Abstract 3770: Comparison of ctDNA detection methods for monitoring minimal residual disease in patients with bladder cancer: Insights from the TOMBOLA Trial

Abstract Introduction: There is a critical need for biomarkers to assess treatment efficacy and detect minimal residual disease (MRD), enabling timely treatment initiation and response monitoring. Tumor-informed detection of mutations in cell-free DNA (cfDNA) has shown promise in monitoring MRD. In the TOMBOLA trial (NCT04138628), circulating tumor DNA (ctDNA) is detected using patient-specific droplet digital PCR (ddPCR) to guide immunotherapy. However, due to the design process of personalized ddPCR assays, standardization and clinical usability may be challenging. Whole-genome sequencing (WGS) is an alternative approach that enables tumor-informed ctDNA detection and may improve overall clinical utility. To comprehensively evaluate the performance of WGS compared to ddPCR for ctDNA detection, we analyzed ctDNA data from ddPCR and WGS for 1, 282 paired plasma samples from 100 patients enrolled in the TOMBOLA trial. Methods: Plasma was collected in 10mL EDTA tubes at diagnosis, during neoadjuvant chemotherapy, at cystectomy (RC), post-RC, and during immunotherapy. To detect ctDNA, we used 2 tumor-informed approaches: ddPCR and WGS. For ddPCR, cfDNA was extracted from 8 mL of plasma (n = 1282) and 4 assays were designed per patient based on whole-exome sequencing of tumor/germline DNA (mean coverage: 380x/239x). WGS was conducted on cfDNA extracted from 2 mL of plasma (mean coverage: 30x) and tumor/germline DNA to build patient-specific, genomic-wide tumor signatures (mean coverage: 59x/24x). Results: We compared paired plasma samples and found an 82.9% (1063/1282) concordance in ctDNA status between ddPCR and WGS. Among the inconsistent ctDNA calls, 4.2% (54/1282) were only positive with WGS, while 12.9% (165/1282) were only positive with ddPCR. Inconsistencies were primarily observed in samples with low tumor fractions. Among ctDNA-positive samples (both tests), tumor fraction estimates were strongly correlated (r = 0.82; p<0.0001). Post-RC ctDNA status at a 6-month landmark identified patients with recurrence with 80% sensitivity for both ddPCR and WGS, and specificity of 62% for ddPCR and 87% for WGS. The median lead time over imaging was 99 days for ddPCR and 93 days for WGS. Additionally, ctDNA status at the 6-month landmark was significantly associated with recurrence-free survival (RFS); ddPCR, HR = 4.9 (p = 0.0005):WGS, HR = 11 (p<0.0001). The 12-month RFS for ctDNA negative patients were 97.8% for ddPCR and 98.5% for WGS. Conclusion: ctDNA detection in plasma using ddPCR and WGS methods showed overall high concordance; however, 12.9% of the samples were only positive with ddPCR, indicating some discordance between methods. Despite these inconsistencies, the predictive potential of ctDNA status at the 6-month landmark was comparable between methods. The clinical impact of samples only positive with ddPCR remains to be determined. Citation Format: Iver Nordentoft, Karin Birkenkamp-Demtröder, Sia Viborg Lindskrog, Trine Strandgaard, Rikke Vilsbøll Milling, Stefanie Korsgaard Körner, , Simone Burchardt Brandt, Yarin Cohen, Tomer Lauterman, Gitte Lam, Line Hammer Dohn, Knud Fabrin, Andreas Carus, Astrid Petersen, Ulla Nordström Joensen, Helle Pappot, Per Søndergaard Holt, Niels Viggo Jensen, Tal Katz-Ezov, Danielle Afterman, Boris Oklander, Mads Agerbæk, Jørgen Bjerggaard Jensen, Lars Dyrskjøt. Comparison of ctDNA detection methods for monitoring minimal residual disease in patients with bladder cancer: Insights from the TOMBOLA Trial [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 3770.