Neurogastroenterol Motil 2017; 29: e12926; wileyonlinelibrary.com/journal/nmoDOI: 10.1111/nmo.12926

 |  1 of 10© 2016 John Wiley & Sons Ltd

Received: 13 November 2015  |  Accepted: 21 July 2016

DOI: 10.1111/nmo.12926

AbstractBackground and Purpose: Proton pump inhibitors (PPIs) are commonly used as potent gastric acid secretion antagonists for gastro- esophageal disorders and their overall safety in patients with gastro- esophageal reflux disease (GERD) is considered to be good and they are well- tolerated. However, recent studies have suggested that PPIs may be a potential independent risk factor for cardiovascular adverse events. The aim of our meta- analysis was to examine the association between PPI monotherapy and cardiovascular events in patients with GERD.Methods: A literature search involved examination of relevant databases up to July 2015 including PubMed, Cochrane Library, EMBASE, and ClinicalTrial.gov, as well as selected randomized controlled trials (RCTs) reporting cardiovascular events with PPI exposure in GERD patients. In addition, the pooled risk ratio (RR) and heterogeneity were assessed based on a fixed effects model of the meta- analysis and the I2 statistic, respectively.Key Results: Seventeen RCTs covering 7540 patients were selected. The pooled data suggested that the use of PPIs was associated with a 70% increased cardiovascular risk (RR=1.70, 95% CI: [1.13–2.56], P=.01, I2=0%). Furthermore, higher risks of adverse cardiovascular events in the omeprazole subgroup (RR=3.17, 95% CI: [1.43–7.03], P=.004, I2=25%) and long- term treatment subgroup (RR=2.33, 95% CI: [1.33–4.08], P=.003, I2=0%) were found.Conclusion & Inferences: PPI monotherapy can be a risk factor for cardiovascular adverse events. Omeprazole could significantly increase the risk of cardiovascular events and, so, should be used carefully.

K E Y W O R D S

cardiovascular events, gastro-esophageal reflux disease, meta-analysis, proton pump inhibitors

1Shenyang Pharmaceutical University, Shenyang, China2Department of Gastroenterology, General Hospital of Shenyang Military Area, Shenyang, China

CorrespondenceGang Cheng, Shenyang Pharmaceutical University, Shenhe District, Shenyang, Liaoning Province, China.Email: chenggang63@hotmail.com

O R I G I N A L A R T I C L E

Proton pump inhibitor monotherapy and the risk of cardiovascular events in patients with gastro- esophageal reflux disease: a meta- analysis

S. Sun1 | Z. Cui2 | M. Zhou1 | R. Li1 | H. Li1 | S. Zhang1 | Y. Ba1 | G. Cheng1

1  | INTRODUCTION

Proton pump inhibitors (PPIs) are effective antagonists of gastric acid secretion and are used for many common gastro- esophageal disorders such as dyspepsia, gastro- esophageal reflux disease, Zollinger–Ellison syndrome, Barrett’s esophagus and Helicobacter pylori (H. pylori) infec-tion of the upper gastrointestinal tract.1–3 Each year, over $13 bil-lion is spent on PPIs globally4 because of generally good safety and

tolerability based on previous trials and analyses.5–10 However, an approximately 3% adverse reaction rate has been reported, which includes headache, dizzines, diarrhea, constipation, and cutaneous reactions.11 In addition, recent reports have identified PPIs as poten-tial risk factors for interstitial nephritis, osteoporosis and clostridium difficile- associated diseases.

Although controversy about whether such cardiovascular adverse events are due to a potential drug interaction between clopidogrel

2 of 10  |     Sun et al.

and PPIs remains unresolved,12–18 the updated guidelines of GERD have strongly recommended that PPI therapy does not need to be changed for patients receiving clopidogrel, since no data supporting an increased risk of adverse cardiovascular events have been found clinically.19 Safety concerns about PPI monotherapy were raised in 2007, and a communication about the ongoing safety review of omeprazole and esomeprazole was issued by the Food and Drug Administration (FDA); nevertheless, there was no consensus due to the discrepancy in baseline comorbidity between PPI users and non- users. However, a small number of observational studies have recently suggested that PPIs might be independently related to cardiovascular adverse events.20–22 A similar conclusion has been reached based on a population- based study in Canada.23 However, owing to serious pub-lic health implications, most previous studies have focused on patients with existing coronary artery disease (CAD) and the outcomes of recur-ring cardiovascular events. Therefore, the exclusive impact of PPIs on the risk of cardiovascular events, such as patients with no previous history of cardiovascular disease, is still unclear. In consequence, we decided to conduct a meta- analysis based on relevant and available RCTs to assess the cardiovascular risks of PPIs in patients with GERD.

2  | MATERIALS AND METHODS

We used the preferred reporting items for systematic reviews and meta- analysis for guideline in this study.24

2.1 | Eligibility criteria

In this meta- analysis, we selected randomized controlled trials (RCTs) in which patients with gastro- esophageal reflux disease were random-ly divided into two groups: PPI users and non- PPI users.

The titles and abstracts were independently screened by two reviewers (RFL and HXL), and studies were chosen for meta- analysis if they met the following inclusion criteria: (i) adult patients with GERD; (ii) comparison of the safety of PPI monotherapy with other acid- suppressive drugs, placebo, or surgery; (iii) clear reporting of adverse events; (iv) studies involving more than 100 patients; (v) RCTs. There was no restriction on the language of publication. Also, the exclusion criteria were: (i) no human subjects in the study; (ii) another PPI in the control group; (iii) combination drug therapy; (iv) duplicate publication. Disagreements were resolved through consensus and arbitration by the third author (ZMC).

2.2 | End points and subgroup analyses

The clinical outcomes of interest were reported for cardiovascular events including cardiovascular abnormalities, heart diseases, and vas-cular diseases (i.e. acute myocardial infarction, angina pectoris, arteri-osclerosis coronary artery, cardiac failure, CAD, myocardial infarction, myocardial ischemia, coronary artery stenosis, or cardiac disorders). We also did subgroup analyses involving different types of PPIs, treat-ment duration, and therapeutic approaches.

2.3 | Search strategy

We searched databases from PubMed, Cochrane Library, EMBASE up to July 2015 to identify RCTs that reported PPIs vs other acid- suppressive drugs (such as ranitidine), placebo or surgery in patients with gastro- esophageal reflux disease. Both MeSH and free text terms were used to identify relevant articles. The ClinicalTrials.gov was searched to identify additional eligible clinical trials without language restriction. The details of the search strategy are provided in the “Appendix”. As a result, 948 studies from the Cochrane Library, 556 studies from the PubMed data-base and 324 studies from the EMBASE database were obtained using our search strategy. We also examined the bibliographies of the includ-ed trials and recent review articles for relevant studies.

2.4 | Study selection and data extraction

Independently, two reviewers (SZ and YJB) extracted data using a standardized data extraction form which included the following items: the first author, the year of publication, the mean age of participants, gender, intervention type, study duration, medication duration, com-pleteness of participants, and relevant outcomes. The data extraction was performed by two of the authors independently (SZ and YJB), and any conflicts were resolved by the third investigator (ZMC). When relevant outcomes were not reported in interest, we tried to contact the authors to obtain any relevant missing information.

2.5 | Assessment of risk of bias

In the quality evaluation, two reviewers (RFL and HXL) assessed the quality (risk of bias) of each study independently using the Cochrane Risk of Bias Tool (http://handbook.cochrane.org/).25,26 Biases were mainly classified into seven domains: random sequence generation, allocation concealment, blinding of participants and personnel, blind-ing of outcome assessment, incomplete outcome data, selective reporting, and other bias.

Quality evidence of this meta- analysis was assessed using Grading of Recommendations Assessment, Development, and Evaluation (GRADE)27 by two investigators, and the GRADE profiler was used to create a GREDE evidence profile (Version 3.6.1, GRADE Working Group). Any disagreements were resolved by discussion.

Key Points

• Recently, a small number of studies have suggested that PPIs might not be as safe as we think.

• In the light of the results of our meta-analysis, PPI mono-therapy can be a risk factor for cardiovascular adverse events in patients with GERD, and omeprazole could significantly increase the risk of cardiovascular events.

• Thus, we believe that doctors should advised to carefully consider the use of PPIs in clinical situations, and try to choose the best treatment option for each patient..

     |  3 of 10Sun et al.

2.6 | Data analysis

Review Manager ([RevMan] Version 5.2.3 Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2012) was used to estimate the pooled risk ratio (RR) and its 95% confidence intervals based on a fixed effect model meta- analysis. The Cochrane Q test was used to evaluate the heterogeneity among studies. A P value of ≤0.05 was considered statistically significant and the I2 test was also performed to evaluate the magnitude of heterogeneity. Studies with an I2 statistic of 25–50% were considered to have low heterogeneity, while those with an I2 statistic of 50–75% were considered to have moderate heterogeneity and those with an I2 statistic of >75% were considered to have high heterogeneity.

A funnel plot of each trial’s effect size against the standard error was used to check the publication bias. Funnel plot asymmetry was evaluated by Begg’s and Egger’s tests (P<.05 defined as having a pub-lication bias), performed using STATA 11.0 software (Statacorp LP, Colege Station, TX, USA). A sensitive analysis of a single study on the overall pooled estimate was conducted by sequentially omitting one study in the STATA 11.0 software.

3  | RESULTS

3.1 | Study identification and selection

As shown in the following searching strategy (Fig. 1), we identified 1828 studies, and a total of 1014 records remained after removal of duplicates. After reading the titles and abstracts, 906 records were excluded due to unrelated topics or unsuitable control groups. One

hundred and eight potentially relevant full- text articles were identi-fied and 92 studies were excluded after reviewing the full articles. Most of these were not used because they did not mention serious adverse events or report matters of interest. Finally, 16 studies which met the inclusion criteria were included in the meta- analysis.5–10,28–37

3.2 | Study characteristics

These 16 RCTs, which were all multi- center studies, involved 7540 patients with 4512 patients in the PPI group and 3028 patients in the control group. The characteristics of all 16 studies are summarized in Table 1.

3.3 | Risk of bias of included studies and quality of evidence

A total of five studies described the random sequence generation method mainly by a computer- generated randomization schedule or an interactive voice response system,5,6,29,32,34 and others were regarded as an unclear risk. Two studies were performed by a phar-macy (not involved in recruiting or evaluating patients) to maintain concealment of allocation,9,10 and two studies were performed using sequentially numbered, opaque, sealed envelopes,6,28 and one study maintained concealment of allocation using a numbered bot-tle.6 These studies were considered to have a low risk of bias in allocation concealment and others were regarded as an unclear risk. A total of seven studies were registered in the ClinicalTrial.gov and an NCT number was found.5,6,9,28,34,35,37 According to their proto-col on ClinicalTrial.gov, these were considered to have a low risk of

F IGURE  1 Flow diagram of literature search to identify RCTs evaluating the risk of cardiovascular of proton pump inhibitor in patients with gastro- esophageal reflux disease

4 of 10  |     Sun et al.

bias in selective reporting and others were regarded as an unclear risk. For incomplete outcome data, two studies5,29 with a higher rate of discontinuation in the placebo group was primarily though to be because of a higher rate of erosive oesophagitis recurrence and resulted in higher mean lengths of study drug exposure to a PPI compared with placebo. So, we considered these two studies

as a high risk of incomplete outcome data, whereas others had a similar rate of discontinuation. We could not determine whether there were unknown factors that might bias each study, so all stud-ies were regarded as an unclear risk in the domain of other risk. The risk of bias domain of all 16 studies is summarized in supplemen-tary Figure 1. The quality of evidence was assessed by the GRADE

TABLE  1 Characteristics of the randomized controlled clinical trials included in this analysis

Study Design Year Conditions Arms of treatmentMale (%)

Mean age Population

Therapeutic duration (weeks)

Completeness (%)

Howden5 Double- blind RCT

2009 Esophagitis, reflux Dexlansoprazole MR 60 mg

52 49.7 159 27 56.5

Dexlansoprazole MR 90 mg

54 48.8 152

Placebo 50 48.2 140

Peura6 Double- blind RCT

2013 Gastroesophageal reflux disease

Dexlansoprazole MR 30 mg

27 47.6 315 4 94.6

Dexlansoprazole MR 60 mg

34 47.5 315

Placebo 26 47.6 317

Armstrong7 Double- blind RCT

2001 Gastroesophageal reflux disease

Pantoprazole 40 mg 57 47 106 4 93.8Nizatidine 51 48 102

Friedlander28 Double- blind RCT

2010 Esophagitis, reflux Dexlansoprazole MR 30 mg

49 47.1 140 27 81.1

Dexlansoprazole MR 60mg 47 47.9 158Placebo 49 49.5 147

Joel E. Richter29

Open- label RCT

1996 Gastroesophageal reflux disease

Omeprazole20 mg 61 49 100 8 83.8Ranitidine 61 49 97

Joel E. Richter30

Double- blind RCT

2000 Non- erosive gastroesophageal reflux disease

Omeprazole10 mg 51.7 49.5 118 4 97.8Omeprazole20 mg 55.1 50.0 118Placebo 54.5 49.7 123

Lundell31 Open- label RCT

2009 Gastroesophageal reflux disease

Omeprazole20 mg 75 55 154 540 67.8Surgery 76 51 144

Lundell8 Open- label RCT

2008 Gastroesophageal reflux disease

Esomeprazole 20 mg 75 45 266 8 82.5Surgery 69 45 248

Bigard32 Double- blind RCT

2005 Gastroesophageal reflux disease

Lansoprazole 15 mg 56 52.5 84 4 74.0Placebo 36 54.0 97

Talley33 Double- blind RCT

2001 Gastroesophageal reflux disease

Esomeprazole 20 mg 55 49 170 27 68.1Placebo 57 49 172

Flook9 Double- blind RCT

2013 Gastroesophageal reflux disease

Esomeprazole 40 mg 52 47 297 4 86.5Placebo 53 47 302

Sugisaki34 Double- blind RCT

2011 Non- erosive gastroesophageal reflux disease

Rabeprazole sodium 5 mg 41 46.3 93 4 95.1Rabeprazole sodium 10 mg 50 46.8 101Placebo 44 49.7 91

Attwood35 Open- label RCT

2015 Gastroesophageal reflux disease

Esomeprazole 20 mg 75 45.3 266 270 70.3Surgery 69 44.8 288

Venables10 Double- blind RCT

1997 Gastroesophageal reflux disease

Omeprazole 10 mg 52 51 330 4 NMOmeprazole 20 mg 49 51 338Ranitidine 50 50 326

Thomas36 Double- blind RCT

2010 Healed erosive esophagitis

Lansoprazole 15 mg 72 49.6 100 54 57.8Ranitidine 63 50.3 106

Toni O. Kiljander37

Double- blind RCT

2010 Gastroesophageal reflux disease, asthma

Esomeprazole 40 mg 25 44 319 4 87.7Esomeprazole 80 mg 23 45 313Placebo 25 45 328

RCT, randomized controlled trial; NM, not mentioned.

     |  5 of 10Sun et al.

profiler, and the quality of the evidence and their causes are shown in Supplementary Figure 3.

3.4 | PPI group vs control group with regard to the risk of cardiovascular events

There were 16 studies5–10,28–37 which compared PPIs with placebo, H2 receptor antagonists or surgery in terms of the risk of cardiovas-cular events of GERD. A total of 4512 patients received PPIs while 3028 patients did not. The results of a fixed effects meta- analysis showed that taking PPIs was associated with a 70% increased risk of cardiovascular events (RR=1.70, 95% CI (1.13–2.56), P=.01). A fixed effect model was chosen because no obvious statistical heterogene-ity was identified (I2=0%; P=.72) (Fig. 2). For this outcome, the funnel plot was symmetrical (supplementary Figure 2), and Egger’s (P=.513) and Begg’s test (P=.392) also did not find any publication bias. We also performed subgroup analyses, according to the different type of PPIs (Fig. 3) or different treatment durations (Fig. 4) or compared with a different type of therapy in the control group (Fig. 5) on the risk of cardiovascular events. The subgroup analyses showed that ome-prazole increased the risk of adverse cardiovascular events (RR=3.17 [1.43–7.03], P=.004), for the long- term treatment subgroup (more than 8 weeks), a significant increase in cardiovascular risk (RR=2.33, 95% CI [1.33–4.08], P=.003) being found and the PPIs increased the risk of cardiovascular events compared with an H2- receptor antago-nist (RR=4.50 [1.03–19.67], P=.05) or surgery (RR=1.82 [1.09–3.03], P=.02), but this risk was not statistically significant for the esomepra-zole subgroup and the short- term treatment subgroup. Sensitivity analysis indicated that no study apparently influenced the result after excluding each study.

4  | DISCUSSION

To date, numerous studies have found that PPIs may affect vascu-lar function independently,20–22,38,39 and this might be associated with an increased risk of cardiovascular adverse events. However, most of these were population- based, case cross- over or pharmaco- epidemiological studies with results being easily limited by protopath-ic bias or confounded by indication. The association between PPIs and cardiovascular events in the studies did not, in itself, proved causation. Up to now, no meta- analysis has investigated the cardiovascular risk of PPI monotherapy for patients without any cardiovascular diseases.

We performed a meta- analysis of cardiovascular outcomes in 7540 patients with GERD from 16 RCTs taking PPIs or not, and a 70% increase in cardiovascular risks was obtained, demonstrating that PPIs might be associated with an elevated risks of cardiovascular events in patients with GERD. Moreover, subgroup analyses based on the PPI type, treatment duration and other common GERD treatment meth-ods were also carried to evaluate the risk of cardiovascular events, and the results obtained showed a higher risk of cardiovascular adverse events in the omeprazole subgroup and the long- term treatment subgroup. Compared with surgery or H2- receptor antagonists, PPIs significantly increased cardiovascular risks. However, there was no connection with the risk of cardiovascular events in the esomeprazole and short- term treatment subgroups.

The last published systematic review and meta- analysis40 of the incidence of cardiovascular events in patients receiving clopido-grel, with and without proton pump inhibitors, showed that PPIs are markers of increased cardiovascular risk in patients taking clopido-grel, although there are a number of differences between our studies.

F IGURE  2 Meta- analysis for PPI group vs control group on the risk of cardiovascular

6 of 10  |     Sun et al.

Firstly, the former meta- analysis included studies in patients taking clopidogrel and prespecified subgroup analyses were performed to search for potential factors associated with a clopidogrel- PPI interac-tion, such as patients with PCI or ACS that describes the cardiovascu-lar condition of the patients. It is well known that patients who take clopidogrel have cardiovascular diseases. In our meta- analysis, the 16 studies involved are all RCTs in which patients with gastro- esophageal reflux diseases were randomized into those who were PPI users or non- PPI users. No potential interactions were present. In addition, patients were diagnosed with gastro- esophageal reflux diseases rather than cardiovascular diseases, and their vascular function is more similar to that of the general population. Secondly, the former meta- analysis

has considerable heterogeneity (I2=79%) and it was considered that inconsistency in each study may not be fully explained by differences in populations, interventions, or outcomes. The cardiovascular conditions of patients on clopidogrel were inconsistent and there was uncertainty about the underlying effects due to this inconsistency. However, in our study, no heterogeneity was found in the studies we included (I2=0%). Therefore, there was no influence on the results of different severity in patients with GERD. Thirdly, most studies included in the former meta- analysis were non- randomized and inherently more susceptible to bias. The quality of evidence assessed by the GRADE approach in the former meta- analysis was rather low with regard to outcomes and we assessed the quality of outcomes as being moderate evidence.

F IGURE  3 Subgroup analyses of different proton pump inhibitors on the risk of cardiovascular

     |  7 of 10Sun et al.

Similar to our findings, a large US data mining study39 indicates people taking PPIs have a 16% higher risk of myocardial infarction. However, there were significant variations between our meta- analysis and this US research. Firstly, all data quoted in our studies are infor-mation from RCTs in place of clinical data for pharmacovigilance. Secondly, our results were obtained using RevMan software process-ing rather than descriptive analysis. More importantly, several con-victive evaluation methods, such as the Cochrane Risk of Bias Tool, funnel plot, Begg’s and Egger’s test as well as the GRADE approach, were also used to ensure the reliability of results.

According to our results, PPIs appeared to be associated with an increased risk of cardiovascular adverse events in the general popula-tion. Subgroup analyses showed that different results were obtained for omeprazole and esomeprazole, the biological mechanism of which is not yet completely understood and this therefore highlights the impor-tance of choosing the most suitable type of drug in clinical practice.

However, the potential mechanism can be deduced from the very long treatment duration of omeprazole in our research and the length of treatment and dosage may be the proper explanation. Recently, Ghebremariam et al.41 conducted a controlled open- label cross- over pilot study suggesting that the use of PPIs may not cause vascular dysfunction over the short 4 week course studied in the pilot, which

further emphasizes the influence of the length of PPI treatment on adverse reactions.

However, it is still difficult to reach a meaningful conclusion about other types of PPIs (rabeprazole sodium, lansoprazole, pantoprazole, dexlansoprazole) owing to limited number of studies performed. In addition, we found that the long- term treatment subgroup (more than 8 weeks) showed that the risk of cardiovascular adverse events in people taking PPIs was nearly double that in people not taking PPIs, while there was no direct cardiovascular risk associated with PPIs in the subgroup on short- term therapy. PPIs showed elevated cardiovas-cular risks compared with other GERD treatments. These conclusions should remind doctors that PPIs may not be entirely safe and clinicians should consider both the pros and cons of PPIs, and choose the best treatment option for long- term therapy in GERD patients.

The mechanism of this possible adverse effect might be that PPIs increase the plasma asymmetric dimethylarginine (ADMA) lev-el and reduce the nitric oxide level.42 The increase in ADMA would be expected to impair vascular nitric oxide synthase activity, increase oxidative stress, reduce vasodilator function, and impair vasopro-tective mechanisms especially in the case of omeprazole.42–48 This plausible biological mechanism could explain the association of PPIs with increased cardiovascular events in patients with GERD. However,

F IGURE  4 Subgroup analyses of different duration of PPI on the risk of cardiovascular

8 of 10  |     Sun et al.

there remain several limitations in our meta- analysis. The quality of evidence assessed by the GRADE approach was moderate or low for outcomes. Cardiovascular adverse events as outcomes were non- essential for the quoted studies that mostly were about the efficacy and remission rate, which was the main reason for a downgrade in the quality of evidence. In addition, we were unable to identify the impact of different doses of PPIs on cardiovascular events by virtue of the on- demand strategy applied to long- term treatment programs. Finally, several potential confounding factors, for instance, obesity, smoking, alcohol, and family history of cardiovascular disease, were not covered by our analyses.

The detailed investigation and analysis of cardiovascular adverse events produced by several PPIs in the context of safe use may lead to better designed, larger, prospective RCTs for patients without a history of cardiovascular diseases, and reporting results in detail regarding the cardiovascular characteristics of participants.

In conclusion, our meta- analysis showed that taking a proton pump inhibitor was related to a 70% increased risk of cardiovascular events in patients with GERD. Among PPIs, omeprazole could significantly increase cardiovascular risks, and long- term PPI therapy is associated with an increased risk of adverse cardiovascular events.

FUNDING

No funding declared.

DISCLOSURE

The authors have no competing interests.

AUTHOR CONTRIBUTION

SHS studied the concept and design, acquired data, analyzed and interpreted data, drafted the manuscript, and approved final submit-ted draft; ZMC, MJZ, RFL, HXL, SZ, YJB analyzed, and interpreted the data, critically revised the manuscript for important intellectual content, and approved final submitted draft; GC supervised, criti-cally revised the manuscript for important intellectual content, and approved the final submitted draft. GC will act as the guarantor for this paper.

REFERENCES

1. Kahrilas PJ. Clinical practice. Gastroesophageal reflux disease. N Engl J Med. 2008;359:1700–1707.

2. Berardi RR, Welage LS. Proton- pump inhibitors in acid- related dis-eases. Am J Health Syst Pharm. 1998;55:2289–2298.

3. Stedman CA, Barclay ML. Review article: comparison of the pharma-cokinetics, acid suppression and efficacy of proton pump inhibitors. Aliment Pharmacol Ther. 2000;14:963–978.

4. Katz MH. Failing the acid test: benefits of proton pump inhibitors may not justify the risks for manyusers. Arch Intern Med. 2010;170:747–748.

5. Howden CW, Larsen LM, Perez MC, et al. Clinical trial: efficacy and safety of dexlansoprazole MR 60 and 90 mg in healed erosive oesophagitis – maintenance of healing and symptom relief. Aliment Pharmacol Ther. 2009;30:895–907.

6. Peura DA, Pilmer B, Hunt B, et al. Distinguishing the impact of dexlansoprazole on heartburn vs. regurgitation in patients with gastro- oesophageal reflux disease. Aliment Pharmacol Ther. 2013;38:1303–1311.

7. Armstrong D, Pare P, Pericak D, et al. Symptom relief in gastroesoph-ageal reflux disease: a randomized, controlled comparison of pan-

F IGURE  5 Subgroup analyses of different type of therapy in the control group on the risk of cardiovascular

     |  9 of 10Sun et al.

toprazole and nizatidine in a mixed patient population with erosive esophagitis or endoscopy- negative reflux disease. Am J Gastroenterol. 2001;96:2849–2857.

8. Lundell L, Attwood S, Ell C, Fiocca R, et al. Comparing laparoscopic antireflux surgery with esomeprazole in the management of patients with chronic gastro- oesophageal reflux disease: a 3- year interim anal-ysis of the lotus trial. Gut. 2008;57:1207–1213.

9. Flook NW, Moayyedi P, Dent J, et al. Acid- suppressive therapy with esomeprazole for relief of unexplained chest pain in primary care: a randomized, double- blind, sal. Am J Gastroenterol. 2013;108:56–64.

10. Venables TL, Newland RD, Patel AC, et al. Omeprazole 10 milligrams once daily, omeprazole 20 milligrams once daily, or ranitidine 150 mil-ligrams twice daily, evaluated as initial therapy for the relief of symp-toms of gastro- oesophageal reflux disease in general practice. Scand J Gastroenterol. 1997;32:965–973.

11. Mearin F, Ponce J. Potent acid inhibition: summary of the evidence and clinical application. Drugs 2005;65(suppl):113–126.

12. Juurlink DN, Gomes T, Ko DT, et al. A population- based study of the drug interaction between proton pump inhibitorsand clopidogrel. CMAJ. 2009;180:713–718.

13. Ho PM, Maddox TM, Wang L, et al. Risk of adverse outcomes associ-ated with concomitant use of clopidogrel and proton pump inhibitors following acute coronary syndrome. JAMA. 2009;301:937–944.

14. Dunn SP, Macaulay TE, Brennan DM, et al. Abstract 3999: Baseline proton pump inhibitor use is associated with increased cardiovascular events with and without the use of clopidogrel in the CREDO Trial. Circulation. 2008;118:S815.

15. Rassen JA, Choudhry NK, Avorn J, et al. Cardiovascular outcomes and mortality in patients using clopidogrel with proton pump inhibitors after percutaneous coronary intervention or acute coronary syn-drome. Circulation. 2009;120:2322–2329.

16. Food and Drug Administration. Public Health Advisory 17 November 2009:updated Safety Information about a drug interaction between Clopidogrel Bisulfate (marketed as Plavix) and Omeprazole (marketed as Prilosec and Prilosec OTC). http://www.fda.gov/Drugs/Drug-Safety/PublicHealthAdvisories/ucm190825.htm. Accessed Decem-ber 15, 2009.

17. Sibbing D, Kastrati A. Risk of combining PPIs with thienopyridines: fact or fiction? Lancet. 2009;374:952–954.

18. Laine L, Hennekens C. Proton pump inhibitor and clopidogrel interac-tion: fact or fiction? Am J Gastroenterol. 2010;105:34–41.

19. Katz PO, Gerson LB, Vela MF. Guidelines for the diagnosis and man-agement of gastroesophageal reflux disease. Am J Gastroenterol. 2013;108:308–328.

20. Charlot M, Grove EL, Hansen PR, et al. Proton pump inhibitor use and risk of adverse cardiovascular events in aspirin treated patients with first time myocardial infarction: nationwide propensity score matched study. BMJ. 2011;342:d2690.

21. Charlot M, Ahlehoff O, Norgaard ML, et al. Proton- pump inhibitors are associated with increased cardiovascular risk independent of clopidogrel use. Ann Intern Med. 2010;53:378–386.

22. Goodman SG, Clare R, Pieper KS, et al. Association of proton pump inhibitor use on cardiovascular outcomes with clopidogrel and tica-grelor: insights from PLATO. Circulation. 2012;125:978–986.

23. Juurlink DN, Dormuth CR, Huang A, et al. Proton pump inhibitors and the risk of adverse cardiac events. PLoS ONE. 2013;8:e84890.

24. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta- analyses: the PRISMA statement. Ann Intern Med. 2009;151:264–269.

25. Cochrane Collaboration. Cochrane Handbook for Systematic Reviews of Interventions, 2009. Available at www.cochrane-handbook.org. Accessed August 28, 2013.

26. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.0.0 [updated February 2008]. The Cochrane

Collaboration, 2008. Available from www.cochrane-handbook.org randomised trials. BMJ. 2011;343:d5928.

27. Guyatt G, Oxman AD, Akl EA, et al. GRADE guidelines: 1. Introduction- GRADE evidence profiles and summary of findings tables. J Clin Epide-miol 2011;64:383–394.

28. Friedlander EA, Pallentino J, Miller SK, et al. The evolution of proton pump inhibitors for the treatment of gastroesophageal reflux disease. J Am Acad Nurse Pract. 2010;22:674–683.

29. Richter JE, Sabesin SM, Kogut DG, et al. Omeprazole versus raniti-dine or ranitidine/metoclopramide in poorly responsive symptomatic gastroesophageal reflux disease. Am J Gastroenterol. 1996;91:1766–1772.

30. Richter JE, Peura D, Benjamin SB, et al. Efficacy of omeprazole for the treatment of symptomatic acid reflux disease without esophagitis. Arch Intern Med. 2000;160:1810–1816.

31. Lundell L, Miettinen P, Myrvold HE, et al. Comparison of outcomes twelve years after antireflux surgery or omeprazole maintenance ther-apy for reflux esophagitis. Clin Gastroenterol Hepatol 2009;7:1292–1298.

32. Bigard MA, Genestin E. Treatment of patients with heartburn without endoscopic evaluation: on- demand treatment after effective contin-uous administration of lansoprazole 15 mg. Aliment Pharmacol Ther. 2005;22:635–643.

33. Talley NJ, Lauritsen K, Tunturi-Hihnala H, et al. Esomeprazole 20 mg maintains symptom control in endoscopy- negative gastro- oesophageal reflux disease: a controlled trial of ‘on- demand’ therapy for 6 months. Aliment Pharmacol Ther. 2001;15:347–354.

34. Sugisaki N, et al. A comparative study on the efficacy and safety of E3810 in patients with non- erosive gastroesophageal reflux disease.ClinicalTrials.gov. https://clinicaltrials.gov/ct2/show/NCT00165646? term=NCT00165646&rank=1. (June 1, 2015)

35. Attwood SE, Ell C, Galmiche JP, et al. Long- term safety of proton pump inhibitor therapy assessed under controlled, randomised clinical trial conditions: data from the SOPRAN and LOTUS studies. Aliment Pharmacol Ther. 2015;41:1162–1174.

36. Kovacs TO, Freston JW, Haber MM, et al. Long- term quality of life improvement in subjects with healed erosive esophagitis: treatment with lansoprazole. Dig Dis Sci. 2010;55:1325–1336.

37. Kiljander TO, Junghard O, Beckman O, et al. Effect of esomepra-zole 40 mg once or twice daily on asthma: a randomized, placebo- controlled study. Am J Respir Crit Care Med. 2010;181:1042–1048.

38. Shih CJ, Chen YT, Ou SM, et al. Proton pump inhibitor use represents an independent risk factor for myocardial infarction. Int J Cardiol. 2014;177:292–297.

39. Shah NH, LePendu P, Bauer-Mehren A, et al. Proton pump inhibitor usage and the risk of myocardial infarction in the general population. PLoS ONE. 2015;10:e0124653.

40. Cardoso RN, Benjo AM, DiNicolantonio JJ, et al. Incidence of cardio-vascular events and gastrointestinal bleeding in patients receiving clopidogrel with and without proton pump inhibitors: an updated meta- analysis. Open Heart. 2015;2:e000248.

41. Ghebremariam YT, Cooke JP, Khan F, et al. Proton pump inhibitors and vascular function: a prospective cross- over pilot study. Vasc Med. 2015;20:309–316.

42. Ghebremariam YT, LePendu P, Lee JC, et al. An unexpected effect of proton pump inhibitors: elevation of the cardiovascular risk factor ADMA. Circulation. 2013;128:845–853.

43. Mittermayer F, Krzyzanowska K, Exner M, et al. Asymmetric dimeth-ylarginine predicts major adverse cardiovascular events in patients with advanced peripheral artery disease. Arterioscler Thromb Vasc Biol. 2006;26:2536–2540.

44. Wilson AM, Shin DS, Weatherby C, et al. Asymmetric dimethylargi-nine correlates with measures of disease severity, major adverse car-diovascular events and all- cause mortality in patients with peripheral arterial disease. Vasc Med. 2010;15:267–274.

10 of 10  |     Sun et al.

45. Lu TM, Chung MY, Lin MW, et al. Plasma asymmetric dimethylarginine predicts death and major adverse cardiovascular events in individuals referred for coronary angiography. Int J Cardiol. 2011;153:135–140.

46. Ari H, Ari S, Erdogan E, et al. A novel predictor of restenosis and adverse cardiac events: asymmetric dimethylarginine. Heart Vessels. 2010;25:19–26.

47. Boger RH, Sullivan LM, Schwedhelm E, et al. Plasma asymmetric dimethylarginine and incidence of cardiovascular disease and death in the community. Circulation. 2009;119:1592–1600.

48. Wanby P, Teerlink T, Brudin L, et al. Asymmetric dimethylarginine (ADMA) as a risk marker for stroke and TIA in a Swedish population. Atherosclerosis. 2006;185:271–227.

SUPPORTING INFORMATION

Additional Supporting Information may be found online in the support-ing information tab for this article.

APPENDIX Search Strategy

PUBMED

1. proton pump inhibitor[Title/Abstract] OR ppi[Title/Abstract] OR omeprazole [Title/Abstract]) OR esomeprazole[Title/Abstract] OR lansoprazole[Title/Abstract] OR pantoprazole[Title/Abstract] OR rabeprazole[Title/Abstract] OR dexlansoprazole [Title/Abstract]=21,604

2. gastric acid reflux[Title/Abstract] OR gastroesophageal reflux[Title/Abstract] OR gastro oesophageal reflux[Title/Abstract] OR Laryn-gopharyngeal reflux[Title/Abstract] OR Respiratory Aspiration of Gastric Content[Title/Abstract]=18,843

3. randomized controlled trial[Publication Type] OR controlled clinical trial[Publication Type] OR randomized[Publication Type] OR randomized[Title/Abstract] OR placebo[Title/Abstract] OR drug

therapy[MeSH Major Topic] OR randomly[Title/Abstract] OR trial[Title/Abstract] OR groups[Title/Abstract]=2,457,435

4. #1 AND #2 AND #3 Filters: Controlled Clinical Trial; Randomized Controlled Trial; Clinical Trial; Humans=556

COCHRANE LIBRARY

1. gastroesophageal reflux disease: ti,ab,kw OR gastric acid reflux ti,ab,kw OR esophageal reflux: ti,ab,kw OR gastro oesophageal reflux OR Laryngopharyngeal reflux: ti,ab,kw OR Respiratory Aspiration of Gastric Content: ti,ab,kw OR GERD: ti,ab,kw=2,527

2. proton pump inhibitor: ti,ab,kw OR ppi: ti,ab,kw OR omeprazole: ti,ab,kw OR esomeprazole: ti,ab,kw OR lansoprazole: ti,ab,kw OR pantoprazole: ti,ab,kw OR rabeprazole: ti,ab,kw OR dexlansopra-zole: ti,ab,kw=5,454

3. #1 AND #2=948

EMBASE

1. ‘gastroesophageal reflux’:ab,ti OR ‘gastric acid reflux’:ab,ti OR ‘gastro oesophageal reflux’:ab,ti OR ‘Laryngopharyngeal reflux’:ab,-ti OR ‘GERD’:ab,ti=28,355

2. ‘proton pump inhibitor’:ab,ti OR ‘PPI’:ab,ti OR ‘omeprazole’:ab,ti OR ‘esomeprazole’:ab,ti OR ‘lansoprazole’:ab,ti OR ‘pantoprazole’:ab,ti OR ‘rabeprazole’:ab,ti OR ‘dexlansoprazole’:ab,ti=30,771

3. ‘random$’:ab,ti OR ‘factorial’:ab,ti OR ‘crossover’:ab,ti OR ‘placebo’:ab,ti OR ‘doubl’:ab,ti OR ‘blind’:ab,ti OR ‘singl’:ab,ti OR ‘randomized controlled trial’:ab,ti=626,711

4. #1 AND #2 AND #3 AND ([controlled clinical trial]/lim OR [rand-omized controlled trial]/lim) AND [humans]/lim AND [embase]/lim=324