Abstract Only about 40% of patients with non-muscle invasive bladder cancer (NMIBC) respond to Bacillus Calmette-Guérin (BCG) therapy. While immune cell exhaustion has been linked to BCG resistance, early disease dissemination (i.e., occult muscle-invasive disease) may also contribute to BCG failure. We assessed whether serial analysis of urine and plasma samples could reveal signs of immune exhaustion, residual disease, and early dissemination, predictive of BCG outcome. In a prospective cohort of 102 BCG-treated NMIBC patients (median follow-up 28 months; inclusion ended April 2024), whole-genome DNA sequencing (WGS) was used to generate patient-specific genomic-wide tumor-based signatures (Veracyte) from paired buffy coat and tumor or urine pellet (UPEL) DNA. These were applied to cfDNA WGS data from 271 urine and 329 plasma samples collected before, during, and after treatment to detect urinary tumor (ut)DNA and circulating tumor (ct)DNA. In parallel,344 urine supernatants were profiled using Olink immuno-oncology panels, and total RNA sequencing was performed on 55 tumors. Updated follow-up will be presented at the meeting. Plasma ctDNA was detected pre-BCG in 5 (6%) patients; three later developed high grade (HG) recurrence or progressed. Pre-treatment ctDNA positivity was associated with higher recurrencerisk (HR = 6.88) but not with tumor stage (p 0.42) or grade (p =0.91). Urine utDNA was detected in 58%, 51% and 27% of patients before, during, and after BCG, respectively. utDNA-positivity during or after BCG correlated with shorter recurrence-free survival. Recurrence was observed in 10/22 (45%) patients with detectable post-BCG utDNA and in 11/49 (22%) patients without measurable utDNA (p=0.022). On average, utDNA preceeded clinical detection with 133 days (range: 4-405). UPEL DNA enabled patient-specific model generation and ctDNA/utDNA detection in cases with limited tumor material (CIS only). BCG therapy induced increased urinary immuno-oncology proteins. Tumor-associated proteins (MMP-12, IL6, IL8) were elevated pre-BCG, while eight immunemodulatory proteins - including HO-1, GZMA, GZMB, LAP TGF-beta-1 and TRAIL were upregulated early during treatment in HG-recurrent versus non-HG-recurrent cases. Pre-treatment ctDNA identified patients at high risk of HG recurrence or progression, supporting a role as a potential restaging tool. UPEL DNA was a reliable tumor proxy and useful when tissue was limited (e.g., CIS). utDNA detection identified patients at high risk of recurrence and earlier than cystoscopy. Urine proteomics mirrored BCG-induced immune responses, with distinct early on-treatment profiles in recurrent versus non-recurrent patients. Together, ctDNA/utDNA and proteomic monitoring may enable early detection of treatment failure and guide timely therapeutic interventions. Citation Format: Trine Strandgaard, Tine Ginnerup Andreasen, Iver Nordentoft, Nathalie Demuth Fryd, Philippe Lamy, Julie Pedersen Harrits, Boris Oklander, Danielle Afterman, Tal Katz-Ezov, Imane Bourzgui, Jørn Jakobsen, Kristian Juul, Jørgen Bjerggaard Jensen, Lars Dyrskjøt. Early detection of treatment failure and disease dissemination inBCG-treated NMIBC using liquid biopsies and urine proteomics 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 1139.
Strandgaard et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: