Response analysis for injected and non-injected lesions and of the safety and efficacy of superficial and deep/visceral RP1 injection in the registrational cohort of anti–PD-1–failed melanoma patients of the IGNYTE trial.

9537 Background: The IGNYTE trial (NCT03767348) primary analysis of RP1 (vusolimogene oderparepvec) plusnivolumab (nivo) showed clinically meaningful durable efficacy (ORR, 32.9%; median DOR, 33.7 mos, by RECIST 1.1 and independent central review) in patients (pts) with advanced melanoma, including deep responses in non-injected visceral lesions, demonstrating systemic efficacy. Here we present an analysis of efficacy in injected and non-injected lesions and safety and efficacy in pts receiving superficial and/or deep/visceral RP1 injections. Methods: Pts with confirmed progression during anti–PD-1 ± anti–CTLA-4 for ≥8 weeks were enrolled. RP1 (1×10 6 PFU/mL x1, then Q2W 1×10 7 PFU/mL x7, up to 10 mL) was injected into superficial and/or deep/visceral tumors using imaging guidance. Nivo was given (240 mg Q2W) from the 2 nd dose of RP1 through dose 8, then alone (240 mg Q2W or 480 mg Q4W) for 2 yrs, with additional RP1 injections allowed if indicated. Results: For the 46 responding patients by RECIST 1.1 (of the 140 enrolled) 197 lesions were measured, 78 injected, 119 non-injected of which 98.7% and 96.6% had any reduction and 93.5% and 79.0% >30% reduction, respectively. For visceral lesions, 85.7% of injected and 96.2% of non-injected lesions had any reduction and 85.7% and 65.4% had >30% reduction, respectively. 104 patients had superficial only injections, and 36 had deep/visceral +/- superficial injections. Treatment-related adverse event (TRAEs) rates were comparable in patients who were injected superficially compared to patients who received deep/visceral injections, except for chills, influenza-like illness, and injection-site pain, which were numerically higher in the deep/visceral +/- superficial group. Grade ≥3 TRAEs occurred in 14.4% of pts by superficial injection and 8.3% by deep/visceral +/- superficial injection. Grade 1/2 pneumothorax occurred in 3/52 (5.8%) lung injections. No liver function abnormalities or significant bleeds were reported after liver injections. The ORR for pts with superficial injection only was 29.8%, and 41.7% for deep/visceral +/- superficial. Conclusions: Meaningful systemic responses were observed independent of the injection status of individual lesions or their anatomical site. Overall response was therefore driven by the response of both injected and non-injected lesions. The safety profile of deep/visceral injection was comparable to that of superficial injections, with efficacy also being similar. Clinical trial information: NCT03767348 .