A randomized phase 2 study of trametinib with or without GSK2141795 in patients with advanced uveal melanoma.

9511 Background: There are few treatment options for patients (pts) with advanced uveal melanoma (UM). MEK inhibition (MEKi) has demonstrated efficacy in a randomized Phase 2 trial (Carvajal et al, 2014), but resistance may be mediated via Akt signaling (Ambrosini et al, 2013). We hypothesized that adding an Akt inhibitor (Akti), GSK2141795 (GSK795), to the MEKi trametinib (TRAM) would improve progression-free survival (PFS) in pts with UM. Methods: We conducted an open-label Phase 2 trial where pts with advanced UM and no prior systemic or liver-directed therapy were randomized to one of two arms stratified by liver disease and LDH: TRAM 2mg daily (Arm A) or TRAM 1.5mg + GSK795 50mg daily (Arm B). Efficacy was assessed every 8 weeks (wk) by RECIST 1.1. Pts whose tumors progressed on TRAM could receive TRAM + GSK795. With 40 patients per arm, this design had 80% power to detect a PFS hazard ratio of 0.56 with a 5% one-sided alpha. A pre-planned futility analysis after 40 enrolled pts would cease accrual to an arm if <2/20 pts had objective responses. Matched biopsies at baseline and Day 15 were mandatory. Results: 40 pts were enrolled, of which 39 received ≥1 dose of study drug (N=18 TRAM; N=21 TRAM + GSK795). Median age 61, 55% male, 85% liver disease, 50% elevated LDH. One partial response was observed in each arm (16+ wk in Arm A, 8 wk in Arm B); thus, accrual was held on both arms. We did not detect a difference in median PFS between Arm B and Arm A (15.6 vs 15.7 wk; p=0.74, log rank). Median PFS for N=11 who crossed over to TRAM + GSK795 was 7.9 wk (range: 3.7-41+ wk). All pts had ≥1 adverse event (AE) at least possibly related to drug(s). AEs for TRAM included rash (100%, all Grade (G)1-2), diarrhea (72%, all G1-2), elevated AST/ALT (55%, 6% G3), mucositis (50%, all G1-2). 7/18 required dose reduction. AEs with TRAM + GSK795 included rash (81%, 14% G3), nausea (67%, all G1-2), diarrhea (62%, all G1-2), elevated AST/ALT (43%, 14% G3). 7/21 required dose reduction (2 TRAM, 3 GSK795, 2 both). There were no G4-5 AEs. Conclusions: The addition of Akti GSK795 to TRAM did not improve PFS, and objective responses were uncommon. Dose reductions for AEs were frequent. Ongoing analysis of 31 matched biopsies at baseline and day 15 is investigating therapeutic resistance. Clinical trial information: NCT01979523.