A process of certification to establish clinical center uniformity required repeated certification of each investigator, clinical coordinator, plaquing surgeon, enucleating surgeon, examining ophthalmologist, radiation oncologist, radiation physicist, echographer, photographer, ophthalmic pathologist, and visual acuity examiner.

Patient eligibility and enrollment were determined by a rigid protocol that included

Overall protocol compliance in the COMS requiring radiation therapy for large tumors was >98% and nearly 95% for medium tumors. Data were collected on standardized forms and submitted within a specific time frame to the COMS central coordinating center and to specific resource centers.

This component of the study was an observation series of otherwise healthy adults with small choroidal melanomas. The primary endpoints were 5-year mortality and tumor growth. The overall 5-year mortality was low. Five-year all-cause mortality was 6.0% (95% confidence interval [CI], 2.7% to 9.3%) and 8-year all-cause mortality was 14.9% (95% CI, 9.6% to 20.2%.¹

Of the small choroidal melanomas initially observed, 21% demonstrated growth by 2 years and 31% by 5 years.² Factors associated with time to growth were greater initial tumor thickness and diameter, presence of orange pigment, absence of drusen, and absence of retinal pigment epithelial alteration adjacent to the tumor. These observations have also been noted in other large case series of small choroidal melanomas and atypical choroidal nevi. In a separate study, Shields et al. found increased tumor thickness (greater than 1 mm), proximity to the optic nerve, visual symptoms, presence of orange pigment, and presence of subretinal fluid to be predictive of future growth (Fig. 15.1).³

The medium tumor trial compared enucleation to radioactive I¹²⁵ plaque application. Patients included in the study met the following criteria (Fig. 15.2):

Patients evaluated in the COMS were excluded for the following indications (Fig. 15.3):

A total of 22% of patients evaluated for the study were deemed to be ineligible for enrollment. The most common reasons for exclusion were proximity of the tumor to the optic disk (40%), one or more primary cancers (20%), and melanoma that was primarily in the ciliary body (11%).

Plaque radiotherapy remains in widespread use throughout the world as the most common method of delivering therapeutic doses of radiation to intraocular tumors, including choroidal melanoma. High-energy sources such as cobalt have been replaced by lower energy radioisotopes including iridium-192, ruthenium-106, palladium-103, and iodine-125. There are a number of reasons why iodine-125 was the radioisotope chosen for the COMS. Eleven millimeters of lead is needed to block 50% of emitted photons from a plaque employing cobalt, whereas only 1 mm of gold can arrest all radioactive emission from an iodine-125 containing plaque. Although some investigators utilize different radioisotopes for various tumor shapes and volumes, the most commonly employed radioactive isotope is iodine-125. With a half-life of 60 days, this isotope is relatively safe for both the patient and the operating surgeon and can be incorporated into the plaque structure on site with a high degree of accuracy and reproducibility. Iodine-125 can provide sufficient dose rates to tumors without exposing the sclera to excessive radiation.

Plaque dosimetry for individual tumors is provided by the ocular oncologist and radiation oncologist in conjunction with the radio-physicist, and according to precise clinical measurements obtained through ophthalmoscopy, ocular imaging, and echography, plaques are constructed to cover the tumor base with a 2.0 mm margin greater than the largest tumor base dimension (Fig. 15.4). At the time of surgery, plaque placement is facilitated by either transillumination of the tumor margins or standard indirect ophthalmoscopic localization techniques with or without ultrasound assistance. The plaque is then secured to the sclera and the final position verified by either intraoperative echography or an indirect ophthalmoscopic-fiberoptic system (Fig. 15.5).

Brachytherapy is usually applied to medium category tumors with the aim of treating the tumor apex with ˜8,500 cGy, with dose delivery over 3 to 7 days (Fig. 15.6). Brachytherapy in the COMS employed iodine-125 plaques.

Complications of plaque therapy include treatment failure due to tumor recurrence, radiation vasculopathy, cataract formation, vitreous hemorrhage, optic neuropathy, and diplopia resulting from ocular muscle manipulation during plaque placement and removal.

Of the 1,317 patients enrolled in the study of medium-sized choroidal melanomas, 660 were assigned to enucleation and 657 to iodine-125 brachytherapy. The primary outcome was mortality at 5 and 10 years (Table 15.1). Eighty-one percent of patients had been followed for 5 years and 32% for 10 years at the end of the 11.5-year accrual period. The unadjusted estimated 5-year survival rates were 81% and 82%, respectively, with no clinical or statistical difference in survival rates overall (Table 15.2). Five-year rates of death with histopathologically proven melanoma metastasis were 11% and 9% following enucleation and brachytherapy, respectively (Table 15.3).⁴ The power of the study was sufficient to indicate that neither treatment was likely to increase or decrease mortality rates by as much as 25% relative to the other.

While there was no significant survival advantage conferred by removal of the eye, the visual results of patients receiving treatment with radioactive I¹²⁵ plaque did show significant visual impairment due to radiation-related complications, particularly radiation retinopathy. At 3 years, 50% of patients had a loss of 6 or more lines of visual acuity and 43% had a visual acuity less than 20/200.⁵ The risk factors for visual loss at 3 years included

By 5 years following brachytherapy for choroidal melanoma, treatment failure resulting in enucleation occurred in 12.5% of patients. Treatment failure was the most common reason for enucleation within 3 years, and beyond 3 years, ocular pain was most common. The
risk factors for treatment failure included older age at enrollment, larger tumors, and increasing proximity of the tumor to the fovea. Treatment failure was associated weakly with poorer survival. Almost all surviving patients retained good visual acuity in the fellow eyes throughout 5 years following treatment for choroidal melanoma.⁶

Enucleation of eyes containing choroidal melanoma does not necessarily prevent metastatic disease from presenting years later. The observations of Zimmerman et al.⁷ and McLean et al.⁸ raised the question of whether globe manipulation during enucleation disseminated viable tumor cells to produce metastases. This premise is often referred to as “The Zimmerman hypothesis.” Adjunctive treatments have been proposed for large choroidal melanoma to reduce the possibility of tumor dissemination during enucleation. These have included pre- and post-enucleation irradiation of the globe and orbit as well as cryotherapy, chemotherapy, immunotherapy, and combinations thereof. The COMS was the first prospective randomized management strategy to assess preoperative radiation preceding enucleation. The primary outcome measure was mortality at 5 and 10 years.