Objectives: Based on reports of antitumour properties of sodium-valproate, we hypothesised that valproate has a cancer-protective effect in people with epilepsy. We aimed to determine cancer risk in people with epilepsy using sodium-valproate.
Materials and methods: Continuous data for 2997 people with epilepsy who had been prescribed valproate for at least two years, and for 11,988 unexposed people were provided by the UK General Practice Research Database. Hazard ratios (HRs) for all cancers and individual cancers between the exposed and unexposed groups, with smoking and alcohol consumption and age as covariates, were calculated using the Cox proportional hazards method.
Results: Exposure to valproate had no influence on the incidence of the composite of all cancers [HR: 1.19, 95% CI: 0.97-1.47, P = 0.10]; there was, however, a significant excess of colon cancers [HR: 3.95, 95% CI: 1.97-7.92, P = 0.001] and a trend towards an excess of prostate neoplasms [HR: 2.15, 95% CI: 0.92-5.02, P = 0.08] and in addition, a trend towards reduced incidence of breast cancer [HR: 0.40, 95% CI: 0.14-1.30, P = 0.08] in the exposed group.
Conclusions: The lack of an inverse association between valproate use and hazard ratios for all cancers and several individual cancer sites does not lend support for a cancer-protective role for valproate.
© 2011 John Wiley & Sons A/S.