Introduction:

Current guidelines recommend radical cystectomy (RC) for patients with BCG-unresponsive non-muscle invasive bladder cancer (NMIBC) willing and fit to undergo surgery. However, RC is associated with significant morbidity and quality of life implications, leading many patients and clinicians to pursue bladder sparing therapies (BST). Limited data exists on the oncologic outcomes following BST in BCG-unresponsive NMIBC, most of which are from single-institution series. Furthermore, given the FDA and SUO definition of BCG-unresponsiveness NMIBC represents a heterogenous group of patients with various stages and burden of disease, it is critical to explore the impact of tumor stage on oncologic outcomes in this population. The purpose of this study was to compare oncologic outcomes in patients with T1 or Ta BCG-unresponsive NMIBC to those with pure carcinoma in situ (CIS) disease in a multi-center, international cohort.

Methods:

Patients undergoing BST for BCG-unresponsive NMIBC as defined by FDA and SUO criteria were identified at 10 academic institutions spanning three countries and compiled into a singular database. Demographic variables; date of last follow up; pathologic outcomes for those undergoing RC; and incidence and date of intravesical high grade (HG) recurrence, progression to muscle-invasive bladder cancer, cystectomy, metastasis, cancer-specific mortality, and all-cause mortality were collected. Survival for each oncologic event was defined as length of time from BCG-unresponsiveness to either the respective event or date of last follow up. The Kaplan-Meier method was utilized to determine survival differences in T1 +/- CIS, Ta +/- CIS, and pure CIS patients, and statistical comparisons were performed using log-rank tests. Pathologic stage differences for those undergoing subsequent RC were determined using Chi-squared tests.

Results:

401 patients underwent BST for BCG-unresponsive NMIBC, including 137 T1, 121 Ta, and 143 patients with pure CIS alone (Table). At median follow up of 47 months, death from bladder cancer increased from 5% to 25% from 24 to 48 months in T1 patients compared to 4% to 7% and 5% to 9% in Ta and CIS alone patients, respectively (p=0.010) (Figure). All-cause mortality increased from 12% to 36% from 24 to 48 months in T1 patients compared to 10% to 21% and 7% to 15% in Ta and CIS alone patients, respectively (p=0.018). Median time to RC was 14 months in T1 group vs. 11.5 months in Ta vs. 16 months in CIS alone with 26% of total T1 group undergoing RC by 48 months compared to 15% and 32% of Ta and CIS (p<0.001). Pathologic outcomes at time of RC were similar between all groups.

Conclusion:

Rates of disease progression, metastasis, cancer-specific mortality, and all-cause mortality were worse in patients with concomitant T1 disease at BCG-unresponsiveness compared to Ta or CIS alone. Specifically, bladder cancer mortality within the T1 group increased most dramatically from 24 to 48 months but remained comparable to the Ta and CIS groups prior. These findings suggest that those with T1 invasive disease electing to proceed with BST may have a brief window of comparable oncologic outcomes to those with non-invasive disease but, over time, have worse oncologic outcomes. While T1 and CIS patients were more likely to undergo RC, median time to RC for all three groups and pathologic outcomes were similar. The findings from this cohort of select, non-randomized patients support close monitoring of the carefully selected T1-staged BCG-unresponsive NMIBC patient. Future studies could investigate the impact of pathologic stage on oncologic outcomes following delayed RC in this population.

Funding: N/A