Independent validation of time to treatment as a prognostic factor in uveal melanoma.

BACKGROUND: An earlier study suggested that longer intervals between diagnosis and treatment of uveal melanoma are associated with worse prognosis. We evaluated whether a similar association could be observed in an independent cohort. METHODS: Retrospective cohort of 336 patients with posterior uveal melanoma treated at a single referral center (2000–2011), assembled by investigators independent of the original report. We compared disease-specific (DSS) and overall survival (OS) across prespecified diagnosis-to-treatment intervals (≤ 15, 30, 45 days) using cumulative-incidence (Gray’s) and Kaplan–Meier analyses (unmatched and propensity-matched). Fine–Gray regression with treatment day as time zero estimated the subdistribution hazard ratio per 10-day delay. RESULTS: Tumor size and stage at diagnosis did not differ between patients treated within 30 days (prompt) and those treated after 30 days (delayed). In competing-risk analysis, delayed treatment was associated with a higher cumulative incidence of metastatic death. Kaplan–Meier OS—but not DSS—differed significantly between prompt and delayed treatment. A model of exponential tumor growth indicated that deferring treatment of a lesion of a medium-sized tumor by one month is associated with an estimated absolute increase in the 10-year competing-risk incidence of metastatic death by ≈ 0.8%. CONCLUSIONS: While treatment delay is unlikely to be among the strongest prognostic factors in uveal melanoma, this independent validation cohort supports an association between diagnosis-to-treatment intervals beyond about one month and worse survival.