Results of the Extension Trial of Optim, a Multicenter, Randomized Phase 3 Trial of Talimogene Laherparepvec (T-Vec) Vs Gm-Csf for Unresected Stage Iiib-Iv Melanoma

Abstract

Aim: T-VEC is an HSV-1-derived oncolytic immunotherapy designed to selectively replicate in tumors and produce GM-CSF to enhance systemic antitumor immune responses. Compared to GM-CSF alone, T-VEC significantly improved durable response rate (DRR; partial response [PR] or complete response [CR] lasting continuously for ≥ 6 months [m]) from 2% to 16% (p < 0.0001) in patients (pts) with stage IIIB-IV melanoma (Andtbacka et al, ASCO 2013). Here we report the results of an extension trial of OPTiM as of Mar 2013. Methods: Pts were eligible for the extension trial if: 1) they had received the maximum treatment (tx) cycles allowed on OPTiM and did not have disease progression (PD) associated with reduced performance status or 2) had a CR on OPTiM and then developed new lesions within 12 m. In the extension study, pts could continue their randomized tx: T-VEC (intralesional ≤ 4 mL × 10 8 pfu/mL q2w) or GM-CSF (sc 125 µg/m 2 qd × 14 days q4w) until CR, additional 12 m of tx, PD or unacceptable toxicity. The primary endpoint was safety; efficacy was also evaluated. Results: 30 (27 T-VEC, 3 GM-CSF) of the 436 pts enrolled in OPTiM entered the extension trial. Including OPTiM tx, median duration of tx was 91 wks (range: 29-132) for T-VEC and 100 wks (54-120) for GM-CSF. Most common adverse events (AE) on each arm were chills and pyrexia. There were no grade 5 tx-related AEs. For T-VEC, 5 new CR occurred (best OPTiM response was 3 PR and 2 stable disease [SD]). Best overall response was maintained in 16 pts (1 CR, 6 PR, and 9 SD), 2 pts progressed from PR to SD; PD occurred in 4 pts. A new DR occurred in 1 T-VEC-treated pt. For GM-CSF, PR was maintained in 1 pt and 2 pts progressed from PR to SD. For both trials combined, CR, OR, and DR rates per investigator for T-VEC were 16.3%, 31.5%, and 19.3%, respectively, and 0.7%, 6.4%, and 1.4% for GM-CSF. Conclusions: Continued tx with T-VEC but not GM-CSF was associated with improved response rates, with 5 additional CR and one DR (ie. previously a response of <6 m). A tolerable safety profile consistent with that seen in the main OPTiM study was observed.