Abstract
Atezolizumab (anti-PD-L1), combined with carboplatin and etoposide (CE), is now a standard of care for extensive-stage small-cell lung cancer (ES-SCLC). A clearer understanding of therapeutically relevant SCLC subsets could identify rational combination strategies and improve outcomes. We conduct transcriptomic analyses and non-negative matrix factorization on 271 pre-treatment patient tumor samples from IMpower133 and identify four subsets with general concordance to previously reported SCLC subtypes (SCLC-A, -N, -P, and -I). Deeper investigation into the immune heterogeneity uncovers two subsets with differing neuroendocrine (NE) versus non-neuroendocrine (non-NE) phenotypes, demonstrating immune cell infiltration hallmarks. The NE tumors with low tumor-associated macrophage (TAM) but high T-effector signals demonstrate longer overall survival with PD-L1 blockade and CE versus CE alone than non-NE tumors with high TAM and high T-effector signal. Our study offers a clinically relevant approach to discriminate SCLC patients likely benefitting most from immunotherapies and highlights the complex mechanisms underlying immunotherapy responses.
Original language | English (US) |
---|---|
Pages (from-to) | 429-443.e4 |
Journal | Cancer cell |
Volume | 42 |
Issue number | 3 |
DOIs | |
State | Published - Mar 11 2024 |
Keywords
- atezolizumab
- immune checkpoint blockade
- immunotherapy
- IMpower133
- molecular subtyping
- small cell lung cancer
- transcriptomics
- tumor-associated macrophages
ASJC Scopus subject areas
- Oncology
- Cancer Research