Login to Access Video or Poster Abstract: MP15-19
Sources of Funding: none

Introduction

Previous literature has suggested that metformin may affect recurrence of non-muscle invasive bladder cancer (NMIBC), but data could be confounded by differences in cohorts. The purpose of this study was to evaluate the association of metformin among common prognostic factors for bladder cancer recurrence in a multivariate model and evaluate whether Metformin demonstrates a dose dependent effect.

Methods

An institutional database identified 503 unique patients treated with transurethral resection (TUR) for NMIBC. These patients were followed longitudinally and had an additional 682 recurrences and subsequent TURs. A total of 1185 TURs were performed on these 503 patients. 144 cases of NMIBC, in 60 unique patients, met inclusion criteria and were taking Metformin at the time of TUR. Cox proportional hazards models were used to evaluate associations with recurrence-free survival (RFS).

Results

Median time to recurrence was 15 months (IQR 6.18-35.6). Median age was 70.6 years. On univariate analysis, factors associated with statistically significant improved RFS included: metformin use at TUR (p=0.01, HR 0.61, 95% CI 0.42-0.89), metformin dose ≥2000 mg (p=0.03, HR 0.50, 95% CI 0.28-0.90), age, multifocality, tumor size, perioperative Mitomycin-C, bacillus Calmette-Guerin therapy, and intravesical chemotherapy. _x000D_ _x000D_ Multivariate analysis demonstrated improved RFS when comparing diabetic patients on metformin at TUR to diabetic patients not on metformin (p=0.0002, HR 0.51, 95% CI 0.36-0.72) and improved RFS even when comparing diabetic patients on metformin to non-diabetic patients not on metformin (p=0.0001, HR 0.60, 95% CI 0.46-0.77). A separate multivariate analysis, demonstrated improved RFS when comparing patients taking ≥2000 mg of metformin to patients taking <2000 mg at the time of TUR (p=0.0054, HR 0.39, CI 0.20-0.76). The 5-year RFS rate was 42.3% for diabetic patients on metformin, 35.1% for non-diabetics not on metformin, and 9.7% for diabetic patients not treated with metformin (p=0.0001).

Conclusions

Metformin use at the time of TUR is associated with improved 5 year RFS in a multivariate model. Metformin dose ≥2000 mg is independently associated with improved RFS.

Funding

none