PARV4
DNA in nasal or fecal specimens, correlation of infection
with age groups was not possible. A limitation of our
study was the lack of blood specimens from children with
current respiratory or fecal PARV4 shedding, and serologic
studies are needed to evaluate susceptibility of different
age groups to PARV4 infection. Furthermore, detection
of PARV4 in patients with respiratory disease does not
indicate that PARV4 was the cause of the disease. In 5 of
8 PARV4-positive nasal swabs, typical respiratory viruses
Figure 2. Phylogenetic analysis of a 483-nt fragment of the (parainfluenza virus, influenza A virus, rhinovirus) were
parvovirus 4 (PARV4) capsid-encoding open reading frame (ORF) also detected and the pattern of symptoms in PARV4-
2 for PARV4 strains identified in children, Ghana. Neighbor-joining positive children did not differ from symptoms in PARV4-
phylogeny was conducted in MEGA5.05 (www.megasoftware.net) negative children. Similarly, 3 of 5 children with PARV4-
by using a gap-free ORF2 fragment corresponding to positions
2,432–2,914 in the PARV4 genotype 3 prototype strain NG-OR
positive feces did not have gastrointestinal symptoms at
(GenBank accession no. EU874248) with a nucleotide percentage the time of fecal sampling. One child had vomiting and
distance substitution model and 1,000 bootstrap replicates. another child had vomiting and diarrhea. Moreover, in 3 of
Scale bar indicates percentage uncorrected nucleotide distance. these 5 children, in addition to PARV4, Giardia lamblia, a
Previously published PARV4 sequences are given with strain potential cause of diarrhea, was also detected.
names (if available) and GenBank accession numbers. Viruses
newly identified are in boldface. The source of PARV4 strains
Although data for exposure and risk factors and paired
identified in the study is indicated by capital letters (N, nasal samples were not available, suggested transmission routes
specimen; F, fecal specimen). PARV4 genotypes are given to the might explain the high infection rates in western Africa.
right of taxa. A chimpanzee partetravirus was used as the outgroup. Further studies are needed to assess the effect of PARV4
excretion on virus epidemiology and the chronology of
PARV4 infection.
Conclusions
We found PARV4 in 0.8% of nasal swab specimens Acknowledgments
and 0.5% of fecal specimens from 2 groups of children We thank the children and their parents for participating in
in Ghana symptomatic for respiratory illness and with or the study and Carmen Poster for excellent technical assistance.
without diarrheal illness, respectively. Our results provide
This study was supported by grants from the Union Bank of
evidence to suggest that the higher prevalence of PARV4
Switzerland Optimus Foundation to E.T., J.M., C.D. and Y.A.;
reported among adults in countries in western Africa (6)
European Union project European Management Platform for
might be caused by transmission by the respiratory or
Emerging and Re-emerging Infectious Disease Entities (grant
fecal–oral route.
223498); the German Academic Exchange Service to J.A.; the
However, demonstration of PARV4 in the respiratory
German Research Foundation (grant DR 772/3-1); and BONFOR
tract and feces does not identify a transmission route.
to A.M.E.-H. (grant O-151.0021).
PARV4 in the respiratory tract could be caused by high
viremia, which was recently reported in a child in India with Dr Drexler is a physician and clinical virologist at the
a genotype 2 infection (9) and in 2 patients with hemophilia University of Bonn. His research interest is characterization of
in the United Kingdom, 1 with a genotype 1 infection and 1 novel human and zoonotic viruses.
with a genotype 2 infection (14).
It is unclear to what extent the putative nonparenteral References
transmission routes of PARV4 genotype 3 in western Africa
apply to other areas. Markedly lower PARV4 antibody 1. Jones MS, Kapoor A, Lukashov VV, Simmonds P, Hecht F, Delwart
E. New DNA viruses identified in patients with acute viral infection
prevalences observed in Europe (4,5) argue against PARV4
syndrome. J Virol. 2005;79:8230–6. http://dx.doi.org/10.1128/
spread by nonparenteral routes, e.g., from infected injection JVI.79.13.8230-8236.2005

1652 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 PARV4 in Nasal and Fecal Specimens from Children

2. Fryer JF, Lucas SB, Padley D, Baylis SA. Parvoviruses PARV4/5 in 10. Yu X, Zhang J, Hong L, Wang L, Yuan Z, Zhang X, et al. High
hepatitis C virus–infected patient. Emerg Infect Dis. 2007;13:175–6. prevalence of human parvovirus 4 infection in HBV and HCV
http://dx.doi.org/10.3201/eid1301.060856 infected individuals in Shanghai. PLoS ONE. 2012;7:e29474. http://
3. Simmonds P, Manning A, Kenneil R, Carnie FW, Bell JE. Parenteral dx.doi.org/10.1371/journal.pone.0029474
transmission of the novel human parvovirus PARV4. Emerg Infect 11. Lurcharchaiwong W, Chieochansin T, Payungporn S, Theamboonlers
Dis. 2007;13:1386–8. http://dx.doi.org/10.3201/eid1309.070428 A, Poovorawan Y. Parvovirus 4 (PARV4) in serum of intravenous
4. Sharp CP, Lail A, Donfield S, Simmons R, Leen C, Klenerman P, et al. drug users and blood donors. Infection. 2008;36:488–91. http://
High frequencies of exposure to the novel human parvovirus PARV4 dx.doi.org/10.1007/s15010-008-7336-4
in hemophiliacs and injection drug users, as detected by a serological 12. Simmonds P, Douglas J, Bestetti G, Longhi E, Antinori S, Parravicini
assay for PARV4 antibodies. J Infect Dis. 2009;200:1119–25. http:// C, et al. A third genotype of the human parvovirus PARV4 in sub-
dx.doi.org/10.1086/605646 Saharan Africa. J Gen Virol. 2008;89:2299–302. http://dx.doi.
5. Lahtinen A, Kivela P, Hedman L, Kumar A, Kantele A, Lappalainen org/10.1099/vir.0.2008/001180-0
M, et al. Serodiagnosis of primary infections with human parvovirus 13. Fryer JF, Delwart E, Hecht FM, Bernardin F, Jones MS, Shah N,
4, Finland. Emerg Infect Dis. 2011;17:79–82. http://dx.doi. et al. Frequent detection of the parvoviruses, PARV4 and PARV5,
org/10.3201/eid1701.100750 in plasma from blood donors and symptomatic individuals.
6. Sharp CP, Vermeulen M, Nebie Y, Djoko CF, LeBreton M, Tamoufe Transfusion. 2007;47:1054–61. http://dx.doi.org/10.1111/j.1537-
U, et al. Epidemiology of human parvovirus 4 infection in sub- 2995.2007.01235.x
Saharan Africa. Emerg Infect Dis. 2010;16:1605–7. http://dx.doi. 14. Sharp CP, Lail A, Donfield S, Gomperts ED, Simmonds P. Virologic
org/10.3201/eid1701.100750 and clinical features of primary infection with human parvovirus
7. Panning M, Kobbe R, Vollbach S, Drexler JF, Adjei S, Adjei O, 4 in subjects with hemophilia: frequent transmission by virally
et al. Novel human parvovirus 4 genotype 3 in infants, Ghana. inactivated clotting factor concentrates. Transfusion. 2012;52:1482–
Emerg Infect Dis. 2010;16:1143–6. http://dx.doi.org/10.3201/ 9. http://dx.doi.org/10.1111/j.1537-2995.2011.03420.x
eid1607.100025 15. Manning A, Russell V, Eastick K, Leadbetter GH, Hallam N,
8. Yang SJ, Hung CC, Chang SY, Lee KL, Chen MY. Immunoglobulin Templeton K, et al. Epidemiological profile and clinical associations
G and M antibodies to human parvovirus 4 (PARV4) are frequently of human bocavirus and other human parvoviruses. J Infect Dis.
detected in patients with HIV-1 infection. J Clin Virol. 2011;51:64– 2006;194:1283–90. http://dx.doi.org/10.1086/508219
7. http://dx.doi.org/10.1016/j.jcv.2011.01.017
9. Benjamin LA, Lewthwaite P, Vasanthapuram R, Zhao G, Sharp Address for correspondence: Anna Maria Eis-Hübinger, Institute of
C, Simmonds P, et al. Human parvovirus 4 as potential cause of
Virology, University of Bonn Medical Centre, Sigmund-Freud-Str. 25,
encephalitis in children, India. Emerg Infect Dis. 2011;17:1484–7.
http://dx.doi.org/10.3201/eid1708.110165 D-53127 Bonn, Germany; email: anna-maria.eis-huebinger@ukb.uni-
bonn.de

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1653
DISPATCHES

Hepatitis E Virus reflects the total adult population with respect to age,
sex, and geographic region, but persons with migration
Seroprevalence background are underrepresented (non-German citizenship
4.6% in the sample vs. 8.7% in the total adult population).
among Adults, Serum samples were screened for HEV IgG by using

Germany
the recomLine HEV-IgG/IgM immunoassay (Mikrogen,
Neuried, Germany). The assay is based on 7 recombinantly
expressed antigens of genotypes 1 and 3 of open reading
Mirko S. Faber,1 Jürgen J. Wenzel,1 frames 2 and 3. According to the manufacturer’s and our
Wolfgang Jilg, Michael Thamm, Michael Höhle, data (J.J. Wenzel et al., unpub. data), the test is 97%–100%
and Klaus Stark sensitive for detecting acute or previous HEV infections.
Test strips were scanned with the semiautomatic recomScan
We assessed hepatitis E virus (HEV) antibody software (Mikrogen). The intensity of 3 quality assurance
seroprevalence in a sample of the adult population in and other bands was determined by densitometrical
Germany. Overall HEV IgG prevalence was 16.8% (95% CI
detection algorithms. Each antigen band with an intensity
15.6%–17.9%) and increased with age, leveling off at >60
years of age. HEV is endemic in Germany, and the lifetime
greater or equal to the cutoff was assigned a point value. The
risk for exposure is high. final results were classified into 3 categories: no antibodies
detectable (negative), test inconclusive (borderline), and
antibodies detectable (positive). Persons whose test results

I n industrialized countries, hepatitis E virus (HEV)

has long been regarded as a rare imported infection.
However, sporadic cases without travel to disease-endemic
were borderline (n = 70) were excluded from further
analysis.
We poststratified the remaining survey population
areas and caused by genotype 3 are being increasingly (n = 4,352) by age group and location of residence
reported (1,2). Epidemiologic and molecular studies have (16 states) to account for per protocol oversampling in
implicated undercooked pork and wild boar products as eastern Germany and to restore the distribution of age
a source of HEV infection (3–5). An unexpectedly high groups to match the distribution in the total population.
prevalence of HEV-specific antibodies, e.g., among blood Weighted seroprevalence estimates were calculated by
donors, has been shown by several studies in Europe and using survey-weighted logistic regression. Associations
the United States (6–11). between demographic characteristics and seropositivity
In Germany, the number of notified hepatitis E cases were analyzed by using adjusted Wald test p values.
rose from <50 annually in 2001–2003 to 238 in 2011 We also estimated mean annual incidence of HEV
(incidence 0.3/100,000 population); the proportion of seroconversion from the seroprevalence data by using
autochthonous cases increased from 30%–40% to 78%. a catalytic model with age-constant force of infection,
We conducted a study to determine HEV seroprevalence similar to that of Faramawi et al. (12). Detailed methods
in Germany’s adult population and associations with and underlying assumptions are described in the online
sociodemographic characteristics by using an assay highly Technical Appendix (wwwnc.cdc.gov/EID/pdfs/11-1756-
sensitive for HEV genotype 3. Techapp.pdf).
The 4,352 persons who were included in the analysis
The Study were from 108 communities of all federal states in
We assessed HEV seroprevalence in a large Germany (Table 1). Weighted prevalence of HEV IgG was
subsample (n = 4,422) of the 2008–2011 German Health 16.8% (95% CI 15.6%–17.9%); prevalence ranged from
Examination Survey for Adults (Deutscher Erwachsenen 6.1% (95% CI 4.5%–7.8%) in the 18–34-year age group to
Gesundheitssurvey; www.degs-studie.de), a 2-stage >20% in the >50-year groups, with a maximum of 26.4%
national probability sample that assessed the health status (95% CI 21.6%–31.1%) in the 60–64-year group (Figure).
in the general population. The sampling frame comprised In the univariable analysis (Table 2), only age group was
persons 18–79 years of age whose principal residence significantly associated with seropositivity (p<0.01); results
was in Germany and who were fluent in German. Overall were not significant for sex (p = 0.97), residence (northern/
response was 48.4% (7,116 respondents). Our subsample middle/southern Germany, p = 0.29; west/east, p = 0.43),
or population of municipality (4 categories; p = 0.10). In
Author affiliations: Robert Koch Institute, Berlin, Germany
separate multivariable models, each including age group
(M.S. Faber, M. Thamm, M. Höhle, K. Stark); and University of
and 1 other variable, age remained the only significant
Regensburg, Regensburg, Germany (J.J. Wenzel, W. Jilg)

DOI: http://dx.doi.org/10.3201/eid1810.111756 1
These authors contributed equally to this article.

1654 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Hepatitis E Virus Seroprevalence, Germany

Table 1. Comparison of demographic characteristics of persons lower (10,11). Reasons for these differences could be
in study of hepatitis E virus seroprevalence and general adult effects of sample selection, different lifetime exposures
population, Germany, 2008–2011 (e.g., to foods that may serve as transmission vehicles),
% Study % Total
population, n population, n =
or use of different test systems with varying sensitivity.
Characteristic = 4,352 64,139,871* Mansuy et al. recently reported 53% prevalence of HEV
Age group, y antibodies in blood donors in southwestern France (13),
18–19 3.2 2.9 a figure considerably higher than the 17% prevalence
20–29 12.3 15.5
30–39 11.5 15.6 reported earlier for the same geographic region, when a
40–49 18.4 21.7 different test system was used (8). The assay applied in our
50–59 19.5 17.9 study was designed to also detect previous infections with
60–69 19.7 14.3
70–79 15.4 12.2 HEV genotype 3; therefore, it is likely to be more sensitive
Female sex 51.3 50.3 than assays used in previous studies.
State of residence Our data show that HEV exposure is common among
Baden-Württemberg 11.4 13.0
Bavaria 10.7 15.2
the general adult population in Germany, which is consistent
Berlin 3.4 4.4 with increasing evidence for pigs as a reservoir for
Brandenburg 6.2 3.2 foodborne transmission of HEV in industrialized countries.
Bremen 0.9 0.8 HEV seroprevalence is high in domestic pig herds in
Hamburg 1.4 2.2
Hesse 6.4 7.4 Germany and other countries; closely related HEV strains
Mecklenburg-Vorpommern 3.8 2.1 were found in pig livers on retail sale and in autochthonous
Lower Saxony 11.2 9.5 cases of hepatitis E, and HEV seroprevalence was higher in
North Rhine-Westphalia 14.8 21.6
Rhineland-Palatinate 5.1 4.8 persons with occupational exposure to pigs than in control
Saarland 2.2 1.3 groups (4,10,14,15).
Saxony 7.1 5.3 Our data and other studies (11–14) have shown no
Saxony-Anhalt 6.2 3.0
Schleswig-Holstein 2.6 3.4
significant difference in seroprevalence between sexes,
Thuringia 6.7 2.9 despite an assumed higher frequency of alimentary or
*Total no. adults 18–79 years of age as of December 31, 2009 occupational exposure and the higher incidence of clinical
(www.destatis.de).
cases among men. This finding may indicate sex-specific
variable. Mean annual incidence of HEV seroconversion differences in disease development or application of
estimated from the catalytic model was 3.9 (95% CI 3.6%– laboratory testing or that foods frequently consumed by both
4.2%) per 1,000 population. sexes play a substantial role as vehicles for transmission.
These results also may highlight the lack of evidence for 1
Conclusions main risk factor or food vehicle (14).
We found an overall HEV seroprevalence of 16.8% The strong association between age and HEV
among adults in Germany; seroprevalence increased with seroprevalence in our study most likely reflects
age but was not dependent on sex or location of residence. cumulative lifetime exposure to the virus. However, HEV
Similarly high seroprevalence was found in blood donors seroprevalence remains relatively stable from 18 to 34
in Denmark (20.6%), southwestern England (16%), and the years of age, which could indicate a birth cohort effect
United States (18%) (6,7,9). Estimates from Switzerland resulting from a decrease in the overall risk over the past
and the Netherlands, on the other hand, are considerably few decades, as was reported for Denmark (6). The leveling
Figure. Estimated prevalence
of hepatitis E virus (HEV) IgG,
by age group, Germany, 2008–
2011. Error bars indicate 95%
CIs.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1655
DISPATCHES

Table 2. Univariable survey-weighted logistic regression Acknowledgments

estimates of seroprevalence of hepatitis E virus IgG, by age, sex, We thank Monika Erl and Bianka Ehrlich for expert technical
and location of residence, Germany, 2008–2011* assistance and Heribert Stolzenberg and Panagiotis Kamtsiuris for
Seroprevalence, %
Characteristic (95% CI)
critical review of the manuscript.
Total 16.8 (15.6–17.9) No external funding was received for this study. The study
Age group, y
18–19 5.3 (2.5–11.1) was approved by the ethical review board of the Charité University
20–29 6.6 (4.8–9.2) Hospital, Berlin.
30–39 8.0 (5.8–10.8)
40–49 16.9 (14.3–19.9) Dr Faber is a medical epidemiologist at the Robert Koch
50–59 22.6 (19.7–25.7) Institute, Department for Infectious Disease Epidemiology
60–69 26.1 (23.0–29.4)
70–79 23.8 (20.4–27.5) (Gastrointestinal Infections, Zoonoses and Tropical Infections
Sex Unit).
M 16.8 (15.2–18.5)
F 16.7 (15.2–18.5)
Place of residence (north–south) References
Northern states 15.5 (13.5–17.6)
Middle states 16.6 (14.8–18.6) 1. Mushahwar IK. Hepatitis E virus: molecular virology, clinical
Southern states 17.9 (15.8–20.1) features, diagnosis, transmission, epidemiology, and prevention. J
Place of residence (east-west) Med Virol. 2008;80:646–58. http://dx.doi.org/10.1002/jmv.21116
Western states 16.6 (15.2–18.0) 2. Dalton HR, Hazeldine S, Banks M, Ijaz S, Bendall R. Locally
Eastern states 17.6 (15.6–19.7) acquired hepatitis E in chronic liver disease. Lancet. 2007;369:1260.
Population of municipality http://dx.doi.org/10.1016/S0140-6736(07)60595-9
<5,000 17.9 (15.3–20.8) 3. Adlhoch C, Wolf A, Meisel H, Kaiser M, Ellerbrok H, Pauli G.
5,000–<20,000 17.5 (15.2–20.1) High HEV presence in four different wild boar populations in East
20,000–<100,000 14.4 (12.4–16.6)
and West Germany. Vet Microbiol. 2009;139:270–8. http://dx.doi.
>100,000 17.7 (15.7–20.0)
org/10.1016/j.vetmic.2009.06.032
*n = 4,352.
4. Wenzel JJ, Preiss J, Schemmerer M, Huber B, Plentz A, Jilg
W. Detection of hepatitis E virus (HEV) from porcine livers in
off of the seroprevalence above age 60 years could be southeastern Germany and high sequence homology to human HEV
caused by a loss of antibodies in the elderly. isolates. J Clin Virol. 2011;52:50–4. http://dx.doi.org/10.1016/j.
We found a striking difference between the estimated jcv.2011.06.006
5. Wichmann O, Schimanski S, Koch J, Kohler M, Rothe C, Plentz
annual incidence of seroconversion and the relatively A, et al. Phylogenetic and case-control study on hepatitis E virus
low incidence of notified disease in Germany. Besides infection in Germany. J Infect Dis. 2008;198:1732–41. http://dx.doi.
underdiagnosis, a possible explanation is a high proportion org/10.1086/593211
of asymptomatic infections. 6. Christensen PB, Engle RE, Hjort C, Homburg KM, Vach W, Georgsen
J, et al. Time trend of the prevalence of hepatitis E antibodies among
The survey response rate is similar to those achieved in farmers and blood donors: a potential zoonosis in Denmark. Clin
other comprehensive national health examination surveys Infect Dis. 2008;47:1026–31. http://dx.doi.org/10.1086/591970
in Europe. In the statistical analysis, we corrected for 7. Dalton HR, Stableforth W, Thurairajah P, Hazeldine S, Remnarace
differences in the distribution of age and place of residence R, Usama W, et al. Autochthonous hepatitis E in Southwest England:
natural history, complications and seasonal variation, and hepatitis E
between the sample and the general population. However, virus IgG seroprevalence in blood donors, the elderly and patients
foreign-born persons are underrepresented in the survey with chronic liver disease. Eur J Gastroenterol Hepatol. 2008;20:784–
because participation required fluency in German. A 90. http://dx.doi.org/10.1097/MEG.0b013e3282f5195a
high proportion of foreign-born persons in Germany are 8. Mansuy JM, Legrand-Abravanel F, Calot JP, Peron JM, Alric L,
Agudo S, et al. High prevalence of anti–hepatitis E virus antibodies
Muslim, and their avoidance of raw pork products likely in blood donors from south west France. J Med Virol. 2008;80:289–
decreases their risk for HEV infection, which means we 93. http://dx.doi.org/10.1002/jmv.21056
may have slightly overestimated the seroprevalence in the 9. Meng XJ, Wiseman B, Elvinger F, Guenette DK, Toth TE, Engle RE,
adult general population. et al. Prevalence of antibodies to hepatitis E virus in veterinarians
working with swine and in normal blood donors in the United States
Autochthonous HEV infections are common in the and other countries. J Clin Microbiol. 2002;40:117–22. http://dx.doi.
industrialized world, and testing for HEV is indicated in org/10.1128/JCM.40.1.117-122.2002
patients with acute hepatitis, irrespective of history of 10. Bouwknegt M, Engel B, Herremans MM, Widdowson MA, Worm
travel to developing countries. Recommendations against HC, Koopmans MP, et al. Bayesian estimation of hepatitis E virus
seroprevalence for populations with different exposure levels to
eating undercooked pork products exist, but further studies swine in the Netherlands. Epidemiol Infect. 2008;136:567–76.
are needed to specify implicated food products and to http://dx.doi.org/10.1017/S0950268807008941
better target recommendations for safer food production, 11. Kaufmann A, Kenfak-Foguena A, Andre C, Canellini G, Burgisser P,
preparation, and consumption. Moradpour D, et al. Hepatitis E virus seroprevalence among blood
donors in southwest Switzerland. PLoS One. 2011;6:e21150. http://
dx.doi.org/10.1371/journal.pone.0021150

1656 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Hepatitis E Virus Seroprevalence, Germany

12. Faramawi MF, Johnson E, Chen S, Pannala PR. The incidence of 15. Krumbholz A, Mohn U, Lange J, Motz M, Wenzel JJ, Jilg W, et al.
hepatitis E virus infection in the general population of the USA. Prevalence of hepatitis E virus-specific antibodies in humans with
Epidemiol Infect. 2011;139:1145–50. http://dx.doi.org/10.1017/ occupational exposure to pigs. Med Microbiol Immunol (Berl).
S0950268810002177 2012;201:239–44. http://dx.doi.org/10.1007/s00430-011-0210-5
13. Mansuy JM, Bendall R, Legrand-Abravanel F, Saune K, Miedouge
M, Ellis V, et al. Hepatitis E virus antibodies in blood donors, France. Address for correspondence: Mirko S. Faber, Department for Infectious
Emerg Infect Dis. 2011;17:2309–12. http://dx.doi.org/10.3201/
Disease Epidemiology, Robert Koch Institute, DGZ-Ring 1, 13086 Berlin,
eid1712.110371
14. Lewis HC, Wichmann O, Duizer E. Transmission routes and risk Germany; email: faberm@rki.de
factors for autochthonous hepatitis E virus infection in Europe: a
systematic review. Epidemiol Infect. 2010;138:145–66. http://
dx.doi.org/10.1017/S0950268809990847

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1657
DISPATCHES

Scarlet Fever For comparison, we performed a retrospective

review of hospital discharge records kept by public
Epidemic, hospitals. We extracted records of patients hospitalized
during January 2008–July 2011 who had diagnoses that
Hong Kong, 2011 are known complications of scarlet fever, including
toxic shock syndrome, acute rheumatic fever, and acute
Emma Y.Y. Luk, Janice Y.C. Lo, Amy Z.L. Li, glomerulonephritis. These cases were reviewed to
Michael C.K. Lau, Terence K.M. Cheung, determine whether the complications were related to scarlet
Alice Y.M. Wong, Monica M.H. Wong, fever.
Christine W. Wong, Shuk-kwan Chuang, Bacterial culture of S. pyogenes was performed on
and Thomas Tsang diagnostic specimens in hospital laboratories and the
Public Health Laboratory Centre of the Department of
More than 900 cases of scarlet fever were recorded Health; the latter serves as the diagnostic and public health
in Hong Kong during January–July, 2011. Six cases were reference laboratory in Hong Kong. Antimicrobial drug
complicated by toxic shock syndrome, of which 2 were susceptibility testing, emm typing, and detection of various
fatal. Pulsed-field gel electrophoresis patterns suggested virulence genes were performed at the Public Health
a multiclonal epidemic; emm12 was the predominant
Laboratory Centre on S. pyogenes isolates received during
circulating type. We recommend genetic testing of and
antimicrobial resistance monitoring for this reportable
2011 and archived during 2008–2010 (1). Pulsed-field gel
disease. electrophoresis (PFGE) was performed on the basis of the
gram-positive protocol, and PFGE profiles were analyzed
by using BioNumerics 5.0 software (Applied Maths, Sint-

S carlet fever is caused by infection with Streptococcus

pyogenes and mainly affects children. An upsurge of
scarlet fever occurred in Hong Kong, People’s Republic of
Martens-Latem, Belgium).
In June 2011, the Department of Microbiology of
the University of Hong Kong announced the discovery
China, in 2011, exceeding baseline annual incidence rates of a unique 48-kb insertion sequence in the genome of S.
for the previous 2 decades. We investigated possible changes pyogenes isolated from a blood specimen from a 7-year-old
in clinical severity, transmissibility, and characteristics of girl who died of scarlet fever (2). We tested for this insert
the causative pathogen for this outbreak. in a sample of strains collected during 2008–2011 using the
method provided by the University of Hong Kong.
The Study During January 1–July 31, 2011, a total of 996 cases
Scarlet fever is a statutory notifiable disease in Hong of scarlet fever were reported, greatly exceeding the annual
Kong. A clinical case is defined as illness in a person number of cases reported during 2008 (235), 2009 (187),
who has clinical features of scarlet fever (fever and and 2010 (128). Outbreak activity in 2011 peaked at week
fine, sandpaper rash of characteristic distribution that 26 (week ending June 25) (Figure 1). During the outbreak
blanches on pressure, with or without strawberry tongue, period (January–July 2011), the annualized incidence rate
desquamation, or sore throat). A confirmed case is defined was 24.0/100,000 population, ≈9× higher than the average
as a clinical case with positive throat or wound culture for annualized incidence rate of 2.62/100,000 population
S. pyogenes or antistreptolysin titer >200. during the baseline period of 2008–2010. During the
Epidemiologic, clinical, and laboratory data were previous 2 decades, baseline annual incidence rates ranged
collected by standard questionnaire for every reported case. from 0.0351 to 3.37 cases/100,000 population.
A cluster was defined as >2 cases in persons sharing the same Table 1 compares the epidemiologic features, clinical
residential or school address within the incubation period. features, and laboratory results for scarlet fever cases
We compared epidemiologic, clinical, and microbiological reported during 2011 and 2008–2010. Highest incidence
features of the scarlet fever cases from January–July 2011 (547 cases/100,000 population) was reported for children
(outbreak period) with features of those reported during 4–7 years of age (Table 1). Clinical features, complications,
2008–2010 (baseline period). We used SPSS version 14.0 and case-fatality rate for cases reported in 2011 were largely
(SPSS Inc., Chicago, IL, USA) for analyses; p<0.05 was comparable to those reported during the baseline period.
considered significant. The proportion of case-patients requiring hospitalization
during 2011 was lower, and mean duration of hospital stay
was ≈0.5 days shorter than for the baseline period. Details
Author affiliation: Centre for Health Protection, Hong Kong, People’s
of the 9 complicated cases are shown in Table 2.
Republic of China
Among the 996 scarlet fever cases reported during
DOI: http://dx.doi.org/10.3201/eid1810.111900 January–July 2011, S. pyogenes isolates from samples

1658 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Scarlet Fever, Hong Kong, 2011

Figure 1. Weekly number of scarlet fever cases,

by onset date, Hong Kong, January–July 2011.
White bars indicate clinically diagnosed but not
laboratory-confirmed cases; solid bars indicate
laboratory-confirmed cases. Solid triangle in-
dicates May 30 dissemination of press release
about first fatal case (in a 7-year-old girl); open
triangle indicates June 21 dissemination of press
release about second fatal case (in a 5-year-old
boy); circle indicates June 23 launch of health
education campaign.

from 90 patients (mostly throat swab specimens) were were susceptible to penicillin, but 77 (85.6%) strains
characterized. Strains found belonged to the following were resistant to erythromycin. Of 59 strains tested for
emm types (number and percentage of strains): emm12 (70, the 48-kb insert, 78.0% (46 strains) tested positive; 39
77.8%), emm1 (14, 15.6%), emm4 (2, 2.2%), emm22 (2, were emm12 strains and 7 emm1 strains. Antimicrobial
2.2%), emm2 (1, 1.1%), and emm3 (1, 1.1%). All strains drug susceptibility results were available for 39 strains

Table1. Epidemiologic characteristics, clinical features, and laboratory results for scarlet fever cases reported in Hong Kong during
January–July 2011 compared with cases reported during 2008–2010
Characteristic* 2008–2010, n = 550 January 1–July 31, 2011, n = 996 p value
Epidemiology
Sex ratio, M:F 1.6:1 1.5:1 0.50
Age range (median) 9 mo–40 y (5 y) 1 mo–51 y (6 y) 0.40
Local cases 98.0 (539/550) 97.4 (970/996) 0.56
Clustering
Cases in a cluster 5.45 (30/550) 14.4 (143/996) <0.0001†
Cases in home clusters 3.3 (18/550); 9 clusters 6.5 (65/996); 31 clusters
Cases in each home cluster, range 2 (2) 2–3 (2) 0.34
(median)
Cases in school clusters 2.2 (12/550); 4 clusters 7.8 (78/996); 28 clusters
Persons affected in each school cluster, 2–4 (3) 2–7 (2) 0.42
no. (median)
Clinical features
Fever 95.6 (526/548) 93.2 (928/996) 0.065
Sandpaper rash 97.4 (534/548) 95.4 (950/996) 0.13
Strawberry tongue 45.1 (248/550) 51.4 (512/996) 0.020‡
Sore throat 74.4 (409/550) 78.5 (782/996) 0.073
Desquamation 27.8 (153/550) 23.7 (236/996) 0.084
Hospitalization 63.9 (351/549) 56.6 (561/991) 0.005§
Duration of hospitalization, d (mean) 1– 25 (3.8) 1–33 (3.3) 0.005¶
Concomitant chickenpox 5.5 (30/550) 1.9 (19/996) 0.0002#
Complications** 0.73 (4/550) 0.90 (9/996) 0.79
Toxic shock syndrome 0.18 (1/550) 0.60 (6/996) 0.43
Case-fatality rate 0 0.20 (2/996) 0.54
Laboratory results
Laboratory confirmation 46.0 (253/550) 51.8 (533/996) 0.0055††
Positive throat or wound culture 95.3 (241/253) 97.2 (521/533) 0.094
Antistreptolysin O titer >200 IU/mL 4.74 (12/253) 4.37 (12/533) 0.094
*Values are % (no./total no.) unless otherwise indicated. Lower denominators indicate data missing or not applicable.
†F2 = 27.
‡F2 = 5.40.
§F2 = 7.85.
¶t = –2.8 (95% CI of difference 0.15–0.83 d).
#F2 = 14.
**Complications include toxic shock syndrome, septicaemia, parapharyngeal abscess, rheumatic fever, quinsy and hepatitis.
††F2 = 7.7.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1659
DISPATCHES

Table 2. Clinical characteristics of patients with scarlet fever who had medical complications, Hong Kong, January–July 2011*
Case-patient no.
Characteristic 1 2 3 4 5 6 7 8 9
Patient age, y/sex 14/M M/11 8/F 7/F 5/M 6/M 3/M 2/F 12/M
Month of illness onset April April April May June June July July July
Days from symptom onset 1 4 7 7 4 10 1 1 0
to hospital admission
Complications TSS Parapharyngeal TSS TSS TSS Septicemia TSS TSS Septicemia
abscess
Intensive care unit No No Yes Yes Yes Yes No Yes No
admission
Concomitant chickenpox No No No No Yes No Yes No Yes
infection
Recovered Yes Yes Yes No (died) No (died) Yes Yes Yes Yes
S. pyogenes isolates Throat Throat None Blood, Blood and Blood Throat Throat Blood and
lower limb pus pus
blister fluid
emm type NA NA NA emm12 emm1 emm12 NA NA emm12
48-kb insert NA NA NA + + + NA NA –
Virulence geneprofile NA NA NA ++++ +++++ +++++ NA NA + + + + 
(speA¸speB, speC, speF,
speH, ssa)
*All case-patients were healthy before infection. TSS, toxic shock syndrome; NA, not available; +, positive; –, negative.

positive for the 48-kb insert; 3 (7.70%) were susceptible type of virulence gene profile or presence of the 48-kb insert
to erythromycin. Conversely, among all erythromycin- was associated with increased incidence or severity.
resistant emm12 strains in 2011 tested for the 48-kb insert, The reasons for the upsurge remain obscure.
6/42 (14.3%) yielded a negative result. Laboratory findings showed diverse patterns of S. pyogenes
Forty-eight emm12 isolates during January–June 2011 strains, suggesting a multiclonal epidemic. The 48-kb insert
that were subjected to virulence gene profiling showed 5 identified in 2011 was found in S. pyogenes strains isolated
virulence gene profiles. No particular virulence gene profile in 2008–2010, albeit at lower rate (35.7% in 2008–2010
was dominant among the 9 scarlet fever cases associated vs.78% in 2011). Thus, it is difficult to attribute the upsurge
with medical complications (Table 2). Among 26 emm12 to the insert alone. A shift in prevailing emm type that
strains subjected to PFGE, 7 patterns were detected; the occurred in 2011 might have contributed to fluctuations in
emm12 strain from 1 of the 2 fatal cases exhibited a unique the number of cases (3).
PFGE pattern (Figure 2). For the other fatal case, an emm1 A higher rate of erythromycin resistance in S. pyogenes
strain positive for speA was isolated. (>80%) was found in 2011 than in the reported previous
Of the archived S. pyogenes strains collected during
2008–2010, few strains were from patients diagnosed with
scarlet fever; therefore, we analyzed S. pyogenes strains
isolated from throat and superficial wound specimens
from outpatients <15 years of age. Among 28 such strains,
emm28 was detected in 9 strains; emm4 in 4 strains; emm1
in 3 strains; emm12, 22, and 89 in 2 strains each; and 6 other
emm types in 1 strain each. All strains were susceptible
to penicillin; the erythromycin resistance rate was 10.7%
(3/28). The 48-kb insert was found in 10 (35.7%) strains:
3 strains of emm28, and 1 strain each of emm types 1, 2, 4,
22, 44, 89, and stG485.

Conclusions
The 2011 S. pyogenes outbreak in Hong Kong attracted
heightened media coverage, which might have increased
reporting of cases; however, the higher proportion of Figure 2. Pulsed-field gel electrophoresis patterns of 26 emm type
laboratory-confirmed cases in 2011 than those during 2008– 12.0 Streptococcus pyogenes strains, Hong Kong, 2011. Toxin
profile results are shown as corresponding to the genes speA¸
2010 suggests the upsurge was genuine. Overall clinical and speB, speC, speF, speH, and ssa. Strain M11–4380386 was from
epidemiologic profiles in 2011 did not differ from previous a fatal case. ERY suscep., erythromycin susceptibility result; R,
years. We found insufficient evidence that a particular emm resistant; S, susceptible. Scale bar indicates percent similarity.

1660 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Scarlet Fever, Hong Kong, 2011

years (20%–30%) (4). Because all erythromycin-resistant References

strains were also resistant to clindamycin (data not shown),
1. Chan JC, Chu YW, Chu MY, Cheung TK, Lo JY. Epidemio-
we deduced the resistance mechanism to be resistance to logical analysis of Streptococcus pyogenes infections in Hong
macrolides, lincosamides, and streptogramins B system, as Kong. Pathology. 2009;41:681–6. http://dx.doi.org/10.3109/
encoded by the erm genes (5). 00313020903257723
The 48-kb insert provided a mechanism for macrolide 2. University of Hong Kong. The University of Hong Kong finds
genetic mutation in Streptococcus pyogenes possible cause for the
resistance among S. pyogenes in Hong Kong, but our recent community outbreak of scarlet fever [press release]. 2011 Jun
laboratory investigation found macrolide-resistant S. 20 [cited 2011 July 27]. http://www.hku.hk/press/news_detail_6505.
pyogenes strains and the macrolide-susceptible strains html
that bore them negative for this insert. Mutation of the 3. Chiou CS, Wang YW, Chen PL, Wang WL, Wu PF, Wei HL.
Association of the shuffling of Streptococcus pyogenes clones
PCR primer binding site might explain the former strains; and the fluctuation of scarlet fever cases between 2000 and 2006
further investigation is needed to explore this possibility. in central Taiwan. BMC Microbiol. 2009;9:115. http://dx.doi.
The upsurge in scarlet fever cases in Hong Kong during org/10.1186/1471-2180-9-115
2011 likely reflects a regional phenomenon; a marked 4. Ho PL, Johnson DR, Yue AW, Tsang DN, Que TL, Beall B, et
al. Epidemiologic analysis of invasive and noninvasive group a
increase in cases was also observed in mainland China (6) streptococcal isolates in Hong Kong. J Clin Microbiol. 2003;41:937–
and Macao (7) during this period. High resistance rates 42. http://dx.doi.org/10.1128/JCM.41.3.937-942.2003
against macrolides were also observed for the outbreak 5. Livermore DM, Winstanley TG, Shannon KP. Interpretative reading:
in mainland China (8). We recommend close monitoring recognizing the unusual and inferring resistance mechanisms from
resistance phenotypes. J Antimicrob Chemother. 2001;48(Suppl
and surveillance of disease activity, genetic testing, 1):87–102. http://dx.doi.org/10.1093/jac/48.suppl_1.87
antimicrobial susceptibility profiling, and maintaining 6. China Ministry of Health. China Ministry of Health notifiable disease
scarlet fever’s statutory notifiable status. statistics. 2011 Jul [cited 2011 Aug 19]. http://www.moh.gov.cn/
publicfiles/business/htmlfiles/mohjbyfkzj/s3578/201107/52298.htm
7. World Health Organization Western Pacific Region. Scarlet fever
Acknowledgments update, 14 July 2011 [cited 2011 Jul 27]. http://www.wpro.who.int/
We thank all members of the Surveillance and Epidemiology emerging_diseases/ScarletFever/en/index.html
Branch and Public Health Laboratory Services Branch of 8. Liang Y, Shen X, Huang G, Wang C, Shen Y, Yang Y. Characteristics
of Streptococcus pyogenes strains isolated from Chinese children
Department of Health for their contributions to this work.
with scarlet fever. Acta Paediatr. 2008;97:1681–5. http://dx.doi.
Dr Luk is a medical and health officer at the Centre for Health org/10.1111/j.1651-2227.2008.00983.x
Protection, Department of Health, Hong Kong, People’s Republic
Address for correspondence: Emma Y.Y. Luk, Surveillance and
of China. She was in the US Centers for Disease Control and
Epidemiology Branch, 3/F, Centre for Health Protection, Department of
Prevention’s Field Epidemiology Training Program during this
Health, 147C Argyle Street, Kowloon, Hong Kong; email: mo_fetp2@
study. Her research interests include studying the epidemiology
dh.gov.hk
and transmission of communicable disease.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1661
DISPATCHES

Visceral varying incidence levels among villages and hamlets could

also be useful. We therefore studied factors associated with
Leishmaniasis in VL in an area made up of villages with variable levels of
VL incidence and constructed an asset index to control for
Rural Bihar, India confounding by socioeconomic status.

Epco Hasker,1 Shri Prakash Singh,1 The Study

Paritosh Malaviya, Albert Picado, The study area is a geographically continuous area
Kamlesh Gidwani, Rudra Pratap Singh, comprising 50 villages in the Muzaffarpur District of Bihar
Joris Menten, Marleen Boelaert, State, India, a district where VL is highly endemic. The 50
and Shyam Sundar villages can be further subdivided into 200 hamlets, also
known as tolla. We conducted 3 annual surveys in Sep-
To identify factors associated with incidence of visceral tember and October of 2008, 2009, and 2010. In each sur-
leishmaniasis (VL), we surveyed 13,416 households in Bi- vey, we visited all households in the study area, collected
har State, India. VL was associated with socioeconomic sta- demographic information, and asked whether the house
tus, type of housing, and belonging to the Musahar caste. had been covered by indoor residual insecticide spraying
Annual coverage of indoor residual insecticide spraying was
in the year preceding the survey. In each survey, we also
12%. Increasing such spraying can greatly contribute to VL
control.
collected information about VL in the household since the
previous survey. For the first survey, we used a recall pe-
riod of 1.5 years. A case of VL was defined as the combina-
Visceral leishmaniasis (VL), a vector-borne parasitic tion of a clinical history typical for VL (fever of >2 weeks’
disease caused by several Leishmania spp., is nearly al- duration, lack of response to antimalarial drug treatment);
ways fatal if left untreated (1,2). The clinical syndrome is a positive result by the rK39 rapid diagnostic test; and a
characterized by fever, weight loss, splenomegaly, hepato- good response to specific VL treatment, with or without
megaly, and anemia. The disease is endemic in >60 coun- confirmation of parasites. Each case reported was verified
tries, but 90% of all reported cases occur in just 5 countries: from medical records by a study physician. At the time of
Bangladesh, Brazil, India, Nepal, and Sudan (3). On the the second survey in 2009, we also collected information
Indian subcontinent, the disease is assumed to be an an- about assets owned by each household, including domestic
throponosis; the vector is a sand fly, Phlebotomus argen- animals, and we recorded characteristics of the structure
tipes. Approximately 200 million persons on the Indian of the house and the surrounding vegetation. Information
subcontinent are at risk for VL, and the annual incidence is about assets owned (other than domestic animals) was
≈420,000 cases (4). The disease affects mainly poor rural used to subdivide the study population into 5 quintiles of
communities; ≈80% of all cases in the region are reported socioeconomic status. To study potential associations with
from the state of Bihar in India (4). household environment, remote sensing (15-m resolution
Earlier studies on the Indian subcontinent have identi- ASTER images), and ground data were combined to derive
fied several risk factors for VL (5–11). At times, findings a map of the study area showing 7 types of land cover. Data
between studies have been conflicting, particularly in rela- were analyzed in a binomial multilevel model with tolla as
tion to the role of domestic animals (7). The use of bed a random effect.
nets was found to be protective in some studies (2,5), but We enrolled a study population of 81,210 persons, di-
this conclusion could not be confirmed in a recent cluster- vided over 13,416 households (92% of all households in
randomized trial (12). Many of the earlier studies were the study area). During the study period, we registered 207
conducted on fairly small populations, usually 1 or 2 vil- VL cases, equivalent to an average annual incidence of
lages (5,6,8,9); confounding by socioeconomic status was 72.8/100,000 population. Cases were strongly clustered at
controlled to a varying extent. Most studies were conducted the tolla level, with an intraclass correlation of 32%. None
in high-incidence villages or in villages in which a recent of the types of land cover was significantly associated with
outbreak had occurred. Because VL has a strongly clus- disease. With the villages spread out along roads and farm-
tered distribution, understanding the reasons behind widely ing land split up into small plots, the environment experi-
enced by persons from different villages or from different
Author affiliations: Institute of Tropical Medicine, Antwerp, Belgium
locations did not vary greatly.
(E. Hasker, A. Picado, J. Menten, M. Boelaert); and Banaras Hindu
VL was strongly associated with age; the odds of hav-
University, Varanasi, India (S.P. Singh, P. Malaviya, K. Gidwani,
ing VL was lowest for children <5 years of age and highest
R.P. Singh, S. Sundar)

DOI: http://dx.doi.org/10.3201/eid1810.111083 1
These authors contributed equally to this article.

1662 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Visceral Leishmaniasis, India

for children 5–14 years of age (odds ratio [OR] 2.5, 95% near the house are risk factors, but are not strong enough
CI 1.5–4.0). Higher socioeconomic status was associated to warrant specific interventions. Poor housing is a stron-
with reduced risk; comparing the wealthiest to the poorest ger risk factor; thus, housing plans launched by the Indian
quintile, we observed an OR of 0.5 (95% CI 0.3-1.0). Hav- government may positively affect control of VL. Persons
ing at least 1 bed net per 3 household members was protec- in the Musahar caste were at increased risk; they made up
tive on univariate analysis, but the effect was weaker and 2.4% of the study population but had >15% of VL cases.
no longer statistically significant when we controlled for The Musahars are known to be among the poorest of the
confounding (OR 0.8, 95% CI 0.5–1.3). This finding can be poor, but even after we controlled for confounding by so-
explained by a strong association between socioeconomic cioeconomic status, the association remained statistically
status and ownership of bed nets. significant. Some residual confounding cannot be ruled
Of ownership of all the domestic animals investigat- out, but other factors probably play a role. One such fac-
ed, only ownership of goats was weakly, but significantly, tor could be long delays in seeking health care by Musa-
associated with VL (OR 1.4, 95% CI 1.0–1.8). Other fac- hars, which was documented in another recent study (14).
tors at household level that were statistically significant in When devising improved VL control strategies, it would
multivariate analysis were the following: belonging to the certainly be justified to pay special attention to the ham-
Musahar caste (OR 2.9, 95% CI 1.3–6.8); presence of a lets inhabited by the Musahar caste. Overall, however, the
bamboo tree (OR 1.5, 95% CI 1.2–2.0); and type of walls most benefit can be expected from strengthening vector
(OR 1.8, 95% CI 1.0–3.3 for unplastered brick walls and control efforts. In the 1960s, as a byproduct of intensive
OR 2.5, 95% CI 1.3–4.6 for thatched walls, with plastered indoor residual insecticide spraying for malaria eradica-
brick walls as reference for both). (Table) Thatched walls tion campaigns, VL was all but eliminated from the area,
and presence of bamboo trees are likely to provide favor- and biannual indoor residual insecticide spraying is one
able breeding conditions for the sand fly vector (13). of the cornerstones of the regional VL elimination strat-
Indoor residual insecticide spraying coverage was egy (15). Thus, it defies imagination that in this highly
poor. In 2009 (the last year for which data were collected VL-endemic area, the annual indoor residual insecticide
for the full year), only 12% of all households had report- spraying coverage can be as low as 12%.
edly been sprayed at least once.
This work was supported by the National Institute for Aller-
Conclusions
gy and Infectious Diseases, National Institutes of Health, TMRC
In this large cohort study, controlled for potential
grant no. 1P50AI074321.
confounding by socioeconomic status and other contex-
tual factors, we identified several factors associated with Dr Hasker is a scientific collaborator at the Institute of Tropi-
VL. Ownership of goats and presence of bamboo trees cal Medicine in Antwerp, Belgium. His major research interests

Table. Factors associated with visceral leishmaniasis, Bihar State, India, 2008–2010
No. (%) participants
Factor Total, N = 81,210 Case-patients, n = 207 Odds ratio*
Demographic characteristic
Mushahar caste 1,980 (2.4) 32 (15.5) 2.9 (1.3–6.8)
Age group, y
0–4 12,787 (15.8) 20 (9.7) Referent
5–14 21,020 (25.9) 79 (38.2) 2.5 (1.5–4.0)
15–24 14,282 (17.6) 33 (15.9) 1.7 (1.0–3.0)
25–34 10,993 (13.5) 31 (15.0) 2.0 (1.1–3.5)
35–44 8,462 (10.4) 23 (11.1) 1.9 (1.1–3.6)
>45 13,666 (16.8) 21 (10.1) 1.2 (0.6–2.2)
Socioeconomic status, by assets index level
Level 1, poorest 16,515 (20.3) 70 (33.8) Referent
Level 2 16,094 (19.8) 58 (28.0) 0.9 (0.6–1.3)
Level 3 16,124 (19.8) 31 (15.0) 0.7 (0.5–1.1)
Level 4 16,256 (20.0) 35 (16.9) 0.9 (0.6–1.4)
Level 5 16,221 (20.0) 13 (6.3) 0.5 (0.3–1.0)
Other
Ownership of goats 25,703 (31.7) 86 (41.6) 1.4 (1.0–1.8)
Bamboo tree at <10 m 31,554 (38.9) 86 (41.6) 1.5 (1.1–2.0)
Type of walls
Brick, plastered 19,169 (23.6) 15 (7.3) Referent
Brick, unplastered 35,401 (43.6) 96 (46.4) 1.8 (1.0– 3.3)
Thatched 26,640 (32.8) 96 (46.4) 2.5 (1.3– 4.6)
*Based on multivariate model with tolla of residence as random effect.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1663
DISPATCHES

are the epidemiology and control of visceral leishmaniasis, human j.1365-3156.2006.01735.x

African trypanosomiasis, and tuberculosis. 10. Boelaert M, Meheus F, Sanchez A, Singh SP, Vanlerberghe V,
Picado A, et al. The poorest of the poor: a poverty appraisal of
households affected by visceral leishmaniasis in Bihar, India. Trop
Med Int Health. 2009;14:639–44. http://dx.doi.org/10.1111/j.1365-
References
3156.2009.02279.x
11. Rijal S, Uranw S, Chappuis F, Picado A, Khanal B, Paudel IS, et al.
1. Barnett PG, Singh SP, Bern C, Hightower AW, Sundar S. Virgin soil:
Epidemiology of Leishmania donovani infection in high-transmis-
the spread of visceral leishmaniasis into Uttar Pradesh, India. Am J
sion foci in Nepal. Trop Med Int Health. 2010;15(Suppl 2):21–8.
Trop Med Hyg. 2005;73:720–5.
http://dx.doi.org/10.1111/j.1365-3156.2010.02518.x
2. Bern C, Hightower AW, Chowdhury R, Ali M, Amann J, Wagatsuma
12. Picado A, Singh SP, Rijal S, Sundar S, Ostyn B, Chappuis F, et al.
Y, et al. Risk factors for kala- azar in Bangladesh. Emerg Infect Dis.
Longlasting insecticidal nets for prevention of Leishmania donovani
2005;11:655–62.
infection in India and Nepal: paired cluster randomised trial. BMJ.
3. World Health Organization. Control of the leishmaniases. Report
2010;341:c6760. http://dx.doi.org/10.1136/bmj.c6760
of a WHO expert committee. World Health Organ Tech Rep Ser.
13. Poché D, Garlapati R, Ingenloff K, Remmers J, Poché R. Bionom-
1990;793:1–158. http://dx.doi.org/10.3201/eid1105.040718
ics of phlebotomine sand flies from three villages in Bihar, India. J
4. Joshi A, Narain JP, Prasittisuk C, Bhatia R, Hashim G, Jorge A, et al.
Vector Ecol. 2011;36(Suppl 1):S106–17. http://dx.doi.org/10.1111/
Can visceral leishmaniasis be eliminated from Asia? J Vector Borne
j.1948-7134.2011.00119.x
Dis. 2008;45:105–11.
14. Pascual Martinez F, Picado A, Roddy P, Palma P. Low castes have
5. Bern C, Joshi AB, Jha SN, Das ML, Hightower A, Thakur GD, et al.
poor access to visceral leishmaniasis treatment in Bihar, India. Trop
Factors associated with visceral leishmaniasis in Nepal: bed-net use
Med Int Health. 2012;17:666–73. http://dx.doi.org/10.1111/j.1365-
is strongly protective. Am J Trop Med Hyg. 2000;63:184–8.
3156.2012.02960.x
6. Saha S, Ramachandran R, Hutin YJ, Gupte MD. Visceral leishmani-
15. World Health Organization. Regional strategic framework for elimi-
asis is preventable in a highly endemic village in West Bengal, In-
nation of kala-azar from the South-East Asia region (2005–2015).
dia. Trans R Soc Trop Med Hyg. 2009;103:737–42. http://dx.doi.
New Delhi (India): The Organization; 2005 [cited 2011 Jun 30].
org/10.1016/j.trstmh.2008.10.006
http://www.searo.who.int/LinkFiles/Kala_azar_VBC-85_Rev_1.pdf
7. Singh SP, Hasker E, Picado A, Gidwani K, Malaviya P, Singh RP, et
al. Risk factors for visceral leishmaniasis in India: further evidence
on the role of domestic animals. Trop Med Int Health. 2010;15(Sup- Address for correspondence: Epco Hasker, Department of Public Health,
pl 2):29–35. http://dx.doi.org/10.1111/j.1365-3156.2010.02515.x Institute of Tropical Medicine, Nationalestraat 155, B-2000 Antwerp,
8. Ranjan A, Sur D, Singh VP, Siddique NA, Manna B, Lal CS, et al. Belgium; email: ehasker@itg.be
Risk factors for Indian kala-azar. Am J Trop Med Hyg. 2005;73:74–
8.
9. Schenkel K, Rijal S, Koirala S, Koirala S, Vanlerberghe V, Van der
All material published in Emerging Infectious Diseases is in
SP, et al. Visceral leishmaniasis in southeastern Nepal: a cross-sec-
the public domain and may be used and reprinted without
tional survey on Leishmania donovani infection and its risk factors.
special permission; proper citation, however, is required.
Trop Med Int Health. 2006;11:1792–9. http://dx.doi.org/10.1111/

wwwnc.cdc.gov/eid/pages/eCard.htm
1664 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Influenza A(H1N1) 40.0%) (2,3). A(H1N1)pdm09 virus was reported to cause a
reverse zoonosis in pigs (4); thus, a long-term (2009–2011)
pdm09 Virus in serologic and virologic survey was designed to check for
transmission of the virus to pigs on Réunion Island, where
Pigs, Réunion the pork industry is a key economic activity and no live

Island
pigs have been imported since 1978. At 6-month intervals,
a local veterinary surveillance system conducts serologic
surveillance for pathogenic swine influenza viruses (H1N1,
Eric Cardinale,1 Hervé Pascalis,1 Sarah Temmam,2 H1N2, and H3N2) among local herds, and during the last
Séverine Hervé,2 Aure Saulnier, Magali Turpin, 20 years, none have been detected.
Nicolas Barbier, Johny Hoarau,
Stéphane Quéguiner, Stéphane Gorin, The Study
Coralie Foray, Matthieu Roger, Vincent Porphyre, During a first step (November 2009–February 2010),
Paul André, Thierry Thomas, seroprevalence rates for A(H1N1)pdm09 virus were
Xavier de Lamballerie, Koussay Dellagi, assessed in 120 breeding pigs (>4 years old) from 57 farms.
and Gaëlle Simon Blood was obtained from randomly selected pigs at the only
slaughterhouse on the island, where pigs are held for <3
During 2009, pandemic influenza A(H1N1)pdm09 hours. We screened the samples for antibodies to influenza
virus affected humans on Réunion Island. Since then, the A viruses by using the ID Screen Antibody Influenza A kit
virus has sustained circulation among local swine herds,
(ID.vet, Montpellier, France), and titers were determined
raising concerns about the potential for genetic evolution
of the virus and possible retransmission back to humans of
by using hemagglutination-inhibition (HI) assays (5)
variants with increased virulence. Continuous surveillance against all classical swine influenza viruses and A(H1N1)
of A(H1N1)pdm09 infection in pigs is recommended. pdm09 virus (Table 1). Ninety-eight (81.7%; 95% CI
74.7%–88.5%) of the 120 serum samples were positive for
A(H1N1)pdm09 virus (HI titers >20); the range of positive

I nfluenza A(H1N1)pdm09 virus, which caused the

last influenza pandemic among humans, is a unique
reassortant derived from swine influenza viruses of the
titers was 40–640, and 54.2% of the samples expressed high
HI titers (160–640). Of the 98 serum samples, 5 reacted at
low titer and with only 1 European A (H1N1) swine virus
triple reassortant swine North American lineage and the (titer <20), i.e., >4 dilutions lower than for A(H1N1)pdm09
avian-like swine Eurasian lineage (1). Réunion Island, a virus, indicating cross reactivity (6). Thus, pigs from 47
tropical French overseas department in the southwestern (82.4%) of 57 tested farms had been infected by A(H1N1)
Indian Ocean, was struck by the influenza pandemic during pdm09 virus; the seroprevalence rate was 81%–100% for
July–August 2009. The epidemic had a high attack rate in pigs on 79.0% of the farms. Farms with affected pigs were
humans (estimated clinically at 12.5% and serologically at located throughout the island (Figure).
In a second step (June 2010, when A(H1N1)pdm09
Author affiliations: Le Centre de Recherche et de Veille sur les
infection was no longer detected among humans), we tested
Maladies Emergentes dans l’Océan Indien, Sainte-Clotilde, Ile de la
whether the virus was still circulating among pigs born that
Réunion, France (E. Cardinale, H. Pascalis, S. Temmam, M. Turpin,
year. To obtain nasal swab and blood samples for testing,
J. Hoarau, C. Foray, M. Roger, K. Dellagi); Centre de Coopération
we randomly selected 390 fattening pigs (25–27 weeks old)
Internationale en Recherche Agronomique pour le Développement,
at the slaughterhouse; the pigs originated from 45 farms. At
Montpellier, France (E. Cardinale, J. Hoarau, C. Foray, M. Roger,
the time of sampling, the veterinary surveillance system did
V. Porphyre); Institut de Recherche pour le Développement,
not report any clinical signs suggesting virus circulation
Sainte-Clotilde (H. Pascalis, M. Turpin, K. Dellagi); Centre
among herds. However, ≈3.5% of the serum samples (9%
National de la Recherche Scientifique–Université de Lyon, Lyon,
of tested farms) contained antibodies to A(H1N1)pdm09
France (S. Temmam); Agence Nationale de Sécurité Sanitaire, de
virus (HI titers 20–160). Nasal swab specimens from
l’alimentation, de l’Environnement, et du Travail (ANSES), National
6.7% (26/390) of pigs were positive for A(H1N1)pdm09
Reference Laboratory for Swine Influenza, Ploufragan, France (S.
virus as determined by using a specific real-time reverse
Hervé, A. Saulnier, N. Barbier, S. Quéguiner, S. Gorin, G. Simon);
transcription PCR (rRT-PCR); the pigs originated from 13
Clinique Vétérinaire, Saint-Louis, Ile de la Réunion (P. André);
(28.8%) farms (7). Two strains, A/Sw/La Reunion/0164/10
Coopérative des Producteurs de Porcs de la Réunion, Saint-Pierre,
and A/Sw/LaReunion/110348/10, were isolated onto
Ile de la Réunion (T. Thomas); and Université de la Méditerranée,
MDCK cell cultures (5).
Marseille, France (X. de Lamballerie)
1
These authors contributed equally to this article.
DOI: http://dx.doi.org/10.3201/eid1810.120398 2
These authors contributed equally to this article.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1665
DISPATCHES

Table 1. Antigenic characterization of A(H1N1)pdm09-like influenza viruses isolated from pigs, Réunion Island, 2010*
HI titers† obtained with reference swine hyperimmune serum samples directed against
Sw/England/ Sw/Cotes Sw/Scotland/
Virus strain California/04/09‡ 117316/86§ d'Armor/0388/09¶ Sw/Flandres/1/98# 410440/94**
A/Sw/La Reunion/0164/10 1,280 160 40 <10 <10
A/Sw/La Reunion/110348/10 1,280 80 160 10 20
A/Sw/La Reunion/0167/10 640 80 40 <10 <10
A/Sw/La Reunion/110194/10 640 80 20 <10 10
*HI, hemagglutination inhibition.
†HI tests were performed according to standard procedure (5).Titers are expressed as the reciprocal of the highest dilution of serum that inhibits 4
hemagglutination units of virus.
‡A(H1N1)pdm09 lineage.
§Classical swine A(H1N1).
¶Avian-like swine A(H1N1).
#Human-like reassortant swine A(H3N2).
**Human-like reassortant swine A(H1N2).

During July–December 2010, 11 farms reported A(H1N1) viruses from classical and avian-like lineages,
influenza-like clinical signs in pigs, and proof of A(H1N1) although they reacted most strongly with A(H1N1)pdm09
pdm09 virus infection was established on 3 farms (farms virus (Table 1). Genome sequencing of these strains
A–C). In June 2010, fattening pigs on farm A were showed high (>98%) nucleotide sequence homology to
seronegative for A(H1N1)pdm09 virus. In July, when the corresponding genes of A/California/04/09 and 2009
acute respiratory disease was reported among pigs, 12 of human strains from Réunion Island, suggesting human-to-
39 fattening pigs (18–21 weeks old) sampled on farm A swine transmission (H. Pascalis, unpub. data).
were still seronegative for A(H1N1)pdm09 virus; however, In a third step (March, July–August, and October
rRT-PCR results were positive for A(H1N1)pdm09 virus. 2011), 3 other sampling campaigns were conducted at
Four weeks later, when pigs had recovered from influenza, the slaughterhouse, including 831 fattening pigs from
only 7.7% (3/39) of sampled pigs on farm A had rRT- 104 farms. Nasal swab samples for 7 (8.4%) pigs from 3
PCR results positive for A(H1N1)pdm09 virus, and all (2.9%) farms still had rRT-PCR–positive results. However,
39 were seropositive for the virus. High rates of rRT-PCR serologic analyses revealed that pigs on ≈40% of the farms
positivity were also noted for pigs on farms B (17/30 pigs) (distributed throughout the island) were seropositive for
and C (6/15 pigs). Two A(H1N1)pdm09 strains (A/Sw/ A(H1N1)pdm09 virus, indicating continuing circulation of
LaReunion/0167/10 and A/Sw/La Reunion/110194/10) the virus in swine herds (Table 2).
were isolated from pigs on farms A and B, respectively.
Four influenza virus strains were isolated from Conclusions
pigs, and all induced a cytopathic effect and displayed Consistent with findings elsewhere (8), our results
hemagglutinating activity on chicken erythrocytes; all 4 show that A(H1N1)pdm09 virus has substantially affected
were confirmed as A(H1N1)pdm09 virus by specific rRT- swine herds in Réunion Island. Results of our long-term
PCRs. In addition cross-HI assays (5) revealed that these (≈2 years) investigation show that A(H1N1)pdm09 virus
strains exhibit antigenic relationships with swine influenza has circulated in pigs beyond the 5-week epidemic among

Figure. Location of farms tested for antibodies against influenza A(H1N1)pdm09 virus in serologic surveys, Réunion Island, 2009–2011.
Black dots, seronegative farms; gray dots, seropositive farms.

1666 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Influenza A(H1N1)pdm09 Virus in Pigs

Table 2. Seroprevalence rates for influenza A viruses among fattening pigs as determined by ELISA and HI testing, Reunion Island,
2010 and 2011*
Time of sampling campaign (total no. farms; Antinucleoprotein ELISA A(H1N1)pdm09 HI test
total no. pigs) Herd, % Pigs, % Herd†, % Pigs, %
June 2010 (N = 45; n = 252) 13.33 5.55 8.89 3.57
March 2011 (N = 33; n = 256) 48.48 28.12 42.42 16.41
July–August 2011 (N = 38; n = 316) 47.37 34.49 39.47 27.21
October 2011 (N = 33; n = 259) 57.57 40.93 42.42 22.01
*Pigs were randomly sampled at the slaughterhouse during 4 campaigns. Seroprevalence rates are % farms positive at the herd level among all farms (N)
and % pigs positive at the animal level among all animals (n). HI, hemagglutination inhibition.
†Within-herd prevalence of infected pigs ranged from 3% to 23%.

humans during the austral winter 2009 (3) and has become of influenza among humans in 2011 (15); these cases were
a novel enzootic pathogen in Réunion Island. mild, but other, more virulent pathogenic viruses could
Several facts may account for the heavy human-to- emerge. Hence continuous surveillance of A(H1N1)pdm09
swine transmission of A(H1N1)pdm09 virus. First, the infection in pigs is recommended.
reassortant pandemic virus contains genomic segments
originating from swine influenza viruses established in pigs Acknowledgments
(1). Second, pigs are highly susceptible to experimental We are indebted to Nicolas Rose for helpful discussions.
inoculations with A(H1N1)pdm09 virus and support
CRVOI (Le Centre de Recherche et de Veille sur les
high intraspecies transmissibility (9). Third, the pressure
Maladies Emergentes dans l’Océan Indien) was supported in part
of infection caused by A(H1N1)pdm09 virus among
by the Ministry of Agriculture through its local representation,
humans in Réunion Island was high but most infections
DAAF–Réunion (Directorate for Food, Agriculture, and the
were mild or asymptomatic (3); therefore, people pursued
Forest–Réunion); by FEDER (European Funds for Regional
their professional activities, acting as silent spreaders of
Development) convergence–CPER (Contrat Programme Etat/
the virus. Last, pigs on Réunion Island had no history of
Region) funds; and by specific funds from the Institut de
previous passages of swine influenza viruses; thus, the lack
Microbiologie et Maladies Infectieuses; Institut national de la
of specific immunity to influenza A viruses would have
santé et de la recherché medicalé; Agence Inter-établissements
contributed to the high sensitivity of the pigs to infection,
de Recherche pour le Développement; Centre National de la
as described (10,11).
Recherche Scientifique; Institut National de Veille Sanitaire;
Despite serologic proof of large numbers of infected
Agence Nationale de Sécurité Sanitaire, de l’Alimentation, de
pigs during late 2009–early 2010 in Réunion Island,
l’Environnement, et du Travail (ANSES); Institut National de
influenza-like signs were not exhibited and reported until
Recherche Agronomique; and Institut Pasteur. Support for work
July 2010; this finding was similar to that in New Caledonia
at the French National Reference Laboratory for Swine Influenza
(8). In July 2010, several herds showed symptomatic
was provided by the French Ministry of Agriculture (contracts
changes in infection that could indicate either a change in
2009-175/18277 and 2011-53/2200304774).
virulence of the circulating strain or the intervention of co-
infecting pathogens or some other environmental factor(s). Dr Cardinale is in charge of animal and veterinary public
Mycoplasma hyopneumoniae and Pasteurella multocida health at CRVOI, where he is coordinating a surveillance network
were co-detected on farms with pigs with signs of infection on animal diseases in the Indian Ocean entitled “AnimalRisk-OI.”
(data not shown). Co-infection with swine influenza His primary research interests are bacterial and viral zoonoses.
virus and these bacteria is known to contribute to severe
respiratory disorders among pigs; thus, these bacteria may
References
have enhanced pathogenicity of A(H1N1)pdm09 virus on
affected farms (12). 1. Garten RJ, Davis CT, Russell CA, Shu B, Lindstrom S, Balish
Because specific immunity to A(H1N1)pdm09 A, et al. Antigenic and genetic characteristics of swine-origin
virus will decline over time when the virus is no longer 2009 A(H1N1) influenza viruses circulating in humans. Science.
2009;325:197–201. http://dx.doi.org/10.1126/science.1176225
circulating among humans, persistence of the virus in an
2. D’Ortenzio E, Renault P, Jaffar-Bandjee MC, Gauzere BA,
animal reservoir raises concerns about the risk for genetic Lagrange-Xelot M, Fouillet A, et al. A review of the dynamics
evolution of the virus and retransmission back to humans of and severity of the pandemic A(H1N1) influenza virus on Réunion
variants with potentially increased virulence. As an example, Island, 2009. Clin Microbiol Infect. 2010;16:309–16. http://dx.doi.
org/10.1111/j.1469-0691.2010.03171.x
during the 2011 austral winter, only influenza A(H3N2)
3. Dellagi K, Rollot O, Temmam S, Salez N, Guernier V, Pascalis H,
and B viruses were recorded (13). Novel reassortant viruses et al. Pandemic influenza due to pH1N1/2009 virus: estimation
containing genomic segments from A(H1N1)pdm09 and of infection burden in Réunion Island through a prospective
enzootic swine influenza viruses have been isolated in pigs serosurvey, austral winter 2009. PLoS ONE. 2011;6:e25738. http://
dx.doi.org/10.1371/journal.pone.0025738
(14). Such a reassortant was responsible for several cases

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1667
DISPATCHES

4. World Organization for Animal Health. World Animal Health 11. Simon G, Hervé S, Rose N, Marchal C. Transmission of pandemic
Information Database; 2009 [cited 2012 Aug 12]. http://web.oie.int/ influenza A/H1N1 2009 virus to pigs in New Caledonia, an insular
wahis/public.php?page=weekly_report_index&admin=0 region previously free from swine influenza viruses. In: Proceedings
5. World Organization for Animal Health. Manual of diagnostic tests of the 6th International Symposium on Emerging and Re-emerging
and vaccines for terrestrial animals 2010 [cited 2012 Mar 15]. http:// Pig Diseases; Barcelona, Spain; 2011 Jun 12–15; Poster P191.
www.oie.int/en/international-standard-setting/terrestrial-manual/ Barcelona: Centre de Recerca en Sanitat Animal; 2011.
access-online/ 12. Fablet C, Marois-Créhan C, Simon G, Grasland B, Jestin A, Kobisch
6. Kyriakis CS, Olsen CW, Carman S, Brown IH, Brookes SM, Van M, et al. Infectious agents associated with respiratory diseases in 125
Doorsselaere J, et al. Serologic cross-reactivity with pandemic farrow-to-finish pig herds: a cross-sectional study. Vet Microbiol.
(H1N1) 2009 virus in pigs, Europe. Emerg Infect Dis. 2010;16:96– 2012;157:152–63. http://dx.doi.org/10.1016/j.vetmic.2011.12.015
9. http://dx.doi.org/10.3201/eid1601.091190 13. Brottet E, Jaffar-Bandjee MC, Valette M, Polycarpe D, Filleul L.
7. Pol F, Queguiner S, Gorin S, Deblanc C, Simon G. Validation of Surveillance épidémiologique et virologique de la grippe à la
commercial real-time RT-PCR kits for detection of influenza A Réunion: un hiver austral 2011 très calme [cited 2012 Mar 10].
viruses in porcine samples and differentiation of pandemic (H1N1) http://www.ars.ocean-indien.sante.fr/fileadmin/OceanIndien/
2009 virus in pigs. J Virol Methods. 2011;171:241–7. http://dx.doi. Internet/Actualites/BVS_14_decembre_2011.pdf
org/10.1016/j.jviromet.2010.11.007 14. Fan X, Zhu H, Zhou B, Smith D, Chen X, Lam T, et al. Emergence
8. Njabo KY, Fuller TL, Chasar A, Pollinger JP, Cattoli G, Terregino C, and dissemination of a swine H3N2 reassortant influenza virus with
et al. Pandemic A/H1N1/2009 influenza virus in swine, Cameroon, 2009 pandemic H1N1 genes in pigs in China. J Virol. 2012;86:2375–
2010. Vet Microbiol. 2012;156:189–92. http://dx.doi.org/10.1016/j. 8. http://dx.doi.org/10.1128/JVI.06824-11
vetmic.2011.09.003 15. Centers for Disease Control and Prevention. Limited human-to-
9. Brookes SM, Nunez A, Choudhury B, Matrosovich M, Essen SC, human transmission of novel influenza A (H3N2) virus—Iowa.
Clifford D, et al. Replication, pathogenesis and transmission of MMWR Morb Mortal Wkly Rep. 2011;60:1615–7.
pandemic (H1N1) 2009 virus in non-immune pigs. PLoS One.
2010;5:e9068. http://dx.doi.org/10.1371/journal.pone.0009068 Address for correspondence: Eric Cardinale, CIRAD-CRVOI-UMR15,
10. Hofshagen M, Gjerset B, Er C, Tarpai A, Brun E, Dannevig B, et
Cyroi 2 rue Maxime Rivière, Ste Clotilde 97490, Réunion, France; email:
al. Pandemic influenza A(H1N1)v: human to pig transmission in
Norway? Euro Surveill. 2009;14: pii: 19406. eric.cardinale@cirad.fr

1668 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Powassan Virus leukocytes), 5 erythrocytes, and 64 mg/dL of protein. The
patient was given piperacillin/tazobactam and doxycycline.
Encephalitis, The next day, she was less responsive and was trans-
ferred to Abbott Northwestern Hospital in Minneapolis.
Minnesota, USA Shortly thereafter, she became unresponsive and labored
breathing developed. Her temperature reached 102°F
Justin Birge and Steven Sonnesyn (38.9°C), and the following laboratory values were outside
the reference range: leukocyte count (11.3 × 103/mm3),
Powassan virus (POWV) is a rare tick-borne agent of sodium level (131 mmol/L), erythrocyte sedimentation rate
encephalitis in North America. Historically, confirmed cases (49 mm/h), and protein level (2.3 mg/dL). Neurology and
occurred mainly in the northeastern United States. Since infectious disease specialists suspected viral encephalitis.
2008, confirmed cases in Minnesota and Wisconsin have Magnetic resonance imaging (MRI) was deferred because
increased. We report a fatal case of POWV encephalitis in
of the unknown composition of the aneurysm clip, and
Minnesota. POWV infection should be suspected in tick-
exposed patients with viral encephalitis,
the patient underwent a computed tomography angiogram
of the head and neck. Infarction, vasculitis, meningeal
enhancement, and structural abnormalities were not found.
Case Report Twenty-four–hour electroencephalogram monitoring and
A 67-year-old woman from Aitkin, Minnesota, administration of ceftriaxone (2 g intravenously [IV] every
USA, sought treatment at a local hospital on May 30, 24 h), acyclovir (500 mg IV every 8 h), and doxycycline
2011, with a 3-day history of dizziness, fever of up to (100 mg IV every 12 h) were initiated.
103°F (39.4°C), chills, malaise, nausea, and occasional Overnight, the patient became apneic and required
confusion with slurred speech. She had no respiratory or intubation. Examination revealed absent deep tendon
gastrointestinal symptoms and no history of ill contacts, reflexes, ocular deviation, positive Babinski response,
travel, environmental exposures, or other recent illness. and bilateral flaccid paralysis of the extremities. Pupillary
She had not been exposed to animals or vectors, other than light and corneal reflexes remained intact. No independent
those endemic to her area of residence, which included respirations were initiated. Complement levels were within
mosquitoes and deer ticks. She had removed many deer reference range. No evidence of seizure was shown on
ticks after gardening or hiking in the woods. The patient’s electroencephalogram, although epileptiform discharges
past medical history was notable for colon cancer, which were seen. Given the severity of encephalopathy,
had been treated with a partial colectomy in October 2010 prophylactic levetiracetam was initiated.
and chemotherapy through April 2011. She also had a Results of repeat screen for B. burgdorferi antibodies,
history of hypertension, cutaneous lupus, and a remote smear for Ehrlichia spp., and blood and urine cultures
cerebral aneurysm with surgical clipping. Medications were unremarkable. Brain MRI revealed nonspecific
she was taking were atenolol, hydroxychloroquine, and inflammatory changes within the thalamus, midbrain,
valsartan. and cerebellum, with no evidence of meningeal irritation,
On admission, the woman was alert and reported mild temporal lobe abnormality, mass effect, acute infarct, or
neck tenderness. Her temperature was 100°F (37.8°C), hydrocephalus (Figure 1, panel A, Appendix, wwwnc.cdc.
blood pressure 138/77 mm Hg, pulse rate 83 beats/min, gov/EID/article/18/10/12-0621-F1.htm). Acyclovir was
respiratory rate 22 breaths/min, and oxygen saturation dis-continued. Routine bacterial culture of CSF was
98% on room air. Results of neurologic, cardiac, and negative. Ceftriaxone and doxycycline were continued
respiratory examinations were normal. Studies with because acute Lyme disease, which rarely manifests in this
normal test results included comprehensive metabolic manner, could not be ruled out by serologic testing alone.
panel, urinalysis, computed tomographic scan of the The patient remained unresponsive with flaccid
head, and chest radiograph. Results of serum screen paralysis and areflexia. Four days after the initial
for Borrelia burgdorferi antibodies (by ELISA) were examination, repeat MRI showed substantial progression
negative. Leukocyte count was within reference range of signal abnormality in cerebral hemispheres, thalamus,
(10.8 × 103/mm3), with neutrophil (polymorphonuclear and midbrain. Mass effect was evident with crowding
leukocytes) predominance (80%). Her cerebrospinal fluid of structures at the foramen magnum. Lateral and third
(CSF) showed 80 leukocytes (89% polymorphonuclear ventricle dilation, consistent with acute hydrocephalus,
was noted (Figure 1, panel B). A repeat lumbar puncture
Author affiliation: Abbott Northwestern Hospital, Minneapolis,
was not performed, given clinical interpretation of illness,
Minnesota, USA
known imaging findings, key pending results, and lack of
DOI: http://dx.doi.org/10.3201/eid1810.120621 indications for additional testing.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1669
DISPATCHES

Thirteen days after illness onset, a serology panel was

negative for the following viruses: West Nile, La Crosse,
Eastern equine encephalitis, St. Louis encephalitis, and
Western equine encephalitis. PCR of CSF was negative
for enteroviruses and herpesviruses. The Minnesota
Department of Health (MDH) reported positive IgM
serologic results against Powassan virus (POWV).
Ceftriaxone and doxycycline were discontinued. Given the
patient’s clinical deterioration and poor prognosis, she was
electively extubated and then died. The Centers for Disease
Control and Prevention confirmed POWV infection by IgM
serology and reverse transcription PCR of CSF.
POWV is a tick-borne member of the family
Flaviviridae, first reported in 1958 (1). It is the only
tick-borne member of the genus Flavivirus with human
pathogenicity in North America (2). Selection bias in
identifying the infection may exist, diminishing the
reported incidence to only patients with severe disease.
Results of seroprevalence studies in Canada and the
northeastern United States are variable but include
seroprevalence estimates as high as 5.8% in Canada (3)
and 0.7% in New York State (4). Small and medium-sized
mammals are common reservoirs (notably, woodchucks
[Marmota monax] and white-footed mice [Peromyscus Figure 2. Geographic distribution of confirmed Powassan
leucopus]), and several species of tick (4 Ixodes spp., 2 virus (POWV) infections (diagnosis made by serology, reverse
Dermacentor spp.) act as vectors (5–7). Human infection transcription PCR) and counties with POWV–infected ticks in
Minnesota. Data provided by the Minnesota Department of Health.
has been documented in North America and Russia (8).
Both prototypic (POWV) and deer tick virus (DTV)
genotypes exist (6,9). In Canada and the northeastern
United States, I. cookei ticks are the typical vectors for of CSF testing and brain imaging are generally consistent
POWV. In Minnesota and Wisconsin, I. scapularis ticks with viral encephalitis. Reverse transcription PCR of CSF,
are the typical vectors for DTV. Although I. cookei ticks serologic testing of CSF, and serologic testing of serum
are present in these states, they rarely attach to humans, and are the preferred diagnostic tests, but they are not widely
according to MDH data, all sequenced strains in Minnesota available. Diagnostic testing for POWV should be referred
are of the DTV genotype. Transmission time from tick to to state and federal laboratories to ensure accuracy and
host has been documented in mice as quickly as 15 minutes standardization (4,8,9,11–13)
(10). The length of attachment time required for disease- Pathogenesis is due to lymphocytic infiltration of
causing viremia in humans is unknown. Literature estimates perivascular neuronal tissue with a predilection for gray
vary, but <40 cases were documented during 1958–2000 matter, including thalamus, midbrain, and cerebellum
(2,4,11,12). According to the Centers for Disease Control (11). Supportive care is the only therapy. A European
and Prevention and the health departments of Minnesota vaccine against the related tick-borne encephalitis virus
and Wisconsin, 33 US cases were reported during 2001– is available, but although the viruses are antigenically
2010: 12 in New York, 9 in Wisconsin, 8 in Minnesota, 2 similar, its effectiveness against POWV is unknown (12).
in Maine, and 1 each in Michigan and Virginia. In 2011, Approximately 10% of the reported infections have been
of 16 confirmed cases in the United States, 11 occurred in fatal, and an additional 50% have produced long-term
Minnesota and 4 occurred in Wisconsin. Figure 2 illustrates neurologic sequelae, including hemiplegia and headaches
the geographic distribution of human cases and infected (2,9,13).
ticks.
Patients with POWV infection typically exhibit Conclusions
encephalitis after an incubation period of 1–4 weeks. Fever POWV is causing an emerging and potentially severe
and headache are common; the typical viral prodrome lasts tick-borne infection in Minnesota and Wisconsin. POWV
1–3 days. Mental status changes, cerebellar symptoms, and infection should be suspected when tick-exposed patients
hemiplegia are also common and may be severe. Results exhibit viral encephalitis, especially those with cerebellar

1670 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Powassan Virus Encephalitis, Minnesota, USA

symptoms and/or thalamus/midbrain gray matter disease. 5. Brackney DE, Nofchissey RA, Fitzpatrick KA, Brown IK, Ebel
Preventing tick attachment by using chemical prophylaxis GD. Stable prevalence of Powassan virus in Ixodes scapularis in a
northern Wisconsin focus. Am J Trop Med Hyg. 2008;79:971–3.
and vigilance are essential in disease-endemic environments 6. Ebel GD, Spielman A, Telford SR III. Phylogeny of North American
to prevent contraction of POWV infection. Powassan virus. J Gen Virol. 2001;82:1657–65.
7. Orlinger KKK. A tick-borne encephalitis virus vaccine based on
the European prototype strain induces broadly reactive cross-
Dr Birge practices hospital medicine at Bergan Mercy neutralizing antibodies in humans. J Infect Dis. 2011;203:1556–64.
Medical Center in Omaha, Nebraska, USA. His research interests http://dx.doi.org/10.1093/infdis/jir122
8. LaSala PR, Holbrook M. Tick-borne flaviviruses. Clin Lab Med.
include hospital infections, cardiopulmonary resuscitation, and
2010;30:221–35. http://dx.doi.org/10.1016/j.cll.2010.01.002
physical activity. 9. Ebel GD, Kramer LD. Short report: duration of tick attachment
required for transmission of Powassan virus by deer ticks. Am J Trop
Dr Sonnesyn is an assistant clinical professor of medicine Med Hyg. 2004;71:268–71.
at the University of Minnesota. He is also a partner at Infectious 10. Ebel GD. Update on Powassan virus: emergence of a North American
Disease Consultants, P.A., and serves as hospital epidemiologist tick-borne flavivirus.]. Annu Rev Entomol. 2010;55:95–110. http://
at Abbott Northwestern Hospital, Minneapolis. dx.doi.org/10.1146/annurev-ento-112408-085446
11. Gholam BI, Puksa S, Provias JP. Powassan encephalitis: a case report
with neuropathology and literature review. CMAJ 1999;161:1419–
22.
References
12. Günther G, Haglund M. Tick-borne encephalopathies: epidemiology,
diagnosis, treatment and prevention. CNS Drugs. 2005;19:1009–32.
1. McLean DM, Donohue WL. Powassan virus: isolation of virus from
13. Romero JR, Simonsen KA. Powassan encephalitis and Colorado
a fatal case of encephalitis. Can Med Assoc J. 1959;80:708–11.
tick fever. [x.]. Infect Dis Clin North Am. 2008;22:545–59. http://
2. Hinten SR, Beckett GA, Gensheimer KF, Pritchard E, Courtney
dx.doi.org/10.1016/j.idc.2008.03.001
TM, Sears SD, et al. Increased recognition of Powassan encephalitis
in the United States, 1999–2005. Vector Borne Zoonotic Dis.
2008;8:733–40. http://dx.doi.org/10.1089/vbz.2008.0022 Address for correspondence: Justin Birge, Abbot Northwestern Hospital–
3. McLean DM. Powassan virus: surveys of human and animal sera. Medical Education (Rm 11135), 800 Minneapolis, Minnesota, USA;
Am J Public Health Nations Health. 1960;50:1539–44. http://dx.doi. email: birgejustin@gmail.com
org/10.2105/AJPH.50.10.1539
4. Artsob H. Powassan encephalitis. In: Monath TP, editor. The
arboviruses: epidemiology and ecology. Boca Raton (FL): CRC
Press; 1989. p. 29–49.

http://www.facebook.com/CDC

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1671
DISPATCHES

Influenza Virus monkeys might be sold at wet markets, the putative source
of several zoonotic outbreaks (3), where they might be
Infection in caged next to any number of animal species (Figure 1, panel
A) (4). Pet and performing monkeys are likely conduits
Nonhuman Primates for cross-species transmission of respiratory pathogens
like influenza viruses because of their close and long-term
Erik A. Karlsson, Gregory A. Engel, M.M. Feeroz, contact with their owners, audiences, domestic animals,
Sorn San, Aida Rompis, Benjamin P. Y.-H. Lee, wild animals, and birds (Figure 1, panel B) (5). However, the
Eric Shaw, Gunwha Oh, Michael A. Schillaci, breadth and diversity of this interface presents a challenge
Richard Grant, John Heidrich, for monitoring the emergence of infectious diseases. We
Stacey Schultz-Cherry, and Lisa Jones-Engel have approached this challenge by conducting longitudinal
studies at several sites and collecting biological samples
To determine whether nonhuman primates are infected and behavioral data representing various contexts of
with influenza viruses in nature, we conducted serologic human–NHP contact (4–7). We used these historical and
and swab studies among macaques from several parts of newly acquired samples, representing various countries
the world. Our detection of influenza virus and antibodies to and contexts of human–macaque contact, to determine
influenza virus raises questions about the role of nonhuman
whether NHPs are infected with influenza viruses in nature.
primates in the ecology of influenza.

The Study

W orldwide, infections with influenza A viruses are

associated with substantial illness and death among
mammals and birds. Public health and research have
As part of our decade-long longitudinal studies, ≈200
serum samples were collected from macaques. These
included pet macaques (Macaca nigra, M. nigrescens, M.
placed major focus on understanding the pathogenicity hecki) from Sulawesi, Indonesia; performing macaques
of different influenza virus strains and characterizing new from Java, Indonesia (M. fascicularis) and from
influenza vaccines. Nonhuman primates (NHPs), including Bangladesh (M. mulatta); M. fascicularis macaques from
macaques, have become popular experimental models for the Bukit Timah and Central Catchment Nature Reserves in
studying the pathogenesis and immunology of seasonal and Singapore, where they freely interact with wild avian fauna
emerging influenza viruses. NHPs readily seroconvert after and visitors (occasionally entering residential areas) (7); M.
experimental inoculation with seasonal influenza virus and sylvanus macaques from the Upper Rock Nature Reserve
have been used to test candidate vaccines for strains of in Gibraltar, where international tourists frequently use
human and avian origin. Like humans, macaques infected food to entice the macaques to climb about their heads and
with influenza virus exhibit fever, malaise, nasal discharge, shoulders (6); and free-ranging macaques (M. fascicularis
and nonproductive cough; virus replication can be detected and M. nemestrina) at temple shrines or M. fascicularis
in the nasal passages and respiratory tract (1,2). However, macaques that range throughout a wildlife rescue center
whether NHPs are infected with influenza viruses in nature and nearby villages in Cambodia (Figure 2). Serum was
remains unknown. collected and stored as described (8). All samples were
Over the past decade, we have focused on the role stored on cold packs in the field and transferred to dry ice
of pet and performing monkeys in disease transmission for shipment to the United States, where they were then
throughout Asia. Commonly trapped in the wild, these stored at −80°C.
For initial screening for antibodies against influenza
Author affiliations: St. Jude Children’s Research Hospital, virus, serum samples were treated with receptor-destroying
Memphis, Tennessee, USA (E.A. Karlsson, S. Schultz-Cherry); enzyme as described (9) and tested by using a multispecies
University of Washington, Seattle, Washington, USA (G.A. Engel, Influenza A Virus NP Antibody Inhibition Test (Virusys
G. Oh, L. Jones-Engel); Swedish Cherry Hill Family Medicine, Corporation, Taneytown, MD, USA) according to
Seattle (G.A. Engel); Jahangirnagar University, Savar, Bangladesh manufacturer’s instructions. ELISA results indicated
(M.M. Feeroz); National Veterinary Research Institute, Phnom nucleocapsid protein antibodies against influenza in samples
Penh, Cambodia (S. San); University of Udayana, Bali, Indonesia from macaques from Cambodia (29.2%), Singapore
(A. Rompis); Nature Parks, Singapore (B.P.Y.-H. Lee); Gibraltar (16.7%), Sulawesi (16.1%), Bangladesh (13.3%), and Java
Ornithological and Natural History Society, Gibraltar (E. Shaw); (6.0%) (Table 1). Antibodies were detected in animals 1–10
University of Toronto Scarborough, Ontario, Canada (M.A. years of age at the time of sampling. No influenza virus–
Schillaci); and Shin Nippon Biomedical Laboratory, Phnom Penh specific antibodies were detected from the 73 total samples
(R. Grant, J. Heidrich) from Gibraltar, perhaps because persons with influenza
DOI: 10.3201/eid1809.120214 virus infection infrequently travel to the Upper Rock

1672 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Influenza Virus Infection in Nonhuman Primates

Figure 1. The interface between nonhuman primates, birds, and humans. A) A young, recently captured leaf monkey perched on a cage
containing birds in a wet market in Java. B) A man and his performing monkey in Bangladesh. Reprinted with permission from Lynn
Johnson, 2012.

Reserve (healthy-visitor effect) (10) or perhaps because microneutralization studies, which would be useful to
monkeys from Gibraltar are less susceptible to infection. perform with future samples.
Seroprevalence of antibodies against influenza A virus, In 2011, to determine whether any macaques were
by site and collection year, human and NHP population actively infected with influenza virus, we collected oral
densities, and prevalence of avian influenza viruses are swabs from 48 monkeys in Cambodia to test for influenza
shown in Figure 2.
Serum samples that were positive by ELISA were
also screened by hemagglutination-inhibition assay as
described (9). Based on the year and location of NHP
sample collection, the estimated ages of the NHPs at
the time of sample collection, and the presence of avian
H5 and H9 influenza viruses in many of these countries
during the sampling period (11–13), a panel of human
vaccine strains and avian influenza virus strains was used
in the hemagglutination-inhibition assay. Although not
all ELISA-positive serum samples could be subtyped,
antibodies against seasonal subtype H1N1 and H3N2
influenza A strains were detected from macaques in
Bangladesh, Singapore, Java, and Sulawesi (Table 2). Of
the performing macaques in Bangladesh, 2 had antibodies
against A/chicken/Bangladesh/5473/2010, a strain of G1
clade subtype H9N2 avian influenza virus. Subtype H9N2
influenza viruses are prevalent in poultry in Bangladesh Figure 2. Nonhuman primate (NHP) habitat countries and
(14) and have been detected in humans (12). We did not approximate location of sampling sites, with sample size,
year collected, context of human–macaque interaction, and
detect antibodies against highly pathogenic avian influenza seroprevalence of antibodies against influenza virus A. Countries
subtype H5 viruses, which might not be surprising that have reported human influenza infection of avian origin are
given our relatively small sample size (Table 2). Also outlined. A color version of this figure is available online (wwwnc.
given the small sample size, we were unable to perform cdc.gov/EID/article/18/10/12-0214-F2.htm).

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1673
DISPATCHES

Table 1. Anti-influenza nucleocapsid protein antibodies in nonhuman primate populations, 2001–2011*

Location, year No. Macaca species Type(s) No. (%) positive
Singapore, 2003 36 fascicularis Reserve 6 (16.7)
Indonesia
Java, 2002 14 fascicularis Performing, pet 4 (28.6)
Sulawesi, 2001 31 nigra, hecki, nigrescens Pet 5 (16.1)
Gibraltar
2005 37 sylvanus Reserve 0
2009 36 sylvanus Reserve 0
Bangladesh
2006 4 mulatta Performing, pet 0
2010 30 mulatta Performing, pet 4 (13.3)
Cambodia, 2011 48 fascicularis, nemestrina Temple, urban, reserve 14 (29.2)
*Testing was conducted with a multispecies influenza A virus nucleocapsid protein antibody inhibition test for strain-specific antibodies.

virus by real-time reverse transcription PCR as described Conclusions

(8). In brief, the inside of the anesthetized and immobilized Our results indicate that NHPs that have contact with
monkeys’ mouths (cheeks, tongue, and gums) were swabbed. humans can be naturally infected with seasonal endemic
Swabs were immediately placed into viral transport media, human influenza viruses and with emerging pandemic-
stored, and shipped as previously described. RNA was risk avian influenza viruses. We found serologic evidence
isolated by using an Ambion MagMAX-96 AI/ND Viral of infection in several countries, contexts, and macaque
RNA Isolation Kit (Life Technologies Corporation, Grand species. Preliminary real-time reverse transcription PCR
Island, NY, USA) on a Kingfisher Flex system (Thermo results also pointed to the presence of virus in a buccal
Fisher Scientific, Waltham, MA, USA). Viral RNA was swab from an adult macaque from Cambodia, indicating
analyzed in a Bio-Rad CFX96 Real-Time PCR Detection active infection at the time of sampling. On the basis of
System and a C1000 Thermocycler (Bio-Rad, Hercules, CA, results from this study, it seems that pet, performing, and
USA) with TaqMan Fast Virus 1-Step Master Mix (Applied free-ranging macaques are susceptible to influenza virus
Biosystems, Foster City, CA, USA) and the InfA primer/ infection. Given the close relationship between humans
probe sets as described (15). Of the 48 respiratory samples, 1 and NHPs in areas of the world where avian and human
(2.1%) was positive for influenza virus; cycle threshold value influenza viruses cocirculate, further surveillance of these
was 38 (limit of detection is 40). Attempts to amplify longer populations is warranted. The ability to detect and eventually
PCR fragments of matrix, hemagglutinin, or neuraminidase isolate strains of influenza virus currently infecting NHPs
genes or to isolate the virus by blind passage in embryonated and humans at the animal–human interface is paramount to
chicken eggs or MDCK cells were unsuccessful. understanding how NHP–human interactions can affect the

Table 2. Seroprevalence of influenza A virus subtypes in monkeys with nucleocapsid protein–positive ELISAs, by location*
No. tested/no. positive
Virus subtype Years used in Indonesia
Virus strain (H5 clade) vaccine Singapore Java Sulawesi Bangladesh Cambodia
A/Beijing/262/1995 H1N1 1999–2000 0/6 2/4 1/6† NSA NSA
A/Sydney/5/1997 H3N2 1999–2000 2/6‡ 0/4 1/6§ NSA NSA
A/New Caledonia/20/1999 H1N1 2000–2007 0/6 0/4 1/6† 0/4 0/14
A/Panama/2007/1999 H3N2 2000–2004 2/6‡ 0/4 1/6§ 0/4 NSA
A/California/07/2004 H3N2 2005–2006 NSA NSA NSA 0/4 0/14
A/Wisconsin/67/2005 H3N2 2006–2008 NSA NSA NSA 0/4 0/14
A/Brisbane/59/2007 H1N1 2008–2010 NSA NSA NSA 1/4 0/14
A/Brisbane/10/2007 H3N2 2008–2010 NSA NSA NSA 0/4 0/14
A/California/04/2009 H1N1 2010–present NSA NSA NSA 0/4 0/14
A/Perth/16/2009 H3N2 2010–present NSA NSA NSA 0/4 0/14
A/chicken/Bangladesh/5473/2010 H9 G1 NA NSA NSA NSA 2/4 NSA
A/Vietnam/1203/2004 H5 (1) NA NSA NSA NSA NSA 0/14
A/Cambodia/R0H05050/2007 H5 (1) NA NSA NSA NSA NSA 0/14
A/duck/Hunan/795/2002 H5 (2.1) NA NSA NSA NSA 0/4 NSA
A/BHG/Qinghai/01/2005 H5 (2.2.2) NA NSA NSA NSA 0/4 NSA
A/JWE/Hong Kong/1038/2006 H5 (2.3.4.2) NA NSA NSA NSA 0/4 NSA
A/duck/Laos/3295/2006 H5 (2.3.4.2) NA NSA NSA NSA 0/4 NSA
*Samples were only tested for relevant strains based on the collection location, year of collection, and estimated age of the monkey. Monkeys from
Indonesia were not tested for influenza subtype H5N1 viruses because samples were collected in 2001 and 2002 and subtype H5N1 viruses were first
reported in poultry from Indonesia in February 2004. Samples with a hemagglutination inhibition value t1:10 were considered positive. NSA, no samples
available for testing; NA, not applicable for use in vaccine; BHG, bar-headed goose; JWE, Japanese white-eye.
†Individual monkey gave positive results for both strains.
‡Individual monkeys gave positive results for both strains.
§Individual monkey gave positive results for both strains.

1674 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Influenza Virus Infection in Nonhuman Primates

genetics, transmission, and public health risk for infection 2. Bodewes R, Rimmelzwaan GF, Osterhaus AD. Animal models for
with influenza A viruses. the preclinical evaluation of candidate influenza vaccines. Expert
Rev Vaccines. 2010;9:59–72. http://dx.doi.org/10.1586/erv.09.148
3. Webster RG. Wet markets—a continuing source of severe acute re-
Acknowledgments spiratory syndrome and influenza? Lancet. 2004;363:234–6. http://
We thank R. Webby and R. Webster for critical discussions dx.doi.org/10.1016/S0140-6736(03)15329-9
4. Schillaci MA, Jones-Engel L, Engel GA, Paramastri Y, Iskan-
regarding the manuscript. We are particularly grateful to the dar E, Wilson B, et al. Prevalence of enzootic simian viruses
following communities, temple committees, and government among urban performance monkeys in Indonesia. Trop Med
agencies in the areas where we have been sampling monkeys Int Health. 2005;10:1305–14. http://dx.doi.org/10.1111/j.1365-
for years: Lembaga Ilmu Pengetahuan, Indonesia; S. Chan and 3156.2005.01524.x
5. Schillaci MA, Jones-Engel L, Engel GA, Kyes RC. Exposure to hu-
the staff of the Central Nature Reserve, National Parks Board, man respiratory viruses among urban performing monkeys in Indo-
Singapore; M. Pizarro, J. Cortes, and the staff of the Gibraltar nesia. Am J Trop Med Hyg. 2006;74:716–9.
Ornithological & Natural History Society and the Government 6. Fuentes A, Shaw E, Cortes J. Qualitative assessment of macaque
of Gibraltar; S. Begum, K. Hasan, and the students and faculty tourist sites in Padangtegal, Bali, Indonesia, and the Upper Rock
Nature Reserve, Gibraltar. Int J Primatol. 2007;28:1143–58. http://
of the Department of Zoology, Jahangirnagar University; and C. dx.doi.org/10.1007/s10764-007-9184-y
Kimleng, S. Bunnary, T. Sothearos, Sisiket, H. Davun, K. Pal, 7. Sha JC, Gumert M, Lee B, Fuentes A, Chan S, Jones-Engel L. Status
D. Cohn, A. Fuentes, J. Supriatna, R. Babo, Y. Paramastri, E. of the long-tailed macaque (Macaca fascicularis) in Singapore and
Iskandar, J. Froehlich, L. Engel, H. Engel, and L. Johnson for implications for management. Biodivers Conserv. 2009;18:2909–
26. http://dx.doi.org/10.1007/s10531-009-9616-4
supporting and participating in this research. We also thank S. 8. Jones-Engel L, Steinkraus KA, Murray SM, Engel GA, Grant R,
Krauss and M. Ducatez for graciously providing the panel of Aggimarangsee N, et al. Sensitive assays for simian foamy viruses
human vaccine strains and avian influenza virus strains that were reveal a high prevalence of infection in commensal, free-ranging
used in the hemagglutination-inhibition assay. Asian monkeys. J Virol. 2007;81:7330–7. http://dx.doi.org/10.1128/
JVI.00343-07
This work was supported by the National Institutes of 9. Palmer DF, Dowdle WR, Coleman MT, Schild GC. Advanced labo-
ratory techniques for influenza diagnosis. Atlanta: Centers for Dis-
Health National Institute of Allergy and Infectious Diseases,
ease Control, US Department of Health, Education and Welfare.;
contract no. HHSN266200700005C, the American Lebanese 1975.
Syrian Associated Charities, NIH National Center for Research 10. Li CY, Sung FC. A review of the healthy worker effect in occupa-
Resources grant P51 RR000166RR 02S014, National Institutes tional epidemiology. Occup Med (Lond). 1999;49:225–9. http://
dx.doi.org/10.1093/occmed/49.4.225
of Health National Institute of Allergy and Infectious Diseases
11. Perdue ML, Swayne DE. Public health risk from avian influenza vi-
grants R01 AI078229 and R03 AI064865, Defense Advanced ruses. Avian Dis. 2005;49:317–27. http://dx.doi.org/10.1637/7390-
Research Projects Agency N66001-02-C-8072, the University of 060305R.1
Toronto Connaught Fund, the Chicago Zoological Society, and 12. Malik Peiris JS. Avian influenza viruses in humans. Rev Sci Tech.
2009;28:161–73.
the University of New Mexico Research Allocations Committee.
13. Kalthoff D, Globig A, Beer M. (Highly pathogenic) avian influenza
Dr Karlsson is a postdoctoral research associate in the as a zoonotic agent. Vet Microbiol. 2010;140:237–45. http://dx.doi.
org/10.1016/j.vetmic.2009.08.022
Department of Infectious Diseases at St. Jude Children’s Research 14. Negovetich NJ, Feeroz MM, Jones-Engel L, Walker D, Alam SMR,
Hospital, Memphis, Tennessee, USA. His research is focused on Hasan K, et al. Live bird markets of Bangladesh: H9N2 viruses and
surveillance of respiratory and gastrointestinal viruses and on the near absence of highly pathogenic H5N1 influenza. PLoS ONE.
understanding how nutrition can affect virus–host interactions 2011;6:e19311. http://dx.doi.org/10.1371/journal.pone.0019311
15. World Health Organization. CDC protocol of realtime RTPCR for
and emerging infectious diseases. swine influenza A(H1N1). 2009 Apr 30 [cited 2012 Apr 11]. http://
www.who.int/csr/resources/publications/swineflu/CDCrealtimeRT-
PCRprotocol_20090428.pdf
References

1. Berendt RF. Simian model for the evaluation of immunity to influ- Address for correspondence: Lisa Jones-Engel, University of Washington–
enza. Infect Immun. 1974;9:101–5. NRPC, 1705 Pacific St NE, HSB I-039, Seattle, WA 98195, USA; email:
jonesengel@wanprc.org

Search past issues of EID at www.cdc.gov/eid

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1675
DISPATCHES

Human exploring their disease potential, we sought to define their

prevalence in immunocompromised transplant recipients.
Polyomaviruses in To this end, we established a longitudinal cohort of
children undergoing transplantation at St. Louis Children’s
Children Undergoing Hospital, St. Louis, Missouri, USA.

Transplantation, The Study

United States,
We recruited 32 patients who were scheduled to
receive transplants (2 lung, 11 liver, 5 heart, 2 kidney,

2008–2010 1 liver/lung, and 11 bone marrow transplants) during

October 2008–April 2010. The Human Research Protection
Office of Washington University in St. Louis approved
Erica A. Siebrasse, Irma Bauer, Lori R. Holtz, this study. The mean age of enrolled patients was 5.8
Binh-minh Le, Sherry Lassa-Claxton, years, and the median age was 3.1 years. Thirty patients
Charles Canter, Paul Hmiel, Shalini Shenoy, received transplants and were studied for 1 year after
Stuart Sweet, Yumirle Turmelle, Ross Shepherd, transplantation. We collected 716 clinical specimens (160
and David Wang nasopharyngeal swab, 169 urine, 122 fecal, 265 plasma)
Immunocompromised patients are at risk for disease
during 265 patient visits. We collected 298 specimens from
caused by infection by some polyomaviruses. To define patients during symptomatic episodes, which were defined
the prevalence of polyomaviruses in children undergoing as having >1 of the following: fever, respiratory symptoms,
transplantation, we collected samples from a longitudinal or gastrointestinal symptoms. We collected clinical data
cohort and tested for the 9 known human polyomaviruses. using a questionnaire and the medical records.
All were detected; several were present in previously Fecal material was diluted 1:6 in phosphate-buffered
unreported specimen types. saline and filtered through 0.45-μm membranes. For
all specimens, we extracted total nucleic acids using an
Ampliprep Cobas extractor (Roche, Branchburg, NJ, USA).
B K and JC polyomaviruses (BKPyV, JCPyV) cause
disease in immunocompromised persons. Both are
double-stranded DNA viruses in the family Polyomaviridae.
We used published real-time PCRs for WUPyV (9), KIPyV
(9), TSPyV (5), MCPyV (10), BKPyV (11), and JCPyV
Seven additional human polyomaviruses were discovered (12) (Table 1). We developed assays for HPyV6, HPyV7,
during 2007–2011: KI polyomavirus (KIPyV) (1), WU and HPyV9 using Primer Express software (Applied
polyomavirus (WUPyV) (2), Merkel cell polyomavirus Biosystems, Carlsbad, CA, USA) (Table 1). To assess the
(MCPyV) (3), human polyomavirus 6 (HPyV6) (4), human performance of each assay, we used serial dilutions (5 to
polyomavirus 7 (HPyV7) (4), trichodysplasia spinulosa- 5 × 106 copies/reaction) of a plasmid containing the target
associated polyomavirus (TSPyV) (5), and human sequence. All 3 assays demonstrated a sensitivity of ≈5
polyomavirus 9 (HPyV9) (6). copies/reaction and yielded linear curves with R2 values
The 7 novel polyomaviruses have been detected in >0.99.
various specimen types; detection has been extensively Each of the 25-μl quantitative PCRs included 5 μL of
reviewed for KIPyV, WUPyV, and MCPyV (7). extracted sample, 12.5 pmol of each primer, and 4 pmol
Polyomaviruses HPyV6, HPyV7, TSPyV, and HPyV9 of probe. The MCPyV primers and probe were used as
have been detected in skin (4,5,8); TSPyV and HPyV9 described (10). We tested samples in a 96-well plate format,
have also been detected in urine, and HPyV9 was detected with 8 water negative controls and 1 positive control/plate.
in serum (6). However, only 2 of these recently identified Reactions were cycled as recommended using either an
viruses have been specifically implicated in human ABI 7500 real-time thermocycler (Applied Biosystems) or
diseases; MCPyV is associated with Merkel cell carcinoma a CFX96 real-time thermocycler (BioRad, Hercules, CA,
(3), and TSPyV has been linked to trichodysplasia USA). The threshold of all plates was set at a standard
spinulosa (5). Immunosuppression is a likely cofactor in value, and samples were counted as positive if their cycle
both diseases. The potential pathogenicity of the other 5 threshold was <37.00.
novel polyomaviruses is unknown. As a first step toward All 716 specimens were tested for each virus (Table 2).
The most frequently detected virus was BKPyV, which was
found primarily in urine as expected. JCPyV was detected
Author affiliation: Washington University School of Medicine, St. in 1 plasma sample. HPyV6, HPyV7, MCPyV, and TSPyV
Louis, Missouri, USA were detected in specimen types not previously reported.
DOI: http://dx.doi.org/10.3201/eid1810.120359 HPyV6 and TSPyV were detected in fecal samples and

1676 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Polyomaviruses in Children Undergoing Transplantation

Table 1. Real-time PCR assays to detect human polyomaviruses in children undergoing transplants, United States, 2008–2010*
Virus Target Primers, 5ƍ o 3ƍ Probe, 5ƍ o 3ƍ
WUPyV NCCR WU-C-4824-F: GGCACGGCGCCAACT WU-C-4861-TM: FAM-
WU-C-4898-R: TGCCATACCAACACAGCTGCTGAGC-TAMRA-3ƍ
CCTGTTGTAGGCCTTACTTACCTGTA
KIPyV LTAg KI-B-4603-F: GAATGCATTGGCATTCGTGA KI-B-4632-TM: FAM-
KI-B-4668-R: TGTAGCCATGAATGCATACATCCCACTGC-TAMRA
GCTGCAATAAGTTTAGATTAGTTGGTGC
TSPyV LTAg LTF: TGTGTTTGGAAACCAGAATCATTTG LTP: FAM-TTCTTCTTCCTCCTCATCCTCCACCTCAAT-
LTR: TGCTACCTTGCTATTAAATGTGGAG TAMRA
MCPyV LTAg LT.1F: CCACAGCCAGAGCTCTTCCT LT probe: FAM-TCCTTCTCAGCGTCCCAGGCTTCA-
LT.1R: TGGTGGTCTCCTCTCTGCTACTG TAMRA
HPyV6 VP1 ES011F: GCCTGGAAGGGCCTAGTAAAG ES024: FAM-
ES012R: ACCAACCATCTGTTGCAGGCATTAAAGCTA-TAMRA
ATTGGCAGCTGTAACTTGTTTTCTG
HPyV7 VP1 ES017F: ES025: FAM-CCTGCAAGCCCTCAGAAAGCAAGTAAATG-
GGTCCAGGCAATACTGATGTAGCTA TAMRA
ES018R: TCTGCAACCCAGAGCTCTACTG
HPyV9 LTAg ES026F: GAAGACCCTGATCCTGAGGAAGA ES031: FAM-TGGATCATCCCAGAGTTCATTTACCTGCA-
ES027R: TAMRA
CTCTGGAGTATTAGGTTCAGGCTTCT
BKPyV LTAg BK-Deg-F: Bk-Deg-P: FAM-
AGCAGGCAAGRGTTCTATTACTAAAT AAGACCCTAAAGACTTTCCYTCTGATCTACACCAGTTT-
BK-Deg-R: TAMRA
GARGCAACAGCAGATTCYCAACA
JCPyV VP2/3 JL1 (F): JL1 (P): FAM-
AAGGGAGGGAACCTATATTTCTTTTG CTCATACACCCAAAGTATAGATGATGCAGACAGCA-
JL1 (R): TAMRA
TCTAGCCTTTGGGTAACTTCTTGAA
*WUPyV, WU polyomavirus; NCCR, non-coding control region; KIPyV, KI polyomavirus; LTAg, large T antigen; TSPyV, trichodysplasia spinulosa
polyomavirus; MCPyV, Merkel cell polyomavirus; HPyV, human polyomavirus; VP, virion protein; BKPyV, BK polyomavirus; JCPyV, JC polyomavirus.

nasopharyngeal swab samples, and HPyV7 was detected at this time was negative for KIPyV; plasma and urine
in a nasopharyngeal swab and urine. One fecal sample were not available for this study. The patient died of acute
was positive for MCPyV. Because HPyV6, HPyV7, and respiratory failure and extensive pulmonary hemorrhage
MCPyV have been previously detected in skin, we cannot 24 days after collection of this specimen. Despite the
rule out the possibility that their presence in specimens frequent detection of KIPyV in respiratory specimens, no
could have been caused by shedding from skin. studies have definitively linked infection with respiratory
We collected 2 serial nasopharyngeal samples that disease. Titers of KIPyV were high in the nasopharyngeal
were positive for KIPyV from patient 3001 (Table 2), swab sample from this patient 3 weeks before respiratory
a 1-year-old child who had received a bone marrow failure. Although this observation does not necessarily
transplant as treatment for Fanconi anemia. The first implicate KIPyV infection as a contributing factor in the
sample, a nasopharyngeal swab obtained 1 month after death of the patient, it suggests a poorly controlled KIPyV
transplant, had low levels of KIPyV. To determine the infection in the respiratory tract.
viral load of the second nasopharyngeal swab specimen Three specimens collected from patient 4001, a
collected 2 months later, we reanalyzed the sample in 13-year-old heart transplant recipient, were positive for
triplicate; on the basis of extrapolation of the standard TSPyV (Figure), but the patient did not have trichodysplasia
curve run in parallel, we estimated the viral load to spinulosa. At 1 week after transplant, the nasopharyngeal
be 1.3 × 109 genome copies/mL of nasopharyngeal swab and fecal samples were positive for TSPyV. At 1
swab transport media. This patient’s course was month after transplant, the nasopharyngeal swab sample
complicated by graft-versus-host disease of the gut was again positive for TSPyV, with a viral load of ≈2.3
and skin, renal failure requiring dialysis, and recurrent × 104 genome copies/mL of transport media. There is
pulmonary hemorrhage. The patient was critically ill currently only 1 TSPyV sequence in GenBank (accession
and had experienced multiorgan failure at the time of no. GU989205). We used 4 primer pairs to amplify the
the second sampling. Other microbiological test results complete genome of TSPyV from the nasopharyngeal swab
were negative at that time, including PCR for Epstein- taken 1 month after transplant. PCR products were cloned,
Barr virus, cytomegalovirus, human herpesvirus-6, and and the complete genome was sequenced to 3× coverage
adenovirus in the blood; aspergillus antigen detection in (GenBank accession no. JQ723730) and compared with the
blood; and bacterial cultures of blood, tracheal aspirate, other TSPyV sequence. There were 5 nt substitutions: 3 in
urine, and peritoneal fluid. The fecal specimen collected noncoding regions and 2 synonymous mutations.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1677
DISPATCHES

Table 2. Polyomaviruses detected among specimens from children undergoing transplants, USA, 2008–2010*
Virus Specimen type Transplant Ct Patient ID Date of collection, time elapsed from transplant
HPyV6 Feces BMT 32.19 3011 2012 Jun 6, 1 mo after
HPyV6 NP Heart 36.13 4005 2010 Nov 25, 7 mo after
HPyV6 Feces Lung 36.95 5001 2010 Aug 17, 1 mo after
HPyV7 NP Liver 34.57 1002 2009 Jun 16, 7 mo after
HPyV7 Urine Liver 36.54 1002 2009 Jul 15, 8 mo after
HPyV9 Urine Liver 36.72 1009 2010 Feb 9, 1 wk after
KIPyV NP BMT 16.28 3001 2009 Jul 7, 3 mo after
KIPyV NP BMT 36.07 3001 2009 May 19, 1 mo after
KIPyV NP BMT 33.37 3008 2009 Nov 12, before
KIPyV NP BMT 31.04 3009 2010 Jul 3, 6 mo after
MCPyV NP BMT 36.29 3011 2010 Apr 15, before
MCPyV Feces BMT 34.56 3011 2010 Jul 2, 2 mo after
TSPyV NP Heart 32.98 4001 2009 May 29, 1 wk after
TSPyV NP Heart 30.74 4001 2009 Jun 18, 1 mo after
TSPyV Feces Heart 33.89 4001 2009 May 29, 1 wk after
WUPyV NP BMT 36.62 3005 2009 Jul 15, before
WUPyV NP BMT 28.81 3007 2009 Nov 6, 2 mo after
JCPyV Plasma BMT 36.12 3011 2010 Aug 24, 3 mo after
BKPyV Urine BMT 15.83 3010 2010 Apr 15, 1 mo after
BKPyV Urine Kidney 36.67 2022 2010 Jul 1, 10 mo after
BKPyV Urine BMT 30.80 3011 2010 Aug 24, 3 mo after
BKPyV Urine Heart 25.84 4001 2009 Aug 14, 2 mo after
BKPyV Urine Heart 35.89 4003 2009 Dec 23, 2 mo after
BKPyV Urine Heart 24.37 4001 2009 Sep 23, 4 mo after
BKPyV Urine Liver 28.56 1010 2009 Nov 23, 1 wk after
BKPyV Urine Lung 33.13 5002 2011 May 3, 1 y after
BKPyV Urine Lung 25.25 5002 2011 Feb 8, 10 mo after
BKPyV Urine Kidney 9.97 2002 2010 Mar 4, 6 mo after
BKPyV Urine BMT 30.10 3009 2010 Mar 5, 2 mo after
BKPyV Urine Liver 22.89 1001 2009 Jan 7, 3 mo after
BKPyV Urine Kidney 34.41 2002 2010 May 13, 8 mo after
BKPyV NP Kidney 35.93 2002 2010 Mar 4, 6 mo after
BKPyV Feces Kidney 33.15 2002 2010 Mar 4, 6 mo after
BKPyV Feces Liver 33.33 1001 2008 Dec 18, 2 mo after
BKPyV Feces Liver 34.84 1001 2009 Jan 7, 3 mo after
*Ct, cycle threshold; ID, identification; HPyV, human polyomavirus; BMT, bone marrow transplant; NP, nasopharyngeal; KIPyV, KI polyomavirus;
MCPyV, Merkel cell polyomavirus; TSPyV, trichodysplasia spinulosa polyomavirus; WUPyV, WU polyomavirus; JCPyV, JC polyomavirus; BKPyV, BK
polyomavirus

Although serologic studies have demonstrated that apart were positive for TSPyV, suggesting it may persist
≈70% of adults in Europe have been infected by TSPyV for extended periods in the respiratory tract, at least in
(13), its mode of transmission is unknown. The detection immunosuppressed persons.
of TSPyV in nasopharyngeal swab and fecal samples raises
the possibility that it may be transmitted by a respiratory Conclusions
or fecal–oral route. Furthermore, in the current study, 2 Our goals were to establish a longitudinal repository
sequential nasopharyngeal swab samples taken 20 days of different specimen types from transplant recipients
and to define the prevalence of polyomaviruses in these
patients. We detected all 9 polyomaviruses in at least 1
specimen. Although the prevalence of each virus was
generally low, TSPyV, HPyV6, HPyV7, and MCPyV were
detected in specimen types not previously reported. These
observations expand understanding of the recently identified
polyomaviruses and the tissue and organ systems they may
infect and suggest possible modes of transmission. Further
studies to define their possible roles in human diseases are
Figure. Samples tested for TSV (trichodysplasia spinulosa needed.
polyomavirus) during May–June 2009 from patient 4001, a 13-year-
old heart transplant recipient at St. Louis Children’s Hospital,
Acknowledgments
St. Louis, Missouri, USA. Samples tested at each time point are
indicated by white squares. Black squares represent positive We thank Angel Chen and Adira Vinograd for their help in
samples. NP, nasopharyngeal. compiling the clinical data for this study. We also thank M.C.W.

1678 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Polyomaviruses in Children Undergoing Transplantation

Feltkamp for the TSPyV-positive control and Christopher 6. Scuda N, Hofmann J, Calvignac-Spencer S, Ruprecht K, Liman P,
Buck for the HPyV6-, HPyV7-, and MCPyV-positive controls Kuhn J, et al. A novel human polyomavirus closely related to the
African green monkey–derived lymphotropic polyomavirus (LPV).
(AddGene plasmids 24727–24729, respectively; addgene.org). J Virol. 2011;85:4586–90. http://dx.doi.org/10.1128/JVI.02602-10
7. Babakir-Mina M, Ciccozzi M, Perno CF, Ciotti M. The novel
This study was supported by a grant from the Children’s
KI, WU, MC polyomaviruses: possible human pathogens? New
Discovery Institute. E.A.S. was supported by the Department of Microbiol. 2011;34:1–8.
Defense through the National Defense Science & Engineering 8. Sauvage V, Foulongne V, Cheval J, Ar Gouilh M, Pariente K, Dereure
Graduate Fellowship Program. L.R.H. is supported by O, et al. Human polyomavirus related to African green monkey
lymphotropic polyomavirus. Emerg Infect Dis. 2011;17:1364–70.
ULIRR024992 subaward KL2RR024994 from the National
9. Bialasiewicz S, Whiley DM, Lambert SB, Gould A, Nissen MD,
Institutes of Health, National Center for Research Resources. Sloots TP. Development and evaluation of real-time PCR assays for
the detection of the newly identified KI and WU polyomaviruses.
Ms Siebrasse is a graduate student at Washington University J Clin Virol. 2007;40:9–14. http://dx.doi.org/10.1016/j.jcv.2007.
in St. Louis, Missouri. Her research focuses on the discovery and 07.015
characterization of novel polyomaviruses. 10. Goh S, Lindau C, Tiveljung-Lindell A, Allander T. Merkel cell
polyomavirus in respiratory tract secretions. Emerg Infect Dis.
2009;15:489–91. http://dx.doi.org/10.3201/eid1503.081206
References 11. Dumoulin A, Hirsch HH. Reevaluating and optimizing polyomavirus
1. Allander T, Andreasson K, Gupta S, Bjerkner A, Bogdanovic G, BK and JC real-time PCR assays to detect rare sequence
Persson MA, et al. Identification of a third human polyomavirus. J polymorphisms. J Clin Microbiol. 2011;49:1382–8. http://dx.doi.
Virol. 2007;81:4130–6. http://dx.doi.org/10.1128/JVI.00028-07 org/10.1128/JCM.02008-10
2. Gaynor AM, Nissen MD, Whiley DM, Mackay IM, Lambert SB, Wu 12. Pal A, Sirota L, Maudru T, Peden K, Lewis AM Jr. Real-time,
G, et al. Identification of a novel polyomavirus from patients with quantitative PCR assays for the detection of virus-specific DNA in
acute respiratory tract infections. PLoS Pathog. 2007;3:e64. http:// samples with mixed populations of polyomaviruses. J Virol Methods.
dx.doi.org/10.1371/journal.ppat.0030064 2006;135:32–42. http://dx.doi.org/10.1016/j.jviromet.2006.01.018
3. Feng H, Shuda M, Chang Y, Moore PS. Clonal integration 13. van der Meijden E, Kazem S, Burgers MM, Janssens R, Bouwes
of a polyomavirus in human Merkel cell carcinoma. Science. Bavinck JN, de Melker H, et al. Seroprevalence of trichodysplasia
2008;319:1096–100. spinulosa–associated polyomavirus. Emerg Infect Dis. 2011;17:
4. Schowalter RM, Pastrana DV, Pumphrey KA, Moyer AL, Buck 1355–63.
CB. Merkel cell polyomavirus and two previously unknown
polyomaviruses are chronically shed from human skin. Cell Address for correspondence: David Wang, Washington University School
Host Microbe. 2010;7:509–15. http://dx.doi.org/10.1016/j. of Medicine, Campus Box 8230, 660 S. Euclid Ave, St. Louis, MO 63110,
chom.2010.05.006
5. van der Meijden E, Janssens RW, Lauber C, Bouwes Bavinck USA; email: davewang@borcim.wustl.edu
JN, Gorbalenya AE, Feltkamp MC. Discovery of a new human
polyomavirus associated with trichodysplasia spinulosa in an All material published in Emerging Infectious Diseases is in
immunocompromized patient. PLoS Pathog. 2010;6:e1001024. the public domain and may be used and reprinted without
http://dx.doi.org/10.1371/journal.ppat.1001024 special permission; proper citation, however, is required.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1679
DISPATCHES

Preventing and Management of Ships’ Ballast Water and Sediments

(BWM Convention) on February 13, 2004 (6). Regulation
Maritime Transfer D-2 of the convention establishes numeric ballast water dis-
charge standards, to be phased in with a start date of January
of Toxigenic 1, 2012, that limit the concentration of viable organisms and

Vibrio cholerae
human pathogens (including toxigenic V. cholerae, Esch-
erichia coli, and intestinal enterococci). The limit for toxi-
genic V. cholerae is <1 CFU/100 mL or <1 CFU/g (wet
Nicole J. Cohen, Douglas D. Slaten, Nina Marano, weight) zooplankton samples. In the interim, the BWM
Jordan W. Tappero, Michael Wellman, Convention requires that, whenever possible, ships conduct
Ryan J. Albert, Vincent R. Hill, David Espey, ballast water exchange >200 nautical miles from the nearest
Thomas Handzel, Ariel Henry, land and in water >200 m deep. If these requirements can-
and Robert V. Tauxe not be met, the exchange should be performed as far from
the nearest land as possible, but at a minimum >50 nautical
Organisms, including Vibrio cholerae, can be trans- miles from the nearest land and in water >200 m deep. When
ferred between harbors in the ballast water of ships. Zones these requirements cannot be met, areas may be designated
in the Caribbean region where distance from shore and wa-
where ships can conduct ballast water exchange. Ballast wa-
ter depth meet International Maritime Organization guide-
lines for ballast water exchange are extremely limited. Use
ter exchange is based on the principles that 1) organisms
of ballast water treatment systems could mitigate the risk for from coastal areas will not survive in the open ocean and 2)
organism transfer. fewer organisms (including fewer human pathogens) will be
taken up in the open ocean, and these will be less likely to
adapt to coastal waters.
Cholera is an acute diarrheal illness caused by toxi- A cholera epidemic emerged in Haiti in October 2010;
genic strains of the bacterium Vibrio cholerae serogroups lack of safe water and sanitation infrastructure and the dev-
O1 and O139. V. cholerae, like other vibrios, is found com- astation caused by the January 2010 earthquake contributed
monly in marine and estuarine environments, living freely to its spread (7). Concerns were raised that cholera could be
or on surfaces, such as plants and animal shells, and in transferred from Haiti to other countries through contami-
intestinal contents of marine animals (1). V. cholerae in- nation of coastal waters by ship ballast water. Ship traffic to
fection is typically acquired by ingestion of contaminated Haiti (233 vessel calls in Port-au-Prince in 2008) consists
water or food (2). predominantly of cargo vessels with destinations in the
Ballast water is collected in ships to regulate their sta- United States, other Caribbean islands, and Latin Ameri-
bility; the discharge of ballast water can transfer toxigenic ca (8). During an assessment of cholera contamination of
V. cholerae O1 from one harbor to another. (3). During fresh and marine water sources in Haiti during the epidemic
1992, shellfish in Mobile Bay, Alabama, on the US coast of conducted by the US Centers for Disease Control and Pre-
the Gulf of Mexico, were contaminated with an epidemic vention (CDC), US Food and Drug Administration, and
strain of toxigenic V. cholerae O1 from Latin America, al- the Haitian Ministry of Health and Population, water and
though no human illnesses were reported (4). V. cholerae seafood collected from harbors at Port-au-Prince and St.
transfer by cargo ship was documented when the same Marc were tested for viable V. cholerae and for the cholera
strain was isolated from ballast and other nonpotable water toxin gene (ctxA) (9). Toxigenic V. cholerae O1 serotype
samples collected from 5 cargo ships from ports in Latin Ogawa indistinguishable from the outbreak strain was iso-
America that arrived in the US Gulf of Mexico (5). lated from seafood samples from Port-au-Prince. Although
To reduce the risk for transfer of invasive species and V. cholerae was not isolated from marine water samples,
pathogens between harbors by introduction of contaminat- the ctxA gene was detected in broth cultures of seawater
ed ballast water, the International Maritime Organization samples from both harbors, suggesting that harbor waters
(IMO) adopted the International Convention for the Control were contaminated with toxigenic V. cholerae.

Author affiliations: Centers for Disease Control and Prevention, The Study
Atlanta, Georgia, USA (N.J. Cohen, D.D. Slaten, N. Marano, J.W. To further evaluate the risk for cholera transfer through
Tappero, M. Wellman, V.R. Hill, D. Espey, T. Handzel, R.V. Tauxe); ballast water under existing management approaches, we
US Environmental Protection Agency, Washington, DC, USA (R.J. applied the IMO ballast water exchange depth and distance
Albert); and Haitian Ministry of Public Health and Population, Port- criteria to the Caribbean region. Buffers of 50 and 200
au-Prince, Haiti (A. Henry) nautical miles were generated on the basis of the Global
DOI: http://dx.doi.org/10.3201/eid1810.120676 Self-Consistent, Hierarchical, High-Resolution Shoreline

1680 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Preventing Transfer of V. cholerae

first, issued November 10, 2011, asked ship captains not to

exchange ballast water in Haitian harbors. The second, is-
sued November 15, 2011, reminded ship captains to adhere
to IMO guidelines for ballast water exchange. However,
ships operating in the Caribbean Sea face practical diffi-
culties in conducting ballast water exchange at the recom-
mended distances without making large deviations from
usual routes. The BWM Convention recommends that des-
ignated ballast water exchange areas should be on or close
to existing maritime routes and does not require that ships
deviate course or delay voyages to comply with ballast wa-
ter exchange guidelines (6). Ballast water exchange at sea
also presents operational and safety challenges to ships and
might not be completely effective in preventing spread of
aquatic organisms (12).
Figure. Zones in the Caribbean region where distance from As an alternative management measure, ballast water
shore and water depth meet International Maritime Organization treatment systems are being developed to meet the BWM
guidelines for ballast exchange. To exchange ballast >200 nautical Convention’s numeric discharge standards, and several
miles from shore in water 200 m deep, ships must travel 280
systems are available (13). These systems combine filtra-
nautical miles northeast of Haiti (A) or to the Gulf of Mexico (B). To
exchange ballast at the minimum 50 nautical miles from shore in tion with nonchemical (e.g., UV light, shear, heat) and
water >200 m deep, ships must travel >90 nautical miles northeast chemical biocides to remove or kill organisms (14). Ballast
(C) or 50 nautical miles south (D) of Haiti or conduct the exchange water treatment systems are designed to achieve the BWM
in an area <45 nautical miles wide approximately equidistant from Convention efficacy levels, and their future use in the Ca-
Haiti, Cuba, and Jamaica (E). Light gray shading indicates distance
ribbean region would likely be a more effective manage-
from land is <50 nautical miles and/or seawater depth is <200 m.
Medium gray shading indicates distance from land is >50 nautical ment approach than ballast water exchange.
miles but <200 nautical miles, and seawater depth is >200 m. Dark The Haiti cholera outbreak spread to the neighboring
gray shading indicates distance from land is >200 nautical miles Dominican Republic, and cholera cases associated with
and seawater depth is >200 m. travel to Haiti were recognized in the United States (7), but
there is no evidence that V. cholerae was transferred by
ship ballast water in these instances. The Wider Caribbean
Region comprises 28 island and continental countries with
Database (version 2.1; www.ngdc.noaa.gov/mgg/shore- coasts on the Caribbean Sea, the Gulf of Mexico, and adja-
lines/gshhs.html) and overlaid on bathymetric data from cent waters of the Atlantic Ocean. This area is at elevated
the ETOPO1 Global Relief Model (www.ngdc.noaa.gov/ risk for transfer of contaminated ballast water because of
mgg/global/global.html) (10,11). We acquired these datas- the high volume of cargo trade in the region and has been
ets through the National Oceanic and Atmospheric Admin- prioritized by the Global Ballast Water Management Pro-
istration’s National Geophysical Data Center (www.ngdc. gram (http://globallast.imo.org/index.asp), a program ad-
noaa.gov). ministered by IMO, the Global Environment Facility, and
Mapping indicates that waters around Haiti where the United Nations Development Program to assist devel-
the IMO guidelines can be followed are extremely limited oping countries implement ballast water management mea-
(Figure). To exchange ballast >200 nautical miles from sures (15).
shore in water 200 m deep, ships must travel 280 nauti-
cal miles northeast of Haiti (Figure, A) or to the Gulf of Conclusions
Mexico (Figure, B). To exchange ballast at the minimum A comprehensive regional strategic plan has been de-
50 nautical miles from shore in water >200 m deep, ships veloped to promote ratification of the BWM Convention
must travel >90 nautical miles northeast (Figure, C) or 50 and to facilitate its implementation within the Wider Carib-
nautical miles south (Figure, D) of Haiti or conduct the ex- bean Region through regional cooperation, training, com-
change in an area <45 nautical miles wide approximately munication, compliance monitoring and enforcement, and
equidistant from Haiti, Cuba, and Jamaica (Figure, E). promotion of national-level legislation, task forces, action
After discussions with staff at CDC and the Pan plans, and sustainable resources to support activities (15).
American Health Organization, the Director General of Implementation of this plan will help protect public health
the Haitian Ministry of Health and Population issued 2 by reducing the likelihood that V. cholerae and other patho-
memoranda addressing ballast water management. The gens will be transferred by ballast water.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1681
DISPATCHES

Acknowledgments 8. US Department of Transportation Maritime Administration.

We thank André Berro, William Jackson, and Emad Yanni MarView, Marine Transportation System [cited 2011 March 4].
http://hsa0cweb0iwp006.marview.gov/pls/apex/f?p=102:1:
for their contributions to this investigation. We acknowledge the 3801394268415490
invaluable advice and guidance of Daniel Menucci, Agnes Soares, 9. Hill VR, Cohen N, Kahler AM, Jones JL, Bopp CA, Marano N,
Roberta Andraghetti, and the late Yves Chartier. et al. Toxigenic Vibrio cholerae O1 in water and seafood, Haiti.
Emerg Infect Dis. 2011;17:2147–50. http://dx.doi.org/10.3201/
Dr Cohen is associate chief for science, Quarantine and Bor- eid1711.110748
der Health Services Branch, Division of Global Migration and 10. Wessel P, Smith WHF. A global, self-consistent, hierarchical, high-
resolution shoreline database. J Geophys Res. 1996;101:8741–3.
Quarantine, CDC. Her research interests include pediatric infec-
http://dx.doi.org/10.1029/96JB00104
tious diseases and the prevention of spread of disease through 11. Amante C, Eakins BW. ETOPO1 1 Arc-Minute Global Relief Mod-
global migration, travel, and trade. el: procedures, data sources and analysis [cited 2011 Dec 28]. http://
www.ngdc.noaa.gov/mgg/global/relief/ETOPO1/docs/ETOPO1.pdf
12. Committee on Assessing Numeric Limits for Living Organisms in
References Ballast Water; National Research Council. Assessing the relation-
ship between propagule pressure and invasion risk in ballast water.
1. Cabral JPS. Water microbiology. Bacterial pathogens and water. Washington (DC): National Academies Press; 2011.
Int J Environ Res Public Health. 2010;7:3657–703. http://dx.doi. 13. United States Environmental Protection Agency. Efficacy of bal-
org/10.3390/ijerph7103657 last water treatment systems: a report by the EPA Science Advisory
2. Seas C, Gotuzzo E. Cholera. In: Principles and practice of infectious Board. EPA-SAB-11-009. Washington (DC): The Agency; 2011.
diseases. 7th ed. Philadelphia: Elsevier Health Sciences; 2009. p. 14. American Bureau of Shipping. Guide for ballast water treat-
2777–85. ment [cited 2011 Dec 28]. http://www.eagle.org/eagleExternal-
3. Ruiz GM, Rawlings TK, Dobbs FC, Drake LA, Mullady T, Huq A, et PortalWEB/ShowProperty/BEA%20Repository/Rules&Guides/
al. Global spread of microorganisms by ships. Nature. 2000;408:49– Current/187_BWT/Guide
50. http://dx.doi.org/10.1038/35040695 15. United Nations Environmental Program. Regional strategy to mini-
4. Centers for Disease Control and Prevention. Isolation of Vibrio mize the transfer of harmful aquatic organisms and pathogens on
cholerae O1 from oysters—Mobile Bay, 1991–1992. MMWR Morb ships’ ballast water and sediments, Montego Bay, Jamaica [cited
Mortal Wkly Rep. 1993;42:91–3. 2011 Dec 28]. http://cep.unep.org/racrempeitc/globallast-partner-
5. McCarthy SA, Khambaty FM. International dissemination of epi- ships-1/WG.32%20REF.5%20Rev1_BWM%20Strategic%20Ac-
demic Vibrio cholerae by cargo ship ballast and other nonpotable tion%20Plan_ENG.pdf
waters. Appl Environ Microbiol. 1994;60:2597–601.
6. International Maritime Organization. International convention for Address for correspondence: Nicole J. Cohen, Centers for Disease Control
the control and management of ships’ ballast water and sediments
and Prevention, 1600 Clifton Rd NE, Mailstop E03, Atlanta, GA 30333,
(BWM) [cited 2011 Dec 28]. www.imo.org/About/Conventions/
ListOfConventions/Pages/International-Convention-for-the-Con- USA; email: hei1@cdc.gov
trol-and-Management-of-Ships’-Ballast-Water-and-Sediments-
(BWM).aspx.
7. Tappero JW, Tauxe RV. Lessons learned during public health re- All material published in Emerging Infectious Diseases is in
sponse to the cholera epidemic in Haiti and the Dominican Repub- the public domain and may be used and reprinted without
lic. Emerg Infect Dis. 2011;17:2087–93. http://dx.doi.org/10.3201/ special permission; proper citation, however, is required.
eid1711.110827

1682 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
Helping CDC Do More, Faster
Established by Congress as an independent, nonprofit
organization, the CDC Foundation connects the Centers for
Disease Control and Prevention (CDC) with private-sector
organizations and individuals to build public health programs
that support CDC’s work 24/7 to save lives and protect people
from health, safety and security threats.

Since 1995, the CDC Foundation has provided more than

$300 million to support CDC’s work, launched more than 500
programs around the world and built a network of individuals
and organizations committed to supporting CDC and public
health.

Each CDC Foundation program involves a talented team of

experts at CDC and at least one outside funding partner.
Sometimes, a program begins with a CDC scientist who has a
great idea and wants to collaborate with an outside partner
to make it happen. At other times, organizations in the
private sector recognize that they can better accomplish their
own public health goals by working with CDC through the
CDC Foundation.

JOIN US www.cdcfoundation.org
Photos: David Snyder / CDC Foundation
ANOTHER DIMENSION

A Natural History of Infective Endocarditis,

Preceded by Decompensated Chronic Liver Disease
and Severe Community-acquired Pneumonia
Indictment in Four Parts and with Right of Reply
Nancy L. Merridew

ONE. Trespass THREE. Progressive Assault

His stalwart skin, sallow armor, His heart, excavated by festering mercenaries,
harbored memories and marks inexorably leached
of a difficult life. their venomous arsenal.
His grave affliction, already perilous, His body, conscripted slave to advancing legion,
he endured alone— appointed satellite outposts—
ostracized by family and society. Streptococcal Empire rising.
His swollen belly, crazed by Medusa, His scent, previously acrid,
pouted taut as artificial breaths drifted metallic—
stretched rankled lungs. life corroding.
His seasoned vessels, illicit euphoria’s labyrinth, His brain, shadowed by grim delirium;
adorned with plastic portals encephalopathic, embolic, prophetic—
infused antidote. entombed within cranial vault.
His cloying blood, claret lately poisoned, His eyelids, fluttering shrouds,
delivered assassin to lair— veiled infinite irises, icteric globes and
Duke in silent carriage alighted at the chamber. retinas maligned by immunity.
His urine, once flaxen,
TWO. Malice Aforethought oozed vermillion—
His haughty foe, ignoble despite the Criteria, tributary of beckoning Rubicon.
stole a vantage of supreme command His skin, chameleon canvas—
at the Greatest Vessel door. flushed lesions and jaundice
His insistent heartbeat, two sounds once discreet, defaced faded tattoos.
in rhythmic phrases murmured His fingers, pulps mottled carmine,
remarks of furtive progress. uniquely etched tips
His ardent doctor, genius clinician, reached to probe oblivion.
lifted stethoscope to precordium— His nail-beds, fleshy Rosetta,
message received. signaled the secret
His pliant pulses, company accommodated— in flecked hieroglyphs.
Watson struck and shuddered, His palms, unaccustomed to prayer,
Corrigan thumped and ebbed. denied redemption—
His teeming veins, mercurial blood extracted, pocked, they too bore stigmata.
yielded Duke’s identity
for precise retaliation. FOUR. Homicidal Ascendancy
His shriveled Liver, nominally ironic,
sacrificed to intoxicated sanctuary— His surgical saviors, glinting mirage,
Child-Pugh C proscribed surgical reprieve. could offer no cure
His adept assailant, though exposed for battle, except hastened demise.
abetted by fickle liver His thwarted physicians, ashen surrender—
retained advantage for the coup. our elixirs no match
for the adversary.

Author affiliation: The Alfred Hospital, Carlton, Victoria, Australia

DOI: http://dx.doi.org/10.3201/eid1810.AD1810

1684 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 A Natural History of Infective Endocarditis

His recalcitrant Duke, relentless, His only visitor, a community case worker,
accomplished unsavory absent that day.
feats of cardiac passage. Alone when he succumbed to the darkness.
His core engine, chambers thus connected,
ramped to fuel abscessed conduit— Audi Alteram Partem
blood roared relentlessly. Virulent villain—do you admire your conquest?
His slumped neck, kinetically possessed—
terminally mutilated muscle’s Dr. Merridew is a trainee in internal medicine at The Alfred
futile beats rocketed and ricocheted. Hospital, Melbourne, Australia. Her research interests include
His uncanny figure, cadaveric, infectious diseases, public health, and medical education.
inevitably bloated;
Starling forces prevailed. Address for correspondence: Nancy L. Merridew, The Alfred Hospital,
His sodden lungs, emancipated from machines, PO Box 1292, Carlton, Victoria 3053, Australia; email: nancy.merridew@
hissed and spluttered gmail.com
beneath swelling king tide.
His dying breaths, labored then whispered,
dim light obscured final chest fall—
curtains drawn, despite the summer sun.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1685
LETTERS

Trypanososma (Montreal. Quebec, Canada). We was >3-fold higher for persons with
obtained data on case-patients given late-stage cases (8.1%) than for those
brucei rhodesiense a diagnosis of T. b. rhodesiense HAT with early-stage cases (2.4%) (p<0.01,
Sleeping Sickness, at a HAT treatment unit in Uganda. by χ2 test). Given that central Uganda
Uganda These diagnoses were reported to the is the critical zone for convergence
National Sleeping Sickness Control and intervention, such evidence of
To the Editor: The past 2 decades Program of the Ministry of Health presumed diagnosis and treatment
have heralded notable success in efforts (4,6). Data were assigned locations delay is cause for concern.
to control sleeping sickness (human by parish, and analyses focused The SOS phase 1 intervention
African trypanosomiasis [HAT]) in on spatiotemporal trends in case period (2006–2008) coincided with
Africa. HAT is a neglected tropical occurrence. The final cleaned dataset a period of reduced reported pre-
disease with major public health and contained 2,501 reported cases of valence of HAT (Figure). The monthly
economic effects in sub-Saharan Africa, presumed T. b. rhodesiense HAT. average was 27 cases/month before
and its effects on livestock productivity In the past 10 years in Uganda, the intervention and 10 cases/month
and development are considered major 140 cases of fatal T. b. rhodesiense after the intervention (p<0.01, by
constraints to alleviating poverty in this HAT have been reported. However, Mann-Whitney test). However, a
region (1,2). Because of concerted and given estimates of underreporting and substantive component of reduction
coordinated continental control efforts, cessation of active surveillance, actual in incidence occurred in districts not
its incidence has steadily decreased. deaths are likely > 1,700 (170 deaths/ included in the SOS intervention
Despite these successes, concern year) (6). Notably, mortality rates program. This pattern may reflect
has increased recently regarding have increased from an average of 5% reporting bias caused by a transition
potential convergence of the 2 in the early 2000s to ≈10% in later in Uganda in 2005 to a period of
causes of HAT (Trypanosoma brucei years, and rates have been higher in passive surveillance and underfunding
gambiense and T. brucei rhodesiense). recently affected districts. This pattern for national reporting. Therefore,
These organisms differ in transmission is predominantly driven by higher it remains unclear to what extent
and how infections are diagnosed and mortality rates in newly affected increased international attention and
treated, and control, and have never SOS districts in central Uganda, SOS intervention have contributed
coincided in the same area. Uganda in which diagnostic and treatment to HAT prevention and control. The
is the only country with endemic delays are higher, and from which an absence of a clear increase in incidence
distributions of these 2 trypanosome increasing proportion of HAT cases after reinstatement of national data
species, and convergence there are originating. acquisition in 2008 provides an early
represents a major public health Patients in SOS intervention indication that interventions may be
concern, given the potential for districts were more likely to report contributing to the decrease in, or at
overlapping infections to compromise cases in late stages of the disease least to stabilization of, geographic
treatment and control programs and (p<0.01, by χ2 test). The mortality rate spread in central Uganda.
spread into neighboring countries (3,4).
Risk for convergence led to an
international emergency intervention.
In 2006, an international public–
private partnership, Stamp Out
Sleeping Sickness (SOS), was
established to control spread of
this disease in central Uganda (5).
However, despite the continental
effect of convergence of the 2 causes
of HAT, little is known about trends in
incidence and epidemiology of HAT
in central Uganda. We report results of Figure. Human African trypanosomiasis cases and deaths by month, Uganda, 2000–2009.
data analysis for HAT caused by T. b. Bars indicate cases in districts in the Stamp Out Sleeping Sickness (SOS) intervention
rhodesiense during 2000–2009. region and outside the SOS region. Solid line indicates overall 24-month moving average
of deaths, dashed line indicates 24-month moving average of deaths in SOS intervention
This study was approved by
districts, and dotted line indicates 24-month moving average of deaths in non-SOS districts.
the ethics review board for human ISCTRC, International Scientific Council for Trypanosomiasis Research and Control; J M M
subjects at McGill University J S N, Jan, Mar, May, Jul, Sep, Nov.

1686 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

HAT data indicate seasonality References Rickettsia felis in

of this disease; incidence is higher
during January, February, and March
1. Fèvre EM, Wissmann B, Welburn SC, Aedes albopictus
(p = 0.04, by Mann-Whitney test).
Lutumba P. The burden of human African
trypanosomiasis. PLoS Negl Trop Dis. Mosquitoes,
Seasonality of HAT incidence has 2008;2:e333. http://dx.doi.org/10.1371/
journal.pntd.0000333
Libreville, Gabon
been noted elsewhere and linked
2. Welburn SC, Coleman PG, Maudlin I,
to seasonal influences on tsetse Fèvre EM, Odiit M, Eisler EC. Crisis,
To the Editor: Rickettsia felis, an
habitat suitability. We propose that what crisis? Control of Rhodesian sleeping emerging pathogen first identified in
seasonality of cattle trading may also sickness. Trends Parasitol. 2006;22:123–8. the cat flea (1), has been detected in
play a role because cattle purchases http://dx.doi.org/10.1016/j.pt.2006.01.011 other fleas, ticks, mites, and booklice
3. Berrang-Ford L, Berke O, Sweeney S,
increase before the Christmas season, Abdelrahman L. Sleeping sickness in
(2). R. felis can be cultured in mosquito
which promote pathogen spread and south-eastern Uganda: a spatio-temporal cell lines derived from Anopheles
increased transmission. This finding is analysis of disease risk, 1970–2003. Vector gambiae and Aedes albopictus
consistent with research highlighting Borne Zoonotic Dis. 2010;10:977–88. (Asian tiger) mosquitoes (2), so
http://dx.doi.org/10.1089/vbz.2008.0196
the role of livestock markets in the 4. Berrang-Ford L, Berke O, Abdelrahman
its compatibility with mosquitoes
spread of T. b. rhodesiense in central L, Waltner-Toews D, McDermott J. in nature can be suspected. In sub-
Uganda and would further support Spatial analysis of sleeping sickness in Saharan Africa, R. felis bacteremia in
a body of literature suggesting, as south-eastern Uganda, 1970–2003. Emerg humans is common¸ especially during
Infect Dis. 2006;12:813–20. http://dx.doi.
espoused by the SOS initiative, org/10.3201/eid1205.051284
the rainy season, when mosquitoes
that control of animal reservoirs of 5. Kabasa JD. Public-private partnership proliferate. We tested anthropophilic
the disease is a critical component works to stamp out sleeping sickness in mosquitoes for the presence of R. felis
of intervention measures (2,7–9). Uganda. Trends Parasitol. 2007;23:191–2. DNA (3–5).
http://dx.doi.org/10.1016/j.pt.2007.03.006
Implementation and enforcement of 6. Odiit M, Coleman PG, Liu W-C,
During December 2008–January
regulations for treatment of cattle McDermott JJ, Fèvre EM, Welburn SC, et 2010, we randomly selected female Ae.
before sale at markets would also al. Quantifying the level of under-detection albopictus and Ae. aegypti mosquitoes
contribute to limiting spread (9,10); of Trypanosoma brucei rhodesiense (96 each) from specimens obtained by
sleeping sickness cases. Trop Med Int
Interventions in districts in Health. 2005;10:840–9. http://dx.doi.
human-landing collections from 4 sites
central Uganda in which convergence org/10.1111/j.1365-3156.2005.01470.x in Libreville, Gabon (6). Specimens
is predicted have been slow and 7. Fèvre EM, Coleman P, Odiit M, Magona were collected during the rainy season
incomplete. If convergence has J, Welburn S, Woolhouse M. The origins (mid-January–end of May and end of
of a new Trypanosoma brucei rhodesiense
occurred, this finding indicates sleeping sickness outbreak in eastern
September–mid-December); no parity
that a specific region in Africa has Uganda. Lancet. 2001;358:625–8. data were available.
had concurrent infection with both http://dx.doi.org/10.1016/S0140- We extracted 192 DNA samples
causes of HAT, with implications 6736(01)05778-6 from homogenate (abdomen, wings,
8. Kabayo JP. Campaign against tsetse flies.
for prevention, treatment, and Public Health J. 2010;21:23–31.
legs) of each nonengorged, host-
control. Since 2000, Uganda has had 9. Simarro PP, Jannin J, Cattand P. Eliminating seeking, adult mosquito by using
continued northward spread of T. b. human African trypanosomiasis: where the BioRobot 8000 (QIAGEN
rhodesiense infections, reducing the do we stand and what comes next? S.A.S., Courtaboeuf, France) and
PLoS Med. 2008;5:e55. http://dx.doi.
distance with TbG to <100 km, which org/10.1371/journal.pmed.0050055
QIAamp Media MDx Kit (QIAGEN)
we believe is a conservative estimate. 10. Batchelor NA, Atkinson P, Gething P, according to the manufacturer’s
Reinstatement of active surveillance Picozzi K, Fevre E, Kakembo A, et al. instructions. Samples were screened
of HAT and support for central data Spatial predictions of Rhodesian human by quantitative real-time PCR
African trypanosomiasis (sleeping
collection in Uganda are long overdo sickness) prevalence in Kaberamaido
(qPCR) targeting the biotin synthase
and warranted immediately. and Dokolo, two newly affected districts (bioB) gene (4). Positive results were
of Uganda. PLoS Negl Trop Dis. confirmed by qPCR-based molecular
Lea Berrang-Ford, Charles 2009;3:e563. http://dx.doi.org/10.1371/ detection targeting the orfB gene,
Wamboga, and Abbas S.L. journal.pntd.0000563
which codes for a transposition
Kakembo helper protein. This qPCR used a
Address for correspondence: Lea Berrang-Ford,
Author affiliations: McGill University,
Department of Geography, McGill University,
set of primers not previously used
Montreal, Quebec, Canada (L. Berrang-
805 Sherbrooke St. West, Burnside 705,
in our laboratory (R_fel.OrfB_F:
Ford); and Ministry of Health, Kampala,
Montreal, Quebec H3A 2K6, Canada; email:
5′-CCCTTTTCGTAACGCTTTGCT-
Uganda (C. Wamboga, A.S.L. Kakembo)
lea.berrangford@mcgill.ca
3′ and R_fel.OrfB_R: 5′-GGGCTAAA
DOI: http://dx.doi.org/10.3201/eid1810.111213 CCAGGGAAACCT-3′) and the probe

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1687
LETTERS

R_fel.OrfB_P: 6-FAM-TGTTCCGGT Mosquitoes were considered positive of a mosquito after experimental

T T TA A C G G C A G ATA C C C A - for R. felis when the qPCR result was feeding on a bacteremic host
TAMRA. Specificity of the qPCR was <35 cycle thresholds for 1 target gene or bacterial suspension; and
tested in silico and on the 31 Rickettsia and the additional DNA sequence was demonstration of the transmission of
spp. from our laboratory. The final successfully amplified. No sample bacteria to a vertebrate through the
qPCR reaction mixture contained in this study was positive for only 1 bite of a mosquito (10).
extracted DNA (5 μL) and mix (15 target gene or had a qPCR threshold
μL) that contained master mix (10 >35 cycle thresholds for both genes. Cristina Socolovschi,
μL) from the QuantiTect Probe PCR Contamination is a critical Frédéric Pagés,
Kit (QIAGEN, Hilden, Germany), problem for the PCR-based and Didier Raoult
each primer (0.5 μL, 20 pmol), probe identification of microbes. However, Author affiliations: Aix-Marseille Université,
(0.5 μL, 62.5 nmol), and RNase-free the validity of the data we report is Marseille, France (C. Socolovschi, D.
water (3.5 μL). Amplification and based on strict laboratory procedures Raoult); and CIRE/ARS Océan Indien (Les
sequence detection were performed and controls that are commonly used Cellules interrégionales d’épidémiologie/
in a CFX96 Touch thermocycler (Bio- in the World Health Organization Agence Régionale de Santé), La Réunion,
Rad, Marnes-la-Coquette, France) as Reference Center for Rickettsial France (F. Pagés)
follows:15 min at 95°C followed by Diseases, including rigorous positive DOI: http://dx.doi.org/10.3201/eid1810.120178
40 cycles of 1 s at 95°C, 40 s at 60°C, and negative controls to validate the
and 40 s at 45°C. test. Each positive qPCR result was References
Test results for all Ae. aegypti confirmed by another R. felis–specific
1. La Scola B, Meconi S, Fenollar F,
homogenates were negative for R. qPCR (orfB) not previously used in Rolain JM, Roux V, Raoult D. Emended
felis DNA. Of the 96 Ae. albopictus our laboratory (to avoid contamination description of Rickettsia felis (Bouyer
specimens, 3 (3.1%) had positive test with other amplicons). et al. 2001), a temperature-dependent
results for the R. felis species–specific Ae. albopictus mosquitos are cultured bacterium. Int J Syst Evol
Microbiol. 2002;52:2035–41. http://
real-time qPCR and the confirmatory native to Southeast Asia, colonizing dx.doi.org/10.1099/ijs.0.02070-0
qPCR, with mean cycle thresholds ± rural and peri-urban sites. In Gabon, 2. Parola P. Rickettsia felis: from a rare
SDs of 37.34 ± 1.7 (bioB gene; mean Ae. albopictus was the vector for disease in the USA to a common cause
copies/mosquito 5 × 102 [minimum outbreaks of chikungunya and of fever in sub-Saharan Africa. Clin
Microbiol Infect. 2011;17:996–1000.
1.2 × 102, maximum 1.4 × 103]) and dengue virus infections (6). Our study h t t p : / / d x . d o i . o r g / 1 0 . 1111 / j . 1 4 6 9 -
33.64 ± 1.4 (orfB gene; mean copies/ indicates that mosquitoes can carry 0691.2011.03516.x
mosquito 5 × 102 [minimum 1.5 × R. felis, and the prevalence and load 3. Richards AL, Jiang J, Omulo S, Dare
102, maximum 1 × 103). One of the 3 (1.8%–70% and 1.3 × 103–1.6 × 107, R, Abdirahman K, Ali A, et al. Human
infection with Rickettsia felis, Kenya.
samples was collected in January and respectively) detected in mosquitoes Emerg Infect Dis. 2010;16:1081–6. http://
2 in March. The samples came from 3 in this study are consistent with the dx.doi.org/10.3201/eid1607.091885
different districts of Libreville (Akebe low-end range of those detected in cat 4. Socolovschi C, Mediannikov O,
Poteau, Alibandeng, Camp des Boys) fleas, the confirmed biological vector Sokhna C, Tall A, Diatta G, Bassene
H, et al. Rickettsia felis–associated
and were tested by nested PCR and reservoir (8,9). uneruptive fever, Senegal. Emerg Infect
targeting the citrate synthase (gltA) We investigated the presence of Dis. 2010;16:1140–2. http://dx.doi.
gene (7). Rickettsia montanensis Rickettsia spp. in mosquitoes neglected org/10.3201/eid1607.100070
DNA was used as a positive control. as possible vectors of rickettsial 5. Maina AN, Knobel DL, Jiang J,
Halliday J, Feikin DR, Cleaveland S,
Sequencing was performed as diseases (2). Other Aedes spp. and et al. Rickettsia felis infection in febrile
described (7), and ChromasPro other genera of mosquitoes should patients, western Kenya, 2007–2010.
version 1.34 (Technelysium Pty Ltd., be tested. The role of mosquitoes Emerg Infect Dis. 2012;18:328–31. http://
Tewantin, Queensland, Australia) as Rickettsia spp. vectors remains dx.doi.org/10.3201/eid1802.111372
6. Mourou JR, Coffinet T, Jarjaval F,
was used to analyze sequence data. to be demonstrated in additional Cotteaux C, Pradines E, Godefroy L, et al.
Sequences of the bioB (120/120) studies that use the Mitchell criteria. Malaria transmission in Libreville: a one
and gltA (1,230/1,230) amplicons These studies should include the use year survey. Malar J. 2012;11:40. http://
at the nucleotide level were 100% of cell culture to isolate or detect R. dx.doi.org/10.1186/1475-2875-11-40
7. Mediannikov OY, Sidelnikov Y, Ivanov
homologous to sequences for felis in salivary glands of specimens L, Mokretsova E, Fournier PE, Tarasevich
R. felis URRWXCal2 (GenBank from wild-caught mosquitoes, I, et al. Acute tick-borne rickettsiosis
accession no. CP000053). The gltA PCR, immunofluorescence, and the caused by Rickettsia heilongjiangensis
fragment sequence was deposited in fluorescence in situ hybridization in Russian Far East. Emerg Infect Dis.
2004;10:810–7. http://dx.doi.org/10.3201/
GenBank (accession no. JQ674484). technique; demonstration of infection eid1005.030437

1688 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

8. Reif KE, Macaluso KR. Ecology which is common in Europe, may nlm.nih.gov/Blast.cgi) comparison to
of Rickettsia felis: a review. J Med also transmit zoonotic Bartonella spp. published sequences.
Entomol. 2009;46:723–36. http://dx.doi.
org/10.1603/033.046.0402 Evidence of possible tick transmission On the basis of the amplicon-
9. Reif KE, Stout RW, Henry GC, Foil LD, of bartonellae to humans under natural specific melting temperature and DNA
Macaluso KR. Prevalence and infection conditions was provided by Eskow et bands representing the specific size
load dynamics of Rickettsia felis in al. (3) and Angelakis et al. (4), who of 249-bp after gel electrophoresis,
actively feeding cat fleas. PLoS ONE.
2008;3:e2805. http://dx.doi.org/10.1371/ identified Bartonella spp. in tissue results of qPCR showed 100 (4.76%)
journal.pone.0002805 samples of patients who were recently infected I. ricinus ticks (Table).
10. Mitchell CJ. The role of Aedes albopictus bitten by ticks. We determined the Positive results did not vary by
as an arbovirus vector. Parassitologia. prevalence of Bartonella spp. in developmental tick stages; 4.84%
1995;37:109–13.
questing I. ricinus ticks in the city (18/372) adult ticks (5.08% [9/177]
Address for correspondence: Didier Raoult, of Hanover, Germany, which is female and 4.62% [9/195] male),
URMITE, UMR CNRS 7278, IRD 198, nicknamed The Green Metropolis and 4.71% (80/1,698) nymphs, and 6.67%
INSERM 1095, Faculté de Médecine, 27 Bd was selected the German Capital of (2/30) larvae were infected (Table).
Jean Moulin, 13385 Marseille Cedex 5, France; Biodiversity in 2011. Because Bartonella spp. do not seem
email: didier.raoult@gmail.com During April–October 2010, to be transmitted transovarially (6),
we collected 2,100 questing ticks, it is likely that larvae had interrupted
consisting of 372 adults (177 female blood meals and thus did not take
and 195 male), 1,698 nymphs, and up enough blood to develop into the
30 larvae, from 10 recreation areas nymphal stage.
in Hanover. Tick DNA was extracted Seasonal changes in Bartonella
by using the NucleoSpin 8 Blood kit spp. infection rates resulted in a higher
(Macherey-Nagel, Düren, Germany). peak in May (38/300 [12.67%]) than in
Plasmid DNA constructed from B. the other months (Table). For sampling
Bartonella spp. henselae reference strain ATCC49793 locations, infection rates for grassy
Infection Rate and containing the 249-bp target sequence sampling location 6 (4/210 [1.90%]
B. grahamii in Ticks of the gltA gene was used as positive infected ticks) differed significantly
control. Bartonella spp. in ticks was (Bonferroni-Holm adjusted p<0.001;
To the Editor: Bacteria of the detected by quantitative PCR (qPCR) *<0.0011) from that of densely
genus Bartonella are transmitted by by using the Mx3005 Multiplex wooded sampling location 9 (22/210
arthropods and are often implicated Quantitative PCR System (Stratagene, [10.48%] infected ticks).
in human disease. Even though ticks Heidelberg, Germany) according to Sequencing of the gltA fragment
are known to transmit a variety of the protocol described by Mietze et al. resulted in Bartonella species
pathogens, vector competences for (5), with minor modifications. Samples identification for 56/100 positive
transmission of Bartonella spp. positive by qPCR were verified by gel samples; 52 of these samples (from
by ticks were speculative (1) until electrophoresis. Bartonella species 38 nymphs, 13 adults, and 1 larva)
recently, when in vivo transmission were differentiated by sequencing were identified as infected with B.
of B. birtlesii by Ixodes ricinus ticks (Eurofins MWG Operon, Ebersberg, henselae. In 51 samples (92.86%),
was demonstrated in mice (2). This Germany), and obtained sequences maximum identity with the BLAST
finding suggests that this tick species, underwent BLAST (http://blast.ncbi. top hit sequence was 99% because of

Table. Seasonal distribution of Ixodes ricinus ticks infected with Bartonella spp., Hanover, Germany, 2010*
Ticks April May June July August September October Total
No. infected ticks/no. tested (%) 5/300 38/300 7/300 10/300 5/300 17/300 18/300 100/2,100
(1.67) (12.67) (2.33) (3.33) (1.67) (5.67) (6.00) (4.76)
No. (%) adults positive 1/88 8/48 0/39 0/41 2/56 3/53 4/47 18/372
(1.14) (16.67) (3.57) (5.66) (8.51) (4.84)
No. (%) females 1/32 3/19 0/20 0/17 1/32 2/28 2/29 9/177
(3.13) (15.79) (3.13) (7.14) (6.90) (5.08)
No. (%) males 0/56 5/29 0/19 0/24 1/24 1/25 2/18 9/195
(17.24) (4.17) (4.00) (11.11) (4.62)
No. (%) nymphs 3/203 30/248 7/261 10/259 3/244 14/240 13/243 80/1,698
(1.48) (12.10) (2.68) (3.86) (1.23) (5.83) (5.35) (4.71)
No. (%) larvae 1/9 0/4 ND ND ND 0/7 1/10 2/30
(11.11) (10.00) (6.67)
*ND, testing not done.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1689
LETTERS

nucleotide substitutions in position 198 Elisabeth Janecek, 9. Serratrice J, Rolain JM, Granel B, Ene N,
(T→C) and in position 136 (A→G) of Conrath J, Avierinos JF, et al. Bilateral
Andreas Mietze, Ralph Goethe,
retinal artery branch occlusions revealing
the 249-bp fragment. The remaining Thomas Schnieder,1 Bartonella grahamii infection. Rev Med
sample showed 100% identity with and Christina Strube Interne. 2003;24:629–30. http://dx.doi.
B. henselae strains Brazil-1 and 45- Author affiliation: University of Veterinary org/10.1016/S0248-8663(03)00224-8
00249 (GenBank accession nos. 10. Kerkhoff FT, Bergmans AM, van Der
Medicine Hannover, Hanover, Germany
Zee A, Rothova A. Demonstration of
HQ012580 and GQ225709). DOI: http://dx.doi.org/10.3201/eid1810.120390 Bartonella grahamii DNA in ocular fluids
Four of the 56 successfully of a patient with neuroretinitis. J Clin
sequenced samples (7.14%; all References
Microbiol. 1999;37:4034–8.
samples from nymphs) showed the
sequence pattern of B. grahamii. One 1. Billeter SA, Levy MG, Chomel BB, Address for correspondence: Christina Strube,
sample revealed 100% identity with Breitschwerdt EB. Vector transmission Institute for Parasitology, University of
of Bartonella species with emphasis on Veterinary Medicine Hannover, Buenteweg
B. grahamii (GenBank accession no. the potential for tick transmission. Med
EU014266); the remaining 3 samples 17, 30559 Hanover, Germany; email: christina.
Vet Entomol. 2008;22:1–15. http://dx.doi.
showed an identity of 98% with org/10.1111/j.1365-2915.2008.00713.x strube@tiho-hannover.de
the B. grahamii strain Hokkaido-1 2. Reis C, Cote M, Le Rhun D, Lecuelle
B, Levin ML, Vayssier-Taussat M,
(GenBank accession no. AB426652) et al. Vector competence of the tick
and 99% (T→C in position 93) Ixodes ricinus for transmission of
with a sequence described as B. Bartonella birtlesii. PLoS Negl Trop Dis.
grahamii–like (GenBank accession 2011;5:e1186. http://dx.doi.org/10.1371/
journal.pntd.0001186
no. AY435122). Sequences obtained 3. Eskow E, Rao RV, Mordechai E.
in this study (deposited in GenBank Concurrent infection of the central
under accession nos. JQ770304
and JK758018) support the genetic
nervous system by Borrelia burgdorferi
and Bartonella henselae: evidence for a
Human
variability of Bartonella spp., as novel tick-borne disease complex. Arch Parvovirus 4
Viremia in Young
Neurol. 2001;58:1357–63. http://dx.doi.
demonstrated by others (5,7,8). org/10.1001/archneur.58.9.1357
It remains unclear whether
ticks are involved in transmission
4. Angelakis E, Billeter SA, Breitschwerdt
EB, Chomel BB, Raoult D. Potential for
Children, Ghana
tick-borne bartonelloses. Emerg Infect
of pathogenic Bartonella spp. to Dis. 2010;16:385–91. http://dx.doi.org/ To the Editor: Establishment
humans under natural conditions. 10.3201/eid1603.091685 of viremia is a characteristic feature
However, the total prevalence rate of 5. Mietze A, Morick D, Kohler H, Harrus S, of infection with human parvovirus
4.76% (100/2,100) questing I. ricinus Dehio C, Nolte I, et al. Combined MLST
4 (PARV4). In northern Europe,
and AFLP typing of Bartonella henselae
ticks infected with B. henselae and isolated from cats reveals new sequence PARV4 (human partetravirus) is
B. grahamii highlights the need for types and suggests clonal evolution. primarily transmitted by blood-borne
public awareness and draws attention Vet Microbiol. 2011;148:238–45. http:// routes (1,2). In other areas (southern
to the possibility of an infection dx.doi.org/10.1016/j.vetmic.2010.08.012
Europe, western Africa, South Africa,
6. Cotté V, Bonnet S, Le Rhun D, Le
with zoonotic Bartonella spp. after Naour E, Chauvin A, Boulouis HJ, et Asia) infection seems to be more
a tick bite (3,4). B. henselae, the al. Transmission of Bartonella henselae widespread, suggesting alternative
predominantly identified species, by Ixodes ricinus. Emerg Infect Dis. modes of virus acquisition (3–6).
has been associated with cat scratch 2008;14:1074–80. http://dx.doi.org/10.
We reported PARV4 genotype
3201/eid1407.071110
disease, peliosis hepatis, and bacillary 7. Arvand M, Schubert H, Viezens 3 viremia in young children with
angiomatosis in humans. Eskow et al. J. Emergence of distinct genetic unknown parenteral blood exposure
(3) also connected chronic symptoms variants in the population of primary from the rural Ashanti region of
of Lyme disease to co-infections with Bartonella henselae isolates. Microbes
Ghana (7). In that study, 2 (2.1%) of
Infect. 2006;8:1315–20. http://dx.doi.
Borrelia burgdorferi and B. henselae. org/10.1016/j.micinf.2005.12.015 94 children (median age 14.9 months)
B. grahamii has been associated 8. Ehrenborg C, Handley S, Ellis B, Mills and 22 (11.9%) of 185 children
with neuroretinitis and ocular artery J, Holmberg M. Bartonella grahamii (median age 24.0 months) were virus
thrombosis in humans (9,10). The infecting rodents display high genetic
positive (ages of the 2 virus-positive
diversity over short geographic distances.
potential risk for zoonotic Bartonella Ann N Y Acad Sci. 2003;990:233–5. http:// children from the younger cohort
spp. infection in urban recreation areas dx.doi.org/10.1111/j.1749-6632.2003. 14.9 and 15.6 months). Because the
should not be underestimated. tb07369.x number of infants was small in that
study, we extended our investigations
1
Deceased. on PARV4 viremia to an additional

1690 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

cohort of 15-month-old children from only short genomic regions (780 nt hemophilia screened over 5 years for
the same study group. for 1 child, 599 nt for a second child, PARV4 viremia and seroconversion
Plasma samples from 361 and 95 nt for a third child) could be (IgG and IgM) (9). Nine patients who
randomly selected children (191 amplified and sequenced because of seroconverted were identified, and
girls) were tested. Specimens were low viral loads. Four children had 1 had PARV4 viremia (genotype 1)
collected during January–December positive results in the first sample, and over an 8-month period. Viral loads
2004 during a trial of intermittent 3 had positive results in the second for this patient were low (<103 copies/
preventive malaria treatment in the sample. mL), a finding similar to ours for the
rural Afigya Sekyere District, Ashanti Because comparison of large 3 children. However, negative IgM
Region, Ghana (7). Plasma samples and contiguous parts of the viral results in the person with hemophilia
were analyzed because of limited genomes within each sample pair suggest that the sampling window
availability of whole blood samples. was not possible and serologic data might have missed the acute infection.
Median age of children was 14.9 were lacking, PARV4 positivity over Comparison of results of our
months (range 13.8–17.5 months, a 9-month period can be interpreted study with those of our previous
interquartile range 14.5–15.2 months). by 3 hypotheses. First, detection study (7) showed 2 differences. First,
Nucleic acid was prepared from of PARV4 DNA over time might frequency of viremia in children tested
200-μL plasma samples by using represent long-term viremia after previously at 15 months of age was
the NucliSENS EasyMAG system infection, similar to observations in lower (2.1%, 2/94) than in the children
(bioMérieux, Nürtingen, Germany). human parvovirus B19 infection. in this study (8.9%). Second, median
All samples were analyzed by using 2 Second, demonstration of PARV4 viral loads differed by nearly 1 log10,
real-time PCRs and primers described during widely spaced intervals might with the higher concentrations in the
elsewhere (7,8). The limit of detection indicate endogenous reactivation of previous study analyzing EDTA whole
was ≈200 plasmid copies/mL. Strict viremia. Third, exogenous reinfection blood. Whether these differences
precautions were applied during might have occurred. were caused by the relatively small
plasma handling and amplification to PARV4 viremia was detected in a number of children included by or
avoid false-positive results. study in the United Kingdom among by the fact that whole blood samples
PARV4 genotype 3 DNA was 110 PARV4-negative persons with were compared with plasma samples
detected in plasma of 32 (8.9%) of 361
children. Viral load ranged from ≈200
copies/mL to 3.0 × 104 copies/mL
(Figure). Median viral load was 453
copies/mL. Nucleotide sequencing
of screening PCR amplicons and
additional genomic regions amplified
from 6 plasma samples identified
the viruses as PARV4 genotype 3
(GenBank accession nos. JN183933–
JN183938). There was no association
between history of fever, anemia, or
erythema in children with or without
PARV4 viremia (p>0.05, by χ2 test).
PARV4 viremia status was
already known for 78 children 24
months of age (7). These data enabled
comparison of viremia at 2 time points
(24 months and 15 months of age). Of
these 78 children, 10 showed viremia Figure. Parvovirus 4 DNA loads in virus-positive plasma specimens from children compared
(viral load range 4.0 × 102–1.4 × 104 with those in whole blood samples previously tested (7), Ghana. Virus concentrations
copies/mL) and 3 (3.8%) had viremia are given on a log scale on the y-axis. Each dot represents 1 specimen. Horizontal lines
represent median values for each sample group. Children whose plasma was tested had
at both time points and identical viral a median age of 15 months, and children whose whole blood was tested had a median
nucleotide sequences (time between age of either 15 or 24 months. Viral load data (i. e., median viral load and range) for the 2
bleedings 8.7 months for 2 children groups of whole blood samples have been reported (7) and were included for comparison
and 9.0 months for 1 child). However, with plasma data from this study.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1691
LETTERS

remains to be clarified. However, our 2. Lahtinen A, Kivela P, Hedman L, Kumar

A, Kantele A, Lappalainen M, et al.
Multidrug-Resistant
previous hypothesis that prenatal or
perinatal transient infection was an
Serodiagnosis of primary infections with Salmonella enterica,
Democratic
human parvovirus 4, Finland. Emerg
unlikely mode of virus acquisition Infect Dis. 2011;17:79–82. http://dx.doi.
needs to be modified because PARV4
3.
org/10.3201/eid1701.100750
Touinssi M, Brisbarre N, Picard C, Frassati
Republic of the
infection in newborns has recently
been demonstrated (10).
C, Dussol B, Uch R, et al. Parvovirus 4 in Congo
blood donors, France. Emerg Infect Dis.
Although we lacked IgM and IgG 2010;16:165–6. http://dx.doi.org/10.3201/ To the Editor: Salmonella enteri-
serologic data to interpret our findings, eid1601.090517
ca serotype Typhi and the nontyphoid
our study suggests that PARV4 genotype 4. Sharp CP, Vermeulen M, Nébié Y, Djoko
CF, LeBreton M, Tamoufe U, et al. S. enterica (NTS) are leading causes
3 infection might be characterized Epidemiology of human parvovirus 4 of bacteremia in sub-Saharan Africa,
by viral persistence, reactivation, or infection in sub-Saharan Africa. Emerg but little information is available from
reinfection. Additional longitudinal Infect Dis. 2010;16:1605–7. http://dx.doi.
central Africa (1,2). We describe an
studies, including serologic testing for org/10.3201/eid1610.101001
5. Benjamin LA, Lewthwaite P, epidemic increase of S. enterica bacte-
short intervals, are needed to determine Vasanthapuram R, Zhao G, Sharp C, remia in Kisantu in southwestern Dem-
the pathogenesis and potential public Simmonds P, et al. Human parvovirus 4 as ocratic Republic of the Congo (DRC).
health role of PARV4 infection. potential cause of encephalitis in children,
The Hospital of Saint Luc in
India. Emerg Infect Dis. 2011;17:1484–7.
http://dx.doi.org/10.3201/eid1708.110165 Kisantu is a 274-bed referral hospi-
This study was supported by grants 6. Yu X, Zhang J, Hong L, Wang J, Yuan tal serving a community of 150,000
from the Union Bank of Switzerland Z, Zhang X, et al. High prevalence of inhabitants. As part of an ongoing
Optimus Foundation to J.M. and C.D., the human parvovirus 4 infection in HBV and
HCV infected individuals in Shanghai.
microbiological surveillance study
European Union FP7 project European in DRC (3), we identified pathogens
PLoS ONE. 2012;7:e29474. http://dx.doi.
Management Platform for Emerging org/10.1371/journal.pone.0029474 grown from blood cultures (BacT/
and Re-emerging Infectious Disease 7. Panning M, Kobbe R, Vollbach S, Drexler ALERT; bioMérieux, Marcy L’Etoile,
Entities (grant no. 223498), the German JF, Adjei S, Adjei O, et al. Novel human
parvovirus 4 genotype 3 in infants, Ghana.
France) and assessed them for antimi-
Research Foundation (grant DR 772/3- crobial drug susceptibility (Vitek II
Emerg Infect Dis. 2010;16:1143–6. http://
1), and BONFOR to A.M.E.-H. (grant dx.doi.org/10.3201/eid1607.100025 system; bioMérieux) (4) and serotype
O-151.0021). 8. Fryer JF, Delwart E, Hecht FM, Bernardin (Sifin, Berlin, Germany). We deter-
F, Jones MS, Shah N, et al. Frequent
detection of the parvoviruses, PARV4 and
mined MICs for nalidixic acid, cipro-
Jürgen May, Jan Felix Drexler, PARV5, in plasma from blood donors and floxacin, and chloramphenicol using
Ulrike Reber, Nimarko Sarpong, symptomatic individuals. Transfusion. the Etest macromethod (bioMérieux).
Ohene Adjei, Marcus Panning, 2007;47:1054–61. http://dx.doi. For salmonella isolates, we defined
org/10.1111/j.1537-2995.2007.01235.x
Christian Drosten, 9. Sharp CP, Lail A, Donfield S, Gomperts
decreased ciprofloxacin susceptibil-
and Anna Maria Eis-Hübinger ED, Simmonds P. Virologic and clinical ity as an isolate with an MIC >0.064
Author affiliations: Bernhard Nocht Institute features of primary infection with human mg/L (5) and multidrug resistance
for Tropical Medicine, Hamburg, Germany parvovirus 4 in subjects with hemophilia: (MDR) as co-resistance of the isolate
frequent transmission by virally
(J. May); University of Bonn Medical
inactivated clotting factor concentrates.
to ampicillin, chloramphenicol, and
Centre, Bonn, Germany (J.F. Drexler, U. Transfusion. 2012;52:1482–9. http:// trimethoprim/sulfamethoxazole (6).
Reber, C. Drosten, A.M. Eis-Hübinger); dx.doi.org/10.1111/j.1537-2995.2011. Screening for mutations causing de-
Kwame Nkrumah University of Science and 03420.x creased ciprofloxacin susceptibility
10. Chen MY, Yang SJ, Hung CC. Placental
Technology, Kumasi, Ghana (N. Sarpong);
transmission of human parvovirus 4 in
included assessment of the quinolone
Kumasi Centre for Collaborative Research newborns with hydrops, Taiwan. Emerg resistance–determining regions of the
in Tropical Medicine Kumasi (O. Adjei); Infect Dis. 2011;17:1954–6. http://dx.doi. gyrA, gyrB, and parC genes and the
and Freiburg University Medical Center, org/10.3201/eid1710.101841 plasmid-mediated qnrA, qnrB, and
Freiburg, Germany (M. Panning) qnrS genes (7). Multilocus variable-
Address for correspondence: Anna Maria Eis-
DOI: http://dx.doi.org/10.3201/eid1810.111836
number tandem-repeat analysis was
Hübinger, Institute of Virology, University of
performed on a subset of 37 S. enteri-
Bonn Medical Centre, Sigmund-Freud-Strasse
References ca ser. Enteritidis isolates (8).
25, D-53127 Bonn, Germany; email: anna-
The pathogens isolated were S.
1. Simmonds P, Manning A, Kenneil R, maria.eis-huebinger@ukb.uni-bonn.de
enterica ser. Typhi (n = 17, 14.4%),
Carnie FW, Bell JE. Parenteral transmission
of the novel human parvovirus PARV4. Enteritidis (n = 79, 67.0%), and Ty-
Emerg Infect Dis. 2007;13:1386–8. http:// phimurium (n = 22, 18.6%). The in-
dx.doi.org/10.3201/eid1309.070428 creased incidence of S. enterica bac-

1692 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

teremia was caused by an increased um isolate had additional decreased incidences of malaria and malnutrition
incidence of S. enterica ser. Enteritidis ciprofloxacin susceptibility. Most (9) and to contamination of the surface
infection from 1 and 2 isolates report- (16/17, 94.1%) S. enterica ser. Typhi waters caused by floods after heavy
ed in 2008 and 2009, respectively. The isolates were resistant to amoxicil- rainfalls (2).
rate of infection by serotypes Typhi lin and trimethoprim/sulfamethoxa- Case-fatality rates in this study
and Typhimurium had remained con- zole, 4 (23.5%) and 7 (41.2%) were were similar to those reported previ-
stant during this period. MDR and had decreased ciprofloxa- ously from sub-Saharan Africa, de-
During September 2010–May cin susceptibility, respectively. Three scribing mortality rates up to 27% (2),
2011, the proportion of pathogens combinations of resistance genes irrespective of the serotype involved.
isolated from blood cultures in- encoding decreased ciprofloxacin The MDR rates among NTS and
creased to 53.2% (197/370), com- susceptibility were found (Table). among S. enterica ser. Typhi in this
pared with 19.7% (63/319) and No resistance to cefotaxime was ob- study are among the highest reported
25.2% (85/328) for 2008 and 2009, served. Multilocus variable-number in sub-Saharan Africa (10). The find-
respectively (p<0.001). S. enterica tandem-repeat analysis typing of the ing of decreased ciprofloxacin suscep-
isolates represented 59.9% (118/197) S. enterica ser. Enteritidis isolates tibility among isolates in a rural area
of pathogens, compared with 53.3% revealed 2 major profiles (differing in DRC highlights the need for sur-
(70/131) and 30.9% (84/272) for the in 3 tandem repeats in 1 locus) and veillance of antimicrobial drug resis-
same months during 2008–2009 and 3 minor profiles (differing from the tance of S. enterica isolates.
2009–2010, respectively (p<0.001). major profiles by 1 tandem repeat at 1
Of 118 S. enterica samples isolated, and 2 loci, respectively). Considering
The antibiotic resistance surveillance
89 (75.4%) were isolated from speci- the long sample period, we concluded
project in RD Congo is funded by Project
mens from children <5 years old and that 1 clonal type had caused the in-
2.01 of the Third Framework Agreement
17 (15.3%) from children 5–10 years fections in which S. enterica ser. En-
between, the Belgian Directorate General
old. Clinical signs and symptoms teritidis was isolated.
of Development Cooperation and the Insti-
were nonspecific; malaria and gas- A recent literature review of bac-
tute of Tropical Medicine, Antwerp, Bel-
trointestinal infection were the lead- teremia reported aggregated data from
gium. M.-F.P. has a scholarship from the
ing diagnoses on admission. Data 16 studies from eastern (Kenya, Tanza-
Fondation Delacroix, Tienen, Belgium.
for in-hospital deaths (retrieved for nia), western (the Gambia), and south-
87 patients) revealed case-fatality ern Africa (Malawi, Mozambique)
rates of 23.0% (17/74) for children (1). S. enterica ser. Typhi and NTS Marie-France Phoba,1
<5 years old, compared with 1 in 10 represented 0–42% and 9%–84%, re- Octavie Lunguya,1
patients 5–10 years old. Because of spectively, of associated pathogens. Daniel Vita Mayimon,
the retrospective nature of the study, The study reported that most NTS Pantaléon Lewo di Mputu,
it was not possible to assess popula- isolated were of the serotypes Enter- Sophie Bertrand,
tion incidence rates, and we had no itidis and Typhimurium. A sequential Raymond Vanhoof,
estimates of the number of children occurrence of disease caused by NTS Jan Verhaegen, Chris Van Geet,
who were referred to the hospital but serotypes similar to that in this study Jean-Jacques Muyembe,
died before reaching the emergency was recorded in Malawi (6). and Jan Jacobs
department. There was no apparent The reservoir of NTS in sub-Sa- Author affiliations: National Institute for Bio-
geographic clustering, but the epi- haran Africa remains unclear; person- medical Research, Kinshasa, Democratic
demic coincided with the onset of the to-person and zoonotic transmissions Republic of the Congo (M.-F. Phoba, O. Lun-
rainy season, which had started late have been postulated (1,2). The coin- guya, J.-J. Muyembe); University Hospital of
and had unusually heavy rainfall. cidence of the onset of the epidemic Kinshasa, Kinshasa (M.-F. Phoba, O. Lun-
All NTS isolates were MDR; 1 described in this study with the start of guya, J.-J. Muyembe); Saint-Luc Hospital,
(1.3%) S. enterica ser. Enteritidis and the rainy season is a known phenom- Kisantu, Democratic Republic of the Congo
1 (4.5%) S. enterica ser. Typhimuri- enon and might be related to increased (D.V. Mayimona, P. Lewo di Mputu); Institute
of Public Health, Brussels, Belgium (S. Ber-
trand, R. Vanhoof); University Hospital Leu-
Table. MICs and genetic analysis of Salmonella enterica isolates with decreased
ciprofloxacin susceptibility, Democratic Rebublic of the Congo ven, Leuven, Belgium (J. Verhaegen, C. Van
MIC, mg/L No. Geet); and Institute of Tropical Medicine, An-
Serotype Nalidixic acid Ciprofloxacin isolates Mutations twerp, Belgium (J. Jacobs)
Enteritidis 128 0.094 1 Asp82-Asn in gyrA + qnrB
1
Typhimurium >256 0.125 1 Asp87-Tyr in gyrA + qnrB These authors contributed equally to this
Typhi >256 0.19–0.25 7 Ser83-Phe in gyrA article.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1693
LETTERS

DOI: http://dx.doi.org/10.3201/eid1810.120525 Address for correspondence: Jan Jacobs, collected when the children had
Institute of Tropical Medicine, Nationalestraat gastroenteritis. Most of the SAFV-
References 155, 2000 Antwerp, Belgium; email: jjacobs@ positive samples reported in this study
1. Mtove G, Amos B, van Seidlein L, Hen- itg.be were obtained from a randomized
driksen I, Mwambuli A, Kimera J, et al.
Invasive salmonellosis among children
trial in the pediatric department
admitted to a rural Tanzanian hospital of University Hospital of Holbaek
and a comparison with previous studies. (Holbaek) on the effect of probiotic
PLoS ONE. 2010;5:e9244. http://dx.doi. therapy on the incidence of infection
org/10.1371/journal.pone.0009244
2. Morpeth SC, Ramadhani HO, Crump JA.
during early childhood (M. Gyhrs,
Invasive non-typhi Salmonella disease in unpub. data). The study was approved
Africa. Clin Infect Dis. 2009;49:606–11.
http://dx.doi.org/10.1086/603553 Co-Circulation and by the local ethics committee; Den
Regionale Videnskabsetiske Komité
3. Lunguya O, Phoba M-F, Mundeke SA,
Bonebe E, Mukadi P, Muyembe JJ, et
Persistence of for Region Sjaelland, Denmark.
al. The diagnosis of typhoid fever in the Genetically Distinct Nucleic acids were extracted
Democratic Republic of the Congo. Trans
R Soc Trop Med Hyg. 2012;106:348–55. Saffold Viruses, from 200-μL fecal suspension (10%
in phosphate-buffered saline) by
http://dx.doi.org/10.1016/j.trstmh.2012.
03.006
Denmark using the Cobas AmpliPrep Total
4. Clinical and Laboratory Standards Insti- Nucleic Acid Isolation Kit (Roche
To the Editor: Cardioviruses are
tute. Standard institute performance stan- Diagnostics, Ltd., Mannheim,
dards for antimicrobial susceptibility test- positive-sense, single-stranded RNA
Germany) on the MagnaPure LC
ing: twenty-first informational supplement viruses of the family Picornaviridae,
(M100–S21). Wayne (PA): The Institute;
instrument (Roche Diagnostics). We
genus Cardiovirus. Until recently,
2011. used 5 μL of extracted nucleic acid
cardioviruses were primarily known
5. European Committee on Antimicrobial for reverse transcription PCR (RT-
Susceptibility Testing. Breakpoint tables for their ability to infect rodents. In
PCR) (total volume 25 μL) using
for interpretation of MICs and zone di- 2007, findings of a retrospective study
ameters [cited 2012 Feb 1] http://www.
the OneStep RT-PCR Kit (QIAGEN,
of undiagnosed enteric illnesses in
eucast.org/fileadmin/src/media/PDFs/ Hilden, Germany). The samples
the United States were published,
EUCAST_files/Breakpoint_tables/Break- were tested for SAFV by using real-
point_table_v_2.0_120221.pdf including results from analysis of a
time RT-PCR primer/probe, and all
6. Gordon MA, Graham SM, Walsh AL, fecal sample from an infant girl whose
Wilson L, Phiri A, Molyneux E, et al.
positive samples were genotyped
symptoms were diagnosed as fever of
Epidemics of invasive Salmonella en- by partial sequencing of the viral
unknown origin in 1981. The novel
terica serotype Enteritidis and S. enterica protein (VP) 1 gene (6). Overall, 38
serotype Typhimurium infection associ- human cardiovirus that was isolated
(2.8%) of the clinical samples were
ated with multidrug resistance among was designated Saffold virus (SAFV)
adults and children in Malawi. Clin In-
positive for SAFV (online Technical
(1). Eight genotypes of SAFV have
fect Dis. 2008;46:963–9. http://dx.doi. Appendix, wwwnc.cdc.gov/EID/
been described (1–4), and a ninth
org/10.1086/529146 pdfs/12-0793-Techapp.pdf), all of
7. Cavaco LM, Hasman H, Xia S, Aarestrup was recently isolated in Nigeria (O.
which fell into genotype 2 (SAFV-2),
FM. qnrD, a novel gene conferring trans- Blinkova, unpub. data). Serologic
ferable quinolone resistance in Salmonella
which is most prevalent in Western
studies indicate that infection with
enterica serotypes Kentucky and Bovis- nations. Of these samples, 31 had
SAFV genotypes 2 and 3 generally
morbificans of human origin. Antimicrob sequence information of sufficient
Agents Chemother. 2009;53:603–8. http:// occurs in early life (5), although the
length for additional analyses. All
dx.doi.org/10.1128/AAC.00997-08 clinical significance of these findings
8. Hopkins KL, Peters TM, de Pinna E,
SAFV-2 sequences were submitted to
remains unclear.
Wain J. Standardisation of multilocus GenBank (accession nos. JX048000–
The first SAFV infection in
variable-number tandem-repeat analysis JX048030).
(MLVA) for subtyping of Salmonella en- Denmark was recorded in 2009
To determine the evolutionary
terica serotype Enteritidis. Euro Surveill. (6). To elucidate the molecular
2011;16:19942.
history of strains of SAFV identified
epidemiology of SAFV, we performed
9. Green SDR, Cheesbrough JS. Salmonella in persons in Denmark, we combined
a 3-year surveillance study of SAFV
bacteraemia among young children at a the VP1 sequences collected here
rural hospital in western Zaire. Ann Trop in Denmark. During 2009–2011,
with all others available on GenBank.
Paediatr. 1993;13:45–53. we tested 1,393 fecal samples from
10. Graham SM. Non-typhoidal salmonel-
We aligned sequences as described
454 children. Surveillance included
lae: a management challenge for chil- using MUSCLE software (7),
collection of fecal samples from
dren with community-acquired invasive then checked the alignments using
disease in tropical African countries. children at 6, 10, and 15 months of
manual calculations. We performed
Lancet. 2009;373:267–9. http://dx.doi. age; additional fecal samples were
org/10.1016/S0140-6736(09)60073-8
phylogenetic analysis using the

1694 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

maximum likelihood method as identified in samples collected during 0.062), which indicates stronger
described in PhyML 3.0 (8), on the 2009–2010, supporting probable purifying selection on the DK-B
basis of the best-fit GTR+Γ nucleotide sustained viral persistence within group. Ancestral state reconstruction,
model as determined by jModelTest Denmark. Within DK-B, strain 115883 performed by using Datamonkey
(9). Phylogenetic robustness was is phylogenetically distinct from the (10), revealed that the ancestors of
determined by using 1,000 bootstrap other DK-B viruses. DK-A and DK-B differ only at aa 135
replicates. We next measured the selection in VP1: Val in DK-A and Ala in DK-
Our phylogenetic analysis places pressures acting on these lineages B. Notably, aa 135 was positively
the strains isolated in Denmark within through the mean number of selected in DK-A (random effects
the SAFV-2 group (online Technical nonsynonymous (dN) to synonymous likelihood: dN/dS = 3.53, Bayes factor
Appendix). These SAFV-2 strains (dS) nucleotide substitutions per site = 50: fixed effects likelihood: dN/
were further subdivided into 2 strongly using the single-likelihood ancestor dS >>1; cutoff p = 0.1), with more
supported clusters: DK-A, which counting, fixed effects likelihood, tentative evidence for adaptation
comprised viruses isolated during and random effects likelihood at aa 135 in DK-B: dN/dS ratio >>1
2009–2011, demonstrating probable methods available in the Datamonkey by using fixed effects likelihood
persistence in Denmark during this HyPhy package as described (10). (p = 0.2). The functions of aa 135 in
period; and DK-B, a smaller group The DK-A and DK-B groups differed VP1, and what it means for the fitness
that included viruses from United significantly in selection pressure: of SAFV, merit further consideration.
States, Germany, and the Netherlands, DK-A, dN/dS ratio = 0.195 (95% We conclude that SAFV-2 has
indicating widespread viral gene flow CI 0.105–0.328); and DK-B, dN/ been introduced into Denmark in 3
(Figure). The DK-B strains were dS ratio = 0.033 (95% CI 0.015– groups: DK-A, viral strain 115883 and
strains of DK-B reported in Denmark;
all have recently co-circulated in
this country. We have demonstrated
the entry and persistence of 3
phylogenetically distinct lineages of
SAFV-2 in Denmark. That SAFV-2
can persist between years suggests
that it might be common, yet
underreported, in Denmark, which
provides the opportunity for spread
to additional localities. Increased
awareness of improved laboratory
protocols for SAFV detection is
needed among clinicians in Denmark
and neighboring countries.

Alex Christian Yde Nielsen,

Mette Louise Gyhrs,
Edward C. Holmes, and Jie Cui
Author affiliations: Statens Serum Institut,
Copenhagen, Denmark (A.C.Y. Nielsen);
University Hospital of Hobaek, Denmark
(M.L. Ghyrs); University Hospital of
Odense, Odense, Denmark (A.C.Y.
Nielsen); Pennsylvania State University,
University Park, Pennsylvania, USA (E.C.
Holmes, J. Cui); and National Institutes of
Figure. Phylogenetic analyses of Saffold viruses (SAFVs). Phylogenetic analysis of SAFV- Health, Bethesda, Maryland, USA (E.C.
2. Strains from Denmark are named by using the isolation numbers assigned for the study, Holmes)
then the country of origin and year of sampling in parentheses. The 2 subgroups (DK-A
and -B) are shown. The tree is rooted by using SAFV-1 outgroup sequence accession no. DOI: http://dx.doi.org/10.3201/eid1810.120793
EF165067. Bootstrap values >70% are shown. DK, Denmark; US, United States. Scale
bars represent 0.2 nucleotide substitutions per site.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1695
LETTERS

References Pathogenic bacterial infectious agents in island

wild fauna, 129 insectivorous and
1. Jones MS, Lukashov VV, Ganac RD, Leptospira spp. in frugivorous bats were tested for
Schnurr DP. Discovery of a novel human
picornavirus in a stool sample from a Bats, Madagascar Leptospira spp. The bats belonged
pediatric patient presenting with fever
of unknown origin. J Clin Microbiol.
and Union of the to 12 species: 9 from Madagascar
(Mormopterus jugularis, Otomops
2007;45:2144–50. http://dx.doi.org/10. Comoros madagascariensis, Triaenops fur-
1128/JCM.00174-07
2. Abed Y, Boivin G. New Saffold To the Editor: Leptospirosis culus, T. menamena, Miniopterus
cardioviruses in 3 children, Canada.
is a zoonosis of global distribution; gleni, Miniopterus griffithsi,
Emerg Infect Dis. 2008;14:834–6. http:// Miniopterus mahafaliensis, Myotis
dx.doi.org/10.3201/eid1405.071675 incidence rates are particularly high
3. Drexler JF, Luna LK, Stöcker A, in tropical areas (1). Leptospirosis goudoti, Hypsugo anchietae) and
Almeida PS, Ribeiro TC, Petersen N, et is a major public health problem on 3 from Union of the Comoros
al. Circulation of 3 lineages of a novel
islands in the southwestern Indian (Rousettus obliviosus, Chaerephon
Saffold cardiovirus in humans. Emerg pusillus, Miniopterus griveaudi).
Infect Dis. 2008;14:1398–405. http:// Ocean, particularly La Réunion,
dx.doi.org/10.3201/eid1409.080570 Mayotte, and the Seychelles (where Bats were captured in mist nets or
4. Blinkova O, Kapoor A, Victoria J, Jones incidence rates are among the highest harp traps at 8 sites in Madagascar
M, Wolfe N, Naeem A, et al. Cardioviruses
in the world) (1). In contrast, no and 6 sites in Union of the Comoros.
are genetically diverse and cause common Organs were collected in the field and
enteric infections in South Asian children. human case has been reported on the
J Virol. 2009;83:4631–41. http://dx.doi. nearby islands of Madagascar and immediately stored in liquid nitrogen.
org/10.1128/JVI.02085-08 Union of the Comoros. However, the Total nucleic acids were extracted
5. Zoll J, Erkens Hulshof S, Lanke K, Verduyn
recent demonstration of pathogenic from a pool of kidney, spleen, and
Lunel F, Melchers WJ, Schoondermark- lung tissue by using the Biorobot EZ1
van de Ven E, et al. Saffold virus, a human Leptospira spp. in small mammals
Theiler’s-like cardiovirus, is ubiquitous introduced to Madagascar suggests and EZI Virus Mini Kit version 2.0
and causes infection early in life. PLoS possible transmission from free-living (QIAGEN, Les Ulis, France). Reverse
Pathog. 2009;5:e1000416. http://dx.doi.
animals to humans (2). transcription was then performed
org/10.1371/journal.ppat.1000416 with GoScript reverse transcriptase
6. Nielsen ACY, Böttiger B, Banner J, In addition to the fact that
Hoffmann T, Nielsen LP. Serious invasive incidence rates vary among humans, (Promega, Charbonnières-les-Bains,
Saffold virus infections in children, 2009. clinical bacterial isolates from France) to obtain cDNA. We screened
Emerg Infect Dis. 2012;18:7–12. http://
different islands belong to different pathogenic Leptospira spp. with a
dx.doi.org/10.3201/eid1801.110725 probe-specific real-time PCR (7). The
7. Edgar RC. MUSCLE: multiple sequence serogroups and serovars and show
alignment with high accuracy and diverse molecular features (3,4). 25 positive samples were subsequently
high throughput. Nucleic Acids Res. This diversity might be correlated subjected to a PCR procedure that
2004;32:1792–7. http://dx.doi.org/10.
with that of the reservoir hosts; the amplified fragments from 682 to 1,293
1093/nar/gkh340 bp of the 16S rRNA gene (depending
8. Guindon S, Dufayard JF, Lefort V, islands in the southwestern Indian
Anisimova M, Hordijk W, Gascuel O. Ocean are a hot spot of biodiversity on the amplification success) by using
New algorithms and methods to estimate with extraordinary levels of vertebrate published primers (8–10). Resulting
maximum-likelihood phylogenies:
endemism. Most studies investigating PCR products from 7 samples were
assessing the performance of PhyML 3.0. sequenced and compared with
Syst Biol. 2010;59:307–21. http://dx.doi. wild-animal reservoirs of Leptospira
org/10.1093/sysbio/syq010 spp. on the islands in the southwestern available sequences in GenBank by
9. Posada D. jModelTest: phylogenetic model Indian Ocean have focused on small using phylogenetic construction with
averaging. Mol Biol Evol. 2008;25:1253–
mammals that had been introduced PhyML 3.0.
6. http://dx.doi.org/10.1093/molbev/msn Of the 12 bat species tested, 11
083 to the islands (2,5), although bats
10. Delport W, Poon AF, Frost SD, Kosakovsky infected with pathogenic Leptospira were positive for Leptospira spp.
Pond SL. Datamonkey 2010: a suite of spp. have been identified in other (the only H. anchietae bat tested
phylogenetic analysis tools for evolutionary
regions (6). Whether bats are a was negative). Among 52 bats
biology. Bio-informatics. 2010;26:2455–7. from Madagascar, 18 (34.6%) were
http://dx.doi.org/10.1093/bioinformatics/ reservoir of Leptospira spp. on these
btq429 islands remains unknown. Therefore, infected; detection rates were often
we looked for this bacterium in bats high, e.g., 8 (80%) of 10 T. menamena
Address for correspondence: Jie Cui, from Madagascar and Union of the bats. In contrast, among 77 bats from
Department of Biology, Millennium Science Comoros and characterized associated Union of the Comoros, only 9 (11.7%)
Complex, Pennsylvania State University, genetic diversity. were infected. Leptospira spp. seem
University Park, PA 16802, USA; email: As part of an ongoing program to be ubiquitous in the study areas;
jiecui@yahoo.com aimed at identifying viral and infected bats were found at 7 of 8

1696 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

sites in Madagascar and 3 of 6 sites which are endemic to Madagascar. attic (Pomoni, Anjouan, Union of the
in the Union of the Comoros. Of the Potentially pathogenic Leptospira Comoros), and bat scats were visible
7 Leptospira spp. sequences obtained spp. were found in bats of a wide on the floor within the classroom.
from bats in this study, 3 were closely variety of species in Madagascar and Bats from Madagascar and
related to L. borgpetersenii, 1 grouped Union of the Comoros, at most study Union of the Comoros harbor a
with L. interrogans, and 3 were sites, and at levels notably higher than notable diversity of Leptospira spp.;
not associated with any described those reported from similar studies in this finding is in accordance with
species (Figure). L. interrogans and other regions (6). Some of the bats the diversity found in a comparable
L. borgpetersenii were identified that were Leptospira spp.–positive investigation of bats in the Amazon
from R. obliviosus bats from the same by PCR, particularly the genera region (6). Although leptospirosis
cave in the Union of the Comoros, Mormopterus and Chaerephon, often in humans is suspected only on the
and the L. borgpetersenii sequence occupy synanthropic day roost sites. islands associated with this study (10),
was closely related to that identified For example, we sampled 1 positive incidence among humans in Mayotte,
from the O. madagascariensis bats, colony of C. pusillus bats in a school part of the Union of the Comoros
archipelago, has been shown to be
high and mainly associated with L.
borgpetersenii (3). The use of more
polymorphic markers combined with
the sequencing of clinical isolates
should provide better characterization
of Leptospira spp. diversity and
the potential role of bats in human
leptospirosis.

Permits for this research were

kindly issued by Direction du Système
des Aires Protégées, Direction Générale
de l’Environnement et des Forêts
(Madagascar), and the Centre National
de Documentation et de Recherche
Scientifique (Union of the Comoros). This
work was supported by Fonds Européen
de Développement Régional Programme
Opérationnel de Coopération Territoriale
Réunion, Pathogènes Associés à la Faune
Sauvage Océan Indien #31189. M.D. is
supported by the European Community
FP7 Capacity RegPot Run-Emerge
Program.

Erwan Lagadec,1
Yann Gomard,1
Vanina Guernier,
Muriel Dietrich,
Hervé Pascalis, Sarah Temmam,
Beza Ramasindrazana,
Figure. Maximum-likelihood phylogenetic tree of Leptospira spp.16s rDNA in bats from Steven M. Goodman,
Madagascar and the Union of the Comoros. The dendrogram was constructed with a Pablo Tortosa,
fragment of 641 bp, with the TIMef+I+G substitution model, and with 1,000 replications. and Koussay Dellagi
Only bootstrap supports >70% are shown (circles). The precise geographic information of
the sampled bats can be accessed through the GenBank accession numbers indicated in
parentheses at branch tips. Host bat species for the sequences generated in this study are 1
These authors contributed equally to this
shown in boldface. Scale bar indicates number of nucleotide substitutions per site. article.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1697
LETTERS

Author affiliations: Centre de Recherche 7. Smythe LD, Smith IL, Smith GA, Dohnt endemic in Africa, southern Asia,
et de Veille sur les Maladies Émergentes MF, Symonds ML, Barnett LJ, et al. and northern Australia, and only
A quantitative PCR (TaqMan) assay
dans l’Océan Indien, Ste. Clotilde, La
for pathogenic Leptospira spp. BMC sporadic cases or small epidemics are
Réunion, France (E. Lagadec, Y. Gomard, Infect Dis. 2002;2:13. http://dx.doi. seen in Europe (2). In 1999, WNV
V. Guernier, M. Dietrich, H. Pascalis, S. org/10.1186/1471-2334-2-13 emerged in North America. By 2010,
Temmam, P. Tortosa, K. Dellagi); Université 8. Fenner JS, Anjum MF, Randall LP, ≈1.8 million persons had become
Pritchard GC, Wu G, Errington J, et al.
de Lyon, Villeurbanne, France (E. Lagadec,
Analysis of 16S rDNA sequences from infected, with 12,852 reported cases
S. Temmam); Université de La Réunion, St. pathogenic Leptospira serovars and use of meningoencephalitis and 1,308
Clotilde (M. Dietrich, P. Tortosa); Université of single nucleotide polymorphisms deaths (2). In Europe, the last notable
d’Antananarivo, Antananarivo, Madagascar for rapid speciation by D-HPLC. Res outbreak of WNV infection occurred
Vet Sci. 2010;89:48–57. http://dx.doi.
(B. Ramasindrazana); Association Vahatra,
org/10.1016/j.rvsc.2009.12.014 in Greece in 2010; 197 persons were
Antananarivo (B. Ramasindrazana, S.M. 9. Lourdault K, Aviat F, Picardeau M. infected, and 33 died (3). The Czech
Goodman); Field Museum of Natural Use of quantitative real-time PCR Republic, Denmark, France, and the
History, Chicago, Illinois, USA (S.M. for studying the dissemination of Netherlands reported laboratory-
Leptospira interrogans in the guinea pig
Goodman); and Institut de Recherche
infection model of leptospirosis. J Med confirmed WNV infections in travelers
pour le Développement, Ste. Clotilde Microbiol. 2009;58:648–55. http://dx.doi. returning from North America (1).
(Y. Gomard, V. Guernier, H. Pascalis, S. org/10.1099/jmm.0.008169-0 We report a case of WNV
Temmam, K. Dellagi) 10. Silverie R, Monnier M, Lataste-Dorolle meningoencephalitis in a 28-year-old
C. Recent survey of leptospirosis on
DOI: http://dx.doi.org/10.3201/eid1810.111898 Madagascar. Contribution to the study of German woman, who sought treatment
human, bovine and porcine leptospirosis the emergency department of a hospital
References
in the southern region [in French]. Bull in Potsdam, Germany, on September 7,
Soc Pathol Exot Filiales. 1968;61:346–59. 2011. She had a 3-day history of fever
1. Pappas G, Papadimitriou P, Siozopoulou V, of up to 40°C and mental confusion.
Christou L, Akritidis N. The globalization Address for correspondence: Pablo Tortosa,
Six days before admission, she had
of leptospirosis: worldwide incidence CRVOI, Plateforme de Recherche CYROI,
trends. Int J Infect Dis. 2008;12:351–7. returned from a 2-week holiday trip
2 Rue Maxime Rivière, 97490 St. Clotilde,
http://dx.doi.org/10.1016/j.ijid.2007. to Ottawa, Ontario, Canada. She had
France; email: pablo.tortosa@univ-reunion.fr
09.011 spent most of her time in the city of
2. Rahelinirina S, Leon A, Harstskeerl RA, Ottawa.
Sertour N, Ahmed A, Raharimanana C,
et al. First isolation and direct evidence The patient’s medical history was
for the existence of large small-mammal unremarkable. She was in a reduced
reservoirs of Leptospira sp. in Madagascar. general condition because of a severe
PLoS One. 2010;5:e14111. http://dx.doi. encephalitic syndrome characterized
org/10.1371/journal.pone.0014111
3. Bourhy P, Collet L, Clément S, Huerre by somnolence, meningism, fever,
M, Ave P, Giry C, et al. Isolation and and mental confusion. Laboratory
characterization of new Leptospira investigations revealed leukocytosis
genotypes from patients in Mayotte West Nile Virus with 15.000 leukocytes/μL (reference
(Indian Ocean). PLoS Negl Trop Dis.
2010;4:e724. http://dx.doi.org/10.1371/
Meningoencephalitis range 4,400–11,300 leukocytes/
journal.pntd.0000724 Imported into μL) and elevated C-reactive protein
4. Yersin C, Bovet P, Mérien F, Wong of 14.8 mg/L (reference <3 mg/L).
T, Panowsky J, Perolat P. Human Germany Cerebrospinal fluid (CSF) analysis
leptospirosis in the Seychelles (Indian
Ocean): a population-based study. Am J To the Editor: West Nile virus on the day of admission showed
Trop Med Hyg. 1998;59:933–40. (WNV) is a single-stranded RNA pleocytosis, 430 cells/μL (72%
5. Desvars A, Cardinale E, Michault A.
virus in the family Flaviviridae that is granulocytes, 27% lymphocytes. and
Animal leptospirosis in small tropical 1% monocytes); elevated levels for
areas. Epidemiol Infect. 2011;139:167– transmitted to humans by mosquitoes.
88. http://dx.doi.org/10.1017/S09502688 Approximately 80% of WNV total protein, 1,023 mg/L (reference
10002074 infections in humans are asymptomatic, range 150–450 mg/L); an albumin
6. Matthias MA, Díaz MM, Campos KJ,
whereas ≈20% of infected persons level of 637 mg/L (normal range 0–350
Calderon M, Willig MR, Pacheco V, et mg/L); and a moderately elevated level
al. Diversity of bat-associated Leptospira experience fever, often accompanied
in the Peruvian Amazon inferred by by a rash. Less than 1% of infections of the albumin quotient of 20 (reference
Bayesian phylogenetic analysis of 16S are manifested as neuroinvasive range <6.5). The CSF/serum diagrams
ribosomal DNA sequences. Am J Trop
disease, such as meningoencephalitis, demonstrated a moderate disturbance of
Med Hyg. 2005;73:964–74. the blood–CSF barrier and a substantial
polyradiculoneuritis, and polio-
like flaccid paralysis (1). WNV is intrathecal IgM synthesis of 27.6%

1698 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

(6.15 mg/L) but no intrathecal IgA or than the titers of antibodies against imported case adds to these cases and
IgG synthesis. Results of magnetic the other flaviviruses tested (Table), suggests that travelers are at risk, even
resonance imaging of the brain were indicating that the antibodies resulted if they visit only Ottawa. Physicians in
unremarkable. No parenchymal lesions from a WNV infection. The serologic Germany should be aware of the risk
were found. diagnosis was further substantiated for WNV infection among travelers
Antimicrobial drug therapy by detection of WNV neutralizing returning from Canada, especially
was initiated with ceftriaxone and antibodies at day 11(VNT titer 640). during late summer.
ampicillin. Acyclovir was administered The VNT titer further increased to
empirically for herpes simplex 2,560 on day 26 after onset of disease. Jörg Schultze-Amberger,1
encephalitis until this diagnosis Results of reverse transcription PCR Petra Emmerich,1
was excluded. Molecular and were negative for WNV and members Stephan Günther,
serologic testing of serum and CSF of genus Flavivirus in serum and CSF and Jonas Schmidt-Chanasit
samples revealed no acute infection samples taken 4 days after disease onset. Author affiliations: Klinikum Ernst von
with herpesviruses, enteroviruses, Attempts to isolate WNV from serum Bergmann, Potsdam, Germany (J. Schultze-
alphaviruses, orthobunyaviruses, and and CSF samples in cell culture failed Amberger); and Bernhard Nocht Institute for
arenaviruses or with mycobacteria, as well. The patient recovered slowly Tropical Medicine, Hamburg, Germany (P.
Borrelia spp., Toxoplasma gondii, and was discharged from the hospital Emmerich, S. Günther, J. Schmidt-Chanasit)
Chlamydia spp., Leptospira spp., in Potsdam on September 15, 2011. DOI: http://dx.org/10.3201/eid1810.120204
and Mycoplasma pneumoniae. CSF She was then referred to a neurologic
and blood cultures were negative rehabilitation center in Berlin and was References
for fungi and bacteria, including discharged from there after 2 months
mycobacteria. An encephalitic with a characterization of restitutio ad 1. Artsob H, Gubler DJ, Enria DA, Morales
MA, Pupo M, Bunning ML, et al. West
syndrome caused by N-methyl-D- integrum (i.e., full recovery, restoration Nile Virus in the New World: trends
aspartate antibodies was also excluded. to original condition). in the spread and proliferation of West
On the basis of the patient’s travel We report a case of WNV Nile virus in the Western Hemisphere.
history, the clinical symptoms, and infection imported into Germany Zoonoses Public Health. 2009;56:357–
69. http://dx.doi.org/10.1111/j.1863-
the initial laboratory findings, WNV that was unambiguously confirmed 2378.2008.01207.x
infection was suspected. Indirect by laboratory testing. WNV 2. Kilpatrick AM. Globalization, land use,
immunofluorescence assays and virus meningoencephalitis was diagnosed and the invasion of West Nile virus.
neutralization tests (VNT) for WNV on the basis of strict serologic criteria Science. 2011;334:323–7. http://dx.doi.
org/10.1126/science.1201010
and other flaviviruses were performed established by the Centers for Disease 3. Danis K, Papa A, Theocharopoulos
as described (4). IgM and IgG against Control and Prevention (Atlanta, GA, G, Dougas G, Athanasiou M, Detsis
WNV were detected in serum and in USA) (5). In 2011, 69 clinical cases of M, et al. Outbreak of West Nile virus
CSF by indirect immunofluorescence WNV infection were reported from the infection in Greece, 2010. Emerg Infect
Dis. 2011;17:1868–72. http://dx.doi.
assay with an 8-fold (IgM) and 32- province of Ontario, although no cases org/10.3201/eid1710.110525
fold (IgG) increase in serum titer in the city of Ottawa were reported
from day 4 to day 26 (Table). WNV (Public Health Agency of Canada; 1
These authors contributed equally to this
IgG and WNV IgM titers were higher www.eidgis.com/wnvmonitorca). This article.

Table. Results of indirect immunofluorescence assays performed on serum and CSF samples from patient with suspected WNV
infection, Germany, 2011*
Antibody titer in serum Antibody titer in CSF
Virus used as antigen Ig class Day 4 Day 6 Day 11 Day 26 Day 4 Day 6
WNV IgG 320 1,280 5,120 10,240 20 40
WNV IgM 160 160 1,280 1,280 10 20
SLEV IgG 80 80 1,280 2,560 <10 <10
SLEV IgM 20 20 40 <20 <10 <10
JEV IgG ND <20 ND 1,280 ND <10
JEV IgM ND <20 ND <20 ND <10
DENV IgG ND 80 ND 640 ND <10
DENV IgM ND 20 ND <20 ND <10
YFV IgG ND <20 ND ND ND <10
YFV IgM ND. <20 ND ND ND <10
TBEV IgG ND <20 ND ND ND <10
TBEV IgM ND <20 ND ND ND <10
*CSF, cerebrospinal fluid; WNV, West Nile virus; SLEV, St. Louis encephalitis virus; JEV, Japanese encephalitis virus; ; ND, not done; DENV, dengue
virus; YFV, yellow fever virus; TBEV, tick-borne encephalitis virus. Titers <20 for serum and <10 for CSF are considered negative.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1699
LETTERS

4. Tappe D, Schmidt-Chanasit J, Ries A, is mainly transmitted by direct con- cdc.gov/EID/pdfs/12-0062-Techapp.

Ziegler U, Müller A, Stich A. Ross River tact with saliva and nasal fluids from pdf).
virus infection in a traveller returning
from northern Australia. Med Microbiol infected persons (3). Many children In 2011, scarlet fever notification
Immunol (Berl). 2009;198:271–3. http:// can also carry GAS or be asymptom- rates were elevated in all 4 regions
dx.doi.org/10.1007/s00430-009-0122-9 atically infected (4). A recent study in of Hong Kong: New Territories East,
5. Centers for Disease Control and China showed that GAS is commonly New Territories West, Kowloon, and
Prevention. Epidemic/epizootic West
Nile virus in the United States: guidelines resistant to macrolides and tetracy- Hong Kong Island at 27.2, 21.7, 18.9,
for surveillance, prevention, and control cline but sensitive to penicillin, chlor- and 19.6 cases per 100,000 popula-
[cited 2012 Aug 4]. http://www.cdc.gov/ amphenicol, cefradine, and ofloxacin tion, respectively. However, a distinct-
ncidod/dvbid/westnile/resources/wnv- (5). In Hong Kong, GAS emm type 12 ly higher proportion of imported cases
guidelines-aug-2003.pdf
dominated among the isolates cultured before July 2011 (12 of 14, p value for
Address for correspondence: Jonas Schmidt- during 2011 (6). Most of the cases re- exact binomial test = 0.01) were noti-
Chanasit, Bernhard Nocht Institute for ported were in children <10 years of fied in New Territories East and New
Tropical Medicine, Department of Virology, age (range 1 month–51 years; me- Territories West, where the main bor-
WHO Collaborating Centre for Arbovirus and dian 6 years [interquartile range 4–7 der crossings to mainland China are
Haemorrhagic Fever Reference and Research, years]). The age distribution is similar located. This finding suggests a link
Bernhard Nocht Strasse 74, D-20359 Hamburg, to that reported during previous years to the outbreak in Guangdong in these
Germany; email: jonassi@gmx.de (data not shown). regions during the early phase of the
In the United Kingdom during local outbreak.
the mid-19th century, scarlet fever We estimated the instantaneous
epidemics were found to follow a 5- reproduction number (Rt), which
to 6-year cycle, but this pattern disap- measures the time-dependent fre-
peared as incidence decreased (7). An- quency of transmission per single
nual scarlet fever notifications in Hong primary case (online Technical Ap-
Kong remained low during 2001– pendix) (9). An Rt consistently >1
Scarlet Fever 2010 (<4 cases/100,000 population) would indicate sustained local trans-
Outbreak, and did not demonstrate any apparent mission. We estimated Rt on the basis
long-term pattern. The recent increase of the daily scarlet fever notification
Hong Kong, 2011 in scarlet fever notifications might be data in different periods, adjusted for
To the Editor: Scarlet fever is attributable to antigenic drift, increase imported cases. For 19 cases (1.2%
a notifiable disease in Hong Kong, in virulence of GAS (8), or increased of all cases), we could not determine
Guangdong Province, and Macau in circulation of GAS. However, other whether infection was local or im-
the People’s Republic of China. All 3 than mandatory notification of medi- ported. We estimated Rt in 2 differ-
areas reported substantial increases in cally attended case-patients, system- ent ways: either by assuming that all
cases during 2011 (Figure, panel A). In atic laboratory testing of GAS isolates of these cases were local or by as-
Hong Kong, individual data, including was not conducted in Hong Kong, and suming that they all were imported,
age, geographic location, date of no- these possibilities could not be further to represent possible extreme values
tification, and travel history within investigated. of Rt. Rt fluctuated between 0.6 and
the incubation period, were collected Notifications of scarlet fever usu- 2.0 and was consistently >1 from
from all locally notified scarlet fever ally peak during December–March in mid-May through the end of June. Rt
case-patients. As of December 31, Hong Kong, but the outbreak in 2011 fell quickly to <1 beginning in early
2011, a total of 1,535 cases (21.7 cas- peaked in June (Figure, panel B). The July after 2 fatal scarlet fever cases
es/100,000 population) were reported, rise in scarlet fever cases in Guang- were reported on May 29 and June
which was ≈10× higher than the aver- dong Province and Macau slightly 21, which raised widespread concern
age number of annual cases reported preceded that in Hong Kong; cases in in the community (Figure, panel C).
during the preceding 10 years (1). Of Guangdong peaked in April (Figure, Heightened surveillance, publicity,
those, 730 cases were laboratory con- panel A). Maximum cross-correlations health education to the public (on-
firmed; 46 cases were imported; and 2 between spline-interpolated weekly line Technical Appendix) were im-
cases, 1 each in a 7-year-old girl and a scarlet fever notifications in Guang- plemented by the Centre for Health
5-year-old boy co-infected with chick- dong and Macau and those in Hong Protection in early June and could
enpox, resulted in death (2). Kong were found at 1- and 2-week have contributed to the reduction in
Group A Streptococcus (GAS), lags, respectively (ρ = 0.45 and 0.58) transmissibility. The health education
the bacterium that causes scarlet fever, (online Technical Appendix, wwwnc. measures included guidance on pre-

1700 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

vention and control measures, such investigated spatiotemporal spread- information so that future disease con-
as updates of antimicrobial drug re- ing patterns of the disease with certain trol efforts can be made at multiple
sistance profile of GAS issued to all time lags in Hong Kong, Macau, and geographic levels.
doctors and strengthening reporting Guangdong. The estimated Rt in 2011
Acknowledgments
of scarlet fever cases by child care indicated the potential for local trans-
We thank S.K. Chuang and Thomas
centers and schools for prompt epide- mission and persistence. Such a bor-
Tsang of the Centre for Health Protection
miologic investigations, derless spread indicates a critical need
in Hong Kong for their kind support and
In summary, we analyzed the to enhance cross-border communica-
assistance in collating the notification data.
notification data of scarlet fever and tion and timely sharing of epidemic
We thank Peng Wu for technical support.
This project was supported by the
Harvard Center for Communicable Dis-
ease Dynamics from the National Institute
of General Medical Sciences (grant no.
U54 GM088558) and the Research Fund
for the Control of Infectious Diseases,
Food and Health Bureau, Government of
the Hong Kong Special Administrative Re-
gion (grant no. HKU-11-04-02). H.N. re-
ceived funding support from JST PRESTO
program. B.J.C. received research funding
from MedImmune Inc.. D.K.M.I. received
research funding from F. Hoffmann-La
Roche Ltd.

Eric H.Y. Lau, Hiroshi Nishiura,

Benjamin J. Cowling,
Dennis K.M. Ip, and Joseph T. Wu
Author affiliations: The University of Hong
Kong School of Public Health, Hong Kong
Special Administrative Region, People’s
Republic of China (E.H.Y. Lau, H. Nishiu-
ra, B.J. Cowling, D.K.M. Ip, J.T. Wu); and
Japan Science and Technology Agency,
Saitama, Japan (H. Nishiura)

DOI: http://dx.doi.org/10.3201/eid1810.120062

References
Figure. Trends in scarlet fever during outbreak in Hong Kong, Guangdong, and Macau,
People’s Republic of China, 2011. A) Monthly scarlet fever notifications in Hong Kong, 1. Centre for Health Protection, Government
Guangdong (data obtained from Department of Health Guangdong Province, www.gdwst. of the Hong Kong Special Administrative
gov.cn/a/yiqingxx), and Macau (data obtained from Health Bureau, Government of the Region. Number of notifications for notifi-
Macau Special Administrative Region (www.ssm.gov.mo/news/content/ch/1005/statistic. able infectious diseases, 2011 [cited 2012
aspx). Vertical tick marks indicate January of each year. Data from Guangdong and Macau Jan 11]. http://www.chp.gov.hk/en/notifi-
were available beginning in 2005. Black line indicates data from Hong Kong; gray line, data able1/10/26/43.html
from Guangdong; broken line, data from Macau; gray bar, number of imported cases in 2. Centre for Health Protection, Government
Hong Kong, 2005-2011. B) Weekly notifications of scarlet fever cases in Hong Kong and of the Hong Kong Special Administrative
monthly notifications in Guangdong and Macau. Black line indicates data from Hong Kong; Region. Update on scarlet fever in Hong
gray line, data from Guangdong; broken line, data from Macau. C) Estimated instantaneous Kong [cited 2012 Jun 30]. http://www.
reproduction number (Rt) and 95% pointwise confidence intervals (CIs) based on scarlet chp.gov.hk/files/pdf/ltd_20110622_scar-
fever notifications in Hong Kong, February–December, 2011. Black line indicates estimate let_fever_update.pdf
calculated by grouping patients with unknown importation status with patients with imported 3. Centers for Disease Control and Preven-
cases; gray line, estimate calculated by grouping patients with unknown importation status tion. Scarlet fever: a group A streptococcal
with local case-patients; broken lines, the upper and lower limits of the 95% CIs for Rt. For infection. 2011 [cited 2011 Sep 1]. http://
better presentation, CIs are shown only for the former estimates. Horizontal line indicates www.cdc.gov/Features/ScarletFever
the critical value of Rt, under which transmission of disease will not be sustainable.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1701
LETTERS

4. Shaikh N, Leonard E, Martin JM. Preva-

lence of streptococcal pharyngitis and
Hand, Foot, and the exanthem. The rash resolved over
7–14 days with no residual scarring.
streptococcal carriage in children: a meta- Mouth Disease Samples from the oropharynx, rectum,
Caused by
analysis. Pediatrics. 2010;126:e557–64.
http://dx.doi.org/10.1542/peds.2009-2648 and vesicles from these patients were
5. Liang Y, Shen X, Huang G, Wang C,
Shen Y, Yang Y. Characteristics of Strep-
Coxsackievirus A6 sent to the Centers for Disease Control
and Prevention (Atlanta, GA, USA)
tococcus pyogenes strains isolated from
Chinese children with scarlet fever. Acta To the Editor: Coxsackievirus for analysis. Reverse transcription
Paediatr. 2008;97:1681–5. http://dx.doi. A6 (CVA6) is a human enterovirus PCR and sequencing by using primers
org/10.1111/j.1651-2227.2008.00983.x associated with herpangina in infants. specific for a portion of the viral
6. Lau MCK. Increase in scarlet fever cases In the winter of 2012, we evaluated a protein 1 coding region identified
in 2011. Communicable Diseases Watch.
2011; 8:48–9 [cited 2012 Jun 30]. http:// cluster of 8 patients, 4 months–3 years CVA6 (1) (Table, Appendix).
www.chp.gov.hk/files/pdf/cdw_v8_12.pdf of age, who were brought for treatment Outbreaks of HFMD caused
7. Duncan CJ, Duncan SR, Scott S. The at Boston Children’s Hospital (Boston, by CVA6 have been described in
dynamics of scarlet fever epidemics in MA, USA) with a variant of hand, Singapore, Finland, Taiwan, and
England and Wales in the 19th century.
Epidemiol Infect. 1996;117:493–9. http:// foot, and mouth disease (HFMD) most recently in Japan; most cases
dx.doi.org/10.1017/S0950268800059161 that has now been linked to CVA6 have occurred in the warmer months
8. Köhler W, Gerlach D, Knöll H. Strepto- (Table, Appendix, wwwnc.cdc.gov/ (2–6). Cases in the cluster described
coccal outbreaks and erythrogenic toxin EID/article/18/10/12-0813-T1.htm). here are likely related to an emerging
type A. Zentralbl Bakteriol Mikrobiol Hyg
[A]. 1987;266:104–15 10.1016/S0176- During this same period, the Boston outbreak of CVA6-associated HFMD
6724(87)80024-x. Public Health Commission’s syndromic in the United States (7). The atypical
9. Nishiura H, Chowell G. The effective surveillance system detected a 3.3-fold seasonality of the outbreak, during the
reproduction number as a prelude to sta- increase in emergency department winter in Boston, could be related to
tistical estimation of time-dependent epi-
demic trends. In: Chowell G, Hayman JM, discharge diagnoses of HFMD. In the the unusually mild temperatures in the
Bettencourt LMA, editors. Mathematical United States, HFMD typically occurs winter of 2012.
and statistical estimation approaches in in the summer and early autumn and Recent CVA6 outbreaks have
epidemiology. New York: Springer; 2009. is characterized by a febrile enanthem been characterized by a febrile illness
p. 103–21.
of oral ulcers and macular or vesicular associated with an oral enanthem
Address for correspondence: Eric H.Y. Lau, lesions on the palms and soles; the and lesions on the palms, soles, and
School of Public Health, The University of etiologic agents are most often CVA16 buttocks. CVA6 infections in Taiwan
Hong Kong, Pokfulam, Hong Kong Special and enterovirus 71. during 2004–2009 were associated
Administrative Region, People’s Republic of In contrast to the typical mani- with HFMD in 13% of cases, with
China; email: ehylau@hku.hk festation, the patients in the Boston disease defined as oral ulcers on the
cluster exhibited symptoms in tongue or buccal mucosa and vesicular
late winter (Table, Appendix) and rashes on the palms, soles, knees, or
Letters had perioral (Figure, panel A) and buttocks (2). In Singapore, where
Letters commenting on recent articles
perirectal (Figure, panel B) papules CVA6 accounted for 24% of HFMD
as well as letters reporting cases,
and vesicles on the dorsal aspects cases, patients had oral lesions and <5
outbreaks, or original research are
of the hands and feet (Figure, panel peripheral papules, placing them on
welcome. Letters commenting on ar-
C). Patients experienced a prodrome a spectrum closer to the herpangina
ticles should contain no more than
lasting 1–3 days, consisting of fever more typically observed in CVA6
300 words and 5 references; they are
(8 patients), upper respiratory tract infection (8).
more likely to be published if submitted
symptoms (4 patients), and irritability The patients we report in this
within 4 weeks of the original article’s
(7 patients). This prodrome was cluster most typically had perioral
publication. Letters reporting cases,
followed by the development of a and perirectal papules in addition to
outbreaks, or original research should
perioral papular rash (8 patients), vesicles on the dorsum of their hands.
contain no more than 800 words and
which was often impetiginized with Two reports of CVA6-associated
10 references. They may have 1
secondary crusting; a prominent HFMD outbreaks describe cases
Figure or Table and should not be di-
papulovesicular rash on the dorsum that more closely resemble patients
vided into sections. All letters should
of the hands and feet (6 patients); in the Boston outbreak. In a series
contain material not previously pub-
and a perirectal eruption (7 patients). from Finland in 2008, representative
lished and include a word count.
Half of the patients had intraoral patients had both perioral lesions and
lesions. Fever abated in most of the vesicles on the dorsum of their hands
patients within a day after onset of (6). In a large series of patients with

1702 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

Figure. Manifestations of hand, foot, and mouth disease in patients, Boston, Massachusetts, USA, 2012. Discrete superficial crusted
erosions and vesicles symmetrically distributed in the perioral region (A), in the perianal region (B), and on the dorsum of the hands (C).
A color version of this figure is available online (wwwnc.cdc.gov/EID/article/18/10/12-0813-F1.htm).

HFMD in Taiwan in 2010, patients Because of the atypical presentation DOI: http://dx.doi.org/10.3201/eid1810.120813
with CVA6 had perioral lesions in of CVA6-associated HFMD, clinical References
addition to an enanthem (3). vigilance is needed to recognize
Outbreaks of CVA6-associated emerging regional outbreaks. More 1. Nix WA, Oberste MS, Pallansch MA.
HFMD in Finland, Taiwan, and Japan detailed epidemiologic and genetic Sensitive, seminested PCR amplification
of VP1 sequences for direct identification
were associated with onychomadesis, analyses will be required to of all enterovirus serotypes from original
with the loss of nails occurring 1–2 characterize the role of CVA6 in US clinical specimens. J Clin Microbiol.
months after initial symptoms (3,4,6). outbreaks of HFMD. 2006;44:2698–704. http://dx.doi.org/10.
The association between more typical 1128/JCM.00542-06
2. Lo SH, Huang YC, Huang CG, Tsao KC,
HFMD and onychomadesis has Acknowledgments Li WC, Hsieh YC, et al. Clinical and
additionally been described in the We thank Richard Rossi and Renee epidemiologic features of coxsackievirus
United States and Europe but without Roy for clinical sample preparation and A6 infection in children in northern Taiwan
a link to specific serotype or with a between 2004 and 2009. J Microbiol
processing and Kenneth McIntosh for Immunol Infect. 2011;44:252–7. http://
small percentage of CVA6-associated critical review of this manuscript. dx.doi.org/10.1016/j.jmii.2011.01.031
cases (9). Cases from the Boston 3. Wei SH, Huang YP, Liu MC, Tsou TP,
A.A.A. is supported by National
epidemic may fit into an emerging Lin HC, Lin TL, et al. An outbreak
Institutes of Health grant no. 5 K08 of coxsackievirus A6 hand, foot,
clinical phenotype of CVA6, and it
AI093676-02. and mouth disease associated with
will be interesting to see whether nail onychomadesis in Taiwan, 2010. BMC
loss develops in those patients. Infect Dis. 2011;11:346. http://dx.doi.
Kelly Flett,1 Ilan Youngster,1
Given the numerous CVA6 org/10.1186/1471-2334-11-346
Jennifer Huang, 4. Fujimoto T, Iizuka S, Enomoto M,
outbreaks in multiple countries in
Alexander McAdam, Abe K, Yamashita K, Hanaoka N, et al.
2008 and a US population that may be Hand, foot, and mouth disease caused by
Thomas J. Sandora,
relatively naïve to this serotype, CVA6 coxsackievirus A6, Japan, 2011. Emerg
Marcus Rennick,
is likely to spread throughout North Infect Dis. 2012;18:337–9. http://dx.doi.
Sandra Smole, org/10.3201/eid1802.111147
America. Clinicians should be aware
Shannon L. Rogers, 5. Blomqvist S, Klemola P, Kaijalainen S,
that, although standard precautions Paananen A, Simonen ML, Vuorinen T,
W. Allan Nix, M. Steven Oberste,
are routinely recommended for et al. Co-circulation of coxsackieviruses
Stephen Gellis, and Asim A. Ahmed
managing enteroviral infections A6 and A10 in hand, foot and mouth
Author affiliations: Children’s Hospital disease outbreak in Finland. J Clin
in health care settings, contact
Boston, Boston, Massachusetts, USA (K. Virol. 2010;48:49–54. http://dx.doi.
precautions are indicated for children org/10.1016/j.jcv.2010.02.002
Flett, I. Youngster, J. Huang, A. McAdam,
in diapers to control institutional 6. Osterback R, Vuorinen T, Linna M, Susi
T.J. Sandora, S. Gellis, A.A. Ahmed);
outbreaks (10). In addition, the P, Hyypia T, Waris M. Coxsackievirus A6
Boston Public Health Commission, and hand, foot, and mouth disease, Finland.
presence of perioral lesions and
Boston (M. Rennick); Massachusetts Emerg Infect Dis. 2009;15:1485–8. http://
peripheral vesicles on the dorsum dx.doi.org/10.3201/eid1509.090438
Department of Public Health, Jamaica
rather than palmar/plantar surface of
Plain, Massachusetts, USA (S. Smole);
the hands and feet represents a unique
and Centers for Disease Control and
phenotype of HFMD that could be
Prevention, Atlanta, Georgia, USA (S.L.
confused with herpes simplex or 1
These authors contributed equally to this
Rogers, W.A. Nix, M.S. Oberste)
varicella-zoster virus infections. article.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1703
LETTERS

7. Centers for Disease Control and the presence of the parasite in these gorillas (i.e., Democratic Republic
Prevention. Notes from the field: severe regions, the most convincing of of the Congo [3], Uganda [4], and
hand, foot, and mouth disease associated
with coxsackievirus A6—Alabama, which were the steady and consistent Equatorial Guinea [5]). During our
Connecticut, California, and Nevada, numbers of non-African travelers who seroepidemiologic study from the
November 2011–February 2012. MMWR returned to their countries of origin Democratic Republic of the Congo,
Morb Mortal Wkly Rep. 2012;61:213–4. infected with malarial parasites that in which P. vivax sporozoite–specific
8. Wu Y, Yeo A, Phoon MC, Tan EL, Poh CL,
Quak SH, et al. The largest outbreak of were subsequently identified as P. antibodies were detected in ≈10% of
hand; foot and mouth disease in Singapore vivax (2). the population, we found that women
in 2008: the role of enterovirus 71 and More recently, evidence has were significantly more likely than
coxsackievirus A strains. Int J Infect emerged regarding the transmission men to have been exposed to P. vivax
Dis. 2010;14:e1076–81. http://dx.doi.
org/10.1016/j.ijid.2010.07.006 of P. vivax in regions of Africa sporozoites (3). Women in this region
9. Bracho MA, Gonzalez-Candelas F, Valero where the local human population is typically spend more time than men
A, Cordoba J, Salazar A. Enterovirus co- predominantly Duffy negative (3–6). near the forest fringe, where they work
infections and onychomadesis after hand, In 4 (3.5%) of 155 patients from in crop fields. This forest is within the
foot, and mouth disease, Spain, 2008.
Emerg Infect Dis. 2011;17:2223–31. western Kenya (6), 7 (0.8%) of 898 known habitat range of the chimpanzee
http://dx.doi.org/10.3201/eid1712.110395 persons from Angola (4), and 8 (8.2%) Pan troglodytes and the gorilla, Gorilla
10. Siegel JD, Rhinehart E, Jackson M, of 97 persons from Equatorial Guinea gorilla gorilla, both of which have
Chiarello L. 2007 Guideline for isolation (4), P. vivax parasites were detected been reported to be natural hosts of the
precautions: preventing transmission of
infectious agents in health care settings. in the blood of apparently Duffy- malaria parasite P. schwetzi, which is a
Am J Infect Control. 2007;35(Suppl negative persons, suggesting that the P. vivax–like or P. ovale–like parasite
2):S65–164. http://dx.doi.org/10.1016/j. parasite might not be as absolutely that might also be unable to invade
ajic.2007.10.007 dependent on the Duffy receptor for the erythrocytes of persons who are
erythrocyte invasion as previously Duffy negative (8). These animals have
Address for correspondence: Asim A. Ahmed,
thought. These findings are supported recently been shown to be infected
Children’s Hospital Boston–Infectious
by a report from Madagascar (where occasionally with parasites that have
Diseases, 300 Longwood Ave, Boston, MA
the human population is composed mitochondrial genomes closely
02115, USA; email: asim.ahmed@childrens.
of a mixture of Duffy-positive and resembling those of P. vivax (9,10).
harvard.edu
Duffy-negative persons), in which 42 We have argued that, given the
(8.8%) of 476 Duffy-negative persons high malaria transmission rates in sub-
who had symptoms of malaria were Saharan Africa, it is plausible that the
reported to be positive for P. vivax 1%–5% of the human population who
by both microscopy and PCR (7). are Duffy positive might maintain
The prevalence of P. vivax in Duffy- the transmission of the parasite (2).
negative persons was significantly The discovery of P. vivax parasites
Duffy Phenotype lower than its prevalence in Duffy- (or P. vivax–like parasites) in the
and Plasmodium positive persons residing in the same
area, suggesting that Duffy negativity
blood of African great apes leads to
a question: could nonhuman primates
vivax infections in is a barrier to the parasite to some in Africa be acting as Duffy-positive
Humans and Apes, degree. Given the extremely high rates reservoirs of P. vivax in regions
Africa of malaria transmission in western and
central Africa, a P. vivax parasite that
where the human population is almost
entirely insusceptible? This possibility
To the Editor: Benign tertian could efficiently invade Duffy-negative warrants further investigation. Given
malaria, caused by Plasmodium vivax, erythrocytes would, presumably, the increasing rarity of the great
has long been considered absent, or become highly prevalent very rapidly. apes, however, their capacity to act as
at least extremely rare, in western With the exception of the cases zoonotic reservoirs could be limited.
and central Africa. In these regions, reported from Angola and Kenya, It would be informative, in any case,
95%–99% of humans are of the Duffy which lie outside the area where to determine how the regions that P.
negative phenotype, a condition that is the proportion of the population vivax–positive travelers visit during
thought to confer complete protection with Duffy negativity is highest, their stay in Africa correspond with the
against the parasite during the blood the reports of the transmission of P. ranges of chimpanzees and gorillas.
stages of its life cycle (1,2). Sporadic vivax within predominantly Duffy- If African great apes do, indeed,
reports throughout the latter half of the negative populations all come from constitute a zoonotic reservoir of
20th century, however, have hinted at regions inhabited by chimpanzees and P. vivax parasites, what are the

1704 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

repercussions for human health? Given 6. Ryan JR, Stoute JA, Amon J, Dunton RF, reported as <42% in the United States,
that 95%–99% of humans possibly Mtalib R, Koros J, et al. Evidence for which is higher than reported rates of R.
transmission of Plasmodium vivax among
exposed to such a reservoir are Duffy a Duffy antigen negative population in rickettsii (the cause of Rocky Mountain
negative, and therefore resistant to western Kenya. Am J Trop Med Hyg. spotted fever) in Dermacentor species
the parasite, these would appear to be 2006;75:575–81. ticks. Misdiagnosis among SFGR
slight. However, as humans encroach 7. Ménard D, Barnadas C, Bouchier C, infections is not uncommon, and
Henry-Halldin C, Gray LR, Ratsimbasoa
more frequently into ape habitats, the A, et al. Plasmodium vivax clinical malaria R. parkeri rickettsiosis can cause
chances of humans encountering the is commonly observed in Duffy-negative symptoms similar to those for mild
parasite will increase. In the short Malagasy people. Proc Natl Acad Sci U Rocky Mountain spotted fever (1). We
term, the risks are probably limited S A. 2010;107:5967–71. http://dx.doi. evaluated infection rates of R. parkeri
org/10.1073/pnas.0912496107
to Duffy-positive persons who enter 8. Coatney GR, Collins WE, Warren M, and Candidatus Rickettsia andeanae,
areas where apes are present, such as Contacos PG. Plasmodium schwetzi. a recently identified but incompletely
tourists and migrant workers. In: Coatney GR, Collins WE, Warren characterized SFGR, in Gulf Coast
M, Contacos PG, editors. The primate ticks in Mississippi, USA.
malarias. Bethesda (MD): US Department
Richard Leighton Culleton During May–September of 2008–
of Health, Education, and Welfare; 1971.
and Pedro Eduardo Ferreira p. 141–52. 2010, we collected adult Gulf Coast
Author affiliations: Nagasaki University, Na- 9. Krief S, Escalante AA, Pacheco MA, ticks from vegetation at 10 sites in
gasaki, Japan (R.L. Culleton, P.E. Ferreira); Mugisha L, Andre C, Halbwax M, et Mississippi. We extracted genomic
al. On the diversity of malaria parasites
University of Algarve, Faro, Portugal (P.E.
in African apes and the origin of DNA from the ticks using the
Ferreira); and Karolinska Institutet, Stock- Plasmodium falciparum from Bonobos. illustra tissue and cells genomicPrep
holm, Sweden (P.E. Ferreira) PLoS Pathog. 2010;6:e1000765. http:// Mini Spin Kit (GE Healthcare Life
dx.doi.org/10.1371/journal.ppat.1000765 Sciences, Piscataway, NJ, USA). We
DOI: http://dx.doi.org/10.3201/eid1810.120120 10. Liu W, Li Y, Learn GH, Rudicell
RS, Robertson JD, Keele BF, et al. tested amplifiable tick DNA by PCR of
References Origin of the human malaria parasite the tick mitochondrial 16S rRNA gene
Plasmodium falciparum in gorillas. (2). We tested for molecular evidence
1. Miller LH, Mason SJ, Clyde DF, Nature. 2010;467:420–5. http://dx.doi.
McGinniss MH. The resistance factor to
of any SFGR species by nested PCR
org/10.1038/nature09442
Plasmodium vivax in blacks. The Duffy- of rompA (rickettsial outer membrane
blood-group genotype, FyFy. N Engl
Address for correspondence: Richard Leighton protein A gene) (1). Samples positive
J Med. 1976;295:302–4. http://dx.doi. for SFGR were subsequently tested
org/10.1056/NEJM197608052950602 Culleton, Malaria Unit, Institute of Tropical
2. Culleton RL, Mita T, Ndounga M, Unger Medicine, Nagasaki University, 1-12-4 by using species-specific rompA PCR
H, Cravo P, Paganotti G, et al. Failure Sakamoto, Nagasaki 852-8523, Japan; email: for R. parkeri (3) and Candidatus R.
to detect Plasmodium vivax in West and
richard@nagasaki-u.ac.jp andeanae (4). All PCRs included 1) a
Central Africa by PCR species typing. positive control of DNA from cultured
Malar J. 2008;7:174. http://dx.doi.
org/10.1186/1475-2875-7-174
R. parkeri– (Tate’s Hell strain) or
3. Culleton R, Ndounga M, Zeyrek FY, Candidatus R. andeanae–infected Gulf
Coban C, Casimiro PN, Takeo S, et Coast ticks and 2) a negative control
al. Evidence for the transmission of of water (nontemplate). PCR products
Plasmodium vivax in the Republic of
the Congo, West Central Africa. J Infect
were purified by using Montage
PCR Centrifugal Filter Devices
Dis. 2009;200:1465–9. http://dx.doi.
org/10.1086/644510 Rickettsia parkeri (Millipore, Bedford, MA, USA) and
4. Dhorda M, Nyehangane D, Renia L, Piola
P, Guerin PJ, Snounou G. Transmission
and Candidatus sequenced by using Eurofins MWG
of Plasmodium vivax in south-western Rickettsia Operon (Huntsville, AL, USA). We
generated consensus sequences using
Uganda: report of three cases in pregnant
women. PLoS One. 2011;6:e19801. http:// andeanae in ClustalW2 (www.ebi.ac.uk/Tools/
5.
dx.doi.org/10.1371/journal.pone.0019801
Mendes C, Dias F, Figueiredo J, Mora
Gulf Coast Ticks, msa/clustalw2/) alignment and
VG, Cano J, de Sousa B, et al. Duffy Mississippi, USA identified the sequences using
negative antigen is no longer a barrier to GenBank BLAST searches (www.ebi.
Plasmodium vivax–molecular evidences To the Editor: Rickettsia parkeri, ac.uk/Tools/clustalw2/).
from the African West Coast (Angola and a spotted fever group Rickettsia Proportions of ticks infected
Equatorial Guinea). PLoS Negl Trop Dis.
2011;5:e1192. http://dx.doi.org/10.1371/ (SFGR) bacterium, is transmitted by with SFGR, by region and year,
journal.pntd.0001192 Amblyomma maculatum, the Gulf were compared separately by using
Coast tick (1). The prevalence of R. Fisher exact test followed by pairwise
parkeri in Gulf Coast ticks has been comparisons with a Bonferroni

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1705
LETTERS

Table. PCR results for adult Rickettsia parkeri– and Candidatus Rickettsia andeanae–infected Gulf Coast ticks (Amblyomma
maculatum) collected from 10 sites in Mississippi, USA, 2008–2010*
No. (%; 95% CI)
Location (no. No. No. (%; 95% CI) SFG No. (%; 95% CI) R. Candidatus R. Expected no. (%) No. (%; 95% CI)
collection sites) ticks rompA parkeri only andeanae only co-infected ticks co-infected ticks
North (4) 257 49 (19.1; 14.5–24.4)† 48 (18.7; 14.1–24)‡ 0§ 0.19 (0.07) 1 (0.4; 0–2.1)¶
Central (1) 38 4 (10.5; NA) 1 (2.6; NA) 2 (5.3; NA) 0.16 (0.42) 1 (2.6; NA)
South (5) 403 75 (18.6; 14.9–22.8)† 57 (14.1; 10.9–17.9)‡ 8 (2.0; 0.9–3.9)§ 2.99 (0.74) 10 (2.5; 1.2–4.5)¶
Total (10) 698 128 (18.3; NA) 106 (15.2; NA) 10 (1.4; NA) 3.65 (0.52) 12 (1.7; NA)
*The estimated value of co-infection caused by chance alone (E) was calculated by using the formula E = (a + b)(a + c) / (a + b + c + d) (5), where a = no.
ticks infected with both Rickettsia species, b = no. ticks infected only with R. parkeri, c = no. ticks infected only with Candidatus R. andeanae, and d = no.
ticks not infected with either Rickettsia species. SFG rompA, spotted fever group rickettsial outer membrane protein A gene. NA, not applicable.
†p = 0.9187.
‡p = 0.1275.
§p = 0.0257.
¶p = 0.0578 (comparison of prevalence from northern and southern sites only).

adjustment (PROC FREQ, SAS for (p = 0.03). The infection rate for co- significantly higher than expected from
Windows, V9.2; SAS Institute, Cary, infected ticks in southern sites was chance alone. The biologic role of co-
NC, USA). For all analyses, p<0.05 higher than that in northern sites (p = infections of Gulf Coast ticks with R.
was considered significant. An index 0.06). Among the 3 collection years parkeri and Candidatus R. andeanae
of co-infection was calculated by using for northern and southern sites, only remains to be determined.
the formula IC = ([O – E]/N) × 100, in the prevalence of R. parkeri in singly
which IC is index of co-infection, O is infected ticks differed significantly Acknowledgments
number of co-infections, E is expected (p = 0.01) (data not shown); the We acknowledge Gail Moraru, Diana
occurrence of co-infection caused by infection rate was significantly greater Link, Claudenir Ferrari, Marcia Carvalho,
chance alone, and N is total number during 2010 than during 2009 (p = and Ryan Lawrence for assisting with tick
of ticks infected by either or both 0.003, α/3 = 0.02). The overall index collection; Robert Wills for assisting with
Rickettsia species. A χ2 test was used of co-infection with R. parkeri and statistical analyses; and Erle Chenney and
to determine statistical significance (5). Candidatus R. andeanae was 6.5, Whitney Smith for contributing to the
A total of 707 adult Gulf Coast statistically higher than expected by molecular analysis of ticks.
ticks were collected during the 3 years chance alone (Table) (p<0.0001).
This work was supported by a
(350 in 2008, 194 in 2009, and 163 in The overall prevalence of infection
Southeastern Center for Emerging Biologic
2010). Tick mitochondrial 16S rRNA with SFGR species in Gulf Coast ticks
Threats grant awarded to A.S.V.-S. in 2009
gene was detected in 698 (98.7%), of sampled was 18.3%; 15.2% of ticks
and funding from the College of Veterinary
which 128 (18.3%) were positive for were singly infected with R. parkeri,
Medicine at Mississippi State University.
SFGR DNA, comprising 106 (15.2%) and 1.7% were infected with R. parkeri
positive only for R. parkeri, 10 (1.4%) and Candidatus R. andeanae. As
positive only for Candidatus R. reported, the frequency of R. parkeri Flavia A.G. Ferrari,
andeanae, and 12 (1.7%) co-infected in Gulf Coast ticks is generally high, Jerome Goddard,
with R. parkeri and Candidatus R. ranging from ≈10% to 40% (3,4,6–8). Christopher D. Paddock,
andeanae (Table). Positive test results We found approximately 1 R. parkeri- and Andrea S. Varela-Stokes
from 22 ticks singly or co-infected infected Gulf Coast tick for every 6 Author affiliations: Mississippi State
with Candidatus R. andeanae were ticks tested, suggesting that infected University, Mississippi State, Mississippi,
confirmed by sequencing. Gulf Coast ticks are commonly USA (F.A.G. Ferrari, J. Goddard, A.S.
Most (94.6%) ticks were from encountered in Mississippi. Because Varela-Stokes); and Centers for Disease
northern (n = 260) and southern (n = Gulf Coast ticks are among the Control and Prevention, Atlanta, Georgia,
409) Mississippi (online Technical most common human-biting ticks in USA (C.D. Paddock)
Appendix, Figure, wwwnc.cdc.gov/ Mississippi (9), awareness of R. parkeri DOI: http://dx.doi.org/10.3201/eid1810.120250
EID/article/18/10/12-0250-F1.htm). rickettsiosis should be increased in this
No significant difference in the number state. We identified Candidatus R. References
of R. parkeri–infected ticks between andeanae in ≈3% of Gulf Coast ticks 1. Paddock CD, Sumner JW, Comer JA,
northern and southern Mississippi was in Mississippi; this frequency is similar Zaki SR, Goldsmith CS, Goddard
J, et al. Rickettsia parkeri: a newly
observed (p = 0.13) (Table). However, to those reported in other studies of recognized cause of spotted fever
significantly more ticks were singly Gulf Coast ticks in the southern United rickettsiosis in the United States. Clin
infected with Candidatus R. andeanae States (4,6). Our finding of co-infected Infect Dis. 2004;38:805–11. http://dx.doi.
in southern sites than in northern sites Gulf Coast ticks is at a frequency org/10.1086/381894

1706 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

2. Black WC, Klompen JS, Keirans JE.

Phylogenetic relationships among tick
Attributing Cause were 11 infectious disease and 2
public health physicians involved with
subfamilies (Ixodida: Ixodidae: Argasidae) of Death for CDI surveillance at their respective
Patients with
based on the 18S nuclear rDNA gene. Mol
Phylogenet Evol. 1997;7:129–44. http:// hospitals but not this hospital. The

3.
dx.doi.org/10.1006/mpev.1996.0382
Varela-Stokes AS, Paddock CD,
Clostridium difficile median (minimal, maximal) κ statistics
for comparison of external reviews
Engber BMT. Rickettsia parkeri in Infection with surveillance classification were
Amblyomma maculatum ticks, North
Carolina, USA, 2009–2010. Emerg To the Editor: Hota et al. report 0.495 (0.252, 0.607) for directly
Infect Dis. 2011;17:2350–3. http://dx.doi.
that for deceased patients who had attributable, 0.182 (−0.091, 0.182) for
org/10.3201/eid1712.110789 contributed, and 0.321 (0.124, 0.614)
4. Paddock CD, Fournier PE, Sumner JW, Clostridium difficile infection (CDI),
Goddard J, Elshenawy Y, Metcalfe MG, agreement is poor between causes of for unrelated. Comparison within
et al. Isolation of Rickettsia parkeri and death reported on death certificates and external reviewers yielded 0.697
identification of a novel spotted fever
those categorized by a review panel (0.394, 1.0), 0.233 (−0.294, 0.703),
group Rickettsia sp. from Gulf Coast and 0.542 (0.154, 0.909), respectively.
ticks (Amblyomma maculatum) in the (1). Our data support the difficulty of
United States. Appl Environ Microbiol. attributing cause of death for patients Complete agreement was found for
2010;76:2689–96. http://dx.doi.org/10. with CDI. only 6 cases (4 directly attributable and
1128/AEM.02737-09
In 2004 in Quebec, Canada, a 2 unrelated) (Figure).
5. Ginsberg HS. Potential effects of mixed Variation among reviewers sug-
infections in ticks on transmission mandatory CDI surveillance program
dynamics of pathogens: comparative was implemented. Deaths that occurred gested that categorizations reported to
analysis of published records. Exp Appl within 30 days after CDI diagnosis were surveillance were inaccurate. Number
Acarol. 2008;46:29–41. http://dx.doi.
classified as 1) directly attributable of deaths among patients with CDI,
org/10.1007/s10493-008-9175-5 regardless of the cause of death,
6. Sumner JW, Durden LA, Goddard J, to CDI (e.g., toxic megacolon, septic
Stromdahl EY, Clark KL, Reeves WK, shock), 2) having a CDI contribution seemed to better indicate CDI severity.
et al. Gulf Coast ticks (Amblyomma (e.g., acute decompensation of Since 2008, only the crude numbers
maculatum) and Rickettsia parkeri, United
chronic heart failure), or 3) unrelated of deaths, not subjected to individual
States. Emerg Infect Dis. 2007;13:751–3. interpretation, have been reported
http://dx.doi.org/10.3201/eid1305.061468 to CDI (e.g., terminal cancer) (2). To
7. Fornadel CM, Zhang X, Smith JD, determine accuracy of the surveillance to surveillance. A questionnaire
Paddock CD, Arias JR, Norris DE. High classifications, we compared cause-of- addressing concurrent medical
rates of Rickettsia parkeri infection
death classification of 22 deceased CDI conditions, prognosis, level of care,
in Gulf Coast ticks (Amblyomma and circumstances of death is being
maculatum) and identification of patients reported to surveillance by 1
“Candidatus Rickettsia andeanae” from hospital in 2007 with causes of death implemented in Quebec hospitals
Fairfax County, Virginia. Vector Borne reported by 13 external reviewers who participating in CDI surveillance and
Zoonotic Dis. 2011;11:1535–9. http://
examined summaries of medical files should help determine the role of CDI
dx.doi.org/10.1089/vbz.2011.0654 in deaths.
8. Wright CL, Nadolny RM, Jiang J, of the deceased patients. Reviewers
Richards AL, Sonenshine DE, Gaff HD,
et al. Rickettsia parkeri in Gulf Coast
ticks, Southeastern Virginia, USA. Emerg
Infect Dis. 2011;17:896–8. http://dx.doi.
org/10.3201/eid1705.101836
9. Goddard J. A ten-year study of tick biting
in Mississippi: implications for human
disease transmission. J Agromedicine.
2002;8:25–32. http://dx.doi.org/10.1300/
J096v08n02_06

Address for correspondence: Andrea S. Varela-

Stokes, Mississippi State University, Wise
Center, 240 Wise Center Dr, Mississippi State,
MS 39762, USA; email: stokes@cvm.msstate.
edu
Figure. Classification of cause of death among 22 patients with Clostridium difficile
All material published in Emerging
infection (CDI), by 13 external reviewers, Quebec, Canada, 2007. Bars indicate the number
Infectious Diseases is in the public domain
and may be used and reprinted without of reviewers who assigned each category. Gray bars indicate that CDI was unrelated to
special permission; proper citation, death, white bars indicate that CDI contributed to death, and black bars indicate that death
however, is required. was directly attributable to CDI.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1707
LETTERS

Rodica Gilca, Charles Frenette, Characterization Apart from this, we found that
Nathanaëlle Thériault, the sequence type (ST) 587 was not
Élise Fortin,
of Mycobacterium the only spoligotype specific for M.
and Jasmin Villeneuve orygis orygis. In our study, the variant type
Author affiliations: Institut National de Santé ST701 (annotated as M. africanum in
Publique du Québec, Québec City, Québec,
To the Editor: In a recently the spolDB4 database) (4) is also an
Canada (R. Gilca, E. Fortin); Québec
published study, van Ingen et M. orygis–specific type and exactly
University Hospital Centre, Québec City
al. (1) described the molecular matches that of a previous isolate of
(R. Gilca); Laval University, Québec City
characterization and phylogenetic the oryx bacillus (SB0319) from the
(R. Gilca); McGill University Health Center,
position of the oryx bacillus, a M. bovis spoligotype database (5).
Montreal, Québec, Canada (C. Frenette, E.
member of the Mycobacterium This spoligotype differs from ST587
Fortin); and Direction Régionale de Santé
tuberculosis complex, and proposed a by the presence of spacer 18, and
Publique, Québec City (N. Thériault, J.
long overdue name for the organism: the spoligotype was not found in the
Villeneuve)
Mycobacterium orygis. The authors extensive sample set of van Ingen et
described oryx bacillus as a separate al. (1).
DOI: http://dx.doi.org/10.3201/eid1810.120202 taxon; the aim was for this description
to be used in the future to identify Nicolaas C. Gey van Pittius,
References the subspecies. Thus, we thought Paul D. van Helden,
1. Hota SS, Achonu C, Crowcroft NS, it pertinent to provide additional and Robin M. Warren
Harvey BJ, Lauwers A, Gardam MA. information that would be useful in Author affiliations: Stellenbosch University
Determining mortality rates attributable speciating isolates of the oryx bacillus. Faculty of Medicine and Health Sciences,
to Clostridium difficile infection. Emerg In a recent study, we genotyped an Tygerberg, South Africa
Infect Dis. 2012;18:305–7. http://dx.doi.
org/10.3201/eid1802.101611 isolate of oryx bacillus obtained from
DOI: http://dx.doi.org/10.3201/eid1810.120569
2. Gilca R, Fortin E, Hubert B, Frenette C, an African buffalo in South Africa (2).
Gourdeau M. Surveillance of Clostridium This isolate was typed by using 16S
difficile–associated diarrhea in Quebec. References
rDNA, M. tuberculosis complex–
Report for August 22, 2004, to August
20, 2005 [in French]. Québec: Institut specific multiplex-PCR, regions-of- 1. van Ingen J, Rahim Z, Mulder A,
National de Santé Publique du Québec; difference analyses, gyrase B gene Boeree MJ, Simeone R, Brosch R, et al.
2005. p. 4 [cited 2012 Jul 20]. http:// single nucleotide polymorphism Characterization of Mycobacterium orygis
www.inspq.qc.ca/pdf/publications/434- as M. tuberculosis complex subspecies.
(SNP) analysis, spoligotyping, and Emerg Infect Dis. 2012;18:653–5. http://
BilanCdifficile-22aout2004-20aout2005.
pdf mycobacterial interspersed repetitive dx.doi.org/10.3201/eid1804.110888
units–variable number tandem repeat 2. Gey van Pittius NC, Perrett KD, Michel
typing. We showed that, in addition to AL, Keet DF, Hlokwe T, Streicher EM, et
Address for correspondence: Rodica Gilca,
al. Infection of African buffalo (Syncerus
Institut National de Santé Publique du Québec, the markers described by van Ingen et caffer) by oryx bacillus, a rare member of
2400 d’Estimauville, Québec City, Québec, al. (1), regions of difference 701 and the antelope clade of the Mycobacterium
Canada, G1E 7G9; email: rodica.gilca@ssss. 702 were also intact in M. orygis. tuberculosis complex. J Wildl Dis. In
In addition, van Ingen et al. press.
gouv.qc.ca
3. Huard RC, Fabre M, de Haas P, Oliveira
identified the Rv204238 GGC mutation Lazzarini LC, van Soolingen D, Cousins
as a novel, useful genetic marker to D, et al. Novel genetic polymorphisms
identify M. orygis. However, such a that further delineate the phylogeny of
marker already exists in the form of the Mycobacterium tuberculosis complex.
J Bacteriol. 2006;188:4271–87. http://
the very specific gyrBoryx G to A SNP dx.doi.org/10.1128/JB.01783-05
at position 1113, which was described 4. Brudey K, Driscoll JR, Rigouts L,
by Huard et al. (3). On its own, SNP Prodinger WM, Gori A, Al-Hajoj SA,
detection in the gyrB gene allows et al. Mycobacterium tuberculosis
complex genetic diversity: mining
differentiation of at least 6 of the 9 the fourth international spoligotyping
M. tuberculosis complex species from database (SpolDB4) for classification,
each other (M. canettii, M. tuberculosis, population genetics and epidemiology.
M. orygis, M. microti, M. caprae, and BMC Microbiol. 2006;6:23. http://dx.doi.
org/10.1186/1471-2180-6-23
M. bovis) (3). Thus, the SNP at position 5. Mbovis.org. M. bovis spoligotype
1113 is more useful than the Rv204238 database. 2008 [cited 2012 Jun 29]. http://
mutation as a novel and distinct genetic www.mbovis.org/spoligodatabase/intro.
marker to identify M. orygis. htm

1708 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 LETTERS

Address for correspondence: Nicolaas C. Gey explanation for this discrepancy Address for correspondence: Brigid Lucey,
van Pittius, DST/NRF Centre of Excellence is that Cornelius et al. “did not Department of Medical Microbiology, Cork
for Biomedical TB Research/MRC Centre specifically exclude volunteers who University Hospital, Wilton, Cork, Ireland;
for Molecular and Cellular Biology/Division had had gastrointestinal disturbances email: brigid.lucey@cit.ie
of Molecular Biology and Human Genetics, in the 10 days before sampling,”
Stellenbosch University Faculty of Medicine Campylobacter can be shed in
and Health Sciences, PO Box 19063, Tygerberg feces for <4 weeks after infection.
7505, South Africa; email: ngvp@sun.ac.za Also, Cornelius et al. (1) noted the
possibility of “genetically distinct
but phenotypically indistinguishable In Response: In response to the
genomospecies differing in their letter by Bullman et al. (1), a major as-
pathogenic potential” to account pect of our study (2) was to compare
for the presence of the emerging epsilonproteobacterial populations in
pathogen C. concisus in healthy healthy persons and those who have
diarrhea. We have not examined as
Epsilonproteo- volunteers and patients with diarrheal
illness. This may also apply for C. many diarrheal samples as Bullman et
bacteria in ureolyticus. al. (3). However, in contrast with their
Humans, We reported a strong seasonal study, we have examined samples
from persons with no evident disease
New Zealand prevalence of C. ureolytcius and a
bimodal age distribution (2). The manifestations. Because the presence
To the Editor: Cornelius et lack of any related details from of an agent during disease is not proof
al. (1) addressed the potential of Cornelius et al. may undermine their of causation, we believed that a base-
Campylobacter ureolyticus as an reported detection rates. These factors line for comparison was needed. Cam-
emerging pathogen by conducting strongly suggest that the statement, pylobacter ureolyticus was found in
a molecular study on 128 diarrheal “these species are unlikely causes of a greater proportion of samples from
specimens and 49 fecal samples diarrhea,” should, at the very least, be healthy persons (24%) than samples
from healthy volunteers. Reporting taken under advisement. from persons who had diarrhea (11%)
the identification of C. ureolyticus in (p = 0.041, by χ2 test).
12 (24.5%) of 49 healthy volunteers, Susan Bullman, Samples from healthy persons
a number that they compared with Daniel Corcoran, were tested on 2 occasions: 18 sam-
our finding of 349 (23.8%) from James O’Leary, Deirdre Byrne, ples in September 2007 (New Zea-
Campylobacter spp.–positive samples Brigid Lucey, and Roy. D. Sleator land spring) at the Institute of En-
(2), the authors concluded that C. Author affiliations: Cork Institute of vironmental Science and Research,
ureolyticus species “are unlikely Technology, Cork, Ireland (S. Bullman, B. Christchurch, in the workplace, and
causes of diarrhea,” an assertion with Lucey, R.D. Sleator); and Cork University 31 samples in June 2009 (New Zea-
which we take issue. Hospital, Cork (D. Corcoran, J.O’Leary, D. land winter), at Christchurch Hospi-
This interpretation does not Byrne, B. Lucey) tal under the guidance of a clinician.
take into account that our screening DOI: http://dx.doi.org/10.3201/eid1810.120369
We have no reason to believe any of
involved 7,194 symptomatic patients: the workplace samples were provided
a sample size 40× greater than that of when volunteers had diarrhea, par-
Cornelius et al. In this context, the ticularly considering our workplace
likely carriage rate for C. ureolyticus References guidelines and staff characteristics.
is 1.15%. Also, our assay, which has In each testing round, 6 fecal samples
a limit of detection in the picomolar 1. Cornelius AJ, Chambers S, Aitken had positive test results for C. ureo-
J, Brandt SM, Horn B, On SL. lyticus. These periods equate to the
range, is likely comparable with, if Epsilonproteobacteria in humans,
not greater than, that of Cornelius et New Zealand. Emerg Infect Dis. peak and trough periods described by
al. (1). 2012;18:510–2. http://dx.doi.org/10.3201/ Bullman et al. (3). We were unable to
Accounting for variations in eid1803.110875 provide many details regarding sam-
2. Bullman S, Corcoran D, O’Leary J, pling in our paper because of space
geographic location and detection O’Hare D, Lucey B, Sleator RD. Emerging
methods, a detection rate of 24.5% in dynamics of human campylobacteriosis in constraints.
healthy volunteers (overall detection southern Ireland. FEMS Immunol Med Considering our baseline com-
rate 14.7%) is high in contrast to our Microbiol. 2011;63:248–53. http://dx.doi. parisons of healthy persons with those
org/10.1111/j.1574-695X.2011.00847.x who had diarrhea, we affirm our con-
reported rate of 1.15%. One possible

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1709
LETTERS

clusions are reasonable and that C. DOI: http://dx.doi.org/10.3201/eid1810.121122 5. Burgos-Portugal JA, Kaakoush NO, Raf-
ureolyticus is an unlikely cause of tery MJ, Mitchell HM. Pathogenic poten-
tial of Campylobacter ureolyticus. Infect
acute diarrheal disease. Similarly, C. References
Immun. 2012;80:883–90. http://dx.doi.
ureolyticus (previously classified as org/10.1128/IAI.06031-11
1. Bullman S, Corcoran D, O’Leary J, Byrne
Bacteroides ureolyticus) has been de- D, Lucey B, Sleator R. Epsilonproteobac-
6. Vandamme P, Debruyne L, De Brandt E,
tected in patients with Crohn’s disease Falsen E. Reclassification of Bacteroides
teria in humans, New Zealand [letter].
ureolyticus as Campylobacter ureolyti-
and in controls (4). However, different Emerg Infect Dis 2012;18:1709. http://
cus comb. nov., and emended description
dx.doi.org/10.3201/eid1810.120369
subtypes or undescribed subspecies of the genus Campylobacter. Int J Syst
2. Cornelius AJ, Chambers S, Aitken J,
may be pathogenic: some strains ex- Brandt SM, Horn B, On SLW. Epsi-
Evol Microbiol. 2010;60:2016–22. http://
hibit certain pathogenic characteristics dx.doi.org/10.1099/ijs.0.017152-0
lonproteobacteria in humans, New
in vitro (5) and others yield amplified Zealand. Emerg Infect Dis. 2012;18:
510–2. http://dx.doi.org/10.3201/eid1803.
fragment length polymorphism pro- Address for correspondence: Stephen L.W.
110875
files that are visually quite distinct 3. Bullman S, Corcoran D, O’Leary J, O’Hare On, Institute of Environmental Science and
from others (6). Host factors also can- D, Lucey B, Byrne D, et al. Emerging dy- Research, Food Safety Programme, 27 Creyke
not be discounted. namics of human campylobacteriosis in
southern Ireland. FEMS Immunol Med Road , Christchurch, Ilam, PO Box 29 181, New
Microbiol. 2011;63:248–53. http://dx.doi. Zealand; email: stephen.on@esr.cri.nz
Stephen L.W. On
org/10.1111/j.1574-695X.2011.00847.x
and Angela J. Cornelius 4. Zhang L, Man SM, Day AS, Leach ST,
Author affiliation: Institute of Environmental Lemberg DA, Dutt S, et al. Detection and
Science and Research, Christchurch, New isolation of Campylobacter species other
than C. jejuni from children with Crohn’s
Zealand
disease. J Clin Microbiol. 2009;47:453–5.

1710 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 ABOUT THE COVER

Mori Sosen (1747–1821) Monkey Performing the Sanbasō Dance (Dated 1800, the first day of the Monkey Year) Scroll painting,
ink on paper (49.5 cm × 115.6 cm) Pacific Asia Museum Collection, Pasadena, California, USA, Gift of Mr. and Mrs. Bruce Ross www.
pacificasiamuseum.org

Human minus Three Pieces of Hair

Polyxeni Potter

A t age 61, Mori Sosen changed the first character of

his name to one meaning “monkey.” So close had he
become to the subject of his paintings. To learn how to
coats and the movement of their muscles underneath. So
well did he depict the nature of monkeys that he was ac-
cused of being their descendent. He established, with his
paint the animals convincingly, he lived for a time in the brother Shūhō, a school of animal painting along the lines
mountains, their natural environment, not relying as oth- of the Maruyama-Shijō school in Kyoto. Shūhō’s son stud-
ers before him on copying their images from Chinese art. ied there under Maruyama Ōkyo, an expert in a style influ-
His mastery in depicting the Japanese macaque earned him enced by Western realism and direct observation. The Shijō
the title “undisputed master” from Dutch orientalist Robert school promoted synthesizing this modern development
van Gulik (1910–1967). In his book, Gibbon in China, van with the local trend toward the decorative and stylized.
Gulik translated Confucian scholar Kimura Kenkadō’s ac- Monkey Performing the Sanbasō Dance, on this
count of the animal’s arrival in Japan. “In the winter of the month’s cover, showcases both Mori Sosen’s favorite sub-
sixth year of the era (1809), a gibbon was shown in Osaka ject matter and his artistic style, a blend of realism and ex-
…. Although we have heard the word ‘gibbon’ since olden pressiveness. The action is set against a vacant background,
times and seen pictures of him, we never had seen a live the viewer drawn toward the main figure. Bold deliberate
specimen, and therefore a large crowd assembled to see this strokes outline the facial features, right hand and both feet
gibbon. Generally he resembled a large macaque, and fig- of the animal, and folds in the kimono. Smudged strokes
ure and fur are very similar.” Mori Sosen created a graphic from a dry brush draw the ruffled texture of the fur against
record of this sensational event. the black cap and smooth, mostly unpainted surface of the
Not much is known about his life but that he grew up clothing. The monkey, mouth pursed with concentration,
with and around artists, started his training with his father eyes fixed on some point outside the painting, holds a fan in
Mori Jokansai, and lived most of his life in Osaka. Mori one hand, and with the other, it raises a cluster of bells. The
Sosen’s legacy is his painting of animals, particularly right leg is lifted in a dance step, while the left, toes curled
monkeys, their personalities and attitudes as well as their inward for better balance, is rigid. During the Edo period
(1603–1868), Kabuki theater programs began at dawn with
Author affiliation: Centers for Disease Control and Prevention, At- a dance. In the final of three scenes in this dance, “the bell-
lanta, Georgia, USA tree,” the dancer would shake a wand covered with small
DOI: http://dx.doi.org/10.3201/eid1810.AC1810 bells. Along these lines, the monkey in Mori Sosen’s work

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1711
ABOUT THE COVER

lifts the bells in performing the Sanbasō, a dance celebrat- In this issue of the journal, a colony of Japanese ma-
ing the New Year, the first day of the Monkey Year. caques saw a mass die-off attributed to severe soil contami-
Macaques, more than 20 species of Macaca, a genus nation by Clostridium tetani in the facility maintaining the
of Old World monkeys mostly found in Asia, occupy a geo- animals. In China, Cryptosporidium spp., Giardia duode-
graphic range second only to that of humans in its extent. nalis, and Enterocytozoon bieneusi organisms were detect-
Their habitat varies from near desert to rainforest, from ed in free range rhesus monkeys in a popular public park.
sea level to snow-covered mountain tops. The Japanese Most genotypes and subtypes detected were anthroponotic,
macaque, also known as snow monkey, which has been which indicates that these animals, after becoming infected
reported at an elevation of 3,180 m, representing the north- from exposure to infected humans, may have become res-
ernmost nonhuman primate population in the world, has ervoirs for human cryptosporidiosis, giardiasis, and micro-
gained some notoriety for visiting a hot spring in Nagano sporidiosis. In Afghanistan, bites from macaques may have
to find comfort from the cold in winter. exposed US troops and presumably the Afghans to serious
Culturally, this monkey has been a metaphor, a polyse- infections, among them rabies, B virus, and tetanus. In Af-
mic symbol throughout Japanese history―now a mediator rica, nonhuman primates may be acting as a zoonotic res-
between gods and humans, now a scapegoat, now a clown. ervoir of P. vivax in regions where the human population is
Because of its unique role as similar to yet different from almost entirely refractory. If so, with human encroachment
humans, the Japanese macaque was used to define what it into nonhuman primate habitats, the chances of susceptible
means to be human and alternatively what it means to be a humans encountering the parasite will increase.
monkey: “human minus three pieces of hair,” to the Japa- As it lifts up the bells to ring in the New Year 1800,
nese. This definition satisfied both the affinity between hu- Mori Sosen’s beloved monkey in the flawless kimono
mans and monkeys and the animal’s local status just below continues the age-old dance celebrating our phylogenetic
grade. closeness. Because of this closeness, humans and nonhu-
The monkey’s role as mediator between gods and hu- man primates are susceptible to many of the same infec-
mans was long lived and well established. It implied posses- tions, minus three pieces of hair or not.
sion of supernatural powers, which were often expressed in
ritual dances with music. The monkey was believed a guard- Bibliography
ian that could cure disease in horses and secure good crops
1. Culleton RL, Mendes PE. Duffy phenotype and Plasmodium
as mediator between the Mountain Deity and humans. This vivax infections in humans and apes, Africa. Emerg Infect Dis.
status diminished gradually. The monkey was secularized 2012;18:1704–5.
and demoted, becoming the object of ridicule, a scapegoat, 2. Gardner MB, Luciw PA. Macaque models of human infectious dis-
for lacking (even if only by 3 pieces of hair) the essence of ease. ILAR J. 2008;49:220–55.
3. van Gulik RH. The gibbon in China. An essay in Chinese animal
humanness. Though a monkey dance performance still like- lore. Leiden (the Netherlands): EJ Brill; 1967.
ly showcases human superiority, the powerful metaphorical 4. Mease LE, Baker KA. Monkey bites among US military members,
presence persists, despite the animal’s virtual disappearance Afghanistan, 2011. Emerg Infect Dis. 2012;18:1647–9.
from everyday human contact outside the zoo. 5. Nature of the beast [cited 2012 Jun 26]. http://www.pacificasiamu-
seum.org/japanesepaintings/html/essay2.stm
Around the world, the status of macaques and their 6. Ohnuki-Tierney E. The monkey as self in Japanese culture. In: Cul-
connection with humans continues to evolve. The Japanese ture through time. Ohnuki-Tierney E., editor. Palo Alto (CA): Stan-
tradition that the monkey was a scapegoat for a human vic- ford University Press; 1991. p. 128–53.
tim of smallpox or of another disease, which persisted for 7. Poster AG. Japanese paintings of the Shijō school. New York: The
Brooklyn Museum; 1981.
centuries, is no longer held. In more recent times the ani- 8. Smith L, Harris V, Clark T. Japanese art: masterpieces in the British
mals have served instead as models for human disease, pro- Museum. London: The British Museum Press; 1990.
viding through their own infections or experimental studies, 9. Nakano T, Nakamura S, Yamamoto A, Takahashi M, Une Y. Teta-
insight into pathogenic mechanisms, treatments, and vac- nus as cause of mass die-off of captive Japanese macaques, Japan,
2008. Emerg Infect Dis. 2012;18:1633–5.http://dx.doi.org/10.3201/
cine approaches for human infectious agents, among them, eid1810.120503
hepatitis B, influenza virus, flaviviruses, Plasmodium spp. 10. Ye J, Xiao L, Ma J, Guo M, Liu L, Feng Y. Anthroponotic enter-
Some infections (HIV-2, P. knowlesii) have been transmit- ic parasites in monkeys in public park, China. Emerg Infect Dis.
ted from nonhuman primates to humans, suggesting that 2012;18:1640–3.
the role of these primates as “mediators” persists, but some,
Address for correspondence: Polyxeni Potter, EID Journal, Centers for
including measles and tuberculosis, can go both ways, with
Disease Control and Prevention, 1600 Clifton Rd NE, Mailstop D61,
infected humans compromising the health of nonhuman
Atlanta, GA 30333, USA; email: pmp1@cdc.gov
primates and because of the infections in the monkeys, an
employee health vaccination program was launched, poten-
tially preventing tetanus among workers.

1712 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 NEWS & NOTES

Upcoming Infectious
Disease Activities
October 22–26, 2012
4th ASM Conference on Beneficial
Microbes

Upcoming Issue
San Antonio, TX, USA
http://conferences.asm.org

Investigation of Outbreaks Complicated by Universal Exposure October 23–24, 2012

“Emerging Viruses: Disease
Seroprevalence of Chikungunya, Dengue, and Rift Valley Fever Models and Strategies for Vaccine
Viruses after Febrile Illness Outbreak, Madagascar Development”
A symposium in honor of
Sources of Dengue Importation to Areas at Risk for Epidemic CJ Peters, MD
Spread, Australia Galveston, TX, USA
http://www.utmb.edu/wrce
Risk for Travel-associated Legionnaires’ Disease, Europe, 2009
Phylogeography of Dengue Virus Serotype 4, Brazil, October 24–26, 2012
European Scientific Conference
2010–2011
on Applied Infectious Disease
Invasive Pneumococcal Disease and 7-Valent Pneumococcal Epidemiology (ESCAIDE)
Conjugate Vaccine, the Netherlands Edinburgh, Scotland, UK
http://www.escaide.eu
Infectious Disease Mortality Rates, Thailand, 1958–2009
October 27–31, 2012
Unchanged Severity of Influenza A(H1N1)pdm09 Infection APHA 140th Annual Meeting & Expo
in Children during First Postpandemic Season San Francisco, CA, USA
http://www.apha.org/meetings/
Antigenic Diversity of Enteroviruses and Nonpolio Acute AnnualMeeting
Flaccid Paralysis, India, 2007–2009
November 11–15, 2012
Litchi Harvest–associated Acute Encephalitis in Children, ASTMH 61st Annual Meeting
Northern Vietnam, 2004–2009 Atlanta Marriot Marquis
Atlanta, GA, USA
Lack of Evidence for Zoonotic Transmission of http://www.astmh.org
Schmallenberg Virus
December 1–7, 2012
Nasopharyngeal Bacterial Interactions in Children ASLM’s First International Conference
Cape Town, South Africa
Mycoplasmosis in Ferrets http://www.aslm.org
Coccidioidomycosis-associated Deaths, United States,
2013
1990–2008
September 5–10, 2013
Lack of Cross-protection against Bordetella holmesii after Options for the Control of Influenza VIII
Pertussis Vaccination Cape Town, South Africa
http://www.isirv.org
Epidemic Myalgia in Adults and Human Parechovirus Type 3
Infection, Yamagata, Japan, 2008 Announcements
To submit an announcement, send an email message to
EIDEditor (eideditor@cdc.gov). Include the date of the
HIV Infection and Geographically Bound Transmission of event, the location, the sponsoring organization(s), and
a website that readers may visit or a telephone number
Drug-Resistant Tuberculosis or email address that readers may contact for more
information.

Complete list of articles in the November issue at Announcements may be posted on the journal Web

http://www.cdc.gov/eid/upcoming.htm
page only, depending on the event date.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1713
 Earning CME Credit
To obtain credit, you should first read the journal article. After reading the article, you should be able to answer the
following, related, multiple-choice questions. To complete the questions (with a minimum 70% passing score) and earn
continuing medical education (CME) credit, please go to www.medscape.org/journal/eid. Credit cannot be obtained
for tests completed on paper, although you may use the worksheet below to keep a record of your answers. You must
be a registered user on Medscape.org. If you are not registered on Medscape.org, please click on the New Users: Free
Registration link on the left hand side of the website to register. Only one answer is correct for each question. Once you
successfully answer all post-test questions you will be able to view and/or print your certificate. For questions regarding
the content of this activity, contact the accredited provider, CME@medscape.net. For technical assistance, contact CME@
webmd.net. American Medical Association’s Physician’s Recognition Award (AMA PRA) credits are accepted in the US as
evidence of participation in CME activities. For further information on this award, please refer to http://www.ama-assn.org/
ama/pub/category/2922.html. The AMA has determined that physicians not licensed in the US who participate in this CME
activity are eligible for AMA PRA Category 1 Credits™. Through agreements that the AMA has made with agencies in some
countries, AMA PRA credit may be acceptable as evidence of participation in CME activities. If you are not licensed in the
US, please complete the questions online, print the certificate and present it to your national medical association for review.

Article Title:
Methicillin-Resistant Staphylococcus aureus
Sequence Type 239-III, Ohio, USA, 2007–2009
CME Questions
1. You are seeing a 65-year-old woman admitted with newly- 3. Which of the following statements regarding the treatment
diagnosed methicillin-resistant Staphylococcus aureus and prognosis of MRSA ST239-III in the current study is most
(MRSA) bacteremia. Which of the following strains of MRSA accurate?
are most common in the United States?
A. There was broad susceptibility to nearly all antimicrobials
A. USA 100 and MRSA ST239-III tested
B. USA 300 and MRSA ST239-III B. Rates of resistance to vancomycin and linezolid were
C. USA 100 and USA 300 approximately 90%
D. USA 600 and MRSA ST239-III C. Rates of treatment failure of MRSA ST239-III were
significantly higher compared with those associated with USA
2. As you evaluate this patient, what should you consider 100 and USA 300
regarding clinical characteristics of MRSA ST239-III in the D. Over 20% of cases of infection with MRSA ST239-III were fatal
current study?
4. Which of the following statements regarding molecular
A. Most cases of MRSA ST239-III were diagnosed in community
typing of MRSA ST239-III infections in the current study is
hospitals
most accurate?
B. Ninety-five percent of cases of MRSA ST239-III were
healthcare associated A. There was a single repetitive element PCR (rep-PCR) pattern
C. There were no cases of MRSA ST239-III associated with B. There were 9 different pulsed-field gel electrophoresis
implanted medical devices (PFGE) patterns
D. Nearly all cases of MRSA ST239-III were bloodstream C. Traditional PFGE testing could identify all of the bacteria
infections subtypes
D. The cluster groupings from PFGE, rep-PCR, and dru
methods were essentially identical

Activity Evaluation
1. The activity supported the learning objectives.
Strongly Disagree Strongly Agree
1 2 3 4 5
2. The material was organized clearly for learning to occur.
Strongly Disagree Strongly Agree
1 2 3 4 5
3. The content learned from this activity will impact my practice.
Strongly Disagree Strongly Agree
1 2 3 4 5
4. The activity was presented objectively and free of commercial bias.
Strongly Disagree Strongly Agree
1 2 3 4 5

1714 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012
 Earning CME Credit
To obtain credit, you should first read the journal article. After reading the article, you should be able to answer the
following, related, multiple-choice questions. To complete the questions (with a minimum 70% passing score) and earn
continuing medical education (CME) credit, please go to www.medscape.org/journal/eid. Credit cannot be obtained
for tests completed on paper, although you may use the worksheet below to keep a record of your answers. You must
be a registered user on Medscape.org. If you are not registered on Medscape.org, please click on the New Users: Free
Registration link on the left hand side of the website to register. Only one answer is correct for each question. Once you
successfully answer all post-test questions you will be able to view and/or print your certificate. For questions regarding
the content of this activity, contact the accredited provider, CME@medscape.net. For technical assistance, contact CME@
webmd.net. American Medical Association’s Physician’s Recognition Award (AMA PRA) credits are accepted in the US as
evidence of participation in CME activities. For further information on this award, please refer to http://www.ama-assn.org/
ama/pub/category/2922.html. The AMA has determined that physicians not licensed in the US who participate in this CME
activity are eligible for AMA PRA Category 1 Credits™. Through agreements that the AMA has made with agencies in some
countries, AMA PRA credit may be acceptable as evidence of participation in CME activities. If you are not licensed in the
US, please complete the questions online, print the certificate and present it to your national medical association for review.

Article Title:
Epidemiology of Foodborne Norovirus
Outbreaks, United States, 2001–2008
CME Questions
1. You are an infectious disease expert consulting to a US 2. Based on the study by Dr. Hall and colleagues, which of
public health office regarding prevention and reducing the the following statements about sources of US norovirus
impact of norovirus outbreaks. Based on the study by Dr. outbreaks is most likely correct?
Hall and colleagues, which of the following statements about
A. Ground meat was often implicated
general characteristics and outcomes of US foodborne
B. The most common single source was shellfish
norovirus outbreaks during 2001–2008 is most likely to
C. Most foods were likely contaminated during production and
appear in your report?
processing
A. On average, about 1 norovirus outbreak occurred every D. Contact with food handlers during preparation was cited in
week 82% of outbreaks as a possible contributor to contamination
B. The increasing trend that began in the 1990s in the number
of reported foodborne norovirus outbreaks continued through 3. Based on the study by Dr. Hall and colleagues, which of
2008 the following statements about recommended interventions
C. Norovirus outbreaks have been linked to an estimated to reduce the frequency and impacts of foodborne norovirus
average number of 10,324 illnesses, 1,247 health care outbreaks would most likely be correct?
provider visits, and 156 hospitalizations each year
A. Food shipping plants are the best target for intervention
D. No deaths have been attributed to norovirus
B. Food handlers preparing ready-to-eat foods should adhere to
handwashing and gloving recommendations and to ill worker
exclusion policies
C. A certified kitchen manager is unnecessary in most delis and
restaurants
D. Analytic methods to detect norovirus in foods are well
established

Activity Evaluation
1. The activity supported the learning objectives.
Strongly Disagree Strongly Agree
1 2 3 4 5
2. The material was organized clearly for learning to occur.
Strongly Disagree Strongly Agree
1 2 3 4 5
3. The content learned from this activity will impact my practice.
Strongly Disagree Strongly Agree
1 2 3 4 5
4. The activity was presented objectively and free of commercial bias.
Strongly Disagree Strongly Agree
1 2 3 4 5

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 10, October 2012 1715
 Emerging Infectious Diseases is a peer-reviewed journal established expressly to promote the recognition of new and
reemerging infectious diseases around the world and improve the understanding of factors involved in disease emergence, prevention, and elimination.
The journal is intended for professionals in infectious diseases and related sciences. We welcome contributions from infectious disease specialists in
academia, industry, clinical practice, and public health, as well as from specialists in economics, social sciences, and other disciplines. Manuscripts in all
categories should explain the contents in public health terms. For information on manuscript categories and suitability of proposed articles, see below and
visit http://wwwnc.cdc.gov/eid/pages/author-resource-center.htm.
Emerging Infectious Diseases is published in English. To expedite publication, we post some articles online ahead of print. Partial translations of the
journal are available in Japanese (print only), Chinese, French, and Spanish (http://wwwnc.cdc.gov/eid/pages/translations.htm).

Instructions to Authors Synopses. Articles should not exceed 3,500 words and 40 references. Use of sub-
headings in the main body of the text is recommended. Photographs and illustrations are
Manuscript Submission. To submit a manuscript, access Manuscript Central from encouraged. Provide a short abstract (150 words), 1-sentence summary, and biographical
the Emerging Infectious Diseases web page (www.cdc.gov/eid). Include a cover letter sketch. This section comprises concise reviews of infectious diseases or closely related
indicating the proposed category of the article (e.g., Research, Dispatch), verifying the topics. Preference is given to reviews of new and emerging diseases; however, timely
word and reference counts, and confirming that the final manuscript has been seen and updates of other diseases or topics are also welcome.
approved by all authors. Complete provided Authors Checklist.
Research. Articles should not exceed 3,500 words and 40 references. Use of sub-
Manuscript Preparation. For word processing, use MS Word. List the following infor- headings in the main body of the text is recommended. Photographs and illustrations are
mation in this order: title page, article summary line, keywords, abstract, text, acknowledg- encouraged. Provide a short abstract (150 words), 1-sentence summary, and biographical
ments, biographical sketch, references, tables, and figure legends. Appendix materials and sketch. Report laboratory and epidemiologic results within a public health perspective.
figures should be in separate files. Explain the value of the research in public health terms and place the findings in a larger
Title Page. Give complete information about each author (i.e., full name, graduate perspective (i.e., “Here is what we found, and here is what the findings mean”).
degree(s), affiliation, and the name of the institution in which the work was done). Clearly Policy and Historical Reviews. Articles should not exceed 3,500 words and 40 refer-
identify the corresponding author and provide that author’s mailing address (include ences. Use of subheadings in the main body of the text is recommended. Photographs
phone number, fax number, and email address). Include separate word counts for ab- and illustrations are encouraged. Provide a short abstract (150 words), 1-sentence sum-
stract and text. mary, and biographical sketch. Articles in this section include public health policy or his-
Keywords. Use terms as listed in the National Library of Medicine Medical torical reports that are based on research and analysis of emerging disease issues.
Subject Headings index (www.ncbi.nlm.nih.gov/mesh). Dispatches. Articles should be no more than 1,200 words and need not be divided
Text. Double-space everything, including the title page, abstract, references, tables, into sections. If subheadings are used, they should be general, e.g., “The Study” and
and figure legends. Indent paragraphs; leave no extra space between paragraphs. After “Conclusions.” Provide a brief abstract (50 words); references (not to exceed 15); figures
a period, leave only one space before beginning the next sentence. Use 12-point Times or illustrations (not to exceed 2); tables (not to exceed 2); and biographical sketch. Dis-
New Roman font and format with ragged right margins (left align). Italicize (rather than patches are updates on infectious disease trends and research that include descriptions
underline) scientific names when needed. of new methods for detecting, characterizing, or subtyping new or reemerging pathogens.
Developments in antimicrobial drugs, vaccines, or infectious disease prevention or elimi-
Biographical Sketch. Include a short biographical sketch of the first author—both nation programs are appropriate. Case reports are also welcome.
authors if only two. Include affiliations and the author’s primary research interests.
Photo Quiz. The photo quiz (1,200 words) highlights a person who made notable
References. Follow Uniform Requirements (www.icmje.org/index.html). Do not use contributions to public health and medicine. Provide a photo of the subject, a brief clue
endnotes for references. Place reference numbers in parentheses, not superscripts. to the person’s identity, and five possible answers, followed by an essay describing the
Number citations in order of appearance (including in text, figures, and tables). Cite per- person’s life and his or her significance to public health, science, and infectious disease.
sonal communications, unpublished data, and manuscripts in preparation or submitted for
publication in parentheses in text. Consult List of Journals Indexed in Index Medicus for Commentaries. Thoughtful discussions (500–1,000 words) of current topics. Com-
accepted journal abbreviations; if a journal is not listed, spell out the journal title. List the mentaries may contain references but no abstract, figures, or tables. Include biographical
first six authors followed by “et al.” Do not cite references in the abstract. sketch.

Tables. Provide tables within the manuscript file, not as separate files. Use the MS Etymologia. Etymologia (100 words, 5 references). We welcome thoroughly re-
Word table tool, no columns, tabs, spaces, or other programs. Footnote any use of bold- searched derivations of emerging disease terms. Historical and other context could be
face. Tables should be no wider than 17 cm. Condense or divide larger tables. Extensive included.
tables may be made available online only. Another Dimension. Thoughtful essays, short stories, or poems on philosophical is-
Figures. Submit figures in black and white. If you wish to have color figures online, sues related to science, medical practice, and human health. Topics may include science
submit both in black and white and in color with corresponding legends. Submit edit- and the human condition, the unanticipated side of epidemic investigations, or how people
able figures as separate files (e.g., Microsoft Excel, PowerPoint). Photographs should perceive and cope with infection and illness. This section is intended to evoke compassion
be submitted as high-resolution (600 dpi) .jpeg or .tif files. Do not embed figures in the for human suffering and to expand the science reader’s literary scope. Manuscripts are
manuscript file. Use Arial 10 pt. or 12 pt. font for lettering so that figures, symbols, letter- selected for publication as much for their content (the experiences they describe) as for
ing, and numbering can remain legible when reduced to print size. Place figure keys within their literary merit. Include biographical sketch.
the figure. Figure legends should be placed at the end of the manuscript file. Letters. Letters commenting on recent articles as well as letters reporting cases, out-
Videos. Submit as AVI, MOV, MPG, MPEG, or WMV. Videos should not exceed 5 breaks, or original research, are welcome. Letters commenting on articles should contain
minutes and should include an audio description and complete captioning. If audio is no more than 300 words and 5 references; they are more likely to be published if submit-
not available, provide a description of the action in the video as a separate Word file. ted within 4 weeks of the original article’s publication. Letters reporting cases, outbreaks, or
Published or copyrighted material (e.g., music) is discouraged and must be accompanied original research should contain no more than 800 words and 10 references. They may have
by written release. If video is part of a manuscript, files must be uploaded with manu- 1 figure or table and should not be divided into sections. No biographical sketch is needed.
script submission. When uploading, choose “Video” file. Include a brief video legend in Books, Other Media. Reviews (250–500 words) of new books or other media on
the manuscript file. emerging disease issues are welcome. Title, author(s), publisher, number of pages, and
other pertinent details should be included.

Types of Articles
Conference Summaries. Summaries of emerging infectious disease conference ac-
tivities (500–1,000 words) are published online only. They should be submitted no later
Perspectives. Articles should not exceed 3,500 words and 40 references. Use of than 6 months after the conference and focus on content rather than process. Provide
subheadings in the main body of the text is recommended. Photographs and illustra- illustrations, references, and links to full reports of conference activities.
tions are encouraged. Provide a short abstract (150 words), 1-sentence summary, and Online Reports. Reports on consensus group meetings, workshops, and other activi-
biographical sketch. Articles should provide insightful analysis and commentary about ties in which suggestions for diagnostic, treatment, or reporting methods related to infec-
new and reemerging infectious diseases and related issues. Perspectives may address tious disease topics are formulated may be published online only. These should not exceed
factors known to influence the emergence of diseases, including microbial adaptation and 3,500 words and should be authored by the group. We do not publish official guidelines or
change, human demographics and behavior, technology and industry, economic devel- policy recommendations.
opment and land use, international travel and commerce, and the breakdown of public
health measures. Announcements. We welcome brief announcements of timely events of interest to
our readers. Announcements may be posted online only, depending on the event date.
Email to eideditor@cdc.gov.