Digoxin exposure and prostate cancer mortality: A matched cohort study.

Abstract

e13096 Background: Digoxin (DIG) exposure has been associated with reduced prostate cancer (PC) incidence in epidemiological studies. Preclinical data suggests that this anti-cancer effect is mediated through the inhibition of hypoxia-inducible factor 1-α by DIG. This retrospective cohort study examines associations between DIG exposure and mortality in men with PC. Methods: Men diagnosed with PC during 2001–2006 were identified from National Cancer Registry Ireland records and linked prescription claims data. Propensity scores for DIG exposure in the 90 days prior to PC diagnosis were estimated. DIG exposed and unexposed men were matched (1:1) within a calliper of 0.2 standard deviations of the propensity score logit, using greedy matching without replacement. Standardized differences were used to assess covariate balance (z-score <0.1) between matched cohorts. Hazard ratios (HR) for associations between DIG exposure and all-cause (ACM) or PC-specific (PCM) mortality were estimated using Cox proportional hazards models adjusted for age, comorbidity, tumour stage and grade. Categorical exposure-response analyses were carried out using tertiles of exposure (low, intermediate, high) in the 90 days pre-diagnosis. Results: 5734 PC cases were identified from the linked database. 395 cases received DIG in the 90 days pre-diagnosis, of which 391 were matched to unexposed controls. Matched covariate balance was acceptable. In adjusted analyses, DIG exposure was not associated with ACM (HR 1.06, 95% CI 0.88-1.27) but was associated with a small but non-significant reduction in the risk of PCM (HR 0.89, 95% CI 0.68-1.17). In the exposure-response analysis, DIG exposure in the highest tertile, but not in the intermediate or lower tertiles, was associated with a reduced risk of PCM approaching statistical significance (HR 0.69, 95% CI 0.47-1.01, p=0.059). Men in the high DIG exposure group received a supply of DIG for 98% of their eligible follow up in the year post diagnosis. Post diagnostic DIG exposure in the intermediate and lower tertiles was 94% and 80% respectively. Conclusions: DIG exposure was associated with a non-significant decrease in PCM. Stratification by exposure suggests the presence of an exposure-response relationship.