Carcinoma in situ (CIS) is a distinct entity in non-muscle invasive bladder cancer (NMIBC) that has both important clinical and research implications. For NMIBC trials, the FDA considers CIS as a "marker lesion" that can be used to determine drug efficacy and allows for approval based on the 3- or 6-month complete response (CR) in patients with BCG-unresponsive CIS. The classification of CIS +/- papillary cohorts and papillary only cohorts, implies that CIS is a surrogate for minimal residual disease (MRD),;despite having limited supporting evidence. Within the context of a prospective clinical protocol, we sought to determine whether the presence of pre-treatment CIS or a pre-treatment positive urinary cytology are indicators of MRD in NMIBC and whether they are associated with treatment outcomes with BCG. Patients with BCG-naive high grade NMIBC (HGTa, HGT1, Tis) were prospectively enrolled onto a biospecimen protocol developed to identify predictors of response/resistance to BCG (IRB #19-015). Patients routinely underwent;repeat TURBT 4-6 weeks following the initial resection, unless not clinically indicated. Pre-treatment CIS status was based on consensus review of the initial and/or repeat TURBT as determined by a GU pathologist. Pre-treatment urinary cytology was collected after the patient's last TURBT but prior to the initiation of BCG and reported based on the Paris System for Reporting Urinary Cytology.;All patients were treated with a standard induction course of BCG +/- maintenance. Patients were monitored for recurrence with cystoscopy and urinary cytology every 3 months for the first 2 years, then every 6 months or as indicated. Recurrences were defined as any high-grade recurrence after BCG treatment. We performed Kaplan Meier survival estimates to compare recurrence free survival (RFS). Among 177 patients identified, 174 (98%) completed 6 instillations of induction BCG and 76 (44%) had at least one course of maintenance. Overall, 56 (32%) patients had high grade recurrence and 17 (10%) underwent cystectomy. The overall median follow up was 25.5 months. 68 (46%) patients had CIS on either initial or repeat TURBT. There was no difference in HG-RFS between those that had CIS compared to those that did not (p=0.85) (Figure 1a). 121 patients had cytology results after TURBT but prior to BCG of which 63 (52%) were negative, 26 (21%) were atypical, 13 (11%) were suspicious, and 19 (16%) were positive. Patients with a negative urinary cytology had significantly improved RFS compared to the other groups (p<0.001) (Figure 1b). Abnormal cytology, irrespective of CIS, had the worst RFS (Figure 2). Pre-treatment CIS on TURBT is currently considered an indicator of MRD for patients with NMIBC and it is assumed by FDA that CIS cannot be eradicated through TURBT alone. However, utilizing a prospective cohort of high grade NMIBC patients, we found that pre-treatment CIS was not predictive of RFS. Contrarily, patients that had either a positive, suspicious, or atypical cytology after TURBT but prior to BCG had significantly worse RFS compared to those with a negative urinary cytology. Our results suggest that CIS on pre-treatment TURBT is a poor surrogate for MRD and pre-treatment urinary cytology is more predictive of disease recurrence. Using this same cohort, we are evaluating whether urinary tumor DNA sequencing is an even better indicator of MRD for NMIBC. [ABSTRACT FROM AUTHOR]