Introduction -Prior studies evaluating the efficacy of first line (1L) immune checkpoint inhibitor (ICI) monotherapy or combined chemoimmunotherapy in advanced NSCLC found improved outcomes with chemoimmunotherapy independent of PD-L1 and KRAS mutation status. As such, combined chemoimmunotherapy was proposed as the preferred comparator arm for 1L trials in KRAS mutated (KRASmt) NSCLC. Herein, we report a multimodal, real world data (RWD) analysis for outcomes with 1L therapy in patients with advanced NSCLC stratified by KRAS mutation status and PD-L1 levels.

Patients and Methods – Deidentified, multimodal RWD from the Tempus database was utilized to retrospectively analyze 1,980 patients with advanced NSCLC receiving 1L ICI containing therapy. Patients were stratified by tumor KRAS mutational status. Subgroup analyses were performed using Cox model stratified by KRAS mutational status, PD-L1 levels, and the presence of pathogenic alterations in STK11, KEAP1 and TP53.

Results – KRAS mutations were identified in 33.4% (662/1980) of patients. There was a higher proportion of PD-L1 high tumors in the KRASmt to KRAS wild-type cohort. Among KRASmt NSCLC, median overall survival (mOS) was longest in the PD-L1 high cohort. Patients with KRAS G12C and PD-L1 high tumors had the longest mOS at 30.28 months. Finally, pathogenic mutations in KEAP1 and STK11 correlated with worse outcomes in KRASmt tumors.

Conclusions – Outcomes following 1L therapy in KRASmt advanced NSCLC varied based on PD-L1 levels and the presence of STK11 and KEAP1 co-mutations, indicating that KRASm NSCLC represents a heterogeneous disease.

Micro abstract – KRAS mutations are the most common oncogenic driver alterations in NSCLC and influence response to immune checkpoint inhibitors. We utilized a multimodal, real world database to retrospectively analyze outcomes in patients with advanced NSCLC receiving first-line ICI containing therapy stratified by KRAS mutation status. We found that outcomes in KRAS mutant advanced NSCLC varied based on PD-L1 levels and the presence of co-mutations in STK11 and KEAP1.

VIEW THE PUBLICATION