Welcome to the CJU website » LOG IN


Impact of diabetes and metformin use on prostate cancer outcome of patients treated with radiation therapy: results from a large institutional database
Department of Radiation Oncology, University of Montreal Health Center, Montreal, Quebec, Canada
Oct  2018 (Vol.  25, Issue  5, Pages( 9509 - 9515)
PMID: 30281009


Text-Size + 


    Conflicting data exists on the influence of metformin on prostate cancer. We investigated the importance of metformin in patients treated with radiotherapy or brachytherapy.


    All patients from a large institutionalized database, treated for primary localized prostate cancer with either brachytherapy or external-beam radiotherapy ± androgen deprivation therapy were identified. Groups were compared by Kaplan–Meier analyses and Cox regression models. Multivariate analysis was adjusted for CAPRA-Score, type of treatment and age.


    A total of 2441 patients with complete data was identified. Among the 382 patients (16% of total) were diabetic. Two-hundred and eighty-one of the 382 diabetics (74%) were treated with metformin and 101 were treated with other anti-diabetic medication. Median follow up was 48 months (interquartile range [IQR] 24-84). Two-hundred eighteen patients (9%) died and 150 (6%) experienced biochemical recurrence (BCR). On unadjusted univariate analysis for BCR-free survival, metformin users showed a 50% reduction in BCR compared to non-metformin users. The results remained significant on multivariate analysis comparing diabetic metformin users to non-metformin users (diabetics and non-diabetics combined) (hazard ratio [HR] 0.5-0.6, p = 0.03-0.04) but lost its significance when adjusting for cancer aggressiveness. On multivariate analysis, diabetics had worse overall survival (OS) than non-diabetics (HR 1.5, 95% confidence interval [CI] 1.08-2.06, p = 0.01), but diabetics on metformin fared better than diabetics not taking metformin (HR 0.5, 95% CI 0.26-0.86, p = 0.01).


    Metformin use in this analysis appears to be associated with better BCR and OS. Larger datasets and prospective trials are warranted to validate these results.