Effects of regular aspirin on long-term cancer incidence and metastasis: a systematic comparison of evidence from observational studies versus randomised trials

Background Long-term follow-up of randomised trials of aspirin in prevention of vascular events showed that daily aspirin reduced the incidence of colorectal cancer and several other cancers and reduced metastasis. However, statistical power was inadequate to establish effects on less common cancers and on cancers in women. Observational studies could provide this information if results can be shown to be reliable. We therefore compared effects of aspirin on risk and outcome of cancer in observational studies versus randomised trials. Methods For this systematic review, we searched for case-control and cohort studies published from 1950 to 2011 that reported associations between aspirin use and risk or outcome of cancer. Associations were pooled across studies by meta-analysis and stratified by duration, dose, and frequency of aspirin use and by stage of cancer. We compared associations from observational studies with the effect of aspirin on 20-year risk of cancer death and on metastasis in the recent reports of randomised trials. Findings In case-control studies, regular use of aspirin was associated with reduced risk of colorectal cancer (pooled odds ratio [OR] 0.62, 95% CI 0.58-0.67, p(sig) < 0.0001, 17 studies), with little heterogeneity (p(het)=0.13) in effect between studies, and good agreement with the effect of daily aspirin use on 20-year risk of death due to colorectal cancer from the randomised trials (OR 0.58, 95% CI 0.44-0.78, p(sig)=0.0002, p(het)=0.45). Similarly consistent reductions were seen in risks of oesophageal, gastric, biliary, and breast cancer. Overall, estimates of effect of aspirin on individual cancers in case-control studies were highly correlated with those in randomised trials (r(2)=0.71, p=0.0006), with largest effects on risk of gastrointestinal cancers (case-control studies, OR 0.62, 95% CI 0.55-0.70, p<0.0001, 41 studies; randomised trials, OR 0.54, 95% CI 0.42-0.70, p<0.0001). Estimates of effects in cohort studies were similar when analyses were stratified by frequency and duration of aspirin use, were based on updated assessments of use during follow-up, and were appropriately adjusted for baseline characteristics. Although fewer observational studies stratified analyses by the stage of cancer at diagnosis, regular use of aspirin was associated with a reduced proportion of cancers with distant metastasis (OR 0.69, 95% CI 0.57-0.83, p(sig)<0.0001, p(het)=0.89, five studies), but not with any reduction in regional spread (OR 0.98, 95% CI 0.88-1.09, p(sig)=0.71, p(het)=0.88, seven studies), consistent again with the findings in randomised trials. Interpretation Observational studies show that regular use of aspirin reduces the long-term risk of several cancers and the risk of distant metastasis. Results of methodologically rigorous studies are consistent with those obtained from randomised controlled trials, but sensitivity is particularly dependent on appropriately detailed recording and analysis of aspirin use.