A phase 2 study of CMB305 and atezolizumab in NY-ESO-1+ soft tissue sarcoma: Interim analysis of immunogenicity, tumor control and survival

Abstract

Background: CMB305 is an active immunotherapy designed to generate and expand anti‐NY‐ESO‐1 immune response (IR). CMB305 consists of a dendritic cell‐targeting lentiviral vector encoding NY‐ESO‐1 (LV305), and a boost with an NY‐ESO‐1 recombinant protein plus GLA‐SE (G305), a TLR‐4 agonist. Phase 1 studies of LV305 and CMB305 showed this approach is safe, generates IR and appears to impact survival with 81% 1‐yr survival in NY‐ESO‐1+ sarcoma patients (pts) following LV305 treatment. We evaluated efficacy and IR for combination of CMB305 (C) and atezolizumab (A) or A alone in NY‐ESO‐1+ synovial sarcoma (SS) and myxoid round cell liposarcoma (MRCL). Methods: A prospective randomized open label phase 2 study of C (LV305 Intradermal Days 0, 14, 42, 70+ G305 Intramuscular Days 28, 56, 84 then q6wk up to one year) + A (1200mg IV q3wk) vs. A alone in locally advanced or metastatic NY‐ESO‐1+ SS/ MRCL. Primary endpoints are progression free survival (PFS) and overall survival (OS) with secondary endpoints of safety, IR, and response rate. Results: As of December 30, 2016, 58 patients were enrolled. A prespecified interim analysis of PFS included the first 36 pts with median 7.0 mos follow up (Arm A+C: median age 47 yrs, 78% SS, 100% metastatic, 78%=>2 chemotherapy; Arm A: median age 44 yrs, 56% SS, 67% metastatic, 56%=>2 chemotherapy). Combination A+C was well tolerated. Clinical benefit was similar between arms (Arm A+C: 8/18 pts with SD, 1 pt unconfirmed PR, 6 mos PFS rate 17%; Arm A: 10/18 pts SD, 6mos PFS rate 22%). In addition, anti‐NY‐ESO‐1 IR seen in 10/19 (53%) pts Arm A+C vs. 3/12 (25%) pts Arm A by T Cell ELISpot, and 9/22 (41%) pts Arm A+C vs. 0% Arm A by antibody ELISA. Pts with IR had target lesion increase of 2% compared to 18% in pts without IR based on preliminary ANOVA‐model based analysis. No deaths observed in pts with induced anti‐NY‐ESO‐1 T cell IR (0/13 deaths IR+ pts vs. 5/18 deaths IR‐ pts). Conclusions: In the interim analysis, Arm A+C resulted in a higher level of anti‐NYESO‐ 1 IR when compared to Arm A; pts with IR tend to have better target lesion control. Early data indicate that induction of anti‐NY‐ESO‐1 IR may be associated with better survival.