THE JULES FRANÇOIS MEMORIAL LECTURE *

THE COLLABORATIVE OCULAR MELANOMA

STUDY AND MANAGEMENT OF CHOROIDAL
MELANOMA
BRADLEY R. STRAATSMA, M.D.**

Baron Jules François was the preeminent oph-

thalmologist in a generation that included a
number of notable colleagues. Throughout his
distinguished tenure as Professor and Director
of Ophthalmology at the University of Ghent,
he combined intellectual and scholarly achieve-
ment with extraordinary energy and global vi-
sion.
Scientific interests ranging from genetics1 to
glaucoma2 culminated in authorship of hun-
dreds of articles, chapters, and books. His
boundless energy permitted an intense work
ethic at the University of Ghent, leadership as
President of the Royal Academy of Medicine of
Belgium (1969) and a broad commitment to
international ophthalmology. At the conclusion of his term, grateful col-
Central to all activities, Jules François had a vi- leagues elected Jules François Honorary Life
sion of ophthalmology as a global community. President of the International Council of Oph-
While President of the International Council of thalmology.3
Ophthalmology from 1974 to 1982, he pre- In 1975, Jules François initiated formation of
sided over the XXIII International Congress of the Academia Ophthalmologica Internationa-
Ophthalmology in Tokyo, Japan, during 1978 lis to encourage interaction among academic
and the XXIV International Congress of Oph- ophthalmologists. In recognition, he was ac-
thalmology in San Francisco, United States, corded the title of Honorary Life President of
during 1982. At the Closing Ceremony of the the Academia Ophthalmologica Internationa-
latter, Jules François stated that ’’friendship has lis.
gained as much as science and we have felt Illustrating the global vision of Jules François,
close to one another, both in heart and in mind.’’ in April 2001, the International Council of Oph-
thalmology and the Academia Ophthalmolog-
ica Internationalis, working together, published
zzzzzz the International Ophthalmology Strategic Plan
* presented at “Ophthalmologia Belgica”, Brussels,
to Preserve and Restore Vision - Vision for the
24th November 2001. Future.4 Presented simultaneously in print and
full-text on the Internet (www.icoph.org and
** From the Jules Stein Eye Institute and Department
of Ophthalmology, UCLA School of Medicine,
www.acad-ophthal-int.org), this is a multiyear
University of California, Los Angeles, California, blueprint for improving eye care throughout the
USA world.

Bull. Soc. belge Ophtalmol., 283, 5-12, 2002. 5

The international leadership of Jules François nals. Computerized search engines such as Med-
and his participation in countless ophthalmo- line may not assemble information from an on-
logy meetings in all parts of the world forged going or completed clinical trial. Recognizing
friendships that united the profession of oph- this, the American Journal of Ophthalmology in-
thalmology. With respect and warm memories augurated the Clinical Trials Digest as an In-
of association, it is a distinct honor and a great ternet feature available to all viewers at ww-
personal pleasure to present the Jules François w.ajo.com in October 2001.7 This Digest is de-
Memorial Lecture. signed to include ongoing and completed trials
In this presentation, I will focus on the role of that evaluate the effect of treatment (drug, pro-
randomized clinical trials in determination of cedure, device, or regimen) in ophthalmic pa-
evidence-based medical intervention, results of tients. For each listed trial, the Digest presents
the Collaborative Ocular Melanoma Study, and the trial design, names of participating cen-
forward-looking goals of choroidal melanoma ters, and a list of all publications from the trial
research. The first major publication from a by title, date, and authorship. An electronic link
large, multicenter randomized clinical trial in permits the viewer to click on the title of any
ophthalmology was published in the American publication and immediately review the ab-
Journal of Ophthalmology in 1976, exactly 25 stract of the article. This electronic informa-
years ago. This report by the Diabetic Retino- tion resource will enable ophthalmologists to lo-
pathy Study Group changed the global stan- cate evidence-based information, alert physi-
dard of care of proliferative diabetic retinopa- cians to new treatment opportunities for their
thy.5 Since then, randomized clinical trials have patients, and foster collaborations within the
influenced the management of nearly every major global research community.
pathologic condition in ophthalmology, includ- Illustrating a randomized clinical trial, the Col-
ing corneal disease, cataract, glaucoma, laborative Ocular Melanoma Study (COMS), in
endophthalmitis, complications of acquired im- progress since 1985, is sponsored by the Na-
munodeficiency syndrome (AIDS), retinopathy tional Eye Institute and National Cancer Insti-
of prematurity, retinal vein occlusion, age-re- tute. The COMS is Chaired by Stuart L. Fine,
lated macular degeneration, and choroidal mel- M.D.; John D. Earle, M.D. (radiation oncolo-
anoma.6 gist), and Bradley R. Straatsma, M.D., are Co-
In the quest for evidence-based medical inter- Chairs; Barbara S. Hawkins, Ph.D., directs the
vention, a clinical trial may be appropriate when Coordinating Center; and Natalie Kurinij, Ph.D.,
(1) the disease to be studied is a substantial represents the National Eye Institute. This study
public health concern, (2) there is preliminary is being conducted at 43 clinical centers in the
evidence that a ’’new’’ test treatment may be United States and Canada.
superior to the current ’’standard’’ treatment, During 2001, the COMS Randomized Trial of
(3) a valid trial outcome may be obtained in a Iodine 125 Brachytherapy for Choroidal Mela-
reasonable time, and (4) ethical patient pro- noma presented results in three publications:
tection is in place. When these conditions are characteristics of patients enrolled and not en-
met, it is more ethical to conduct a clinical tri- rolled8, initial mortality findings9, and visual
al than to pursue either the ’’new’’ test treat- acuity three years after brachytherapy10. To as-
ment or the ’’standard’’ treatment on the basis sess the extent to which findings of the trial can
of the incomplete or inconclusive evidence. Prop- be generalized to future patients, one report
erly designed, a randomized clinical trial can considered the characteristics of patients en-
control for bias or unascertained factors by ran- rolled and not enrolled. At the 43 COMS Clin-
domization, reduce the variability in patient out- ical Centers, 8,712 choroidal melanoma pa-
come by a suitable sample size, and protect the tients were screened between February 1987
participating patients by appropriate ethical safe- and completion of accrual in July 1998, 5,046
guards.6 had melanoma of eligible size, 2,882 fulfilled
Results of clinical trials are central to evidence- all eligibility criteria, and 1,317 enrolled in the
based medical intervention. However, results trial.8 Enrolled patients had choroidal melano-
are often published over several years in nu- ma from 2.5 mm to 10.0 mm in apical height,
merous articles that appear in multiple jour- no more than 16 mm in longest basal diame-

6
 Fig 1. Standardized gold plaque with insert for I-125 seeds and eyelets for fixation sutures.

ter, and with exceptions, a proximal border 2

mm or more from the optic disk. Eligible pa-
tients were adults, free of clinically detectable
melanoma metastasis and with no history of
other primary cancer.
Approximately half (1,371 (46%) of 2,882) of
the eligible patients enrolled. Baseline demo-
graphic characteristics and tumor characteris-
tics of enrolled and not-enrolled patients were
well-balanced for all factors judged to be im-
portant in determining outcome. Thus, results
of enrolled patients may be generalized to in-
clude eligible but not-enrolled patients. Fig 2. 62-year-old man. Medium choroidal melanoma, right
A further question is whether the enrolled pa- eye, with basal diameters of 9 mm and 10 mm and apical
tients are broadly representative of patients who thickness of 3.8 mm prior to brachytherapy. Visual acuity:
would be candidates for treatment with I-125 20/25.
brachytherapy. Based on a choroidal melano-
ma incidence of six to eight new cases per mil- ies, but COMS patients were consistently with-
lion population per year and a 280 million po- in the parameters of patients in other studies.
pulation in the United States and Canada, there The COMS has high internal validity by design
were about 19,320 to 25,760 new cases of and by comparison of enrolled and not-enrolled
choroidal melanoma during the 11.5 years of patients. The COMS has high external validity
COMS patient accrual. The 8,718 patients based on the proportion of melanoma cases
screened at COMS Centers were 34% to 45% screened at COMS Centers and comparison of
of all new cases during that time. Moreover, COMS patients with published reports. We may
COMS patients were compared with patients in conclude that results of the COMS randomized
other published studies of I-125 brachythera- trial are generalizable to future patients who
py. There is considerable variability among stud- meet COMS eligibility criteria.

7
Fig 3. Intraoperative ultrasonography confirms that the I-125 plaque extends 3 mm beyond the border of the melanoma
at 90° intervals around the circumference. Cursers identify tumor border and plaque edge. 85-year-old man with collar-
button shape melanoma measuring 10 mm x 10 mm x 6.3 mm.

In the randomized trial of I-125 brachythera- assure a 3 mm margin of treatment beyond the
py, 1,317 patients were enrolled; 660 were as- tumor at 90° intervals around the circumfer-
signed randomly to enucleation, and 657 to ence of the melanoma (Figure 3). After confir-
I-125 brachytherapy.9 When histopathology mation of plaque placement, fixation sutures
was reviewed centrally, 658 (99.7%) of 660 were tied and conjunctiva was repositioned to
eyes had choroidal melanoma and 2 eyes had the limbus.
carcinoma metastatic to the choroid The I-125 Based on time of enrollment, 1,072 (81%) of
brachytherapy protocol provided a dose of 85 the patients have been followed after treatment
Gy at the tumor apex or 5 mm from the interior for 5 years and 416 (32%) for 10 years; 364
surface of the sclera for tumors with apical height patients have died. The unadjusted estimated
of less than 5 mm. The protocol utilized stan- 5-year survival rates were 81% for enucleation
dard gold plaques that provided a 2 to 3 mm and 82% for brachytherapy; there was no sta-
margin of treatment beyond the tumor border tistical difference in survival rates (P =.48).
(Figure 1). Among COMS patients, survival rates follow-
The I-125 plaque placement procedure may be ing enucleation and brachytherapy did not dif-
illustrated by a typical case (Figure 2). Briefly, fer for up to 12 years after treatment.
under monitored local anesthesia with intrave- In subanalyses and assessment by pathologists,
nous sedation or general anesthesia, a conjunc- younger age and shorter longest basal diame-
tival peritomy was performed and sutures were ter of the tumor were significantly associated
looped around the rectus muscle insertions. The with longer survival. Five year rate of death with
melanoma was localized by transillumination histologically confirmed melanoma metastasis
and ophthalmoscopy, and the sclera was marked was approximately 10%, with no statistically
to position the central meridian of the melano- significant difference between enucleation and
ma and identify the placement of each fixation brachytherapy.
suture. With the plaque in place and slip knots Important visual acuity data were reported for
on the fixation sutures, plaque placement was 623 eyes treated with I-125 brachytherapy and
confirmed by intraoperative ultrasonography to followed for at least one year.10 At baseline be-

8
Fig 4. Prior to I-125 brachytherapy. 77-year-old woman Fig 5. One year after I-125 brachytherapy. Same patient
with medium choroidal melanoma, right eye, measuring as in Figure 4. Decrease in apical thickness of melanoma.
8 mm × 11 mm × 4 mm. Proximal edge of the tumor is Visual acuity: 20/25.
4 mm from the foveola. Visual acuity: 20/20.

fore treatment, median protocol visual acuity

in the eye with melanoma was 20/32; 70% of
eyes had 20/40 or better, and 10% of eyes had
20/200 or worse. Three years after I-125
brachytherapy, median visual acuity was 20/
125; 34% of eyes had 20/40 or better and
45% of eyes had 20/200 or worse; the latter
included eyes that were enucleated. Life-table
estimates of patients who lost six or more lines
of visual acuity (a quadrupling of the minimum
angle of resolution) from baseline were 18% by
1 year, 34% by 2 years, and 49% by 3 years
after treatment. Overall, about half of treated Fig 6. Seven years after I-125 brachytheray. Same patient
eyes had substantial impairment of visual acu- as in Figure 4. Further decrease in melanoma thickness,
ity 3 years after I-125 brachytherapy, defined extensive peritumor radiation-associated retinopathy and
as loss of six or more lines of acuity (49% of choroidopathy. No evidence of melanoma metastasis. Vi-
sual acuity: Count Fingers at one foot.
eyes) or 20/200 or worse (45% of eyes).
Impairment of visual acuity after brachythera-
py is illustrated in an eye with juxtafoveal mel- rapy, there is a general choroidal vasculopathy
anoma and baseline visual acuity of 20/20 (Fig- consisting of choriocapillaris nonperfusion, at-
ure 4). One year after brachytherapy, the tu- tenuation and closure of small choroidal ves-
mor is decreased in thickness and visual acu- sels, and fluorescein staining of residual chor-
ity is 20/25 (Figure 5). Seven years after brachy- oidal vessels. Late choroidal vasculopathy, usu-
therapy, the tumor shows further decrease in ally more than 2 years after brachytherapy, in-
thickness, but the peritumor radiation-associ- cludes closure of large choroidal vessels, ex-
ated retinopathy and choroidopathy are great- tensive retinal pigment epithelial atrophy, pav-
er and visual acuity is Count Fingers at 1 foot ing stone degeneration, and remodeling of the
(Figure 6). choroidal circulation.11
Radiation retinopathy and radiation optic neu- In broader context, survival results of the I-125
ropathy are well-recognized consequences of brachytherapy trial should be compared with
radiation therapy. Radiation choroidopathy is the COMS trial of pre-enucleation radiation of
less fully appreciated. Following I-125 brachy- large choroidal melanoma.12, 13, 14 In the pre-
therapy, there is a progressive decrease in vas- enucleation trial for large melanoma, tumors
cular perfusion of the choroidal melanoma. In were greater than 10 mm in apical height or
addition, during the 2 years after brachythe- greater than 16 mm in longest basal diameter;

9
there was no statistically significant difference lin and co-authors16, tumor doubling time of
between enucleation with or without pre-enu- uveal melanoma metastases ranged from 34 to
cleation radiation, and 5-year survival was 57 220 days, with two-thirds between 30 and 80
- 62%. In contrast, the brachytherapy trial of days, and a median of 63 days. Assuming a
medium choroidal melanoma included tumors constant growth rate, some melanoma metasta-
2.5 mm to 10 mm in apical height and less sis, had initiated as much as 5 years prior to
than 16 mm in longest basal diameter; there primary melanoma treatment. This strength-
was no difference between enucleation or brachy- ens the argument for early treatment, but also
therapy, and 5-year survival was 81 - 82%. indicates the best hope for improving survival
These results emphasize two forward-looking is to combine primary treatment of the choroi-
melanoma research goals. First, how does ra- dal melanoma with adjuvant systemic therapy
diation therapy affect melanoma cells? Some to combat micrometastases. Currently in progress
affect may be from the attenuation and closure are studies of melanoma antigen by Ericsson
of blood vessels. However, recent research by and co-authors18, investigations of ’’tumor kill-
Brantley, Worley, and Harbour15, reported in an er cells’’ and randomized clinical trials of All-
article currently in press in the American Jour- ovectin-7 (a gene therapy product which con-
nal of Ophthalmology and described with the tains the gene for HLA-B7) immunotherapy for
authors’ permission, suggests that radiation in- metastatic disease. To date, however, no adju-
duces ’’reproductive cell death.’’ Although the vant immunotherapy or chemotherapy has been
molecular mechanisms of reproductive cell death documented as effective.
are incompletely understood, most evidence in- In summary, a quarter century after the first
dicates that it results from irreparable damage publication from the Diabetic Retinopathy Study,
to the DNA of the cell. In 3 eyes enucleated for randomized clinical trials have influenced near-
neovascular glaucoma or periocular pain 2 to ly every major pathologic condition in ophthal-
26 months after I-125 brachytherapy and sat- mology. In 2001, the Collaborative Ocular Mel-
isfactory tumor response, melanoma cells could anoma Study trial of I-125 brachytherapy has
be established in culture. Nonetheless, the ir- demonstrated that there is no statistically sig-
radiated melanoma cells had flattened cell bod- nificant difference in survival after enucleation
ies, minimal capacity to divide, and increased or I-125 brachytherapy. Three years after brachy-
cell death compared to non-irradiated melano- therapy, approximately half of treated eyes had
ma cells that were plump, divided rapidly, and substantial impairment of visual acuity. Five-
formed colonies. For example, in one irradiat- year survival after the I-125 brachytherapy tri-
ed eye, estimated tumor doubling time in cul- al for medium melanoma was 81 - 82% but
ture was 375 days with few cells surviving three after the pre-enucleation trial for large mela-
passages, while non-irradiated melanoma cells noma, 5-year survival was 57 - 62%. To im-
had a doubling time of 11 days in culture and prove local tumor control, the molecular mech-
continued to grow indefinitely. We must con- anisms of radiation-induced reproductive cell
tinue to study the molecular mechanisms of ra- death warrant study. To improve survival, the
diation and the process of reproductive cell death best hope is to combine primary treatment of
to develop new therapies that are increasingly choroidal melanoma with adjuvant systemic
effective and associated with fewer adverse side therapy for micrometastasis.
effects.
Second, how can we improve survival of pa- REFERENCES
tients with choroidal melanoma? Certainly, pa-
tients with medium choroidal melanoma have (1) FRANÇOIS J. − Genetic aspects of Ophthal-
a better prognosis than patients with large mel- mology. Ed. Little and Brown, Boston, 1968.
anoma. However, some medium melanoma pa- (2) FRANÇOIS J., KLUYSKENS J. − Le Glaucome
tients who screen negative for melanoma me- Congénital. Ed. Imprimerie Médicale et Scien-
tastasis at the time of brachytherapy or enu- tifique, Brussels, 1950.
cleation develop metastatic melanoma. This (3) FRANÇOIS J. − ACTA: XXIV International Con-
probably relates to early onset of melanoma gress of Ophthalmology, 1982; 1: CLV-CLVi.
metastasis. In a study of 70 patients by Eske- J. B. Lippincott, Philadelphia.

10
 (4) International Council of Ophthalmology and (11) SCHALTTER U.C., MUELLER A.J., BARTSCH
Academia Ophthalmologica Internationalis. In- D.U., FREEMAN W.R., KAMPIK A. − Choroidal
ternational Ophthalmology Strategic Plan to melanoma microcirculation with confocal in-
Preserve and Restore Vision - Vision for the Fut- docyanine green angiography before and one
ure. Los Angeles, CA: 2001. year after radiation brachytherapy. Retina 2000;
(5) Diabetic Retinopathy Study Research Group. 20(6): 627-32.
Preliminary report on effects of photocoagula- (12) Collaborative Ocular Melanoma Study Group.
tion therapy. Am J Ophthalmol, 1976; 81(4): Histopathologic characteristics of uveal mela-
383-396. nomas in eyes enucleated from the Collabora-
(6) BRESSLER S.B., BRESSLER N.M.− Randomi- tive Ocular Melanoma Study: COMS report No.
zed Clinical Trials in Ophthalmology in 2001: 6. Am J Ophthalmol 1998; 125(6): 745-766.
Twenty-fifth anniversary of the first publica- (13) Collaborative Ocular Melanoma Study Group.
tion from the Diabetic Retinopathy Study. Am The Collaborative Ocular Melanoma Study
J Ophthalmol 2000; 129(4): 503-504. (COMS) randomized trial of pre-enucleation ra-
(7) BRESSLER S.B., DIENER-WEST M., TAYLOR diation of large choroidal melanoma, I: Cha-
H.R., STRAATSMA B.R. − Clinical Trials Di- racteristics of patients enrolled and not enrol-
gest on the Internet at AJO.COM: Entering a led. COMS report No. 9. Am J Ophthalmol
new information age. Am J Ophthalmol, 2001; 1998; 125(6): 767-778.
132(4): 566-567. (14) Collaborative Ocular Melanoma Study Group.
(8) Collaborative Ocular Melanoma Study Group. The Collaborative Ocular Melanoma Study
The COMS randomized trial of iodine 125 bra- (COMS) randomized trial of pre-enucleation ra-
chytherapy for choroidal melanoma, II: Cha- diation of large choroidal melanoma, II: Initial
racteristics of patients enrolled and not enrol- mortality findings: COMS report No. 10. Am J
led: COMS report No. 17. Arch Ophthalmol Ophthalmol 1998; 125(6): 779-796.
2001; 119(7): 951-965. (15) BRANTLEY M., WORLEY L., HARBOUR J.W.
(9) Collaborative Ocular Melanoma Study Group. − Altered expression of Rb and p53 in uveal
The COMS randomized trial of iodine 125 bra- melanomas following plaque radiotherapy. Am
chytherapy for choroidal melanoma, III: Initial J Ophthalmol (in press).
mortality findings: COMS report No. 18. Arch (16) ESKLIN S., PHYHONEN S., SUMMANEN P.,
Ophthalmol 2001; 119(7): 969-982. HAHKA-KEMPPINEN M., KIVELA } T. − Tumor
(10) Collaborative Ocular Melanoma Study Ocular doubling times in metastatic malignant mela-
Melanoma Study Group. The COMS randomi- noma of the uvea. Ophthalmology 2000;
zed trial of iodine 125 brachytherapy for me- 107(8): 1443-1449.
dium choroidal melanoma, I: Visual acuity after
3 years: COMS report No. 16. Ophthalmology
2001; 108(2): 348-366.

11