Purpose: Results of the association between non-steroidal anti-inflammatory drugs (NSAIDs) and melanoma risk have been inconsistent. We performed a meta-analysis of relevant studies to investigate the hypothesis of an association between NSAID use and melanoma risk.
Methods: Systematic searches of the PubMed and several other databases up to 23 March 2013 were retrieved. All epidemiologic studies regarding NSAIDs and melanoma risk were included. Fixed- or random-effects meta-analytical models were used to calculate relative risk (RR) and corresponding 95 % confidence intervals (CIs). Sensitivity analyses, Galbraith plots, and subgroup analyses were also performed.
Results: Six case-control studies including 93,432 melanoma cases and 401,251 controls, six cohort studies consisting of 563,380 subjects, and one randomized controlled trial encompassing 39,876 participants were included in this analysis. Compared to non-use, ever use of any NSAIDs was not statistically significantly associated with melanoma risk based on the random-effects models (RR = 0.97, 95 % CI = 0.90-10.4, p = 0.401). No differences were found in the effects on melanoma risk of aspirin, non-aspirin NSAIDs, and cyclooxygenase-2 inhibitor use overall and stratified by gender. However, a slight reduction in the risk of melanoma by taking aspirin was observed in case-control studies (RR = 0.88, 95 % CI = 0.80-0.96, p = 0.004).
Conclusions: Findings from this pooled analysis do not support the hypothesis that NSAID use provides potential benefits in preventing melanoma. More and larger randomized trials, including adequate numbers of patients, are required to further evaluate the relationship between NSAID use and melanoma.