Abstract 2986: Atezolizumab in metastatic TNBC (mTNBC): Long-term clinical outcomes and biomarker analyses

Abstract

Abstract Introduction. Triple negative breast cancer (TNBC) has a poor prognosis and limited treatment options. Atezolizumab (atezo) is a humanized mAb that inhibits the binding of PD-L1 to PD-1 and B7.1, thus restoring tumor-specific T-cell immunity. Atezo was evaluated in an expansion cohort of mTNBC patients (pts) in a Phase Ia study (NCT01375842). Methods. Enrollment was initially limited to TNBC pts with PD-L1 on ≥5% of tumor-infiltrating immune cells (IC2/3), then opened to pts regardless of PD-L1 status. Pts received atezo IV in 1L or 2L+ q3w at 15 or 20 mg/kg or 1200 mg for 1 y with option to be retreated at PD, or until loss of clinical benefit. ORR was assessed by RECIST v1.1 and irRC, to capture non-conventional responses. Baseline PD-L1 expression on IC was centrally scored as IC0/1/2/3 (VENTANA SP142 assay). Pretreatment tumors and on-therapy biopsies were evaluated for TILs, CD8 T cells and macrophages by IHC. Results. As of Mar 31, 2016, 115 mTNBC pts were safety evaluable; atezo was generally well tolerated. There were no additional safety signals from prior report (Emens AACR 2015). 112 pts with FU ≥12 wk were evaluable for response. Based on irRC, ORR in 1L and 2L+ pts were 26% and 11%, respectively. ORR for PD-L1 IC2/3 pts were 17% vs 8% in IC0/1. mDoR was 21.1 mo (3 to 34+). mOS of responders (n=15) was not reached (4+ to 37+ mo) with no deaths as of data cutoff. mOS of non-responders who lived ≥ 6 wk (n=87) was 9 mo (1+ to 19+ mo). OS rates in all pts at 1, 2, and 3 y were 41%, 22% and 22%, respectively. OS rates at 1, 2, and 3 y for PD-L1 IC2/3 were 45%, 28% and 28%, respectively. Pts whose tumors had >10% TILs or ≥1.35% CD8 in the tumor center trended toward higher ORR and longer OS. Atezo increased intratumoral TILs, CD8, macrophages and IC PD-L1 expression, but no response association was observed. Conclusions. In mTNBC, atezo was well tolerated. Responders showed durable clinical benefit. Response rates were higher in 1L or PD-L1 IC2/3 pts. Baseline TILs and CD8 were associated with greater clinical benefit. Table.EfficacyResponses per RECIST v1.1 (irRC efficacy results presented in text)1L2L+PD-L1 IC2/3PD-L1 IC0/1Alln = 19n = 93n = 71n = 37N = 112ORR, % (95%CI)26% (9, 51)7% (2, 14)13% (6, 23)5% (1, 18)10% (5, 17)CR, n21303PR, n35628SD, n31210415CBR, %42%19%27%16%23%Median (range) DoR, mo21 (8+ to 26+)NE (3 to 13+)21 (3 to 26)NE (10 to 10+)21 (3 to 26+)OS1L2L+PD-L1 IC2/3PD-L1 IC0/1All ptsn = 19n = 94n = 71n = 38N = 1131 y OS (95% CI)63% (37, 89)37% (26, 47)45% (32, 58)37% (21, 53)41% (31, 51)2 y OS (95% CI)47% (14, 80)18% (8, 27)28% (15, 41)NE22% (12, 32)3 y OS (95% CI)NE18% (8, 27)28% (15, 41)NE22% (12, 32)CBR; clinical benefit rate, defined as CR, PR or SD for ≥12 weeks. DoR, duration of response; NE, not evaluable.+, censored. Citation Format: Peter Schmid, Cristina Cruz, Fadi S. Braiteh, Joseph Paul Eder, Sara Tolaney, Irene Kuter, Rita Nanda, Cathie Chung, Philippe Cassier, Jean-Pierre Delord, Michael Gordon, Yijin Li, Bo Liu, Carol O’Hear, Marcella Fasso, Luciana Molinero, Leisha A. Emens. Atezolizumab in metastatic TNBC (mTNBC): Long-term clinical outcomes and biomarker analyses [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2986. doi:10.1158/1538-7445.AM2017-2986