Abstract 3262: Whole genome sequencing of ctDNA enables disease monitoring and residual disease detection in head and neck cancer patients

Abstract Detection of circulating tumor DNA (ctDNA) has clinical value as a non-invasive method for therapeutic response monitoring and detection of minimal residual disease (MRD). In this study, we employed a tumor-informed approach for detecting ctDNA from plasma in a cohort of head and neck cancer (HNSCC) patients with matched sets of FFPE tissue (FFPET), peripheral blood, and plasma samples. Patients with locally advanced HNSCC were treated with induction chemoimmunotherapy, followed by surgical resection, and risk-adapted adjuvant therapy. Plasma was collected pre-treatment, post-induction (on-treatment), and after completion of all treatment (post-treatment). By leveraging whole genome sequencing (WGS), we can overcome the sparsity of ctDNA fragments in low tumor fraction (TF) settings via tracking of hundreds to thousands of single nucleotide variants (SNVs). Our approach identifies a patient’s unique compendium of tumor-related SNVs from their FFPET sample (after germline subtraction) and tracks this compendium of variants in their plasma samples to detect the presence of ctDNA. Three patients failed quality control criteria and were excluded from further analysis due to the presence of high proportions of C>T:G>A and C>A:G>T variants, known to be FFPE-related and oxidative damage-related artifacts, respectively. Three plasma samples from the post-induction timepoint were also excluded due to insufficient material for sequencing. Post-induction ctDNA was detected in 13 of 22 evaluable samples. Among the ctDNA positive (ctDNA+) post-induction patients, all of them (13/13) had measurable residual tumors on pathologic review, while (2/9) in the ctDNA negative (ctDNA-) group had measurable tumors. For patients with measurable tumors, mean tumor size in the ctDNA+ group (n=13) was 4 cm, while the mean in the ctDNA- group (n=2) was 2.6 cm. Post-treatment ctDNA was detected in 2 of 18 evaluable samples. Among the ctDNA+ post-treatment patients, 100% (2/2) had disease recurrence within 6 months, while disease recurrence was observed in 12.5% (2/16) of the ctDNA- group (occurring 7- and 32-months post-treatment completion). Lastly, we estimated plasma tumor fraction (TF) following Zivran et al (2020). ctDNA+ post-induction samples had a mean TF of 2.4 x 10-4 (n=13; ranging from 2.3 x 10-5 to 1.1 x 10-3) and ctDNA+ post-treatment samples had an order of magnitude lower mean TF of 5.8 x 10-5 (n=2; ranging from 5.7 x 10-5 to 5.8 x 10-5) indicating partial molecular response to treatment even in the presence of detectable residual ctDNA. In this albeit small clinical study, our tumor-informed ctDNA assay achieved clinical sensitivity of 87% and 50% at post-induction and post-treatment, respectively, and clinical specificity of 100% at both time points. Our findings underscore the utility of ctDNA as a promising biomarker for disease monitoring in head and neck cancer patients. Citation Format: Shetal A. Patel, Melanie Lou, Eddie Ho, Michelle Harwood, Seng Saelee, Dinesh Kumar, Steven Johnson, Siddarth Sheth, Jeffrey Blumberg, Catherine Lumley, Wendell Yarbrough, Michael K. Gibson, Jessica Bauman, Trevor Hackman, Jared Weiss, Sowmi Utiramerur. Whole genome sequencing of ctDNA enables disease monitoring and residual disease detection in head and neck cancer patients [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 3262.