See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/50375086 Evaluation of an autofluorescence based imaging system (VELscope (TM)) in the detection of oral potentially... Article in Oral Oncology · March 2011 DOI: 10.1016/j.oraloncology.2011.02.001 · Source: PubMed CITATIONS READS 68 336 3 authors: Kamran H. Awan Peter R Morgan 36 PUBLICATIONS 176 CITATIONS 103 PUBLICATIONS 2,421 CITATIONS King Saud University SEE PROFILE King's College London SEE PROFILE Saman Warnakulasuriya King's College London 312 PUBLICATIONS 10,674 CITATIONS SEE PROFILE All content following this page was uploaded by Kamran H. Awan on 16 December 2013. The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document and are linked to publications on ResearchGate, letting you access and read them immediately. Oral Oncology 47 (2011) 274–277 Contents lists available at ScienceDirect Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology Evaluation of an autofluorescence based imaging system (VELscope™) in the detection of oral potentially malignant disorders and benign keratoses K.H. Awan a, P.R. Morgan b, S. Warnakulasuriya a,c,⇑ a Oral Medicine, Department of Clinical & Diagnostic Sciences, King’s College London Dental Institute, United Kingdom Head & Neck Pathology, Department of Clinical & Diagnostic Sciences, King’s College London Dental Institute, United Kingdom c WHO Collaborating Centre for Oral Cancer, United Kingdom b a r t i c l e i n f o Article history: Received 24 November 2010 Received in revised form 28 January 2011 Accepted 1 February 2011 Keywords: Autofluorescence Oral potentially malignant disorders Sensitivity and specificity VELscope Early detection s u m m a r y Early detection of oral cancer is crucial in improving survival rate. Identification and detection of oral potentially malignant disorders (OPMD) allow delivery of interventions to reduce the evolution of these disorders to malignancy. A variety of new and emerging diagnostic aids and adjunctive techniques are currently available to potentially assist in the detection of OPMD. The objective of the present study was to evaluate the accuracy of autofluorescence against conventional oral examination and surgical biopsy. A total of 126 patients, 70 males and 56 females (mean age 58.5 ± 11.9 years) who presented to the Oral Medicine Clinics at King’s and Guy’s Hospitals, London with oral white and red patches suspicious of OPMD were enrolled. Following a complete visual and autofluorescence examination, all underwent an incisional biopsy for histopathological assessment. Seventy patients had oral leukoplakia/erythroplakia, 32 had oral lichen planus, 9 chronic hyperplastic candidiasis and rest frictional keratosis (13) or oral submucous fibrosis (2). Of 126 lesions, 105 (83%) showed loss of fluorescence. Following biopsy 44 had oral epithelial dysplasia (29 mild, 8 moderate and 7 severe). The sensitivity (se) and specificity (sp) of autofluorescence for the detection of a dysplastic lesion was 84.1% and 15.3% respectively. While VELscope was useful in confirming the presence of oral leukoplakia and erythroplakia and other oral mucosal disorders, the device was unable to discriminate high-risk from low-risk lesions. Ó 2011 Elsevier Ltd. All rights reserved. Introduction Oral cancer is a growing problem in many European countries including the United Kingdom.1 Delays in diagnosis are frequently reported2 either due to poor symptom recognition3 or missed diagnosis.4 The UK guidelines for the early diagnosis of Head & Neck cancers were published in 2005 setting out criteria for urgent referral for suspicious lesions5 and the British Dental Association and FDI recommend that systematic visual screening examination should be carried out on every patient at the beginning of a new course of treatment. While detection of asymptomatic cancers could be a problem in dental practices due to poor attendance of high-risk patients6, oral potentially malignant disorders7 (OPMD) provide a long preclinical phase during which high-risk patients could be identified to provide interventions. ⇑ Corresponding author. Address: Department of Oral Medicine, King’s College Hospital, Bessemer Rd., London SE5 9RS, United Kingdom. Tel.: +44 20 3 299 2430, fax: +44 20 3 299 3426. E-mail address: s.warne@kcl.ac.uk (S. Warnakulasuriya). 1368-8375/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.oraloncology.2011.02.001 Opportunistic screening by visual clinical examination at dental practices will identify OPMD and other mucosal disorders with similar clinical presentations.8 It is estimated that up to 15% of the population have oral mucosal diseases at any one time, but only very few have the characteristics of OPMD.9 It is therefore a problem for practitioners to identify and refer OPMD with confidence. Several chair-side adjunctive aids have been developed to help practitioners with oral cancer screening with the aim of diagnosing high-risk lesions. None of these have been tested adequately in primary care settings.10 Autofluorescence is one potential technique that may be used to facilitate the visualisation and management of oral cancer and OPMD. As early as in 1924, it was observed that the autofluorescence of tissues could potentially be used for cancer detection.11 Autofluorescence works on the principle that certain biofluorophores present within the tissue become fluorescent on excitation with a suitable wavelength (400–460 nm) light source. However, diseased tissues lose fluorescence (fluorescence visualisation loss – FVL) due to disruption in the distribution of these biofluorophores, and appear darker in colour. 275 K.H. Awan et al. / Oral Oncology 47 (2011) 274–277 The aim of this study was to evaluate the accuracy of autofluorescence examination in its ability to delineate high-risk oral mucosal lesions from other lesions already diagnosed by a specialist, to allow estimates of sensitivity and specificity of the technique. Table 1 Patient characteristics. All Materials and methods One hundred and sixty-four consecutive patients aged over 16 years presenting in oral medicine clinics at two London Hospitals with white, red and mixed white and red patches were invited to participate in the study. One hundred and twenty-six patients (76.8%) consented and were investigated by a standard protocol that involved clinical visual examination and autofluorescence examination followed by biopsy. The study was approved by Institutional Research and Ethics Committees (08/H0808/20). Following a comprehensive clinical examination under an incandescent light source the clinical diagnosis was established by the operator (KHA) and validated by a second experienced examiner (SW). The principal area (site) of morphologically altered mucosa was selected excluding any ulcerated areas (by consensus of both examiners) and photographed. All further investigations were performed on this clinically detected area of mucosal abnormality. Autofluorescence examination was performed using the VELscope™ (Visually Enhance Lesion Scope) under dimmed room light, with protective eye wear worn by the patient throughout the procedure. The possible outcome of the autofluorescence examination was determined by the manufacturer’s literature i.e. FVL – fluorescence visualization loss, FVR – fluorescence visualization retained and FVI – fluorescence visualization increased. Both examiners were calibrated by an experienced professional from the LED Diagnostics (the manufacturer). A surgical biopsy was performed for histopathological assessment and the selection of the biopsy site took into consideration any area of FVL identified by the VELscope within the lesion. The presence or absence of dysplasia in the biopsy specimen was recorded by an experienced oral pathologist (PRM). Data collected was entered through the IBM SPSS 18 (Statistical Package for the Social Sciences). Sensitivity and specificity of the autofluorescence test results, compared to clinical diagnosis by a specialist and dysplasia grade from biopsy, were calculated. Differences and associations between the autofluorescence test and dysplasia grade were examined using either Fisher’s exact test or v2 test with significance set at P < 0.05. All tests were two-sided. A receiver operating characteristic (ROC) curve was used to estimate the diagnostic value of the test. Results The profile of 126 patients enrolled in this study is given in Table 1. Of 126 lesions, more than half (n = 70) were clinically diagnosed as either leukoplakia or erythroplakia. Thirteen lesions were clinically diagnosed as frictional keratoses and 32 as oral lichen planus or lichenoid reaction. The remaining lesions consisted of 9 chronic hyperplastic candidiasis and 2 oral submucous fibrosis. One hundred and sixteen patients underwent surgical biopsy from which oral epithelial dysplasia was confirmed in 44 patients. Autofluorescence examination was performed on all 126 patients. One hundred and five lesions showed FVL (83.3%) whereas 16 retained the fluorescence (12.7%) and appeared apple green in colour. Three of the lesions showed increased fluorescence and 2 had a mixed result showing both loss and increased fluorescence in different areas within the lesion. Of 105 FVL cases, more than 50% (n = 53) showed complete loss of fluorescence whereas 29 * ** Leuko/erythroplakia Dysplasia** n = 126 % n = 70 % n = 44 % Gender Male Female 70 56 55.6 44.4 46 24 65.7 34.3 26 18 59.1 40.9 Ethnicity White Non-white* 76 50 60.3 39.7 51 19 72.8 27.2 28 16 63.6 36.4 Tobacco history Current smokers Ex-smokers Never smoked 61 28 37 48.4 22.2 29.4 41 15 14 58.6 21.4 20.0 24 13 7 54.5 29.5 15.9 Alcohol history Current users Ex-users Never used 92 8 26 73.0 6.4 20.6 57 3 10 81.4 4.3 14.3 33 5 6 75.0 11.4 13.6 Lesion site Buccal mucosa Tongue Floor of mouth Palate Alveolar ridge 54 40 14 11 7 42.9 31.7 11.1 8.7 5.5 21 21 11 11 6 30.0 30.0 15.7 15.7 8.6 13 17 8 4 2 29.5 38.6 18.2 9.1 4.5 Non-white included 26 Asians, 22 Afro-Caribbean and 2 of mixed ethnicity. Mild, 29; moderate, 8; severe, 7. showed partial loss of fluorescence. In 23 lesions, FVL extended beyond the clinically evident oral lesion. Leuko/erythroplakia vs. other group Of 70 leuko/erythroplakia cases, 61 (87.1%) showed FVL whereas only 9 (12.9%) that appeared clinically white had a negative test result (Table 2). All 9 cases of erythroplakias showed FVL. In the case of other oral diagnostic categories, 44 (78.6%) showed FVL with the remaining 12 (21.4%) showing a negative test outcome. In particular, among the 13 frictional keratosis cases, 9 (69.2%) showed FVL. Autofluorescence examination showed a sensitivity and specificity of 87.1% and 21.4%, respectively. Positivity of autofluorescence (FVL) for the leuko/erythroplakia group was not significantly different compared with the other group (v2 = 1.65, P = 0.23). ROC curve for autofluorescence as a tool to detect leuko/erythroplakias showed a poor diagnostic value (AUC = 0.52, 95% CI: 0.42–0.62, P = 0.72). Dysplasia group Autofluorescence showed a sensitivity of 84.1% as 37 out of 44 lesions with dysplasia recorded as FVL in contrast to only 7 which did not show any FVL. Among the 7 FVL negative dysplasias, 5 were graded mild and 2 as moderate. But the autofluorescence was not highly specific for dysplastic oral lesions as FVL was observed in 61 (84.7%) of the non-dysplastic oral lesions, leading to a low specificity (15.3%). No significant difference was noted among the dysplasia group in relation to the autofluorescence test results (v2 = 0.00, P = 1.00). ROC curve for autofluorescence as a tool for the detection of dysplasia group also showed a poor diagnostic value (AUC = 0.49, 95% CI: 0.39–0.61, P = 0.96). Discussion Five-year survival rates for oral cancer have not changed for several decades. Poor survival is at least in part due to the failure in early detection of OPMD and oral cancers. To this end improving diagnostic abilities of primary care dentists/physicians and also 276 K.H. Awan et al. / Oral Oncology 47 (2011) 274–277 Table 2 Autofluorescence in relation to leuko/erythroplakia and dysplasia. Diagnosis Leuko/erythroplakia* Leukoplakia Erythroplakia Others Dysplasia** Non-dysplastic * ** Cases Autofluorescence (n) FVL FVR 70 61 9 56 44 72 61 52 9 44 37 61 9 9 0 12 7 11 Se Sp PPV NPV 87.1 21.4 58.1 57.1 84.1 15.3 37.8 61.1 Based on WHO7 criteria confirmed by a specialist. Based on WHO12 criteria included mild, moderate and severe cases. Table 3 Characteristics of VELscope system observed in the present study. Strengths Limitations 1. 2. 3. 4. 5. 6. 1. 2. 3. 4. Simple to use and non-invasive No consumable reagents needed/no recurrent cost Provides real time results Can be performed by a wide range of operators after short training Limited operator variability High sensitivity for any oral mucosal disorder facilitating less interventional investigations in secondary care units remain important cornerstones in the research agenda. We investigated the utility of autofluorescence as a diagnostic test to evaluate its accuracy in the detection of oral leuko/erythroplakia and oral epithelial dysplasia. FVL was observed in the majority (87.1%) of the clinically diagnosed cases of leuko/erythroplakias, lesions that carry a relatively higher risk of malignant transformation compared to other OPMD. More interestingly, FVL was positive in all 9 cases of erythroplakia giving a sensitivity of 100%. FVL was also observed in the majority of the cases (84.1%) that were histopathologically diagnosed as dysplasia and notably the VELscope detected all severe dysplasia cases (n = 9). These results notably demonstrate the ability of the technique to detect high-risk lesions. However, it was disappointing to note that autofluorescence examination was positive in majority of the other white/red lesions that to a non-specialist could resemble leuko/erythroplakia. This finding re-affirms the lack of specificity of the technique for the detection of leuko/erythroplakia. In addition, VELscope was also unable to detect 7 dysplasias (5 mild, 2 moderate), thus further undermining the utility of the device, if the objective is to pick all dysplasia cases. Limitations of autofluorescence in discriminating between dysplasia and nondysplasia cases have been reported in a recent study13 where 6 cases of dysplasia were not detected by the VELscope. Comparison of the results of the present study with published data proved to be difficult due to limited number of studies in the literature reporting sensitivity and specificity of the device. Only one previous study appears to have employed autofluorescence in a systemic examination on a cohort of patients. The study was conducted at the British Columbia Cancer Agency (BCCA) where a prototype of the VELscope was investigated by the group.14 Using the blue-excitation light, 50 lesions were examined which included 33 oral cancers, 11 severe dysplasia and carcinoma-in-situ and 6 with no oral mucosal lesions. The authors reported a sensitivity of 98% and specificity of 100% against the gold standard (histology). Our data show a low specificity (15.3%) for the technique. We were able to demonstrate this by the inclusion of several benign disorders, thus reducing the ‘spectrum bias’ encountered in published studies – a desirable feature of our study. Two other studies on VELscope reported contrasting results on its utility. Huber et al.15 reported that VELscope failed to detect Needs a dark environment High initial setting-up cost No permanent record unless photographed Low specificity for dysplasia, high referral rate and over-treatment any additional suspicious lesions not identified by conventional oral examination, and Huff et al.16 reported an increase in prevalence of mucosal disorders in a second cohort subjected to VELscope, compared with an earlier cohort examined visually only. Their research was seriously flawed as they did not consider alternative possible reasons for a true increased prevalence of disorders in the later cohort. As sufficient studies had not examined sensitivity and specificity of the VELscope system, our data need to be discussed against the backdrop of sensitivity and specificity reported for clinical visual screening. Moles et al.17 in a meta-analysis of 7 screening studies determined the sensitivity and specificity values of clinical screening in the range of 0.60–0.95 and 0.81–0.94, respectively. It is clear that adjunctive use of the VELscope does not provide added value to visual examination by trained operators.18 This study should not be seen as a screening study as our data are specific to a hospital population referred following the detection of a range of mucosal abnormalities by primary care practitioners. So far no studies have been reported for evaluating VELscope for screening the population. Balevi19 highlighted that the adoption of VELscope as a routine cancer-screening device is premature. Screening for OPMD using this test system warrants further investigation, but ethical aspects of further investigation of an area with FVL detected by VELscope only, in the absence of a clinically evident lesion would need careful consideration. The limitations of the current study include (a) not exploring the differences between the complete vs. partial FVL cases, and (b) lack of refined methodology dictating the biopsy site with reference to large and heterogeneous lesions and VELscope data. Based on our experience of undertaking the largest study so far on the use of autofluorescence on a series of OPMD, Table 3 summarizes the benefits and limitations of the adjunctive tool. Conclusion In conclusion, our study demonstrated a relatively high sensitivity (84%) and a low specificity (15%) in discriminating high-risk (dysplasias) from benign lesions. Further well designed studies are needed to examine the role of VELscope as an oral examination system in primary care. K.H. Awan et al. / Oral Oncology 47 (2011) 274–277 Conflict of interest statement We thank Dr. Connie Yang for assistance in setting up the data entry system and Dr. Derek Cooper for the data analysis. VELscope system for the study was supplied by LED Diagnostics. References 1. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010;127:2893–917. 2. Gómez I, Warnakulasuriya S, Varela-Centelles PI, López-Jornet P, SuárezCunqueiro M, Diz-Dios P, et al. Is early diagnosis of oral cancer a feasible objective? Who is to blame for diagnostic delay? Oral Dis 2010;16:333–42. 3. Grant E, Silver K, Bauld L, Day R, Warnakulasuriya S. The experiences of young oral cancer patients in Scotland: symptom recognition and delays in seeking professional help. Br Dent J 2010;208:465–71. 4. Lovas JG, Daley TD, Kaugars GE, Wright JM. Errors in the diagnosis of oral malignancies. J Can Dent Assoc 1993;59:935–8. 5. Available from: <http://www.nice.org.uk/nicemedia/pdf/CG027publicinfo.pdf#>. 6. Netuveli G, Sheiham A, Watt RG. Does the ‘inverse screening law’ apply to oral cancer screening and regular dental check-ups? J Med Screen 2006;13:47–50. 7. Warnakulasuriya S, Johnson NW, van der Waal I. Nomenclature and classification of potentially malignant disorders of the oral mucosa. J Oral Pathol Med 2007;36:575–80. 8. Lim K, Moles DR, Downer MC, Speight PM. Opportunistic screening for oral cancer and precancer in general dental practice: results of a demonstration study. Br Dent J 2003;194:497–502. View publication stats 277 9. Scully C, Bagan JV, Hopper C, Epstein JB. Oral cancer: current and future diagnostic techniques. Am J Dent 2008;21:199–209. 10. Patton LL, Epstein JB, Kerr AR. Adjunctive techniques for oral cancer examination and lesion diagnosis: a systematic review of the literature. J Am Dent Assoc 2008;139:896–905. 11. Policard A. Etude sur les aspects offerts par des tumeurs expérimentales examinées a la lumie‘re de Wood. C R Soc Biol (Paris) 1924;91:1423–4. 12. World Health Organization. World Health Organization Classification of Tumours. In: Barnes L, Eveson JW, Reichart P, Sidransky D, editors. Pathology & genetics. Head and neck tumours. Lyon: International Agency for Research on Cancer (IARC) IARC Press; 2005, p. 177–79. 13. Mehrotra R, Singh M, Thomas S, Nair P, Pandya S, Nigam NS, et al. A crosssectional study evaluating chemiluminescence and autofluorescence in the detection of clinically innocuous precancerous and cancerous oral lesions. J Am Dent Assoc 2010;141:151–6. 14. Lane PM, Gilhuly T, Whitehead P, Zeng H, Poh CF, Ng S, et al. Simple device for the direct visualization of oral-cavity tissue fluorescence. J Biomed Opt 2006;11:024006. 15. Huber MA. Assessment of the VELscope as an adjunctive examination tool. Tex Dent J 2009;126:528–35. 16. Huff K, Stark PC, Solomon LW. Sensitivity of direct tissue fluorescence visualization in screening for oral premalignant lesions in general practice. Gen Dent 2009;57:34–8. 17. Moles DR, Downer MC, Speight PM. Meta-analysis of measures of performance reported in oral cancer and precancer screening studies. Brit Dent J 2002;192:340–4. 18. Downer MC, Moles DR, Palmer S, Speight PM. A systematic review of test performance in screening for oral cancer and precancer. Oral Oncol 2004;40:264–73. 19. Balevi B. Evidence-based decision making: Should the general dentist adopt the use of the VELscope for routine screening for oral cancer? JCDA 2007;73:603–6.