CELL INTELLIGENCE · 肿瘤细胞治疗研究

BACKGROUND: Despite immunotherapy chemotherapy has transformed the therapeutic landscape for patients with non-small cell lung cancer (NSCLC), critical questions remain regarding how detailed smoking history affects the evolving treatment options and the underlying molecular mechanisms driving these effects. METHODS: We analyzed 4157 patients with advanced NSCLC who were treated with immunotherapy monotherapy (IO alone) (n=2768) or chemoimmunotherapy (chemo-IO) (n=1389) at the Dana-Farber Cancer Institute and Memorial Sloan Kettering Cancer Center (2010-2023). Associations between detailed smoking history (status and cumulative pack-years) and clinical outcomes were assessed using multivariable analyses. First-line chemo-IO versus IO alone was compared in programmed death receptor ligand 1 (PD-L1) Tumor Proportion Score (TPS) of 50% EGFR/ALK wild-type patients. Tobacco smoking-related mutational signature (TSMS) was inferred from the targeted next generation sequencing (NGS) panel. We investigated relationships between detailed smoking history and tumor genomics/transcriptomics, PD-L1 expression, tumor-infiltrating lymphocytes, and circulating plasma proteomics. RESULTS: In patients receiving IO alone, both smoking status and intensity showed dose-dependent associations with improved response and survival outcomes. In contrast, among patients receiving chemo-IO, smoking history did not affect initial response but patients with active (HR=0.73, 95% CI 0.57 to 0.94, p=0.01) and heavy tobacco use (HR=0.76, 95% CI 0.62 to 0.93, p=0.001) showed improved progression-free survival (PFS), and a trend toward improved overall survival (OS). Importantly, these associations remained independent of STK11 , KEAP1 , and KRAS co-mutation status. In patients with PD-L1 TPS of 50% lacking EGFR/ALK alterations, patients who do not smoke derived significant benefit from first-line chemo-IO versus IO alone with higher response rates (70.0% vs 23.9%, p=0.001), prolonged PFS (median PFS 9.5 vs 3.7 months, HR=0.51, 95% CI 0.27 to 0.95, p=0.04), and a trend toward prolonged OS, while patients who smoke showed comparable outcomes with either strategy. TSMS independently predicted improved outcomes of IO alone, even after tumor mutational burden (TMB) adjustment. Molecular analyses revealed associations between tobacco use and higher TMB, increased tumor-infiltrating lymphocytes (CD8 + , PD-1 + , CD8 + PD-1 + , FOXP3 + ) and distinct plasma protein profiles involved in immune signaling pathways (CCL7, CXCL17, CDCP1, TNFRSF6B). CONCLUSIONS: Detailed smoking history provides crucial insights for optimizing IO selection in advanced NSCLC through mechanistic alterations in both the tumor microenvironment and systemic plasma protein profiles.