Abstract 2692: High allele frequency of KRAS functional mutations predicts resistance to MEK inhibitors: Evidence from cell lines and human tumor xenograft models

Abstract

Abstract Introduction KRAS mutant tumors represent a high unmet need for effective therapy. Even though there is no targeted therapy for KRAS, inhibitors of downstream targets like MEK have shown efficacy in preclinical studies. However, clinical trials of MEK inhibitors in KRAS-mutant tumors have not been successful, although there have been some responders. We sought to understand resistance to MEK inhibition in KRAS-mutant tumors and to develop biomarkers for patient stratification. Results Using DNA and RNA sequencing data from cell lines, we found that high variant allele frequency (>50%) of KRAS functional mutations (G12, G13, Q61) was associated with baseline resistance to MEK inhibitors AZD6244 and MEK162 across multiple tumor types. Low allele frequency of KRAS mutation was found to be necessary but not sufficient for sensitivity to MEK inhibitors, suggesting the existence of additional resistance mechanisms. Comparison of copy number and variant allele frequency of KRAS showed that MEK inhibitor-resistant cell lines have either amplified mutant alleles or lost WT alleles. This prediction of high KRAS mutant allele frequency being associated with MEKi resistance was successfully validated in vivo in 19 primary colorectal tumor xenograft models treated with MEK162. High KRAS mutant allele frequency and high KRAS expression were also associated with acquired resistance to BKM120+MEK162 combination treatment in an in vivo model of HeyA8, an ovarian cell line with a KRAS G12D mutation. This prediction was also supported by the observation in a Phase II clinical trial with MEK inhibitor refametinib + gemcitabine in pancreatic cancer [1]. Patients with disease progression had a significantly higher KRAS mutant allele frequency compared to partial responders and stable disease. Conclusion High allele frequency of KRAS functional mutations is associated with resistance to MEK inhibitors. Excluding patients with high KRAS mutant allele frequency has the potential to improve clinical trial outcomes and benefit patients, with the added benefit that it may not require any additional screening or novel biomarkers.