Background: Small cell lung cancer (SCLC) is the most lethal form of lung cancer. A key driver of this near universal lethality is SCLC’s recalcitrance to therapy. The standard of care for SCLC is a chemotherapy doublet of etoposide and cis/carboplatin in combination with immunotherapy (EP+IO). While most tumors respond to these therapies, >80% of SCLCs progress within one year of treatment. Additionally, the aggressive clinical course of SCLC has historically precluded collection of patient-matched treatment naïve and recurrent tumor tissue traditionally required to study evolution. In lieu of tissue, several groups have used circulating tumor DNA (ctDNA) to study SCLC evolution. However, these studies used targeted sequencing panels that analyze <0.04% of the SCLC genome. Using new methods to comprehensively trace how SCLCs rapidly evolve to therapy resistance is a significant unmet need. Methods: Medical records were reviewed to identify treatment naïve SCLC patients treated at MD Anderson that had both pre- and post-EP+IO liquid biopsies collected under IRB approved protocols. ctDNA was isolated from plasma using a MiniMax High Efficiency Cell-Free DNA Isolation Kit (Apostle Bio). Germline DNA was extracted from buffy coat samples using a QIAamp DNA Blood Midi Kit (Qiagen). Deep whole genome sequencing (WGS) (>100X) of ctDNA and WGS of germline DNA (>60X) was performed using an Illumina NovaSeq X sequencer. Reads were aligned to the hg38 reference genome using BWA-mem. Somatic mutations were called using MuTect2, and clonal and subclonal copy number alterations were identified using Battenberg. DPClust was used to reconstruct cancer cell populations, leveraging mutation and copy number calls. Results: We identified several deleterious genomic alterations impacting TP53 and RB1 tumor suppressor genes, known genetic drivers of SCLC development. We also detected recurrent, whole chromosome arm loss of heterozygosity events impacting many chromosomes in treatment naïve samples, some of which are known to be present in >80% of SCLCs. Additionally, we found that both treatment naïve and recurrent SCLCs were composed of multiple subclones, each defined by dozens to thousands of unique somatic mutations. COSMIC signature analysis identified high activity of tobacco carcinogens, APOBEC, and replicative-based processes across ctDNA samples. We also identified emergence of novel subclones carrying unique copy number alterations following EP+IO recurrence. Importantly, the affected regions encode key effectors that regulate anti-tumor immune responses and cell survival following DNA damage, both of which are intimately related to the mechanism of action of EP+IO. Conclusions: Our data shows that deep WGS of ctDNA recapitulates known genetics of SCLC and allows for sensitive characterization of therapy resistant clones. Using this method, we have captured snapshots of both intrinsic and acquired resistance mechanisms to frontline EP+IO. Deep WGS of ctDNA is a promising approach to comprehensively dissect routes of SCLC evolution and therapy resistance.

Citation Format: Benjamin B. Morris, Zhihui Zhang, Simon Heeke, Kyle Concannon, Bingan Zhang, Waree Rinsurongkawong, Vadeerat Rinsurongkawong, J. Jack Lee, Jianjun Zhang, Don L. Gibbons, Ara A. Vaporciyan, Carl M. Gay, Hai T. Tran, John V. Heymach, Lauren A. Byers, Peter Van Loo. Dissecting patterns of small cell lung cancer evolution using deep whole genome sequencing of circulating tumor DNA [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Translating Cancer Evolution and Data Science: The Next Frontier; 2023 Dec 3-6; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_2):Abstract nr A022.