Abstract 3771: Non invasive monitoring of relapsed SCLC using cell free fragmentomes and peripheral blood immunophenotyping

Abstract Background: Small cell lung cancer (SCLC) is the most aggressive type of lung cancer with 5-year survival of 7%. Besides chemotherapy, immune checkpoint blockade (ICB) is part of the treatment regimen but only a subset of patients derive benefit. Tumor tissue can identify distinct subtypes with therapeutic vulnerabilities. Need for prompt treatment and limited access to biopsies at relapse highlight the demand for noninvasive methods to monitor response. Methods: We used sequential blood samples from relapsed/refractory SCLC patients enrolled in 2 clinical trials—durvalumab and olaparib (NCT02484404) or M7824 with/without chemotherapy (NCT03554473). Cell-free DNA (cfDNA) was isolated from plasma, followed by whole-genome sequencing with 30x depth. To infer tumor gene expression profiles, we analyzed genome-wide signals of tissue-specific transcription factors binding sites (TFBS) differentially regulated in SCLC. All the samples were analyzed to obtain tumor fraction (fx), ASCL1 TFBS and Z-scores for copy number variation(CNV). Clinical data, matched tumor transcriptomes, and peripheral blood flow cytometry were integrated in orthogonal analyses. Results: A total of 109 samples from 40 patients were included in the analysis. Low baseline tumor fx was associated with clinical benefit (p = 0.007), longer overall survival (OS) (p = 0.01), and absence of liver metastasis (p = 0.04). Sequential analysis revealed a significant decrease in tumor fx in partial responders (PR) (p = 0.002), while an increase in tumor fx correlated with tumor growth as measured by RECIST (p = 0.003). The ASCL1 TFBS successfully stratified SCLC into high and low neuroendocrine (NE) groups, which was concordant with the tumor transcriptome NE50 score (r = 0.3, p = 0.05) and the SCLC-nonNE score (r = -0.31, p = 0.04). A decrease in ASCL1 TFBS was observed in the PR group, while an increase in ASCL1 TFBS correlated with an increase in tumor size (p = 0.002). Consistent with public SCLC tumor CNV data, we observed losses on chromosomes 3p and 17p, and gains on 5p and 3q. Flow cytometry analysis revealed that patients with high tumor fx at baseline had an immunosuppressed profile, characterized by significantly higher levels of myeloid-derived suppressor cells (MDSCs). Furthermore low ASCL1 TFBS levels were associated with higher dendritic cells and HLA-monocytes. Survival analysis revealed increase in MDSCs was associated with shorter OS, while higher HLA expression was associated with longer OS. Conclusions: Genome-wide cfDNA fragmentome analysis effectively captures genomic and transcriptomic features of SCLC, offering predictive and prognostic insights into SCLC response to ICB. Consistent with the role of tumor intrinsic features in driving tumor immunity, ctDNA derived features show concordance with peripheral blood immune phenotypes and supporting cfDNA as a non-invasive biomarker for treatment monitoring. Citation Format: Chirayu Mohindroo, Brett Schroeder, Min-Jung Lee, Rajesh Kumar, Shraddha Rastogi, Sunmin Lee, Nahoko Sato, Roshan Shrestha, Melissa Abel, Parth Desai, Nobuyuki Takahashi, Anish Thomas. Non invasive monitoring of relapsed SCLC using cell free fragmentomes and peripheral blood immunophenotyping [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3771.