Mehnert, Janice M.

Talimogene laherparepvec (T-VEC) plus ipilimumab has demonstrated greater antitumor activity versus ipilimumab alone, without additional toxicity, in patients with advanced melanoma. Here, we report the 5-year outcomes from a randomized phase II study. These data provide the longest efficacy and safety follow-up for patients with melanoma treated with a combination of an oncolytic virus and a checkpoint inhibitor. Eligible patients with unresectable stage IIIB‒IV melanoma were randomized 1:1 to receive T-VEC plus ipilimumab or ipilimumab alone. T-VEC was administered intralesionally at 10 6 plaque-forming units (PFU)/mL in week 1, followed by 10 8 PFU/mL in week 4 and every 2 weeks thereafter. Ipilimumab (3 mg/kg every 3 weeks; ≤4 doses) was administered intravenously starting at week 1 in the ipilimumab arm and week 6 in the combination arm. The primary end point was investigator-assessed objective response rate (ORR) per immune-related response criteria; key secondary end points included durable response rate (DRR), duration of response (DOR), progression-free survival (PFS), overall survival (OS), and safety. Overall, 198 patients were randomized to receive the combination (n=98) or ipilimumab (n=100). The combination improved the ORR versus ipilimumab (35.7% vs 16.0%; OR 2.9; 95% CI 1.5 to 5.7; p=0.003). DRR was 33.7% and 13.0% (unadjusted OR 3.4; 95% CI 1.7 to 7.0; descriptive p=0.001), respectively. Among the objective responders, the median DOR was 69.2 months (95% CI 38.5 to not estimable) with the combination and was not reached with ipilimumab. Median PFS was 13.5 months with the combination and 6.4 months with ipilimumab (HR 0.78; 95% CI 0.55 to 1.09; descriptive p=0.14). Estimated 5-year OS was 54.7% (95% CI 43.9 to 64.2) in the combination arm and 48.4% (95% CI 37.9 to 58.1) in the ipilimumab arm. Forty-seven (48.0%) and 65 (65.0%) patients in the combination and ipilimumab arms, respectively, received subsequent therapies. No new safety signals were reported. At the 5-year follow-up, the improved response rates observed with T-VEC plus ipilimumab were durable. This is the first randomized controlled study of the combination of an oncolytic virus and a checkpoint inhibitor that meets its primary end point.Trial registration number: NCT01740297 .

10029 Background: This is the first randomized trial testing the addition of an oncolytic virus to an immune checkpoint inhibitor for advanced melanoma. At the 3-year (yr) follow-up, the combination (combo) of T-VEC and ipi demonstrated durable and statistically superior objective response rate (ORR) over ipi alone (36.7% vs. 16.0%; odds ratio, 3.0; 95% Cl, 1.6–6.0; P = 0.002). Complete response (CR) rate was 21.4% with the combo and 6.0% with ipi. Median overall survival (OS) was not reached in either arm. In this post hoc analysis, we utilized the 3-yr landmark data to explore the relationship between CR and OS in the combo arm. Methods: Pts with unresectable, stage IIIB-IV melanoma were randomized 1:1 to receive combo or ipi alone. T-VEC was administered intratumorally on day 1 of week (wk) 1 at 106 plaque-forming units (PFU)/mL followed by subsequent doses at 108 PFU/mL on day 1 of wk 4, and every 2 wks thereafter. Ipi (3 mg/kg) was given every 3 wks starting on day 1 of wk 6 for up to 4 doses. Response was assessed by investigators per immune-related response criteria every 12 wks until disease progression. The primary endpoint was ORR; key secondary endpoints were OS, progression-free survival, and safety. Results: 198 pts were randomized (98 to combo; 100 to ipi). As of February 25, 2019, the median follow-up time was 40.0 mos (range: 0.2–63.7) for the combo arm. Among 98 pts who received combo, 21 (21.4%) had a best overall response of CR including 8 who converted from an initial partial response (PR), 15 (15.3%) had PR, 19 (19.4%) had stable disease, 30 (30.6%) had progressive disease, and 13 (13.2%) were unevaluable. Of 21 pts achieving CR, 17 (81%) had ECOG status of 0, 16 (76.2%) had stage IIIB-IVM1a disease, and 16 (76.2%) had no visceral metastases. Median duration of CR was not reached (range: 5.4[+]–58.2[+] mos); 19 of 21 CRs lasted more than 6 months. The baseline tumor burden was lower in pts with CR than in those with non-CR. Median OS was not reached in pts with CR (range: 25.1[+]–63.7[+] mos) and was 47.6 mos (range: 0.2[+]– 63.7[+] mos) in pts with non-CR (Log-rank P = 0.0005). The Kaplan–Meier estimated 3-year OS rate was 100.0% for patients with CR and 52.3% for those with non-CR. Conclusions: CR rate was higher with T-VEC plus ipi than with ipi alone in pts with advanced melanoma (21.4% vs. 6.0%). In the combo arm, CR was associated with prolonged OS, and pts with CR tended to have better ECOG performance status, earlier-stage disease, and lower baseline tumor burden, as compared with those with non-CR. Clinical trial information: NCT01740297.