Re-analysis of intravesical gemcitabine in the era of Bacillus Calmette-Guérin (BCG)-unresponsive non-muscle-invasive bladder cancer.

PURPOSE

Uncertainty exits over the best treatments for patients whom Bacillus Calmette-Guérin (BCG) has failed. We aim to contextualize novel therapies approved based on single-arm, nonrandomized trials and those supported only by retrospective data by re-analyzing oncologic outcomes achieved with intravesical gemcitabine.

MATERIALS AND METHODS

Patients with non-muscle-invasive bladder cancer (NMIBC) who were treated for BCG failure with single agent intravesical gemcitabine at our institution were analyzed to compare clinical outcomes (high-grade recurrence/progression) across different BCG failure definitions. Associations between clinicopathologic variables with outcomes after gemcitabine were estimated using Cox models.

RESULTS

Of the 127 patients, 57% met the historical definition of BCG-refractory NMIBC and 33% met BCG-unresponsive criteria using a 12-month cut-off. Twelve-month recurrence-free survival was similar between BCG-refractory (47%; 95% CI: 36%, 60%) and BCG-unresponsive (52%; 95% CI: 38%, 71%) definitions. BCG-unresponsive patients who received gemcitabine in a clinical trial had significantly worse recurrence-free survival compared to those receiving the same treatment outside a trial (12-month recurrence-free survival difference: 41%; P = 0.02). A positive pretreatment urine cytology was associated with increased risk of recurrence (P = 0.005) and progression (P = 0.002) and may better indicate minimal residual disease than carcinoma in-situ on pretreatment biopsies.

CONCLUSIONS

Our data raise concern over US Food and Drug Administration approval based on single-arm, nonrandomized trials using expert-based BCG-unresponsive criteria and for the use of combination gemcitabine-docetaxel as a de facto standard treatment based on retrospective data alone. Improved assessments of minimal residual disease, such as pretreatment urinary cytology, are needed to improve risk stratification in NMIBC.