Background: Cisplatin-based neoadjuvant chemotherapy (NAC) followed by cystectomy or chemoradiation is the standard of care for urothelial carcinoma (UC) patients with MIBC. Both cystectomy and chemoradiation have short and long-term toxicity and QOL implications. Mutations in DNA damage repair/response genes are associated with pathologic downstaging after NAC. We hypothesized that a combination of biomarker selection and clinical staging would identify patients prospectively for a cystectomy or chemoradiation avoidance algorithm. Methods: We conducted a single-arm, phase II, non-inferiority trial (NCT02710734) to evaluate a risk-adapted approach for MIBC. Patients with cT2-T3N0M0 UC underwent NAC with accelerated MVAC. Pre-NAC TURBT specimens were sequenced (Caris) for mutations in ATM, ERCC2, FANCC or RB1. Patients with ≥ 1 mutations and no clinical evidence of disease by restaging TUR, urine cytology and imaging post-NAC began pre-defined active surveillance (AS). Remaining patients underwent bladder-directed therapy. The primary endpoint was metastasis-free survival (MFS) at 2 years for the entire cohort. The risk-adapted approach would be declared non-inferior to the standard of care if the lower bound of an exact 1-sided 95% CI for MFS was > 64%. The study required 70 patients with a 4.5% type I error and 81.6% power. Results: 71 (ITT) patients were enrolled over 33 months at four academic centers. Median age was 70 years (47-83), 74% were male, 92% Caucasian, 81% ECOG PS 0 and 79% cT2. 90% completed 3 cycles of MVAC with 17% grade 3-4 TRAEs. In the ITT population, 33 (46%) had a relevant mutation and 26 (37%) began AS. With a median follow-up of 41 months, 47 patients (66%) are metastasis-free (CI 54%-77%). The 2-year MFS for the ITT patients (primary endpoint) was 72% (lower bound exact 1-sided 95% CI=62%). On post hoc analysis, the 2-year MFS was 65% in the AS group (CI 44%-83%) and 76% (CI 60%-87%) in the remaining patients (P=0.42). In the AS group, 18 patients (69%) had some UC recurrence, 8 had a cystectomy, 2 chemoradiation, and 13 (50%) are metastasis-free with an intact bladder. Of the 10 patients (38%) on AS who developed metastatic disease, 9 recurred with localized disease first. The 2-year OS is 83% (CI 72%-90%) and 89% (CI 68%-96%) in the ITT and AS groups, respectively. Associations between mutation presence and MFS or UC recurrence were not observed. Conclusions: The 72% 2-year MFS rate in this MIBC cohort treated with a risk-adapted approach did not satisfy the pre-specified non-inferiority condition. 38% of AS patients developed metastatic disease, with most patients first recurring locally in the bladder. With a median follow-up of 41 months, although 50% of AS patients have been able to avoid cystectomy without metastatic disease, further refinement of this approach is necessary. Clinical trial information: NCT02710734.