Abstract 5737: Precision oncology driven real world clinical genomics data mining of MET alterations, TMB, and PDL-1 to empower indication agnostic patient enrollment and combination strategies

Abstract

Abstract Background: Following MET inhibitor and checkpoint inhibitor approvals, it is essential to identify cancer indications that harbor molecular alterations of interest and a thorough real-world molecular genomics based characterization of those tumors would allow to expand cancer specific indications that might benefit from MET and checkpoint inhibitors and shed light on exploring clinically efficacious combo strategies. Importantly, such an approach provides the end-to-end solutions that shall significantly contribute to clinical translation and/or back translation strategies for innovative cancer therapies. Methods: We systemically interrogated FoundationInsights®, a dataset of real-world molecular genomic tumor profiling from North American patients, and investigated the prevalence of MET amplifications, METex14 skipping and MET fusions prevalence in 62110 unique patient samples across five indications (lung adenocarcinoma [LUAD], esophageal/gastroesophageal-junction/stomach adenocarcinoma [Eso/GEJ/Sto], papillary renal cell carcinoma, hepatocellular carcinoma [HCC] and glioblastoma). MET alteration prevalence was further categorized based on tumor cell PD-L1 IHC status (22C3 pharmDx) and TMB status. Additionally, comprehensive text mining from clinicaltrials.gov and landscape overview were conducted. Results: MET amplification was the most frequent MET alteration examined with a prevalence ranging from 1.4% in HCC to 4.8% in Eso/GEJ/Sto. In HCC and LUAD, MET amplification prevalence was higher in tumors with high TMB (≥10 mutations/megabase, P<0.05). In Eso/GEJ/Sto and LUAD, MET amplification occurrence was higher in PD-L1-high tumor vs. PD-L1-low or PD-L1-neg tumors (P<0.05). METex14 skipping mutations were observed in LUAD (2.3%), rarely in the other five tumor types analyzed (≤0.1%). Lastly in LUAD, METex14 skipping mutations were more frequent in PD-L1 high tumors but not those with high TMB (P<0.05). Text mining results indicate that- while most trials use MET amplification for patient stratification- there is a distinct tendency for MET inhibitor trials to increasingly include METex14 skipping as a major selection criterion as well. Conclusions: Real world genomics corroborated MET amplification as a predominant MET alteration across various indication. The potential clinical benefit to combine MET inhibitors with PD-1/PD-L1 blockades in specific indications and specific subsets of patients is further recognized. Collectively, the comprehensive molecular profiling & text mining approaches will continue to guide application of precision medicine in clinical trial design based on appropriate biomarker selection. Citation Format: Zheng Feng, Jean-Christophe Pignon, Radwa Sharaf, Natalie Romanov, Julien Doudement, Lee A. Albacker, Noriaki Kurata, Neil R. Smith, Nobutoshi Matsushita, Juergen Scheuenpflug. Precision oncology driven real world clinical genomics data mining of MET alterations, TMB, and PDL-1 to empower indication agnostic patient enrollment and combination strategies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5737.