Blood tumor mutational burden (bTMB) and tumor PD-L1 as predictive biomarkers of survival in MYSTIC: First-line durvalumab (D) ± tremelimumab (T) versus chemotherapy (CT) in metastatic (m) NSCLC.

Abstract

9016 Background: MYSTIC, an open-label, Ph3 trial of first-line D (anti-PD-L1) ± T (anti-CTLA-4) vs platinum-based CT, showed an improvement in OS with D vs CT in pts with tumor cell PD-L1 expression ≥25% (PD-L1 TC ≥25%; HR 0.76 [97.54% CI 0.56–1.02], p = 0.036). Exploratory analyses showed bTMB was a predictive biomarker for OS with D±T vs CT. We report further exploratory analyses of OS according to PD-L1 and bTMB. Methods: Immunotherapy/CT-naïve pts with mNSCLC were randomized (1:1:1) to D, D+T or CT. bTMB levels (mut/Mb) were evaluated with the GuardantOMNI platform (Guardant Health), and PD-L1 TC expression with the VENTANA PD-L1 (SP263) IHC assay. Results: D improved OS vs CT in pts with PD-L1 TC ≥25% across bTMB levels (PD-L1 TC ≥25%/bTMB≥20 HR 0.79 [95% CI 0.45, 1.39]; PD-L1 TC ≥25%/bTMB < 20 HR 0.64 [95% CI 0.45, 0.90]). In contrast, D+T improved OS vs CT in pts with bTMB≥20 across different PD-L1 TC expression levels (Table; PD-L1 TC ≥25%/bTMB≥20 HR 0.44 [95% CI 0.23, 0.84]; PD-L1 TC < 1%/bTMB≥20 HR 0.42 [95% CI 0.17, 0.97]). Additional cutoffs and outcomes in subgroups defined by both biomarkers will be presented. Conclusions: These exploratory analyses from MYSTIC support PD-L1 TC expression as an appropriate predictive biomarker for OS with D vs CT, while suggesting bTMB as a predictive biomarker for OS with D+T in mNSCLC. These biomarkers appear to be independent and both may be important for mNSCLC treatment decisions. Interpretation of these data may be limited by small sample sizes; further investigations are warranted. Clinical trial information: NCT02453282. [Table: see text]