KRAS mutational status and efficacy in KEYNOTE-189: Pembrolizumab (pembro) plus chemotherapy (chemo) vs placebo plus chemo as first-line therapy for metastatic non-squamous NSCLC

Abstract

Background : KRAS mutations are observed in ∼25% of lung adenocarcinomas, with some studies suggesting it is a prognostic factor in NSCLC. We assessed the prevalence of KRAS mutations and their association with efficacy in an exploratory analysis of the KEYNOTE-189 study of pembrolizumab plus pemetrexed and platinum chemotherapy vs placebo plus chemotherapy as first-line therapy for metastatic non-squamous NSCLC (NCT02578680). Methods: Whole-exome sequencing (WES) was used to assess KRAS status and tumor mutational burden (TMB) in participants (pts) who had available tumor and matched-normal tissue. Descriptive analyses of the association of KRAS and KRAS G12C status with efficacy were included as part of this exploratory analysis, as was correlation between KRAS mutational status and shifts in TMB and PD-L1 expression distributions. Results: 289 (47%) of 616 pts had evaluable WES data from both tumor and normal DNA. 89 (31%) of 289 pts had KRAS mutation, including 37 (13%) G12C carriers. Pts with vs without KRAS mutations tended to have higher PD-L1 TPS (median [IQR] 30% [1-71] vs 5% [0-60]) and higher TMB (median [IQR] 204 [137-276] vs 141 [85-252] mut/exome). Outcomes of pembrolizumab plus chemotherapy and of placebo plus chemotherapy in pts with and without KRAS mutation and in KRAS G12C carriers are summarized in the table. 95% CIs were wide given the modest frequency of KRAS mutation, particularly KRAS G12C, and the 2:1 randomization that resulted in small populations for placebo plus chemotherapy.