Drug-drug Interactions

After systemic absorption and initial wide distribution within the body, the liver rapidly metabolizes PPIs via the cytochrome P450 enzyme system. The principal isoenzymes involved are 2C19 and, to a lesser extent, 3A4. The potential for individual PPIs to influence P450 enzyme activity has raised concerns about possible drug-drug interactions (DDIs). However, there is little, if any, clinical significance to PPI interactions with other medicines; those commonly included in package inserts with PPIs as having the potential for P450-related interactions include warfarin, diazepam, tacrolimus, cyclosporine, and theophylline. However, dose adjustments are not routinely required in patients taking PPIs and other medications metabolized by the cytochrome P450 system.

Most medicines with which at least some PPIs have been proposed to interact have wide therapeutic indices. That is, their steady-state plasma concentrations associated with toxicity are much higher than those associated with the desired therapeutic effect. A perfect example is diazepam. Although omeprazole interacts with diazepam to slow its metabolism and raise its steady-state plasma concentration, this is not associated with any measurable pharmacodynamic effect. Some drugs (notably warfarin, phenytoin, and theophylline) have much narrower therapeutic indices, so plasma concentrations associated with toxicity are close to those normally associated with their desired therapeutic effects. Theoretically, slowing of the hepatic metabolism of any of these drugs by the concomitant use of a P450 inhibitor (such as a PPI) might increase steady-state plasma concentrations to toxic levels. However, DDIs reported to the United States Food and Drug Administration (FDA) with omeprazole, lansoprazole, and pantoprazole included estimated frequencies of less than 1 per million prescriptions, with no discernable differences among them. The American Gastroenterological Association Institute published a technical review on GERD in 2008, which included an overview of various potential safety concerns with the PPIs, including possible DDIs. They endorsed the view that clinically significant DDIs occurred with a frequency of less than 1 per million prescriptions.

There are a few DDIs common to all the PPIs with respect to their inhibition of gastric acid secretion. All PPIs may reduce the absorption of the antiretroviral agent atazanavir. This reduction may potentially lead to subtherapeutic blood levels of that agent, with risks of treatment failure against human immunodeficiency virus (HIV) infection or the emergence of resistant strains of HIV. All PPIs may increase the bioavailability of digoxin by around 10%, which is unlikely to be clinically relevant.

PPIs and Clopidogrel

Clopidogrel is a prodrug that must be activated in the liver to exert its desired antiplatelet effect. This effect is accomplished in a multistep process that includes cytochrome P450 2C19, the isoenzyme most closely connected with PPI metabolism. Some in vitro studies suggested that omeprazole, and perhaps other PPIs, reduces the rate of conversion of clopidogrel to its active metabolite when used concomitantly. Clopidogrel is generally administered with aspirin, and this combination puts some patients at risk of upper GI hemorrhage. PPI cotherapy is advised for patients with identifiable risk factors such as advanced age, a past history of ulcer or bleeding, or concomitant NSAID or anticoagulant use. Some retrospective studies reported worse cardiovascular outcomes among patients on clopidogrel and a PPI compared with those on clopidogrel alone. However, these were prone to bias; for example, in both studies cited, patients taking a PPI and clopidogrel had worse cardiovascular risk profiles and more comorbidity than those not on a PPI. Because these were not randomized trials, it may be that patients with worse disease, and at presumed greater risk of GI bleeding, were more likely to have been given PPI cotherapy. As of March 2010, there were no published, peer-reviewed data from prospective, randomized trials. However, O’Donoghue and colleagues did make a post hoc analysis of prospectively collected data from the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition with Prasugrel-Thrombolysis in Myocardial Infarction (TRITON-TIMI) 38 trial, which had randomized more than 13,000 patients with acute coronary syndrome (ACS) to clopidogrel or prasugrel. Although TRITON-TIMI 38 was not a randomized comparison of PPI cotherapy with no cotherapy in patients receiving clopidogrel, a large number of patients in the trial had been on PPI cotherapy. O’Donoghue and colleagues did not identify any adverse cardiovascular outcomes in the patients treated with clopidogrel who had received PPI cotherapy compared with those who had not. Furthermore, they identified a small but statistically significant reduction in all-cause mortality among the patients treated with clopidogrel on PPI cotherapy. To date, Clopidogrel and the Optimization of Gastrointestinal Events Trial (COGENT) is the only randomized study to have compared PPI cotherapy with no cotherapy in patients taking clopidogrel. However, this trial was terminated prematurely for financial reasons and, at the time of writing, had not been reported in full. It randomized patients to a fixed-dose combination of clopidogrel and omeprazole or to clopidogrel alone after a myocardial infarction, other ACS, or coronary stent placement. Patients who had received the combination of omeprazole and clopidogrel in COGENT did not have demonstrably worse cardiovascular outcomes than those who received clopidogrel alone. The patients who had received omeprazole had a significantly lower rate of adverse GI outcomes than those who did not.

As of March 2010, there is no definitive clinical evidence that PPI cotherapy with clopidogrel is a risk factor for adverse cardiovascular outcomes. Patients taking clopidogrel who have additional risk factors for GI bleeding should receive PPI cotherapy. However, in November 2009, the FDA announced its analysis of some pharmacodynamic and pharmacokinetic studies conducted by the manufacturers of clopidogrel. These showed that omeprazole cotherapy was associated with reduced serum levels of clopidogrel’s active metabolite and with a reduction in clopidogrel’s in vitro antiplatelet effect. Based on these studies, the FDA recommended that omeprazole (and esomeprazole) be avoided among patients taking clopidogrel. Similar studies are currently being conducted with other PPIs.

Clostridium difficile and Other Enteric Infections

Clostridium difficile is a spore-forming bacterium; its nosocomial acquisition is associated with significant morbidity and it is also emerging as an important community-acquired pathogen. Several published studies have investigated the association between PPI use and C difficile infection. The proposed mechanism for the increased risk is that gastric acid suppression allows ingested spore survival in the upper GI tract and eventual proliferation in the colon. There is also some evidence that PPIs cause leukocyte inhibition, which may contribute to an increase in various enteric infections, including C difficile .

Aseeri and colleagues published a retrospective case-control study of 94 patients with nosocomial C difficile infection. Multivariate analysis in which factors such as antibiotic use, sex, and age were controlled for, determined an odds ratio (OR) for PPI use and C difficile infection of 3.6 (95% confidence interval [CI] 1.7–5.3). In another case-control study, Yearsley and colleagues investigated 155 patients with nosocomial C difficile infection. There was an increased risk of C difficile infection with PPI therapy (OR 1.9; 95% CI 1.1–3.3). To determine the effects of PPIs on community-acquired C difficile , Dial and colleagues conducted a population-based case-control study of 1233 patients and found an adjusted rate ratio of 2.9 (95% CI 2.4–3.4) for C difficile infection among patients using PPIs. Based on the available data, there seems to be a weak association between PPI use and C difficile infection, both nosocomial and community-acquired, although causality has not been established.

Similarly, diminished gastric acid secretion is postulated to increase the risk of other enteric infections. Garcia Rodriguez and colleagues studied a retrospective cohort of 6414 patients with age- and sex-matched controls to investigate the risk of culture-positive bacterial enteric infections with PPI use. They found a dose-dependent relationship with a relative risk (RR) of 2.9 (95% CI 2.5–3.5) between gastroenteritis and PPI use. A meta-analysis evaluating 6 observational studies comprising 11,280 patients found a significant association between PPI use and bacterial enteric infection (OR 2.55; 95% CI 1.53 and 4.26). The investigators also found a quantitatively smaller association between use of H ₂ -receptor antagonists (H ₂ RAs) and bacterial enteric infection. Taken together, these observations suggest an association between acid suppression and enteric bacterial infections. H ₂ RAs are associated with a lower risk than PPIs, so there may be some degree of dose response. However, the magnitude of the increased risk is small. In general, when estimates of effect sizes from observational studies are small, they should not be interpreted as proving causation; they imply an association and might be considered to generate hypotheses.

Pneumonia

Gastric acid suppression has also been associated with the development of both community-acquired and nosocomial pneumonia. The proposed mechanism of infection is through the ability of bacteria to colonize the upper GI tract because of the increase in gastric pH with PPI therapy. The association between PPI use and leukocyte inhibition may also play a role. In a population-based cohort study of 364,683 individuals, Laheij and colleagues reported that PPI use was associated with community-acquired pneumonia (CAP) (RR 1.89; 95% CI 1.36–2.62). This association was dose-dependent, with patients on higher doses of PPI having a higher risk of developing CAP. However, there were significant confounders in this study, which did not control for the presence of GERD or chronic obstructive pulmonary disease in their cohort. Sarkar and colleagues subsequently conducted a nested case-control study of 80,066 patients, in which those who were started on a PPI within the 30 days preceding the diagnosis of CAP had a higher risk than those who had been on long-term PPI treatment. The highest risk of CAP was seen within the first 2 days of PPI therapy (OR 6.53; 95% CI 3.95–10.80). The association between CAP and short-term, but not long-term, PPI use may be an example of protopathic bias, in which patients presenting to their health care providers with certain symptoms were inappropriately prescribed a PPI.

Herzig and colleagues investigated the risk of nosocomial pneumonia and PPI use in a prospective cohort of 63,878 patients who were not ventilated and who were hospitalized for at least 72 hours. The OR for developing nosocomial pneumonia for patients on PPIs was 1.3 (95% CI 1.1–1.4) in a multivariate analysis and the association was strongest with aspiration than for nonaspiration pneumonia.

Based on these studies, there may be a slightly increased risk of both CAP and nosocomial pneumonia associated with PPI use. However, the magnitude of the observed effect is small and at least some of the studies were prone to confounding.

Rebound Hypersecretion

One of the consequences of acid suppression with a PPI is hypergastrinemia, which in turn causes enterochromaffinlike cell hyperplasia and may, ultimately, lead to parietal cell hyperplasia. Thus, the potential exists for increased gastric acid secretory capacity compared with that which existed before the introduction of the PPI. This potential has led to concern about the possibility of rebound acid hypersecretion when stopping PPIs after prolonged use.

Reimer and colleagues conducted a randomized, double-blind, placebo-controlled study in 120 asymptomatic individuals to investigate the possibility of new upper GI symptoms, possibly related to rebound acid hypersecretion, after 8 weeks of therapy with esomeprazole. They found that heartburn and dyspeptic symptoms were statistically significantly increased among the subjects who had discontinued esomeprazole compared with those who had discontinued placebo, with onset approximately 2 to 4 weeks after stopping the PPI. This study claimed that some previously asymptomatic individuals temporarily develop upper GI symptoms (possibly caused by rebound acid hypersecretion, which was not measured) after withdrawal of a PPI. However, these symptoms were temporary and of a minor nature and should not dissuade physicians from prescribing a PPI to patients with legitimate symptoms expected to respond to acid suppression. At least 1 further study has reported similar findings to Reimer and colleagues. Niklasson and colleagues reported more severe symptoms after PPI withdrawal, but these were also more transient than those detected by Reimer and colleagues. The clinical significance of these findings has been questioned.

Bone Health

In 2006, Yang and colleagues published a nested case-control study investigating the risk of hip fracture with PPI use. They identified 13,566 patients with hip fractures and found that greater than 1 year of PPI use was associated with an adjusted OR of 1.44 (95% CI 1.30–1.59). This risk appeared to increase with PPI dose and with duration of treatment. One theoretic mechanism for the increase in hip fractures was that gastric acid suppression decreases dietary calcium absorption and, in time, this may lead to relative demineralization of bone, osteopenia, osteoporosis, and an increased fracture risk. A retrospective matched cohort study of 15,792 patients found that PPI use was only associated with hip fracture after more than 6 years of exposure (adjusted OR 1.92; 95% CI 1.16–3.18). Corley conducted a retrospective case-control study using the Kaiser Permanente database for Northern California. He matched more than 33,000 individuals with hip fracture with more than 130,000 controls. Greater than 2 years of PPI use was associated with a 30% increase in the rate of hip fracture (OR 1.30; 95% CI 1.21–1.39). In addition, greater than 2 years of H ₂ RA use was also associated with a slightly increased risk (OR 1.18; 95% CI 1.08–1.28). Corley was able to demonstrate a positive dose-response relationship between PPI use and hip fracture but did not identify any duration-response relationship.

However, a recent study by Targownik and colleagues conducted both a cross-sectional and longitudinal analysis investigating bone mineral density in 2193 and 2549 patients, respectively. The longitudinal study investigated change in bone mineral density in a 1- to 3-year span. The cross-sectional study found no significant relationship between bone density and PPI use. Similarly, PPI use did not have a significant effect on bone mineral density in patients in the longitudinal analysis. Another nested case-control study by Kaye and Jick found no association between PPI use and hip fracture in 1098 patients.

There are slightly inconsistent results among the different retrospective studies of PPI use and fractures. However, most studies suggest a weak positive association, and some criteria for causation are also fulfilled, notably a dose-response relationship and, at least in most of the studies, a duration-response relationship. Although there was a once-plausible biologic rationale to explain the association, the recent findings by Targownik and colleagues call this into question. Patients who have a genuine requirement for PPI treatment should receive it in the lowest effective dose. At this time, there is no recommendation to provide patients with additional bone mineral density monitoring or more than routine calcium and vitamin D supplementation.

B ₁₂ Deficiency

Vitamin B ₁₂ deficiency, if left untreated or undetected, has potentially serious and irreversible systemic consequences. An early report of a possible association between B ₁₂ deficiency and PPI use (>1 year), was made by Valuck and Ruscin in their case-control study of 53 patients greater than 65 years old and 212 age- and sex-matched controls. However, this association has been questioned based on a cross-sectional study published by den Elzen and colleagues, which recruited 125 patients more than 65 years of age on PPI therapy for greater than 3 years and used their partners who were not on PPI therapy as the control group. The study patients had serum B ₁₂ levels measured and compared with those of their partners. Multivariate analysis, controlling for age, sex, H pylori status, and C-reactive protein, showed no statistical difference in B ₁₂ levels between patients and their partners. Based on these data, there is no evidence of a discernable association with PPI use and the development of B ₁₂ deficiency.

Side Effects, Anaphylaxis, and Allergy

As a class, PPIs are well tolerated with few reported side effects, the most common being headache, constipation, diarrhea, dizziness, and rash. There have been several case reports and series of anaphylactic and allergic reactions to PPI therapy. The largest published case series described 9 patients with adverse reaction to omeprazole: urticaria/angioedema in 7 and anaphylaxis in 2. Most of these patients had positive pin-prick tests suggesting immunoglobulin E–mediated allergy. Four patients also had a positive pin-prick test to pantoprazole. However, lansoprazole was well tolerated, suggesting a lack of a class effect with anaphylactic reactions.

Another reported adverse reaction to PPI therapy is acute interstitial nephritis (AIN). In a systematic review, Sierra and colleagues identified 64 cases of PPI-associated AIN, 59 of which were proven by biopsy. All of the PPIs, except for dexlansoprazole (which was not available at that time), were implicated, suggesting a class effect. In addition, there was improvement in renal function by a mean of 35.5 weeks after withdrawal of PPI therapy, but creatinine levels did not return to baseline. Evidence for causality in the cases reviewed was often poor; in general, patients have not been challenged with reintroduction of the suspected PPI, for obvious reasons. However, these adverse effects are tempered by the extremely low overall rate of adverse reactions, estimated at 0.01% in one case series.

Pregnancy

Given the widespread use of PPIs, there is concern about their possible effects on the developing fetus. Omeprazole is listed by the FDA as a pregnancy class C medication, whereas all other currently available PPIs are class B, which still denotes a possible serious risk to the fetus. Gill and colleagues recently published a systematic review evaluating the risk of PPI use in pregnancy. Among 1530 PPI-exposed subjects compared with 133,410 controls, they found no association between major fetal malformations, spontaneous abortions, or pre-term delivery. This lack of association was also apparent when omeprazole alone was investigated.

However, there is some suggestion that acid suppression alone may have an adverse affect on an exposed fetus. In a recent, large, case-control study investigating children with asthma, Dehlink and colleagues found a relationship between maternal gastric acid suppression with a PPI or H ₂ RA and the development of childhood asthma (OR 1.43; 95% CI 1.29–1.59). The postulated mechanism, based on animal model data, is that acid suppression increases type 2 helper cell bias in their offspring, thus predisposing to increased atopy. Based on these data, the immediate fetal developmental risk of PPIs may be negligible. However, possible risks to the fetus that may only become manifest in childhood require further study.

A summary containing our assessment of the clinical relevance of the various proposed adverse effects of PPIs is given in Table 1 .

Table 1

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Nonsteroidal Antiinflammatory Drug-Related Injury to the Gastrointestinal Tract: Clinical Picture, Pathogenesis, and Prevention Clinical Pharmacology of 5-ASA Compounds in Inflammatory Bowel Disease Therapeutic Potential of Peroxisome Proliferator-Activated Receptors in Chronic Inflammation and Colorectal Cancer Nonsteroidal Antiinflammatory Drug-Related Injury to the Gastrointestinal Tract: Clinical Picture, Pathogenesis, and Prevention Clinical Pharmacology of 5-ASA Compounds in Inflammatory Bowel Disease New Pharmacologic Therapies in Gastrointestinal Disease

Stay updated, free articles. Join our Telegram channel

Join