Extensive-stage small-cell lung cancer in patients receiving atezolizumab plus carboplatin–etoposide: stratification of outcome based on a composite score that combines gene expression profiling and immune characterization of microenvironment

Purpose Small-cell lung cancer (SCLC) is an aggressive disease with a dismal prognosis. The addition of immune checkpoints inhibitors to standard platinum-based chemotherapy in first-line setting achieves a durable benefit only in a patient subgroup. Thus, the identification of predictive biomarkers is an urgent unmet medical need. Experimental design Tumor samples from naive extensive-stage (ES) SCLC patients receiving atezolizumab plus carboplatin–etoposide were analyzed by gene expression profiling and two 9-color multiplex immunofluorescence panels, to characterize the immune infiltrate and SCLC subtypes. Associations of tissue biomarkers with time-to-treatment failure (TTF), progression-free survival (PFS) and overall survival (OS), were assessed. Results 42 patients were included. Higher expression of exhausted CD8-related genes was independently associated with a longer TTF and PFS while increased density of B lymphocytes correlated with longer TTF and OS. Higher percentage of M2-like macrophages close to tumor cells and of CD8+T cells close to CD4+T lymphocytes correlated with increased risk of TF and longer survival, respectively. A lower risk of TF, disease progression and death was associated with a higher density of ASCL1+tumor cells while the expression of POU2F3 correlated with a shorter survival. A composite score combining the expression of exhausted CD8-related genes, B lymphocyte density, ASCL1 tumor expression and quantification of CD163+macrophages close to tumor cells, was able to stratify patients into high-risk and low-risk groups. Conclusions In conclusion, we identified tissue biomarkers and a combined score that can predict a higher benefit from chemoimmunotherapy in ES-SCLC patients.