Abstract 1202: Signatures of selection for biallelic inactivation in tumor suppressor genes across cancer types

Abstract

Abstract Background: The canonical model of tumor suppressor gene (TSG)-mediated oncogenesis, Knudson’s “two hit hypothesis,” posits that loss of both alleles is necessary for TSG inactivation. Many key TSGs accordingly exhibit near universal biallelic loss in associated cancers (e.g., APC in colorectal cancer and RB1 in retinoblastoma). However, because the majority of large-scale cancer genomics studies have utilized non-allele-specific copy number analysis methods, neither the extent of biallelic inactivation across TSGs, nor patterns of selective pressure for biallelic inactivation of TSGs, nor the functional and translational consequences of biallelic inactivation, are well-understood. Methods: Here, through allele-specific analysis of sequencing data from 48,179 cancer patients, we define the prevalence, selective pressure for, and functional consequences of biallelic inactivation across TSGs. We classify TSGs according to their patterns of selection for or against biallelic inactivation across cancer types. Finally, we investigate in detail the association of biallelic status with gene expression and clinical outcome for two TSGs, KEAP1 (in lung adenocarcinoma) and APC (in lung and prostate adenocarcinoma). Results: TSGs largely assort into distinct classes associated with either pan-cancer (Class I) or lineage-specific (Class II) patterns of biallelic loss, while some TSGs, mostly transcription factors, were predominantly monoallelic (Class III/IV). We discover that gene expression and co-mutation patterns indicate selection for biallelic losses in noncanonical contexts, including APC in lung and prostate adenocarcinomas. Selection for biallelic inactivation is also significantly elevated among variants of unknown significance (VUS) of several TSGs, including in KEAP1 in lung adenocarcinoma, nominating these VUS as worthy of in depth functional characterization. Patients with KEAP1 VUS also show similarly poor survival and response to chemoimmunotherapy to patients with known KEAP1 driver mutations, indicating that KEAP1 VUSs phenocopy established KEAP1 oncogenic alleles, and that zygosity, rather than variant classification, is prognostic of therapeutic response. Conclusion: TSGs vary significantly in their patterns of biallelic inactivation, and can be classified according to whether and how consistently, across cancer types, they are selected for loss of both copies. Additionally, taking into account the zygosity of TSG alteration events is often necessary to understand their phenotypic and clinical consequences, and can be useful in identifying novel driver events. Citation Format: Mark R. Zucker, Maria A. Perry, Arielle Elkrief, Anton Safonov, Rohit Thummalapalli, Samuel I. Gould, Debyani Chakravarty, Rose Brannon, Marc Ladanyi, Pedram Razavi, Mark T. Donoghue, Yonina R. Murciano-Goroff, Francisco J. Sánchez-Rivera, Yuan Chen, Ronglai Shen, Sarat Chandarlapaty, David B. Solit, Nikolaus Schultz, Michael F. Berger, Adam J. Schoenfeld, Jason Chang, Ed Reznik, Chaitanya Bandlamudi. Signatures of selection for biallelic inactivation in tumor suppressor genes across cancer types [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1202.