Adjuvant Sunitinib in High-Risk Patients with Uveal Melanoma: Comparison with Institutional Controls.

PURPOSE: To compare overall survival in high-risk patients with primary uveal melanoma who received adjuvant sunitinib with institutional controls. DESIGN: Retrospective cohort. PARTICIPANTS: Selection criteria were (1) monosomy 3 and 8q amplification by cytogenetic or DecisionDx-UM Class 2 and (2) monosomy 3 and large tumor size (T3-4 by American Joint Committee on Cancer classification). Exclusion criteria were date of diagnosis before 2007 or after 2013 and age <18 years. METHODS: A cohort of patients who intended to receive adjuvant sunitinib for 6 months was compared with institutional historical controls with the same risk factors. Kaplan-Meier and Cox proportional hazards models were used to analyze the outcome. Propensity score was used to adjust for nonrandom assignment to sunitinib. MAIN OUTCOME MEASURES: Overall survival. RESULTS: From the Wills Eye Hospital Oncology Service Uveal Melanoma Cytogenetic Database (N = 1172), 128 patients fulfilled the selection and exclusion criteria. Median follow-up was 52.7 months (range, 0.26-108 months). A total of 54 patients received sunitinib. Their median age was 56 years (range, 29-81 years), and 48% were men. A total of 74 historical controls in the same risk category were identified. Their median age was 62 years (21-80 years), and 48% were men. Patients in the sunitinib group had worse cytogenetic or molecular features (monosomy 3 and 8q amplification or class 2 87% vs. 57%; P < 0.001), had smaller tumor sizes (T3-4 56% vs. 83%; P = 0.001), and were younger. There were 51 deaths, 14 (26%) in the sunitinib group and 37 (50%) in the control group. In the univariate analysis, the sunitinib group had longer overall survival (hazard ratio, 0.53; 95% confidence interval, 0.29-0.99; P = 0.041). In multivariate Cox regression analysis, interaction between use of sunitinib and age as a dichotomous variable was highly significant (P = 0.003). The following variables were statistically associated with prediction of overall survival: cytogenetic/molecular status (P = 0.015), T-size category (P = 0.022), gender (P = 0.040), and adjuvant sunitinib in patients aged <60 years (P = 0.004). Results were confirmed by propensity score analysis. CONCLUSIONS: In this retrospective study, the use of sunitinib in the adjuvant setting was associated with better overall survival.