a recent article reviewed the data on the use of aspirin in primary prevention, both cardioprotective and chemoprotective (see doi:10.1093/eurheartj/eht058). basically, no new data challenging the old recommendations for primary and secondary prevention of vascular disease: for secondary prevention there is a 20% decreased risk of vascular complications, translating to an absolute reduction of 10-20/1000 pts in annual incidence of nonfatal events (smaller but important reduction in vascular deaths), with absolute increase in major extracranial bleeding complications which is 20-50 fold smaller. the data on primary prevention is not so impressive. studies difficult to meta-analyze, since different populations and different endpoints. but overall, data insufficient to advocate either for or against aspirin use. there are some data that in certain conditions (eg following CABG, pts with essential thrombocytemia, pts with coronary artery disease who have metabolic syndrome and pts with diabetes), there may be increased platelet activation and may have suboptimal response from aspirin (the POPADAD study was quite striking: men and women with diabetes and peripheral artery disease, at very high risk of other atherosclerotic events, randomized to aspirin 100mg/d vs placebo and no benefit to asprin). the liberal use of statins may provide much of the benefit of aspirin with less toxicity.
a few comments about using aspirin:
1. avoid using enteric-coated aspirin formulations: on the one hand, there is less bioavailability of the enteric-coated and the vast majority of secondary prevention trials used plain aspirin, and on the other hand, it seems that enteric-coated is not gastroprotective (ie, a study found that the relative risks of upper GI bleeding for plain vs enteric-coated vs buffered aspirin (<325 mg) was 2.6, 2.7 and 3.1. if >325 mg, RR was 5.8 for plain and 7.0 for buffered aspirin (not enough data on enteric-coated)
2. risk of bleeding: for extracranial bleeds the risk increases along with all of the major cardiac risk factors, except lipids (ie, increased risk of bleeding if diabetes, male gender, cigarette smoker, hypertension, high BMI). so, the more likely to benefit from cardioprotection, the more likely to bleed. increased bleeding up to 100-fold, depending also on prior history of GI bleeding and age of patient. GI bleed risk is halved by either adding a proton pump inhibitor or treating underlying H. Pylori infection. (there have been a couple of studies showing, for example, that in H pylori positive patients about to begin chronic NSAIDs for rheumatologic diseases, that pretreating the H pylori dramatically decreases GI bleeding, including severe bleeds). in terms of intracranial hemorrhage, less data, but seems to be <1 per 1000 pts per year in high risk trials. meta-anal of primary prevention trials: 5 additional hemorrhagic strokes per 1000 mod risk pts over 5 years, and much less if low-risk patient.
3. cancer risk: trials for vascular protection have pretty consistently found deceased cancer risk with aspirin (after 8-10 years of low-dose aspirin), and decreased incidence and mortality from common cancers. most consistent data for colon cancer, though all of these are secondary analyses and hard to make real recommendations. will send out again a couple of older email/blogs about use of aspirin in breast cancer, etc. but, unlike the diminishing vascular protective effect of aspirin after the initial 3 years, the risk of cancer decreases with time (OR of 1 in 1-3 years, improving to 0.81 for 3-5 years and further to 0.70 after 5 years) and the risk of extracranial bleeds also decreases (increases almost 2-fold for 0-3 years, then not significantly increased). BUT, if person has been on aspirin for secondary prevention and if they are taken off after 3 years, 40% increased risk of ischemic stroke and MI (ie, DO NOT stop the aspirin because the protection is less evident after 3 years!!)
so, no change in recommendations, but i would advise against enteric-coated aspirin. i also have been incorporating the possible cancer benefit if a patient is on the borderline of treating with aspirin (eg, by the old framingham risk score, if the risk is in the 6-10% range)
geoff