Interim analysis of a randomized, open-label phase 2 study of talimogene laherparepvec (T-VEC) neoadjuvant treatment (neotx) plus surgery (surgx) vs surgx for resectable stage IIIB-IVM1a melanoma (MEL).

Abstract

9508 Background: There is no approved neotx for resectable stage IIIB-IVM1a MEL. T-VEC, an HSV-1-based oncolytic virus, may reduce the risk of developing visceral and bone metastases in unresectable Stage IIIB-IVM1a MEL (Andtbacka SSO 2015). We conducted a randomized study to evaluate the effect of neotx T-VEC in high risk resectable MEL (NCT02211131). Methods: Patients (pts) with resectable stage IIIB/C/IVM1a MEL, ≥ 1 injectable cutaneous, subcutaneous, or nodal lesions ≥ 10 mm, and no systemic tx 3 mos prior were randomized 1:1 to 6 doses/12 wks of T-VEC followed by surgx (Arm 1) vs upfront surgx (Arm 2). T-VEC was given at standard dosing until surgx, no injectable tumors, or intolerance. This interim analysis was planned for when the 75th pt in Arm 1 completed the safety follow-up visit (30+ days post surgx). Results: 150 pts were randomized (76 Arm 1, 74 Arm 2). Of all pts, ≥ 94% had no prior radio/systemic tx, 91% had prior surgx and 84% had no plans for adjuvant tx. 75% in Arm 1 and 93% in Arm 2 had surgx as planned. Of the 19 pts who did not have surgx in Arm 1, 11 had progressive disease. In Arm 2, 17 pts recurred within 14 wks post-surgx. For pts who had surgx in Arm 1, the pathological complete response (pCR) rate was 21%. Negative margin resection (R0) rates were 56.1% for Arm 1 and 40.6% for Arm 2 (80% CI: 3-28% for the difference). For all randomized pts, pCR rate in Arm 1 was 15.8%; R0 rates were 42.1% (Arm 1) vs 37.8% (Arm 2). Overall response (OR) rate (CR+PR) in Arm 1 was 14.7% (80% CI: 9-22%). In the safety set (73 pts in Arm 1, 69 pts in Arm 2), tx-emergent adverse events (TEAE) was 93% in Arm 1 (1 grade 4 pain, no grade 5) and 45% in Arm 2 (all ≤ grade 3). In Arm 1, preop AE was 89.5% (most common: pyrexia 35%) and intra/postop AE was 29.8% (most common: seroma 5.3%). In Arm 2, intra/postop AE rate was 45% (most common: pain 7.2%). 17.8% (Arm 1) vs 2.9% (Arm 2) pts had an SAE; of these, 5.5% (Arm 1) and 2.9% (Arm 2) were deemed surgx-related. Conclusions: 12 wks of neo T-VEC produced a pCR rate in stage IIIB-IVM1a MEL higher than observed by ORs and may account for the higher R0 margin in Arm 1. No unexpected toxicities were noted. The primary analysis of RFS is ongoing. Clinical trial information: NCT02211131.