Background: The open-label, single-center phase II study of T-VEC and pembrolizumab in patients with advanced sarcoma met its primary endpoint and demonstrated a best objective response rate of 30% at 24 weeks per RECIST v1.1 (Kelly CM, et al, Jama Oncology, 2020). Responses were seen in undifferentiated pleomorphic sarcoma (UPS), myxofibrosarcoma, (MFS), epithelioid sarcoma (ES), cutaneous angiosarcoma (AS), and undifferentiated sarcoma not otherwise specified. Here, we report the efficacy observed in three histology specific expansion cohorts: 1) UPS/MFS, 2) cutaneous AS and 3) ES. Methods: Patients refractory to ≥ 1 prior line of systemic therapy or declined standard of care systemic therapy received pembrolizumab intravenously and intratumoral T-VEC injections on day 1 of a 21-day cycle. The primary endpoint was best objective response (ORR, complete and partial responses per RECIST v 1.1) by 24 weeks estimated for each subtype-specific cohort. Secondary objectives included: adverse events (AEs, TRAEs), median PFS and correlatives. Results: Twenty-one patients enrolled in the expansion cohorts: median age 72 years (range: 39-85), male 57%, ≤ 1 prior line of therapy (43%). Treatment was well tolerated in twenty patients; one patient discontinued study therapy due to grade 3 immune mediated hepatitis. Nineteen patients were evaluable for efficacy (one patient withdrew from the study and another discontinued treatment before week 24). Subtype specific best ORR by 24 weeks per RECIST v 1.1: UPS/MFS – 11% (n = 1/9)[95% CI: 0.0-0.48]; AS – 43% (n = 3/7)[95% CI: 0.1-0.82]; ES – 0% (n = 0/3). The best ORR overall for the AS cohort was 71% (n = 5/7)[95% CI: 0.03-0.95]. Median PFS (weeks) was 14.9 [CI:7-111] for UPS/MFS and 54 [95% CI: 3- not reached] for AS (Table). Conclusions: TVEC and pembrolizumab demonstrated acceptable safety and promising anti-tumor activity in cutaneous AS [head & neck (n = 4) and Stewart-Treves syndrome involving the upper extremity (n = 1)]. Five AS patients experienced a partial response with durable disease control, remaining on study for 1-2 years or more. Two delayed responses were observed in the AS cohort after the pre-specified 24-week criteria. One AS responder progressed on immune checkpoint inhibition prior to study entry. Correlative analyses are ongoing. Clinical trial information: NCT03069378.

*One patient was evaluable for toxicity but not evaluable for efficacy because they came off study prior to week 24.