The Collaborative Ocular Melanoma Study (COMS) randomized trial of pre-enucleation radiation of large choroidal melanoma: IV. Ten-year mortality findings and prognostic factors. COMS report number 24

PURPOSE: To report rates of death and related outcomes through 10 years after treatment of large choroidal melanoma and to evaluate characteristics of patients and tumors as predictors of relative treatment effectiveness and time to death. DESIGN: Randomized multicenter clinical trial of preenucleation radiation vs enucleation alone conducted as part of the Collaborative Ocular Melanoma Study. 0 METHODS: Eligible patients were free of metastasis and other cancers at enrollment. All patients were followed for 5 years or longer at scheduled examinations and contacts for metastasis, another cancer, or death. Each decedent was classified as having histopathologically confirmed melanoma metastasis, suspected melanoma metastasis without histopathologic confirmation, another cancer but not melanoma metastasis, or no malignancy. RESULTS: Within 10 years after enrollment, 576 of 1,003 patients died. Ten-year all-cause mortality rates were 61% for patients in both treatment arms. Ten-year rates of death with histopathologically confirmed melanoma metastasis were 45% in the pre-enucleation radiation arm and 40% in the enucleation alone arm. Older age and larger maximum basal tumor diameter were the primary predictors of time to death from all causes and death with melanoma metastasis. No differences in unadjusted or adjusted mortality rates were found between treatment arms. Of 448 patients eligible for 10 years of follow,up, 145 patients (32%) were alive and clinically cancer-free 10 years after treatment. CONCLUSIONS: Longer follow-up confirmed the earlier report of no survival advantage attributable to pre-enucleation radiation. Mortality rates by baseline characteristics should facilitate counseling of patients who have large choroidal melanoma and no evidence of metastasis or another malignancy at diagnosis. (C) 2004 by Elsevier Inc. All rights reserved.