20-year assessment of metastatic latency and subsequent time to death after proton therapy for uveal melanomas.

The aim of this study was to evaluate metastatic latency and survival after the occurrence of metastases in patients with choroidal/ciliary body melanoma treated with proton therapy. This was a retrospective cohort study. All consecutive patients with choroidal/ciliary body melanoma treated with proton therapy between 1991 and 2010 were included. Overall survival, specific survival (SS), local recurrence-free interval, and metastasis-free interval (MFI) were calculated. There were 508 patients. The mean follow-up was 239.4 months. Overall survival and SS rates were 57.2 and 67.6% at 10 years. Pre-equatorial tumor location, advanced tumor stage, and initial exudative retinal detachment were associated independently with SS. Thirty-three percent of the patients (n = 169) had metastases. Local recurrence-free interval and MFI were 91.3 and 65.7% at 10 years, respectively. MFI was shorter in pre-equatorial, large tumors, and/or tumors with exudative retinal detachment. After the occurrence of metastases, the median survival time was 1.25 years and survival probabilities were 62.1% at 1 year, 26.0% at 2 years, and 6.0% at 5 years. Except for age, none of the baseline clinical factors was associated with survival after metastasis occurrence. SS after metastasis occurrence was longer for metastasis occurring more than 10 years after tumor diagnosis (P =0.010). Death after metastasis is independent of initial tumor characteristics. Small tumors still have a risk for metastases after 10 years. Thus, lifelong follow-up is necessary for uveal melanoma patients. Larger series of metastatic patients are needed to evaluate aggressive multimodal treatments of metastases. Death after metastasis is independent of the initial tumor characteristics. Small tumors contraintuitively have a long-life risk of metastases. MFI is associated independently with pre-equatorial location, tumor stage, and retinal detachment.