Metformin and prostate cancer: reduced development of castration-resistant disease and prostate cancer mortality

Abstract

Background: In vitro data and early clinical results suggest that metformin has desirable antineoplastic effects and has a theoretical benefit on castration-resistant prostate cancer (CRPC).

Objective: To determine whether the use of metformin would be associated with improved clinical outcomes and a reduction in the development of CRPC.

Design, setting, and participants: Data from 2901 consecutive patients (157 metformin, 162 diabetic non-metformin, and 2582 nondiabetic) with localized prostate cancer treated with external-beam radiation therapy from 1992 to 2008 were collected from a single institution in the United States.

Intervention: Use of metformin in localized prostate cancer.

Outcome measurements and statistical analysis: Univariate and multivariate regression models utilizing k-sample, Fine and Gray, Cox regression, log-rank, and Kaplan-Meier methods to assess prostate-specific antigen-recurrence-free survival (PSA-RFS), distant metastases-free survival (DMFS), prostate cancer-specific mortality (PCSM), overall survival (OS), and development of CRPC.

Results and limitations: With a median follow-up of 8.7 yr, the 10-yr actuarial rates for metformin, diabetic non-metformin, and nondiabetic patients for PCSM were 2.7%, 21.9%, and 8.2% (log-rank p ≤ 0.001), respectively. Metformin use independently predicted (correcting for PSA, T stage, Gleason score, age, diabetic status, and androgen-deprivation therapy use) improvement in all outcomes compared with the diabetic non-metformin group; PSA-RFS (hazard ratio [HR]: 1.99 [1.24-3.18]; p=0.004), DMFS (adjusted HR: 3.68 [1.78-7.62]; p<0.001), and PCSM (HR: 5.15 [1.53-17.35]; p=0.008). Metformin use was also independently associated with a decrease in the development of CRPC in patients experiencing biochemical failure compared with diabetic non-metformin patients (odds ratio: 14.81 [1.83-119.89]; p=0.01). The retrospective study design was the primary limitation of the study.

Conclusions: To our knowledge, our results are the first clinical data to indicate that metformin use may improve PSA-RFS, DMFS, PCSM, OS, and reduce the development of CRPC in prostate cancer patients. Further validation of metformin's potential benefits is warranted.

Copyright © 2012 European Association of Urology. Published by Elsevier B.V. All rights reserved.

Figures

Fig. 1

Unadjusted Kaplan-Meier curves for prostate-specific antigen (PSA) relapse-free survival among metformin, diabetic non-metformin, and nondiabetic groups. The univariate unadjusted p value for comparing groups was statistically significant: metformin versus diabetic non-metformin, p < 0.001; metformin versus nondiabetic, p = 0.04; and diabetic non-metformin versus nondiabetic, p = 0.002.

Fig. 2

Unadjusted Kaplan-Meier curves for distant metastasis-free survival among metformin, diabetic non-metformin, and nondiabetic groups. The univariate unadjusted p value for comparing groups was statistically significant: metformin versus diabetic non-metformin, p < 0.001; metformin versus nondiabetic, p = 0.11; and diabetic non-metformin versus nondiabetic, p < 0.001.

Fig. 3

Unadjusted cumulative incidence function curves for prostate cancer–specific mortality among metformin, diabetic non-metformin, and nondiabetic groups. The univariate unadjusted p value for comparing groups was statistically significant: metformin versus diabetic non-metformin, p = 0.001; metformin versus nondiabetic, p = 0.09; and diabetic non-metformin versus nondiabetic, p < 0.001.