PPIs and recurrent C diff infections

A recent systematic review and meta-analysis found that the use of protein pump inhibitors (PPIs) seems to be associated with increased recurrent Clostridium difficile infections (see ppi cdiff metaanal jamaintmed2017 indropbox, or doi:10.1001/jamainternmed.2017.0212)

Details:

-- literature review from 1995 to 2015, specifically looking at case-control studies, cohort studies, and clinical trials, found 16 observational studies of 7703 patients with C. diff infections, 4038 (53%) were using gastric acid suppressants, and 1525 patients (20%) developed recurrent C. diff​ infections.

-- These studies were performed in the United States, Korea, Israel, Europe, Japan, and Canada

-- 9 studies involved PPIs alone, 5 with PPIs and H2 blockers, and one with H2 blockers alone.

Results:

-- the recurrence rate for C. diff was:

    --22.1% (892 of 4038 patients) in those taking gastric acid suppression

    --17.3% (633 of 3665) in those not taking gastric acid suppression

    -- an overall 38% increased risk, adjusting for age and potential confounders, OR 1.38 (1.08-1.76, p=0.02)

--subgroup analysis found that there was a 66% increased risk of C. diff recurrences with the use of PPIs, OR 1.66 (1.18-2.34), p=0.004, but not an increased risk in those they used both PPIs and H2 blockers, or H2 blockers alone (though only one study looked at H2 blockers by itself)

-- the increased risk was found in both case-control studies and cohort studies

Commentary:

-- Clostridium difficile infections are the most common cause of hospital-acquired diarrhea, and seem to be increasing in incidence, severity, morbidity, and mortality rates. At this point about 40% are community-acquired and in patients felt to be at low risk, suggesting that there may be new risk factors, including such things as gastric acid suppressants, C. difficile in the water and food, and close contacts with those with C. diff infections in the community.

-- The studies are quite mixed on the relationship between gastric acid suppression and C. diff infections overall, especially when controlling for comorbid conditions and age. The connection seem to be somewhat stronger for PPIs versus H2 blockers, leading the FDA to issue a warning that PPIs are associated with an increased risk of C. diff infections (https://www.fda.gov/drugs/drugsafety/ucm290510.htm ).

-- recurrent C. diff infections happen pretty frequently, as high as 50-60% after 3 or more infections. Risk factors for recurrence include older age, concomitant antibiotic use, and comorbid conditions. About 50% of patients with C. diff infections use concomitant gastric acid suppressants. Prior analyses have found an increased risk of reinfection in those on gastric acid suppressants, but these studies were limited by being retrospective or excluding important studies.

-- The current study was methodologically more sophisticated, including prospective and retrospective case-control studies, as well as post hoc analysis of 2 RCTs

-- gastric acid suppressants are known to alter the distal gut microbiota, with substantial decreases in bacterial diversity, and may make the microbiota more prone to both primary and recurrent C. diff infections. And data suggest that PPIs may affect the microbiota more than H2 blockers. On the other hand, the use of antibiotics to treat C. diff infections may further change the microbiota.

-- these data are still observational, limiting our ability to attribute causality. For example, those on gastric acid suppressants are perhaps more likely to be elderly, using concomitant antibiotics, or having comorbid conditions which might increase the likelihood of C. diff infections. The authors did try to control for these issues, but this is difficult in a meta-analysis of 16 studies with different criteria and baseline data on patients.

see PPI cdiff metaanal amjgastro 2012 in dropbox, or doi: 10.1038/ajg.2012.179, which presents a meta-analysis of 300,000 patients, finding a 65% increased risk of C diff  in patients on PPIs, independent of study design (case-control vs cohort), with the unusual p-value of "p<0.000", which on the surface seems pretty strong.....)

see http://gmodestmedblogs.blogspot.com/search/label/c%20difficile for several prior blogs on C diff infections, including the use of probiotics and fecal transplants

see http://gmodestmedblogs.blogspot.com/2016/08/ppi-harms-and-benefits.html for a recent review of some of the benefits and harms of PPIs, including my concerns about their being overused in general, both by being overprescribed initially, and by continuing longterm instead of "stepping down" therapy.