proton pump inhibitors increase risk of aspirin-induced small-bowel injury

so, you might ask, why am i reviewing for primary care providers such a pretty obscure study in a pretty obscure journal (Gastrointestinal Endoscopy)? mostly because i think that proton pump inhibitors (PPIs) are over-prescribed (as well as available over-the-counter) and may well be dangerous, esp with long-term administration. studies have shown that in the treatment of GERD, for example, there have been 2 strategies: the step-up (ie, start with lower potency calcium, then to H2 blockers, then to PPIs as needed) vs step-down (start strong with PPIs, then titrate down). both seem theoretically reasonable, except that in practice the step-down strategy is rarely implemented (ie, patient is doing well on PPI, so just refill the prescription -- why change something that works and makes the patient feel better??). there are several issues with long-term PPI usage, largely related to hypochlorhydria and/or the extreme hypergastrinemia associated. although the data in many cases is not completely robust, there are reasonable associations in several studies with the following: increase in c diff infections (even without antibiotic use); decreased bone mineral density and increased risk of fracture (may be related to decreased calcium absorption); decreased absorption of magnesium, iron, and vitamin B12; increased risk of community-acquired pneumonia (in several studies); increased risk of other enteric infections (esp campylobacter and salmonella); and even increased risk of atrophic gastritis (a potentially pre-malignant state) esp if concurrent H Pylori infection. although all of these issues may not prove to be clinically very important, they do reinforce to me the imperative to use medications with the least potential toxicity whenever possible (ie, causing the least perturbation of endogenous native systems).  

this japanese study involved doing capsule endoscopy (CE) in 198 patients (141 male, mean age 71.9) being evaluated for aspirin-related small bowel lesions or bleeding from unknown site,  and 57.6% of them were found to have at least 1 small-bowel mucosal break (see ppi asa mucosal injury giendosc 2014 in dropbox, or doi.org/10.1016/j.gie.2014.03.024). multivariate analysis was done to see what factors might confer greater risk for mucosal breaks. all patients had been on low-dose aspirin (75-325mg/d) for at least 3 months, with 80-90% on enteric-coated aspirin. all had had UGI endoscopy and colonoscopy soon before the CE. none had been on NSAIDs in the prior 3 months. findings:

--80-90% of patients were on enteric-coated aspirin and rest on buffered aspirin (mean of 60 months), 35% on PPI, 18% on H2-blocker, 30% smokers, 43% drinkers

--PPI use doubled the risk, with OR 2.04 (1.05-3.97), H2-blockers non-significantly increased the risk (??because of smaller numbers of patients or because of less acid suppression that PPIs), with OR 1.65 (0.71-3.81)

--low-dose enteric-coated aspirin (75-325 mg/d), as compared to buffered aspirin, increased the risk the most, with OR 4.05 (1.49-11.0)

a couple of other comments. it has recently become more evident that low-dose aspirin frequently causes small-bowel injury (ie, aspirin may cause GI bleeds even without lesions in the stomach or large intestine -- so having normal EGD and colonoscopy does not rule-out aspirin as the cause of the bleed). and, perhaps some of us have been using PPIs to mitigate the upper GI risk of bleeding. the reason that PPI might confer increased risk includes experimental studies that enterobacteria play a role in NSAID-induced small bowel injury and that PPIs decrease stomach acidity so dramatically that bacteria are not killed in the stomach (ie, inducing "dysbiosis"), and there are some studies suggesting that co-administration of probiotics improves small bowel injury in those on low-dose aspirin plus PPIs. the association with enteric-coated aspirin may be because the enteric coating dissolves in the small intestine, providing aspirin with direct contact to the mucosa there. so, this study also fits in with the myriad of articles i have sent out on the gut microbiome, and also raises further the issue noted in the appended blog below promoting the use of non-enteric-coated aspirin for cardioprotection.