Letters to the Editor
Surgical Removal of Subfoveal distribution charts provides some striking implications.
Neovasculari zation In Figure 1, the untreated subfoveal CNV control group
(MPS untreated controls) and the surgically treated group
bear remarkably similar distributions at similar study time
Dear Editor: points (3-6 months), suggesting similar outcome of visual
We have been tremendously interested in the ongoing acuity distributions. Repeated measures (visual acuity
discussion of surgery for subfoveal choroidal neovascular data) from the MPS subfoveal CNV untreated group could
(CNV) removal for age-related macular degeneration. -
1 3
be compared with repeated measures of surgically treated
Control groups have not been included in any of these populations (if available) to confirm or refute this simi-
studies; therefore, we thought a reassessment based on the larity. In our experience, these patient groups (20/40-20/
most comparable prospective control group might be il- 320) represent a highly motivated population with rapidly
lustrative. deteriorating visual function willing to undergo therapies
As with most studies of this design, an ideal control of any benefit. It remains the responsibility of the
population does not exist. However, to draw conclusions, ophthalmologist to provide an opinion regarding options,
the need for a comparison of the disease's natural history including surgical risks, despite limitations in data avail-
is obvious. In our estimation, the Macular Photocoagu- able.
lation Study4 (MPS) on subfoveallesions provides an un- Of additional interest and possible speculation are those
treated population that represents the most thoroughly patients with subfoveal CNV undergoing surgical CNV
studied and most closely related group for comparison. removal whose visual acuities at initial evaluation were
1
Authors of all three surgical CNV studies - 3 alluded to outside the MPS subfoveal CNV entry criteria visual acu-
differences in entry criteria between the MPS and their ity (better than 20/40 or worse than 20/320). We are un-
patient series, yet none provided full disclosure of their aware of an adequate control group that had initial poor
entry criteria as did the MPS. 5 ·6 Surgically treated patients' visual acuities (20/400 and worse). Distributionally, a
visual acuities were converted to Bailey-Lovie acuity scale subgroup of these patients' measured visual acuities at an
(log Mar) based on prior calculated correlation (R average follow-up of6 months was 20/100 to 20/160 (8%)
= 0.94). Patients' published visual acuity values were
7 and 20/200 to 20/320 ( 19% ). Possible explanations for
8
categorized according to the MPS subfoveal classification this effect have been proposed and include therapeutic
groups of (1) 20/40 to 20/80, (2) 20/100 to 20/160, (3) benefit, measured deviations to the mean, spontaneous
20/200 to 20/320, (4) 20/400 to 20/640, and (5) 20/800 CNV regression, spontaneous decreased CNV exudation,
or worse 4 (see Table 1). Surgically treated patients with biases in visual acuity measurement of known treated pa-
inadequate entry data to be categorized (i.e., counting fin- tients, learned improvement, and/or various other biases
gers vision) were not included. Because of the small patient inherent in nonrandomized, unmasked studies. Unfor-
numbers published to date, adequate statistical compar- tunately, variances between studies and lack of repeated
isons could not be made (chi-square tests were determined measures in the surgically treated patients prevent expla-
to be inconclusive). However, inspection of frequency nation of this effect.
Table 1. Macular Degeneration: Surgical Choroidal Neovascular Removal
Thomas Lambert Berger
et aP et aP et aF Subtotal* Total
Visual Acuity Pre Post Pre Post Pre Post Pre Post Pre Post
20/20-20/30 1 1 1 1
20/40-20/80 3 3 3 0
20/100-20/160 1 4 1 1 2 1 5
20/200-20/320 11 14 2 1 3 1 16 9 16 16
20/400-20/640 13 9 2 2 7 8 6 22 19
20/800 or
worse 4 5 9 10 2 13 15
• Patients who meet Macular Photocoagulation Study entry visual acuity criteria.
795
Ophthalmology Volume 100, Number 6, June 1993
100
7. Lovie-Kitchin JE. Validity and reliability of visual acuity
90 Entry measurements. Ophthal Physiol Opt 1988;8:363-70.
c. 80 8. Abrams GW, 975-6. Discussion of: Berger AS, Kaplan HJ.
"'~ 70 Clinical experience with the surgical removal of subfoveal
"0
(I)
60
50
neovascular membranes: short-term postoperative results.
Ophthalmology 1992;99:967-76.
c"'"'
(I)
40
0 30
G;
a. 20 Author's reply
10
Dear Editor:
100 20/20 -20/80 20/100·20/160 20/ 200-20/320 20/400 -20/640 ~20/ 800 I appreciate the interest Drs. Russell, Crapotta, and Zer-
bolio have shown in the published results of surgical re-
90
3-6 mo. after entry moval of subfoveal choroidal neovascularization (CNV)
c. 80
"'
0
(5 70
for age-related macular degeneration (AMD) and their
60
efforts in identifying an appropriate control group. As
0
(I)
50
these authors indicate, there is no fully comparable set of
c"'"' 40
patients. They compare surgical results with the untreated
(I)
0 30
eyes with subfoveal lesions in the Macular Photocoagu-
G;
a. 20
lation Study (MPS). These results were reported 1•2 as the
10
Foveal Photocoagulation Study (FPS). The majority of
0
our surgical patients would not have met the criteria for
20/20-20/80 20/ 100 -20/160 20/200-20/320 20/400 -20{640 ~ 20/8 0 0 enrollment in the FPS, and thus this comparison has lim-
ited value.
Figure 1. Visual acuity distributions. Solid bar = MPS untreated subfoveal
I agree with the basis of their letter: Surgical removal
patients. Dotted bar = MPS "eligible" surgically treatd CNV.
of subfoveal CNV in AMD has not been proven to be
beneficial, and this technique should not be widely applied
unless it is proven effective by a randomized, prospective
Given the absence of demonstrated surgical effective- trial. Russell et al very appropriately express concern about
ness, we are concerned that subfoveal CNV removal may this technique in a group of patients who may be desperate
be occurring outside institutional review board-approved enough to "try anything."
study protocols. We are concerned that line item presen- We are attempting to address the concerns highlighted
tation of these data without tabulation, graphic, or statis- by these authors. We have submitted for review by skilled
tical validation of effectiveness may be construed as con- angiogram readers who also staff the MPS Fundus Pho-
doning surgical removal of subfoveal CNV in age-related tograph Reading Center all of the cases of subfoveal sur-
macular degeneration. gery performed by us during the past 3 years. Only 12 of
STEPHEN R. RUSSELL, MD our 39 AMD cases were judged by the photograph readers
JOSEPH A. CRAPOTTA, MD as eligible for the FPS. We are in the process of categorizing
DOMINIC J. ZERBOLIO, JR., PHD and establishing appropriate features on which to describe
St. Louis, Missouri the cases and anatomic outcomes. A comparison of out-
comes in surgical cases with outcomes in similar untreated
References cases has been initiated. However, the results of these
I. Thomas MA, Grand MG, Williams DF, et a!. Surgical comparisons are not yet available to share with the readers
management of subfoveal choroidal neovascularization. of Ophthalmology.
Ophthalmology 1992;99:952-68. We remain interested in the potential benefit of
2. Berger AS, Kaplan HJ. Clinical experience with the surgical subfoveal surgery in selected patients, including choroidal
removal of subfoveal neovascular membranes: short-term neovascularization in the presumed ocular histoplasmosis
postoperative results. Ophthalmology 1992;99:969-76. syndrome. We recognize that to prove efficacy of the
3. Lambert HM, Capone A, Aaberg TM, eta!. Surgical excision technique in any subset of patients, a randomized, pro-
of subfoveal neovascular membranes in age-related macular spective clinical trial will be required. We currently are
degeneration. Am J Ophthalmol 1992;113:257-62. engaged in exploring the feasibility of such a trial.
4. Macular Photocoagulation Study Group. Laser photoco-
MATTHEW A. THOMAS, MD
agulation of subfoveal neovascular lesions in age-related
macular degeneration. Results of a randomized clinical trial. St. Louis, Missouri
Arch Ophthalmol 1991; 109:1220-31.
5. Macular Photocoagulation Study Group. Subfoveal neo-
References
vascular lesions in age-related macular degeneration. Arch
Ophthalmol1991;109:1242-57. I. Macular Photocoagulation Study Group. Laser photoco-
6. Blackhurst DW, Maguire MG, The Macular Photocoagu- agulation of subfoveal neovascular lesions in age-related
lation Study Group. Reproducibility of refraction and visual macular degeneration. Results of a randomized clinical trial.
acuity measurement under a standard protocol. Retina Arch Ophthalmol 1991; 109: 1220-31.
1989;9: 163-9. 2. Macular Photocoagulation Study Group. Laser photoco-
796
