Eleven studies (n=1,390,304) were included: one randomised controlled trial (n=39,876), 7 cohort studies (n=1,341,280), 2 hospital-based case-control studies (n=7,100) and one population-based case-control (n=2,048).
Use of aspirin or NSAIDS was not associated with an increased risk of pancreatic cancer when all studies were combined (RR 1.01, 95% CI: 0.91, 1.11). There was weak evidence of heterogeneity (I-squared 38.8%; p=0.09) and no evidence of publication bias (p=0.78). The results remained the same when different study designs were analysed separately.
Frequent aspirin use, defined as six or more tablets/times per week, was not associated with an increased risk of pancreatic cancer (RR 0.86, 95% CI: 0.61, 1.23; based on 4 studies). There was weak evidence of heterogeneity (I-squared 57.4%; p=0.07). Long-term use of aspirin, defined as more than 20 years, was not associated with an increased risk of pancreatic cancer (RR 1.21, 95% CI: 0.74, 1.96; based on 2 studies). There was weak evidence of heterogeneity (I-squared 69.4%; p=0.07).
When the studies of aspirin alone, non-aspirin NSAIDs and all NSAIDs were evaluated separately, there was no association between exposure to medication and risk of pancreatic cancer.