Atezolizumab Plus Best Supportive Care Improves Disease-Free Survival in eNSCLC 

Cancer-associated fibroblasts (CAFs) constitute a critical element of the tumor microenvironment, and TGFβ-induced CAFs have been implicated in poor prognosis across various indications. The IMpower010 study assessed adjuvant atezolizumab (atezo) in resected early-stage non-small cell lung cancer (eNSCLC) post-chemotherapy, leading to approval for patients with PD-L1 expression on tumor cells (TC) ≥1% or ≥50% (SP263) in stage II-IIIA NSCLC after R0 resection. This study delves into exploratory analyses utilizing IMpower010 RNA sequencing (RNAseq) data to identify predictive biomarkers for atezo outcomes. Among the 1,005 randomized patients (intent-to-treat), 500 constituted the RNAseq biomarker-evaluable population. The study focused on assessing various features, including cell-type and -state gene signatures, from baseline tumor samples obtained post-resection. Variable importance scores, calculated through the generalized random forests method, identified the TGFβ CAF gene signature as the most influential. Disease-free survival (DFS) hazard ratios (HRs) were notably favorable in the TGFβ CAF-high (≥ median) population with HRs of 0.54 (95% CI: 0.37, 0.80), while the TGFβ CAF-low (< median) population had HRs of 0.94 (95% CI: 0.63, 1.41). Further analyses revealed that the improved outcomes with atezo versus best supportive care (BSC) in the TGFβ CAF-high population were primarily driven by a shorter DFS in the BSC arm. Importantly, these benefits were consistent across histology and PD-L1 expression levels. Collectively, these findings suggest that patients with an enrichment of the TGFβ CAF gene signature may experience improved DFS after adjuvant atezo following surgical resection, highlighting the potential for this biomarker to guide treatment decisions in the adjuvant setting. The study underscores the importance of understanding the tumor microenvironment for personalized treatment strategies in eNSCLC.