4577 Background: Radical cystectomy (RC) is a life-altering operation associated with substantial morbidity. Despite its curative intent for muscle-invasive urothelial carcinoma (MIUC), approximately 50% of patients experience metastatic recurrence after surgery. Identifying patients destined to relapse despite RC could help spare them from the morbidity of surgery and instead direct them toward alternative systemic therapies. While ctDNA-positive assays post-neoadjuvant therapy (NAT) but pre-cystectomy are associated with a poor prognosis, a subset of patients with ctDNA positive assays pre-cystectomy “clear” ctDNA with cystectomy alone (Powles, ASCO, 2025). We hypothesized that quantitative pre-cystectomy ctDNA thresholds may therefore distinguish local, surgically curable, bladder cancer versus micrometastatic disease. Methods: We conducted a retrospective single-institution cohort study of patients undergoing RC who had tumor-informed ctDNA (Signatera) assays ≤60 days pre-surgery and within ±30 days of a 3-month postoperative landmark. The primary outcome was ctDNA positivity at this landmark. The analyses proceeded in two steps: (1) association between log-transformed pre-cystectomy ctDNA and 3-month ctDNA positivity, assessed using logistic regression; and (2) derivation of the maximal quantitative preoperative ctDNA threshold >0.02 Mean Tumor Molecules per Milliliter (MTM/ml) using constrained optimization (target sensitivity ≥80%, specificity ≥50%), with internal validation via Efron-Gong bootstrap resampling. Results: Of 45 patients with MIUC meeting inclusion criteria, 10 (22.2%) received NAT. Higher pre-cystectomy ctDNA correlated with ctDNA positivity at the post-cystectomy landmark timepoint (odds ratio per 10-fold increase 1.6, 95% CI 1.2–2.4; p=<0.01). A pre-cystectomy ctDNA threshold ≥0.06 MTM/ml (95% CI 0.04–2.475) demonstrated optimal performance characteristics and was confirmed on bootstrap validation (sensitivity 81.6% 95% CI 61.1–88.9%, specificity 66.4% 95% CI 51.9–96.3%, positive predictive value PPV 63.8% 95% CI 55.2–91.7%, and negative predictive value NPV 85.1% 95% CI 78.8–88.9%). Conclusions: Higher pre-cystectomy ctDNA values correlate with a higher likelihood of a persistent ctDNA-positive assay at 3 months post-cystectomy. In this dataset, a ctDNA threshold of ≥0.06 MTM/ml identified patients with high sensitivity and NPV for post-operative micrometastatic disease. Further validation of these findings, and threshold, in larger cohorts (with or without NAT) are needed to refine MIUC care by testing strategies involving (alternative) NAT versus proceeding with cystectomy.
Ayasun et al. (Wed,) studied this question.