Abstract 3536: Quantifying evolutionary dynamics and tumor heterogeneity in oncogene-addicted advanced non-small cell lung cancer

Abstract Non-small cell lung cancer (NSCLC) is characterized by oncogene addiction, where unique genetic, epigenetic, and transcriptomic alterations drive tumor initiation, growth, and survival. Tyrosine kinase inhibitors (TKIs) against actionable oncogenes showed initial promise, but acquired resistance remain significant clinical challenges. To define the evolutionary dynamics underlying resistance and tumor heterogeneity, we integrated WES/WGS and RNA sequencing data from 1,253 longitudinal biopsies and multi-site rapid autopsies from 104 patients, including an ALK-rearranged cohort. We developed a unified framework combining cutting-edge computational tools and novel statistical methods to characterize and quantify tumor heterogeneity under TKI selective pressure. In 2 ALK-rearranged cases with acquired resistance to TKIs, we identified subclones with multiple compound ALK mutations. In pt062, 2 ALK double-mutant subclones shared an ancestral resistance mutation C1156Y, and independently developed secondary hits L1198F and I1171S, where I1171S retains partial sensitivity to Lorlatinib, potentially revealing a transient therapeutic window before full resistance. In pt992, 3 ALK double-mutant subclones shared an original mutation F1193Y, and independently acquired secondary hits G1269A, E1210K, and G1210R, driving differential resistance to Crizotinib and Lorlatinib. In pt992, F1193Y may be a rare activating mutation that is insufficient to confer resistance to ALK TKIs alone, but promoted resistance in combination with secondary hits. Phylogenetic analysis revealed predominantly monoclonal seeding early in disease in pt062 vs polyclonal reseeding after the emergence of resistance in pt992. Metastatic route inference in pt992 suggests that a mix of distinct subclones result in lesions with triple ALK mutations. Intra-tumor heterogeneity (ITH) calculated from the clonal diversity index revealed that TKI exposure increased ITH, resulting in lesion-specific subclones. In pt992, divergent clonal dynamics in G1202R+ clones showed lower ITH than E1210K+ clones. The founding clones persisted at ∼50% (pt992) and 85% (pt062) tumor fraction post-TKI, indicating inadequate selective pressure. Copy number alteration- and transcriptomic-based ITH suggest possible structural differences in clonal populations between primary and metastatic tumors. Bray-Curtis dissimilarity scores of inter-tumor heterogeneity revealed 3 clusters in both cases, suggesting that metastatic lesions evolved from specific subclones. RNA-seq immune deconvolution via CIBERSORT showed variable immune cell infiltration patterns without correlation to clinical outcome. These results demonstrate that TKIs drive clonal diversification and evolution of resistance in oncogene-addicted NSCLC, and offer a quantitative framework for lineage-informed therapeutic strategies. Citation Format: Lyns C. Etienne, Elizabeth E. Martin, Pinar Eser, Kiara Pontious, Junko Tsuji, Natalie Lytell, Nicholas Chevalier, Mandeep Banwait, Jennifer L. Peterson, Michael S. Lawrence, Mari Mino-Kenudsen, Jessica J. Lin, Zofia Piotrowska, Aaron N. Hata, Rebecca Heist, Dejan Juric, Justin Gainor, Gad Getz. Quantifying evolutionary dynamics and tumor heterogeneity in oncogene-addicted advanced non-small cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3536.