Adjuvant Crizotinib In High-Risk Uveal Melanoma Following Definitive Therapy.

10075 Background: Uveal melanomas (UM) measuring at least 12mm in base diameter with a class 2 signature as defined by gene expression profiling (DecisionDx-UM) are characterized by high metastatic risk, with a median time to recurrence of 32 months. No therapy has been shown to reduce this risk. The growth factor receptor Met is highly expressed in UM. We have previously shown that crizotinib, an inhibitor of Met, is an effective adjuvant therapy in preclinical models (Surriga et al, Mol Cancer Ther 2013). We therefore conducted a phase II study of adjuvant crizotinib in high-risk UM. Methods: Eligibility included: primary lesion ≥12mm in base diameter; class 2 by DecisionDx-UM testing; definitive therapy within 120 days before starting crizotinib; and, no evidence of metastatic disease. Patients (pts) received 12 four-week cycles of crizotinib (250 mg twice daily). Surveillance imaging (chest CT and MRI abdomen/pelvis) were performed q3 months. The primary endpoint was distant relapse-free survival (RFS). Secondary endpoints were overall survival (OS) and toxicity. We hypothesized that the addition of crizotinib would increase the 32 month RFS from 50% to 75% (α = 0.05; β = 0.11). Results: As of 1/31/2020, 34 pts had enrolled and received at least one dose of study drug with median age of 60 (range, 26-86); 41% female; and median ECOG PS 0 (range, 0-1). 2 pts could not be evaluated for the primary endpoint due to early withdrawal and loss to follow-up. The median time from primary treatment to crizotinib initiation was 60 days (range, 0-106). All pts experienced a treatment-related adverse event (AE) of any grade. 11/34 (32%) experienced a grade 3 or 4 AE, the most common being transaminase elevation (n = 8/11). 9 pts (28%) did not complete the full 48-week treatment course due to disease recurrence (n = 5) or toxicity (n = 4). An additional 5 pts required dose reduction due to hepatic toxicity or diarrhea. 15/32 evaluable pts developed distant disease relapse, with 14 developing relapse within 32 months. With a median duration of follow up of 28.7 months, the median RFS was 30.6 months (95% CI: 27.8-58.5%). The median OS was not reached. Conclusions: The use of adjuvant crizotinib in patients with high-risk UM did not reduce rates of relapse in this multicenter, single arm trial. 9/32 (28%) pts required dose modification or discontinuation due to AE which may have limited efficacy. Further investigations of adjuvant treatment options are warranted. Clinical trial information: NCT02223819.