Randomized phase II study of adjuvant sunitinib or valproic acid in high-risk patients with uveal melanoma: The final analysis of cohort 1.

9586 Background: Despite successful treatment of primary uveal melanoma (UM), tumors with monosomy 3 and 8q amplification (M3 + 8q amp) or DecisionDx-UM Class 2 have high metastatic death rates. We report the final analysis of Cohort 1 in the randomized phase II clinical trial of 6 months of adjuvant sunitinib or valproic acid (VPA) in high-risk UM patients. Methods: High risk for systemic metastasis was defined as the following: A) M3 + 8q amp; B) Class 2. Patients within 6 months of initial treatment of primary UM were randomized 1:1 to receive either sunitinib 25 mg daily or VPA 750 mg daily for 6 consecutive months. The primary endpoint was to evaluate the improvement of 2-year overall survival (OS) rate from 70% (historical references) to 85% in each arm. Secondary endpoints included 1) systemic relapse-free survival (RFS) rate at 18 months, 2) ability to complete adjuvant treatment and, 3) toxicity assessment. Results: Eighty-eight patients were included in the final analysis. There were no differences in tumor size or T stage between the two treatment arms. Nine of 45 patients in the sunitinib arm and 4 of 43 patients in the VPA arm could not complete the 6-month treatment due to toxicity (sunitinib n = 6, VPA n = 2) or systemic progression (sunitinib n = 3, VPA n = 2). All but 9 patients (death due to metastasis, sunitinib n =4, VPA n = 5) were followed for at least 2 years. With a median follow-up of 52.6 months, both drugs met the primary end point with 2-year OS rates of 95.6% (sunitinib, 90% CI 86.5-98.6%) and 90.7% (VPA, 90% CI 80.1-95.8%). The 18-month RFS rates were 75.6% (sunitinib, 90% CI 63.1-84.3%) and 62.8% (VPA, 90% CI 49.4-73.5%). Although not statistically significant, there was a trend of superior RFS with sunitinib over VPA in primary UM with T-stage 3-4 (p=0.131) or >12mm (p=0.129). There was no significant difference in median RFS in HLA-A*02:01 positive or negative status (24.6 vs. 24.8 months). It is of note that the potential survival benefit of sunitinib over VPA diminished after 3 years, indicating longer duration of sunitinib administration might be required. In the multivariable Cox analysis, the RFS was not significantly different in the two treatment arms, but increase of tumor diameter was associated with increase hazard of progression (HR=1.23, 95% CI: 1.13, 1.33; p<0.001). Conclusions: Six months of adjuvant sunitinib or VPA resulted in 2-year OS of 95.6% and 90.7%, respectively, meeting the primary endpoint of the study. Sunitinib showed a tendency for a better outcome until 3 years after randomization, thus Cohort 2 was created to investigate the safety and prolonged improvement of RFS and OS with 12 months of sunitinib. Additionally, Cohort 3 with adjuvant sunitinib in combination with VPA for 12 months is currently ongoing. The size of primary tumor influenced the survival and should be adjusted for future adjuvant studies. Clinical trial information: NCT02068586.