Transcriptomic profiling of lineage transformation and actionable therapy opportunities in EGFR-mutant non-small cell lung cancer evolving to small cell lung cancer.

BACKGROUND: Neuroendocrine transformation of epidermal growth factor receptor-mutant lung adenocarcinoma (EGFRm LUAD) represents a significant mechanism underlying acquired resistance to EGFR tyrosine kinase inhibitors (TKIs). This histological transformation lacks effective treatments. We comprehensively discussed the treatments of transformed small cell lung cancer (T-SCLC) and analyzed the transcriptomic profile of EGFRm LUAD and paired T-SCLC, aiming to understand the immune landscape of this T-SCLC subtype. METHODS: Patients with pathologically confirmed lung adenocarcinoma (LUAD) at baseline who subsequently developed T-SCLC confirmed by re-biopsy after antitumor therapy resistance were included. For genomic and transcriptomic profiling, paired pre- and post-transformation specimens from EGFR-mutant patients were subjected to next-generation sequencing and RNA sequencing. Additionally, we integrated gene expression data for primary SCLC (P-SCLC) from the GEO database to compare the transformed and primary SCLC types. RESULTS: A total of 32 patients with pathologically confirmed LUAD at baseline who subsequently developed T-SCLC confirmed by re-biopsy after antitumor therapy resistance were included in this study. Among them, the median progression-free survival (mPFS) and overall survival (mOS) of T-SCLC with first-line immunotherapy plus chemotherapy were 5.1 and 21.0 months, respectively. Analysis of post-transformation treatments received after T-SCLC diagnosis revealed that patients exposed to immunotherapy had longer mOS than those who were immunotherapy-naïve (21.0 m vs. 10.0 m, P = 0.03). Our data revealed that paired T-SCLC exhibited an immune-suppressive microenvironment compared with EGFRm LUAD, which included the downregulation of multiple immune regulatory pathways, reduced immune cell infiltration, decreased scores of immune-related signatures and downregulation expression of immune checkpoint-related genes. Additionally, we found that T-SCLC exhibited a more immunologically active microenvironment than P-SCLC, as evidenced by higher scores of immune-related signatures, including MHC signature, T cell-inflamed GEP signature and IFN-gamma signature. CONCLUSIONS: The immune landscape of T-SCLC is more immunosuppressed than EGFRm LUAD, yet more immunologically active than that of P-SCLC. This immune profile provides a rationale for integrating immunotherapy with chemotherapy in T-SCLC.