GI Update 2019
Proton-Pump Inhibitor Therapy is Safe
Over the past few years, safety issues associated with PPIs have come to the forefront of discussion in the field of Gastroenterology.
Many studies on the long-term use of PPIs have reported adverse effects such as fractures, hypomagnesemia, acute and chronic kidney disease, dementia, and Clostridium difficile infection. These have become a source of major concern to both patients and physicians. Some patients discontinued their PPI therapy abruptly or inappropriately, and prescribers withheld or underdosed the treatment of their patients because of this. Up until recently, all the published results were based on retrospective and observational studies. They were unable to establish a definite association, and could not prove cause and effect between PPI use and the development of serious adverse events. It is important to determine the long-term safety of this class of drug because many people depend on it, and yet, we do not want to endanger their well-being because of it.
Retrospective Studies on PPI Use and Development of Adverse Events
Fractures. A meta-analysis of 18 observational studies reported a 33% increased risk of fracture at any site, 26% higher risk of hip fracture, and 58% increase in risk of spine fracture. Both short-term (<1 year) and long-term use of PPI were associated with increased risk of hip fracture. (Zhou B et al, Osteoporos Int. 2016;27(1):339-47). The proposed mechanism is reduced calcium absorption resulting in decreased bone mineral density (BMD). Although a placebo-controlled, double-blind, crossover trial found decreased calcium carbonate absorption after 1 week of Omeprazole treatment (O’Connell M et al, Am J Med 2005;118(7):778-781), there is no evidence to show that PPI are associated with development of osteoporosis. A study on 4,512 patients who received continuous PPI treatment and followed over 10 years, did not show a significant rate of change in BMD (Targownik L et al, Am J Gastroenterol 2012;107(9):1361-69). The association of short-term PPI use with fractures as reported by Zhou B et al in 2016, also lead us to conclude that PPI-associated fractures are unlikely due to osteoporosis.
Hypomagnesemia. A meta-analysis of 9 observational studies that included 109,798 patients, reported a 43% increased risk of hypomagnesemia (Cheungpasitporn W et al, Ren Fail 2015;37(7):1237-41). Magnesium level normalized with discontinuation of PPI treatment. In 2011, the U.S. FDA issued a safety warning regarding this association and recommended monitoring of magnesium levels in patients on long-term PPI therapy.
Acute and chronic kidney disease (CKD). Xie et al (J Am Soc Nephrol 2016;27(10):3153-63) used the Department of Veteran Affairs national databases and assessed the renal outcomes of 173,321 new PPI users compared to 20,270 new H2RA users over 5 years. PPI users were found to have a 28% increased risk of CKD. Antoniou T et al (CMAJ Open. 2015;3(2):E166-171) reported a 2.5-times increased risk of acute kidney injury and 3-times increased risk of acute interstitial nephritis in elderly patients recently started on PPI’s. This was a population-based study of 290,592 patients older than 65-years-old. The mechanism for development of CKD may be due to untreated subclinical acute interstitial nephritis that progressed over time.
Dementia. Two prospective studies investigated the association of PPI use and the risk of dementia. Haenisch et al (Eur Arch Psychiatry Clin Neurosci. 2015;265(5):419-428) studied 3,076 patients 75 years or older with no history of dementia and found a 38% increased risk of dementia in PPI users. Gomm et al (JAMA Neurol 2016;73(4):410-416) studied 73,679 individuals 75 years or older with no history of dementia and found that regular PPI users had a 44% increased risk of dementia. On the other hand, there were two studies that did not show an association between PPI use and dementia. Lochhead P et al (Gastroenterology 2017;153(4):971-979.e4) reported negative results on a prospective analysis of 13,864 middle-aged and older women enrolled in the Nurses’ Health Study II, and Taipale H et al (Am J Gastroenterol 2017 Dec;112(12):1802-08) reported negative results of an analysis of 70,718 newly diagnosed case of Alzheimer’s disease in the Finnish nationwide health care registry. PPIs are suspected to cause dementia by causing an accumulation of amyloid-B peptides in the brain. This was based on the findings of a study where mice that were treated with Lansoprazole were found to have higher levels of amyloid-B peptides in their brain due to an increase in amyloid synthesis and decrease in amyloid degradation in the brain.
Clostridium difficile infection (CDI). A meta-analysis of 42 observational studies revealed an increased risk of both incident and recurrent CDI in patient treated with PPI. All studies were non-randomized observational studies, and the dose and duration of PPI treatment used were highly variable. The hypothesis of this association is the survival of the toxin-secreting vegetative form of C. difficile in the alkaline gastric pH. (Kwok C et al, Am J Gastroenterol 2012;107(7):1011-1019).
All-Cause Mortality Risk Increased. Xie Y et al (BMJ 2019 May;365:l1580) reported results from a retrospective study using data from the U.S. Department of Veterans Affairs to estimate the all-cause and cause-specific mortality among new users of PPIs (n=157,625) and H2RA (n=56,842). Over a median of 10 years of follow-up, there were more deaths among patients taking PPIs (37.92%) than among those taking H2RA (35.69%). PPI use was associated with excess mortality due to cardiovascular disease, CKD, and upper GI cancer.
Prospective Randomized Controlled Trial on PPI Safety
Moayyedi P et al, published the first prospective, double-blind, placebo-controlled randomized trial to evaluate the many long-term safety concerns related to PPI use (Gastroenterology 2019 September;157:682-691). This was a large trial involving 17,598 patients with stable cardiovascular disease and peripheral artery disease. This trial had a median follow-up of 3.01 years. Participants were randomly assigned to receive Pantoprazole 40 mg/day (n=8,791) or placebo (n=8,807). There was no significant difference between the Pantoprazole and placebo groups with regards to the following adverse events: gastric atrophy, C. difficile infection, chronic kidney disease, dementia, pneumonia, fracture, COPD, and diabetes mellitus. Only enteric infections other than C. difficile was more common in the Pantoprazole group (1.4% vs. 1.0%).
Summary and Recommendations
Earlier studies and a few recent ones reported various safety concerns regarding the use of PPIs. However, these results were inconclusive because of inherent biases associated with retrospective and observational studies from which they were derived from. The only large, multi-year, prospective, randomized controlled trial looking into the safety of PPIs by Moayyedi P et al, found that Pantoprazole 40 mg/day is not associated with any adverse events when used for 3 years, with the possible exception of an increased risk of enteric infections. Although this study presented strong evidence that PPI is safe, this conclusion cannot be extrapolated beyond the dose and duration used in this study. There are many questions that are yet to be answered by future studies of various designs. In the meantime, judicious use of PPIs should be practiced. Prescribe PPI only when indicated and give the recommended dose and duration of treatment. When there is a proven indication for PPI therapy, treatment should not be withheld because of concerns of potential long-term harm. Long-term PPI therapy should be given at the lowest effective dose when clinically indicated. PPI dose escalation and continued chronic therapy in those unresponsive to initial empiric therapy should be discouraged. For patients whose symptoms improved on PPIs, and there are no reasons for continued treatment, they should be given a trial of stopping the PPI or use a less potent acid-reducer and see if their symptoms recur.