Abstract A111: An evaluation of the classification and reporting of <i>STK11</i> and <i>KEAP1</i> variants in a real-world US database of metastatic NSCLC tumours

Abstract Introduction: Inactivating mutations in the STK11 (LKB1) and KEAP1 genes have been associated with poor prognosis in non-small cell lung cancer (NSCLC) patients. Currently, there are no approved treatment options aligned to STK11 and KEAP1; however, there is exploratory data from the POSEIDON trial indicating potential clinical utility of tremelimumab + durvalumab + platinum doublet chemotherapy in NSCLC patients with a mutation in one of these genes. STK11 and KEAP1 are included in many commercially available next generation sequencing (NGS) panels but are not yet cited in NSCLC guidelines as emerging biomarkers. Understanding the current test reporting landscape for STK11 and KEAP1, in terms of variant classification, is important in identifying potential challenges for the future use of these genes in treatment decision algorithm tools for metastatic NSCLC (mNSCLC). Methods: A real-world clinical testing database including results for STK11 (n=3535) and KEAP1 (n=3680) in mNSCLC specimens from multiple US laboratories between January 2020 and September 2023 was used to review variant classification and reporting practices. Four online variant databases (cBioPortal, ClinVar, gnomAD, TCGA) were used to retrospectively interrogate ∼200 STK11 and KEAP1 variants within the dataset to highlight variant classification discrepancies across these platforms, whereby a single variant is given different classifications dependent on the database used. Results: Real-world variant classification 22% (209/943) of STK11 variants and 36% (267/746) of KEAP1 variants within the dataset were classified by the US testing laboratories into categories such as pathogenic, likely pathogenic and variant of uncertain significance. No differences in classification were identified in the real-world dataset, meaning all classified variants were consistently aligned to the same classification group by the US testing laboratories contributing to the dataset. Results: Retrospective variant database classification Retrospective STK11 and KEAP1 variant classification using cBioPortal, ClinVar, gnomAD, and TCGA showed inconsistencies between databases in 19% (10/53) variants reviewed, highlighting potential for inconsistencies in variant classification in the real-world dependent on the variant classification processes used within testing laboratories. 146/199 variants interrogated were absent from all four databases. Conclusions: The widespread use of NGS-based testing for mNSCLC in the US, and the frequent inclusion of STK11 and KEAP1 genes within NGS panels, means that STK11 and KEAP1 data may often be available to clinicians. Guidelines regarding evidence-based variant interpretation utilizing a range of data sources, including variant databases, and recommendations to standardize classification terminology should be considered to mitigate potential inconsistency in variant classification for STK11 and KEAP1 and thereby facilitate properly informed clinical decisions. Citation Format: Helen Roberts, Katherine Keating, Christine Quinn, Emre Baser, Mehmet Berktas, Nikoletta Sidiropoulos. An evaluation of the classification and reporting of STK11 and KEAP1 variants in a real-world US database of metastatic NSCLC tumours [abstract]. In: Proceedings of the AACR IO Conference: Discovery and Innovation in Cancer Immunology: Revolutionizing Treatment through Immunotherapy; 2025 Feb 23-26; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Immunol Res 2025;13(2 Suppl):Abstract nr A111.