Scientific Report Evidence for PPI Use in Gastroesophageal Reflux

Scientific Report

Evidence for PPI use in Gastroesophageal Reflux Disease, Dyspepsia and Peptic Ulcer Disease (Draft)

December 6, 2006

CADTH

Consultation Document: Canadian Agency for Drugs and Technologies in Health

COMPUS Expert Review Panel on PPIs

Melissa C Brouwers, PhD, MA, BSc Director of the Program in Evidence-Based Care, Cancer Care Ontario Assistant Professor (PT) in the Department of Clinical Epidemiology and Biostatistics, McMaster University

Marilyn Caughlin, MD, BSc Pharm Family Practice Physician, Regina, Saskatchewan

Ron Goeree, BA, MA (Economics) Assistant Professor, Department of Clinical Epidemiology and Biostatistics, McMaster University Acting Director for the Program for Assessment of Technology in Health (PATH) and faculty member, Centre for Evaluation of Medicines (CEM), St. Joseph's Healthcare

Anne Holbrook, MD, PharmD, MSc, FRCPC (CHAIR) Clinical Pharmacologist and Internal Medicine Specialist, Centre for Evaluation of Medicines, St. Joseph's Hospital and Hamilton Health Sciences Corporation Associate Professor and Director, Division of Clinical Pharmacology & Therapeutics, Department of Medicine, McMaster University

Malcolm Man-Son-Hing, MD, MSc, FRCPC Associate Professor of Medicine and Director of Research of the Division of Geriatric Medicine, University of Ottawa Scientist at both the Elizabeth Bruyere Research Institute and Clinical Epidemiology Unit, Ottawa Health Research Institute Staff Geriatrician, Ottawa Hospital

John K Marshall, MD, MSc, FRCPC Active consultant gastroenterologist, Hamilton Health Sciences Associate Professor of Medicine and Head of Clinical Research, Division of Gastroenterology, McMaster University

Pam McLean-Veysey, BSc Pharm Team Leader and Drug Evaluation Pharmacist, QE II Health Sciences Centre, Halifax, Nova Scotia

John A Rideout, BSc, MDCM, CCFP General Practitioner Physician, Burnaby, British Columbia

Brenda G Schuster, BSP, ACPR, PharmD Pharmacy Educator, Regina Qu’Appelle Health Region Pharmacist, RxFiles Academic Detailing program, Regina

Laura Targownik, BSc, MSc, MD, FRCP(C) Assistant Professor of Medicine, Section of Gastroenterology, University of Manitoba

Alan B. R. Thomson, MD, MSC, PhD, FRCP(C), FACP, FACG

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD ii Consultation Document: Canadian Agency for Drugs and Technologies in Health

Professor of Medicine, Faculty of Medicine, University of Alberta President of the Canadian Association of Gastroenterology

Sander Veldhuyzen van Zanten, MD, FRCPC Director, Division of Gastroenterology, Department of Medicine, University of Alberta

External Consultants

COMPUS staff would like to thank the following people for their time, assistance and expert input throughout the project, including guidance on the approach and methods and constructive feedback on drafts of this report.

Catherine Dubé, MD, MSc, FRCPC Ian D. Graham, PhD Division of Gastroenterology, Associate Professor, School of Nursing, The Ottawa Hospital University of Ottawa

Phil Jacobs, BCom, PhD, CMA Professor and Program Director,Health Policy and Management Program, Faculty of Medicine and Dentistry, University of Alberta

Contributors from COMPUS

Fida Ahmad, MSc Cheryl Arratoon, MSc Research Officer Manager, Implementation Support Annie Bai, MD Greg Bak, MLIS, PhD Manager, Research Information Specialist Denis Bélanger, BScPhm, ACPR. Michelle Fiander, MA, MLIS Director, Topics and Research Information Specialist

Aleksandra Grobelna, MA Phil Avtar Lal, MD, M.Phil (Clinical Pharmacology) Library Technician Research Officer Elaine MacPhail, B.ScPharm, MHP Sameer Rajbhandary, PhD Program Advisor, CDR and COMPUS Health Economist Barb Shea, BSP Vijay Shukla, RPh, PhD, ACPR Vice-President, COMPUS Director, CDR and COMPUS Methods Sumeet R. Singh, BScPhm, MSc, RPh Samantha Verbrugghe, B.Sc. Research Officer Research Assistant Barbara Wells, BScPhm Changhua Yu, MD, MSc Director, COMPUS Research Officer

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD iii Consultation Document: Canadian Agency for Drugs and Technologies in Health

Conflict of Interest

Ron Goeree, BA, MA (Economics), has received post-doctoral funding from Janssen-Ortho Inc.

John K Marshall, MD, MSc, FRCPCD, has received payment to support travel costs from Janssen-Ortho Inc. and honoraria from Abbott Laboratories Ltd. and Solvay Pharma Inc. for educational lectures. He received payment for services on an Advisory Board with Solvay Pharma Inc. and is also serving on AstraZeneca Canada Inc. and Janssen-Ortho Inc. Advisory Boards.

Brenda G Schuster, BSP, ACPR, PharmD, has received an honorarium from AstraZeneca Canada Inc. for an educational lecture.

Laura Targownik, BSc, MSc, MD, FRCP(C), has received honoraria from Janssen-Ortho Inc. for educational talks. She received research grants from Janssen-Ortho Inc. and Altana Pharma Inc. She is also a member of the National Steering Committee for ParietTM with Janssen-Ortho Inc.

Alan B. R. Thomson, MD, MSC; PhD, FRCP(C); FACP; FACG, has received honoraria for various educational lectures from Abbott Laboratories Ltd., Altana Pharma Inc., AstraZeneca Canada Inc., GlaxoSmithKline and Janssen-Ortho Inc. Dr. Thomson also received payment from Janssen-Ortho Inc. in return for consultation services. He was a member of Advisory Boards for Janssen-Ortho Inc. and Altana

Sander Veldhuyzen van Zanten, MD FRCPC, has received funding from AstraZeneca Canada Inc. for travel costs to present studies and has received honoraria from AstraZeneca Canada Inc., Abbott Laboratories Ltd. and Janssen-Ortho Inc. for various educational lectures. He was a member of Advisory Boards for AstraZeneca Canada Inc., Janssen-Ortho Inc. and GlaxoSmithKline. He received research grants from AstraZeneca Canada Inc., Abbott Laboratories Ltd., Janssen-Ortho Inc., Altana Pharma Inc./Solvay Pharma Inc. and GlaxoSmithKline.

None of the other COMPUS Expert Review Panel members declared any conflicts of interest.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD iv Consultation Document: Canadian Agency for Drugs and Technologies in Health

FOREWORD

In March 2004, the Canadian Optimal Medication Prescribing and Utilization Service (COMPUS) was launched by the Canadian Coordinating Office for Health Technology Assessment (CCOHTA) (now the Canadian Agency for Drugs and Technologies in Health (CADTH)) as a service to federal, provincial, and territorial jurisdictions, and other stakeholders. COMPUS is a nationally coordinated program, funded by Health Canada.

The goal of COMPUS is to optimize drug-related health outcomes and cost-effective use of drugs by identifying and promoting optimal drug prescribing and use. Where possible, COMPUS builds on existing applicable Canadian and international initiatives and research. COMPUS goals will be achieved through three main approaches: 1. identifying evidence-based optimal therapy in prescribing and use of specific drugs; 2. identifying gaps in clinical practice, then proposing evidence-based interventions to address these gaps, and; 3. supporting the implementation of these interventions.

COMPUS was asked to identify and promote optimal therapy related to proton pump inhibitors, diabetes management, and antihypertensives. The work in this document addresses the use of proton pump inhibitors (PPIs) for the management of gastroesophageal reflux disease (GERD), dyspepsia, peptic ulcer disease (PUD), H. pylori infection and NSAID-associated ulcer. The COMPUS Advisory Committee identified a number of areas of focus including: double-dose therapy, intermittent vs. chronic and step-up vs. step-down therapy, clinical equivalence of PPIs, evidence in dyspepsia, requirement for objective confirmation of GERD, and gastroprotection for NSAID users.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD v Consultation Document: Canadian Agency for Drugs and Technologies in Health

ABBREVIATIONS

AM in the morning AHRQ Agency for Healthcare Research and Quality AMSTAR A MeaSurement Tool to Assess Reviews ASA acetylsalicylic acid BE Barrett’s Esophagus BID twice daily BMT bismuth, metronidazole and tetracycline CanDys Canadian Dyspepsia Working Group CAP capsule CD consensus document CI confidence interval CMA Canadian Medical Association COMPUS Canadian Optimal Medication Prescribing and Utilization Service COX cyclo-oxygenase CPG clinical practice guideline DD double dose DoD Department of Defence (USA) DR CAP delayed-release capsule DR TAB delayed-release tablet DU duodenal ulcer ENRD endoscopy-negative reflux disease ENT TAB enteric-coated tablet FD functional dyspepsia GERD gastroesophageal reflux disease GI gastrointestinal GU gastric ulcer H2RA histamine H2-receptor antagonist Hp Helicobacter pylori H. pylori Helicobacter pylori HRQOL health related quality of life ICSI Institute for Clinical Systems Improvement IMS IMS Health ITT intention-to-treat LA Los Angeles classification system for the endoscopic assessment of esophagitis LD low dose LNF laparoscopic Nissen fundoplication LSM life style modification LU limited use MA meta-analysis MCG microgram MG milligram MOS months NERD non-erosive reflux disease NICE National Institute for Health and Clinical Excellence

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD vi Consultation Document: Canadian Agency for Drugs and Technologies in Health

NNH number needed to harm NNT number needed to treat NSAID non-steroidal anti-inflammatory drug NUD non-ulcer dyspepsia NZGG New Zealand Guidelines Group OGD oesophago-duodenoscopy OPOT Ontario Program for Optimal Therapeutics OR odds ratio OBS observational study PA prokinetic agent PAC PPI plus amoxicillin and clarithromycin pH potential hydrogen – a measure of acidity and alkalinity PM in the evening PMC PPI plus metronidazole and clarithromycin PP per protocol PPI proton pump inhibitor PRN when needed Pts patients PUD peptic ulcer disease QA quality assessment QD once daily QID four times daily Québec CRUM Québec Comité de revue de l’utilisation des médicaments RBC ranitidine bismuth citrate RCT randomized controlled trial RR relative risk RRR relative risk reduction SIGN Scottish Intercollegiate Guidelines Network SR systematic review TAB tablet TID three times daily VHA Veterans Health Administration (US) ZES Zollinger-Ellison syndrome

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD vii Consultation Document: Canadian Agency for Drugs and Technologies in Health

GLOSSARY

Anti-secretory/ acid suppression therapy: drugs that inhibit or reduce acid secretion. Two classes of drugs belong to this category: Histamine-2 Receptor Antagonists (H2RAs) and Proton Pump Inhibitors (PPIs).

Barrett’s epithelium/esophagus: abnormal esophageal epithelium that demonstrates specialized intestinal metaplasia (esophageal columnar epithelium, intestinal metaplasia-positive) on histological examination.

Case control study: a type of observational study in which past exposures to one or more putative risk factors are measured in a group of subjects with a disease or outcome of interest (cases), and in a group without this outcome (controls), in order to ascertain the degree of association between risk factor and outcome.

Clinical practice guideline: a set of systematically developed statements or recommendations designed to assist practitioner and patient decisions about appropriate health care for specific clinical circumstances.

Consensus document: a statement on the advisable course of action in a particular clinical situation developed collectively by a group of experts through either informal or formal consensus methods.

Continuous maintenance therapy: the daily intake of medication for an indefinite period to prevent or minimize symptoms or risk of disease.

COX-2 inhibitors: non-steroidal anti-inflammatory drugs (NSAIDs) that are highly selective inhibitors of the enzyme cyclo-oxygenase (COX)-2 versus COX-1.

Duodenal ulcer: an ulcer in the lining of the most proximal part of the small intestine (duodenum).

Dyspepsia: a symptom complex of epigastric pain or discomfort thought to originate in the upper gastrointestinal tract, in which heartburn and/or acid regurgitation are not the predominant symptoms. It may include any of the following symptoms: excessive burping/belching, bloating, nausea, a feeling of abnormal or slow digestion, early satiety, or heartburn.

Effectiveness: The extent to which a specific intervention, procedure, regimen, or service, produces the intended outcomes when deployed under routine circumstances.

Efficacy: The extent to which a specific intervention, procedure, regimen, or service produces a beneficial outcome under ideal circumstances.

Empirical therapy: treatment based on symptoms without investigation.

Endoscopy-negative reflux disease (ENRD): also referred to as non-erosive GERD or non-erosive reflux disease (NERD), applies to individuals with GERD who have normal endoscopy results while off treatment.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD viii Consultation Document: Canadian Agency for Drugs and Technologies in Health

Erosive esophagitis: the presence of reflux symptoms and any length of mucosal break in the esophagus as a result of gastroesopageal reflux.

Esophagitis: the minimum requisite for diagnosis of esophagitis is the presence of one erosion at the junction between the columnar and squamous epithelium.

Functional dyspepsia: persistent or recurrent dyspepsia in the absence of organic cause (i.e. upon upper GI endoscopy or GI X-Ray) that is likely to explain the symptoms.

Gastric Ulcer: an ulcer in the lining of the stomach.

Gastroesophageal reflux disease (GERD): the reflux of gastric contents into the esophagus, causing symptoms severe enough to affect the quality of life and/or cause esophageal injury.

Gastrointestinal Symptom Rating Scale (GSRS): is one of the most widely used gastrointestinal- specific symptom scales in various settings and clinical trials. It consists of five sub-scales (reflux, diarrhea, constipation, abdominal pain, indigestion), with scores for each sub-scale ranging from 1 (no discomfort) to 7 (very severe discomfort).

Global Overall Symptom scale (GOS): a 7-point Likert scale ranging from 1 = no problem to 7 = a very severe problem used to measure the severity of dyspepsia symptoms in clinical trials.

H2-Blocker/ H2RA: medicines that reduce the amount of acid the stomach produces by blocking histamine2 receptors. H2-blockers available in Canada include cimetidine, famotidine, nizatidine, and ranitidine.

Helicobacter pylori (H. pylori): a spiral-shaped bacterium found in the stomach that causes gastritis and is implicated in peptic ulcer disease, gastric cancer, and MALT lymphoma.

Intention-to-treat analysis: analysis of a clinical trial in which subjects are analyzed according to the groups to which they were initially assigned, regardless of violations of the study protocol (e.g., crossover, poor treatment compliance, use of disallowed treatments, dropout).

Intermittent therapy: the daily intake of medication for a predetermined, finite period (usually 2 to 8 weeks) to resolve reflux related symptoms or heal esophageal lesions following relapse of a patient’s previous symptoms or condition.

Long-term therapy: treatment given over an indefinite period to minimize reflux-related symptoms or esophageal injury. Continuous, intermittent, and on-demand therapies are all forms of long-term therapy.

Maintenance treatment: long-term treatment administered for the primary or secondary prevention of disease.

Meta-analysis: statistical synthesis of a collection of results from individual studies for the purpose of integrating findings and producing a single estimate of effect.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD ix Consultation Document: Canadian Agency for Drugs and Technologies in Health

Mild GERD: characterized by GERD symptoms that are infrequent (fewer than 3 times/week), of low intensity, short duration, and have minimal long-term effect on the patient’s activities of daily living or health-related quality of life.

Misoprostol: a prostaglandin E1 (PGE1) analogue that protects the gastric mucosa. Used to prevent ulcers.

Moderate or severe GERD: characterized by GERD symptoms that are frequent, associated with intense or prolonged symptoms, and have a significant effect on the patient’s daily activities or health-related quality of life.

Non-erosive GERD: also called non-erosive reflux disease (NERD). Defined as the presence of typical reflux symptoms in the absence of esophageal breaks upon upper gastrointestinal endoscopy.

Non-steroidal anti-inflammatory drugs (NSAIDs): drugs with analgesic, antipyretic and anti- inflammatory effects. Act as inhibitors of cyclo-oxygenase-2 (COX-2); with varying degrees of inhibition of COX-1 across individual agents.

Observational study: epidemiological studies in which the investigator measures and determines associations between one or more exposures and an outcome of interest, without intervening in or manipulating the exposures experienced by study subjects.

On-demand therapy: daily intake of a medication for a period sufficient to achieve resolution of dyspepsia or GERD symptoms. Following symptom resolution, medication is discontinued until symptoms recur, at which point medication intake is resumed until symptoms resolve once again.

PAC triple therapy: H. pylori eradication triple therapy regimen that usually includes a twice-daily course of standard-dose PPI, amoxicillin 1 g and clarithromycin 500 mg.

PMC triple therapy: H. pylori eradication triple therapy regimen that usually includes a twice-daily course of standard-dose PPI, metronidazole 500 mg and clarithromycin 250 mg-500 mg.

Per-protocol analysis: an analysis of a clinical trial from which subjects with major violations of the study protocol are omitted.

Psychological General Well-Being Index (PGWB): a generic quality-of-life instrument consisting of 22 items scored on a 6-point Likert scale with verbal descriptors. Has been used extensively in studies of GERD and upper gastrointestinal diseases and has been shown to have excellent reliability and validity. It primarily focuses on psychological or emotional domains of health-related quality of life (QOL).

Randomized controlled trial: a prospective study designed to test the effectiveness of an intervention in which the investigator randomly allocates subjects to one or more treatment groups and a control group.

Reflux esophagitis: inflammation of the esophageal mucosa resulting from exposure to acidic gastric contents.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD x Consultation Document: Canadian Agency for Drugs and Technologies in Health

36-Item Short-Form General Health Survey (SF-36): a generic quality of life instrument consisting of 36 items. It focuses on physical function, role limitations-physical, bodily pain, vitality, general health perceptions, social functioning, role limitations - emotional, and mental health. SF-36 has been widely used in clinical trials and studies in primary care and chronic disease populations, including GERD and other gastrointestinal diseases.

Standard-doses of PPIs1 Standard-dose Low Dose Double Dose Omeprazole 20 mg once daily 10 mg once daily 40 mg once daily Lansoprazole 30 mg once daily 15 mg once daily 30 mg twice daily Pantoprazole 40 mg once daily 20 mg once daily 40 mg twice daily Rabeprazole 20 mg once daily 10 mg once daily 20 mg twice daily Esomeprazole 20 mg once daily N/A 40 mg once daily

Step-down therapy: the initial use of potent acid suppression, followed by decreased dose or the use of less potent agents to tailor therapy according to individual response.

Step-up therapy: the initial use of less potent agents or lower doses of acid suppressive therapy, followed by increased doses or more potent agents if there is an inadequate response to treatment.

Systematic review: a summary of the medical literature that uses explicit methods to identify, select, appraise, and analyze studies relevant to a particular clinical question.

Triple therapy: a combination of three drugs: two antibiotics and an acid suppressor (H2RA or PPI) or bismuth salt for H. pylori eradication.

Ulcer bleeding: acute or chronic ulcers that erode through a blood vessel, causing clinical evidence of bleeding.

Ulcer relapse/recurrence: re-ulceration after initial healing.

Uncomplicated dyspepsia: dyspepsia that is not accompanied by alarm features or associated with NSAID usage.

Uninvestigated dyspepsia: dyspepsia in a patient who has not undergone radiologic imaging or endoscopy but who may have undergone non-invasive testing for H. pylori infection. Note: Heartburn- dominant uninvestigated dyspepsia is included in the definition of uninvestigated GERD.

Uninvestigated GERD: dominant symptoms of heartburn and /or which may be associated with other symptoms such as epigastric pain/discomfort and not investigated by endoscopy (or upper GI series). Note: Heartburn dominant uninvestigated dyspepsia is included in the definition of uninvestigated GERD.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD xi Consultation Document: Canadian Agency for Drugs and Technologies in Health

Table of Contents

1 THE ISSUE ...... 1 2 OBJECTIVE ...... 1 3 BACKGROUND ...... 1 3.1 Proton Pump Inhibitors ...... 1 3.2 Conditions Related to the Use of Proton Pump Inhibitors ...... 11 4 METHODS FOR CLINICAL EVALUATION ...... 13 4.1 Literature Search Strategy for CPGs and CDs...... 13 4.2 Selection Criteria and Method...... 13 4.3 Selection and Synthesis of Statements from CPGs and CDs ...... 14 4.4 Identification, Evaluation and Presentation of Evidence Cited in CPGs and CDs ...... 14 4.5 Quality Assessment of the Evidence...... 15 4.6 Data Extraction...... 15 4.7 Stakeholder Feedback ...... 15 4.8 Updating Of Evidence ...... 15 4.9 Expert Review Panel...... 16 5. RESULTS ...... 18 Legend...... 18 5.1. GERD...... 18 5.2 Dyspepsia ...... 82 5.3 PUD ...... 98 Appendix 1: Indications for Proton Pump Inhibitors in Canada ...... 134 Appendix 2: Literature Search Strategies...... 138 Appendix 3: AMSTAR Instrument for Systematic Reviews ...... 147 Appendix 4a: Adapted SIGN 50 Checklist for Randomized Controlled Trials149...... 148 Appendix 4b: Adapted SIGN 50 Checklist for Cohort Studies150 ...... 149 Appendix 4c: Adapted SIGN 50 Checklist for Case Control Studies151 ...... 151 Appendix 5: Sample Voting Form...... 153 Appendix 6a: Quality Assessment Summary - Systematic Reviews - GERD ...... 154 Appendix 6b: Quality Assessment Summary - Systematic Reviews - Dyspepsia ...... 155

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD xii Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 6c: Quality Assessment Summary - Systematic Reviews - PUD, H. pylori Infection and NSAID Associated Ulcer ...... 156 Appendix 7a: Quality Assessment Summary – RCTs - GERD ...... 157 Appendix 7b: Quality Assessment Summary – RCTs - Dyspepsia ...... 163 Appendix 7c: Quality Assessment Summary – RCTs - PUD, H .pylori Infection and NSAID Associated Ulcer...... 164

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD xiii Consultation Document: Canadian Agency for Drugs and Technologies in Health

1 THE ISSUE

The Canadian Optimal Medication Prescribing and Utilization Service (COMPUS) was tasked by the federal, provincial and territorial ministries of health to identify and promote the implementation of evidence-based and cost-effective optimal therapy in the prescribing and use of proton pump inhibitors. Since their introduction to the market, the use of this class of drugs continues to grow in Canada. Analysis of IMS data revealed that the estimated total number of prescriptions for proton pump inhibitors (PPIs) dispensed by Canadian retail pharmacies increased by 15% from 10.8 million prescriptions in 2003 to 12.4 million prescriptions in 2004 with a 10% increase in total drug expenditure. (Source: IMS Health Canada, Montreal: personal communication, January 24, 2005). In view of the widespread and growing use of PPIs, healthcare providers, consumers and policy makers require evidence-based information that facilitates optimal therapies in the use of these agents.

2 OBJECTIVE

The objective of this report is to identify evidence-based clinical statements that could be used to support optimal therapy in the prescribing and use of proton pump inhibitors in the management of: 1. GERD; 2. dyspepsia; 3. PUD; 4. H .pylori infection; 5. NSAID associated ulcer.

3 BACKGROUND

3.1 Proton Pump Inhibitors

Proton pump inhibitors are compounds that suppress gastric acid secretion. They are approved for use in the treatment of conditions where the control of gastric acid is needed such as: gastric ulcers, reflux esophagitis, gastroesophageal reflux disease, and eradication of Helicobacter pylori.2 PPIs work by irreversibly inhibiting a gastric enzyme (H+/K+- ATPase). This ‘proton pump’ secretes acid into the lumen of the stomach.

There are five PPI agents currently available on the Canadian market (Table 1). Omeprazole (Losec®) was the first PPI introduced in Canada in 1989 followed by lansoprazole (Prevacid®) in 1995, pantoprazole (Pantoloc®) in 1996, and both rabeprazole (Pariet®) and esomeprazole (Nexium®) in 2001. Apo-Omeprazole, approved for marketing in 2004, is currently the only generic PPI available in Canada.3 The other PPIs are still under patent protection.

The approved indications in Canada for these drugs are described in Appendix 1. Costs related to PPI use for dyspepsia, PUD, H. pylori infection and NSAID associated ulcer are listed in Table 1. Costs related to the use of H2-receptor antagonists, the other major class of acid suppressing agents, for the same conditions are presented in Table 2. The formulary listing status of the available PPIs in each province and territory, as well as for Non-insured Health Benefit (NIHB) recipients, is described in Table 3.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 1 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Table 1: Comparable Adult Doses and Daily Costsa of Proton Pump Inhibitors for Dyspepsia, PUD, H. pylori Infection and NSAID associated Ulcer

Drug OMEPRAZOLE & OMEPRAZOLE LANSOPRAZOLE PANTOPRAZOLE PANTOPRAZOLE ESOMEPRAZOLE RABEPRAZOLE LAN/CLAR/AMOX OMEPRAZOLE Mg (ApoOmeprazole (Prevacid--Abbott) (Pantoloc--Solvay Mg (Nexium-- (Pariet-- Janssen- # (Losec, Losec --Apotex) Pharma) (Pantoloc M— AstraZeneca) Ortho) (Hp-PAC--Abbott) MUPS ALTANA Pharma) --AstraZeneca) Format & Strength & per cap; tab; DR Price ($) tab 20 mg: 15 mg: 2.000/cap 20 mg: 1.7000/tab 40 mg: Price N/A† 20 mg: 2.1000/tab 10 mg: 0.6500/tab Per 7 day pack: 1.2500/cap 30 mg: 2.000/cap 40 mg: 1.9000/tab 40 mg: 2.1000/tab 20 mg: 2.7400/tab 78.2400 10 mg: 1.7500 20 mg: 2.2000 40 mg: 3.0800/cap 10 – 20 mg daily (tab Dyspepsia & MUPS)

Cost ($) 1.7500-2.2000/day ------Duodenal Ulcer Active 20 – 40 mg daily 20 – 40 mg daily 15 mg daily 40 mg daily 40 mg daily 20 mg daily Cost ($) 2.2000-4.4000§/d 1.2500-2.5000/day 2.0000/day 1.9000/day N/A -- 1.3000/dayb -- Duodenal Ulcer 10 – 40 mg daily 15 mg daily

Maintenance § Cost ($) 1.7500-4.4000 /d -- 2.0000/day ------Gastric Ulcer Active 20 – 40 mg daily 20 – 40 mg daily 15 mg daily 40 mg daily 40 mg daily 20 mg daily Cost ($) 2.2000-4.4000§/d 1.2500-2.5000/day 2.0000/day 1.9000/day N/A -- 1.3000/dayb -- Gastric Ulcer Maintenance 20 – 40 mg daily 20 – 40 mg daily Cost ($) 2.2000-4.4000§/d 1.2500-2.5000/day ------NSAID Associated 20 mg daily 20 mg daily Duodenal Ulcers Active Cost ($) 2.2000/day 1.2500/day ------NSAID Associated 20 mg daily 20 mg daily Duodenal Ulcers Maintenance 2.2000/day 1.2500/day ------Cost ($) NSAID Associated Gastric 20 mg daily 20 mg daily 15 – 30 mg daily 20 mg daily Ulcers Active Cost ($) 2.2000/day 1.2500/day 2.000/day -- -- 2.1000/day -- -- NSAID Associated Gastric 20 mg daily 20 mg daily Ulcers Maintenance Cost ($) 2.2000/day 1.2500/day ------Risk Reduction of NSAID- 15 mg daily 20 mg daily 20 mg daily associated Gastric Ulcer Cost ($) -- -- 2.0000/day 1.7000/day -- 2.1000/day -- -- Prevention of NSAID- 20 mg daily induced GI lesions ------1.7000/day ------

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 2 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Eradication Therapy (DU) 20 mg BID±± 30 mg BIDß 40 mg BIDº 40 mg BIDº 20 mg BID** 20 mg BID** Lan/Clar/Amox#BID Cost ($) 4.4000/d -- 4.0000/day 3.8000/day N/A 4.2000/day 2.6000/dayb 11.1771/day To ensure healing after 20 mg daily eradication therapy (DU) Cost ($) 2.2000/day ------Eradication Therapy (GU) 20 mg BID±±

Cost ($) 4.4000/day ------To ensure healing after 20 – 40 mg daily eradication therapy (GU) § Cost ($) 2.2000-4.4000 /d ------Source of Prescribing Losec caps PM4 Information [Product Losec DR tabs Apo-Omeprazole Prevacid PM8 Pantoloc PM9 Pantoloc M PM Nexium PM10 Pariet PM11 Hp-PAC PM12 Monograph (PM)] PM5 PM7 Losec MUPS PM6

Shaded cells: Not an approved indication

a Manufacturer’s list prices used. Dispensing fees and wholesaler mark-up are not included. b Cost for Pariet is based on using the 10 mg tablets as it is the most economical form of Pariet. § Cost for Losec 40 mg is expressed as $4.4000 (2 X 2.2000) as the Losec 40 mg capsule is not commonly used in Canada. † The price of PANTOLOC M has not yet been released publicly. ALTANA Pharma is currently negotiating with various payers and governments around a suitable price for this product. ‡ More than 90% of patients are controlled with doses of 20 to 120 mg daily at a cost of $2.2000-13.2000/d for Losec and $1.2500-7.5000/day for Apo-Omeprazole. # Lan/Clar/Amox—components of Hp-PAC include lansoprazole 30 mg delayed release capsules; clarithromycin 500 mg tablets; and amoxicillin 500 mg capsules ß with amoxicillin 1000 mg and clarithromycin 500 mg BID x 7d, 10d or 14d ** with amoxicillin 1000 mg and clarithromycin 500 mg BID x 7d º with amoxicillin 1000 mg and clarithromycin 500 mg BID OR metronidazole 500 mg and clarithromycin 500 mg BID x 7d ±± with amoxicillin 1000 mg and clarithromycin 500 mg BID (Losec 1-2-3 A) OR metronidazole 500 mg and clarithromycin 250 mg BID (Losec 1-2-3 M) x 7d

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 3 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Table 2: Comparable Adult Doses and Daily Costsa of H2-Receptor Antagonists for PUD, H. pylori Infection and NSAID associated Ulcer

Drug RANITIDINE (Zantac – RANITIDINE (Novo-ranidine – FAMOTIDINE (Pepcid – FAMOTIDINE (Novo-famotidine NIZATIDINE (Novo- Glaxo SmithKline) Novopharm, other generics) Merck Frosst) -- Novopharm, other generics) nizatidine – Novopharm, other generics) Format & Strength & Price ($) per tab per tab per tab per tab per cap 150 mg: 1.145 150 mg: 0.4042 20 mg: 1.002 20 mg: 0.5896 150 mg: 0.5287 300 mg: 2.155 300 mg: 0.7787 40 mg: 1.822 40 mg: 1.0612 300 mg: 0.9580

150 mg twice daily or 300 mg 150 mg twice daily or 300 mg once 150 mg twice daily or 300 mg 40 mg once daily 40 mg once daily Duodenal Ulcer Active once or twice daily or twice daily once daily Cost ($) 2.155 – 4.310/day 0.7787 – 1.5574/day 1.822/day 1.0612/day 0.9580 – 1.0574/day Duodenal Ulcer Maintenance 150 mg or 300 mg once daily 150 mg or 300 mg once daily 20 mg once daily 20 mg once daily 150 mg once daily Cost ($) 1.145 – 2.155/day 0.4042 – 0.7787/day 1.002/day 0.5896/day 0.5287/day Gastric Ulcer Active 150 mg twice daily or 300 mg 150 mg twice daily or 300 mg once 150 mg twice daily or 300 mg 40 mg once daily 40 mg once daily Cost ($) once daily daily once daily 2.155 – 2.290/day 0.7787 – 0.8084/day 1.822/day 1.0612/day 0.9580 – 1.0574/day Gastric Ulcer Maintenance 150 mg once daily 150 mg once daily Cost ($) 1.145/day 0.4042/day ------Treatment of NSAID induced 150 mg twice daily 150 mg twice daily lesions and prevention of recurrence 2.290/day 0.8084/day ------Cost ($) Source of Prescribing Compendium of Compendium of Information [Product Pharmaceuticals and Novo-ranidine PM13 Pharmaceuticals and Novo-famotidine PM14 Novo-nizatidine PM15 Monograph (PM)] Specialties 20062 Specialties 20062

Shaded cells: Not an approved indication a Manufacturer’s list prices used. Dispensing fees and wholesaler mark-up are not included.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 4 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Table 3: Public Drug Plan Benefit Listings for Proton Pump Inhibitors – May 2006

Drug Strength/Dosage BC AB SK MB ON QC NB NS PE NL YK NW NIHB/ form Nunavut Omeprazole 10 mg RESa RESd RESe,f RESi RESm RESn RESo RESp RESq RESq (Losec) 20 mg RESc RESf,g RESh RESj RESk RESm RESn RESo RESp RESq RESq

Omeprazole 20 mg RESf,g RESh RESj RESl RESm RESn RESo RESp (Apotex)

Rabeprazole 10 mg RESb RESf,g RESh RESj RESk RESm RESn RESo RESp (Pariet) 20 mg

Pantoprazole 20 mg RESm RESo RESp (Pantoloc) 40 mg RESc RESf,g RESh RESj RESk RESm RESn RESo RESp RESq RESq

Lansoprazole 15 mg RESc RESf,g RESh RESj RESk RESm RESn RESo RESp RESq RESq (Prevacid) 30 mg RESc RESf,g RESh RESj RESk RESm RESn RESo RESp RESq RESq

Esomeprazole 20 mg RESf,g (Nexium) 40 mg RESf,g

Type of listing decisions:

Full Benefit = Not a Benefit =

RES – restricted benefit with specified criteria (e.g., special authorization, exception drug status, limited use benefit)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 5 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Table 3: continued

British Columbia RESa – GERD maintenance after failure of rabeprazole. GIs exempted.

RESb – GERD or non-bleeding ulcer after failure of H2RA; Barrett’s esophagus, Zollinger-Ellison syndrome, connective tissue disease, Hp triple therapy. GIs exempted.

RESc – GERD or non-bleeding ulcer after failure of H2RA and rabeprazole; Barrett’s esophagus, Zollinger-Ellison syndrome or connective tissue disease after failure of rabeprazole; Hp triple therapy. GIs exempted.

Alberta RESd – For the treatment of patients who are unable to tolerate Losec 20 mg tablets. Special authorization may be granted for 12 months.

Saskatchewan RESe – GERD; maintenance of healed reflux esophagitis; severe erosive esophagitis and Zollinger- Ellison syndrome. f - Reimbursement for PPIs is limited to a maximum of $1.51 per tablet or capsule (subject to the patient's usual co-payment and deductible).

RESg - PUD unresponsive to H2RAs, misoprostol or sucralfate; GERD; severe erosive esophagitis and Zollinger-Ellison syndrome; one-week Hp eradiaction in those with PUD; prevention of gastroduodenal hemorrhage with history of gastrointestinal bleeds where NSAID, glucocorticosteroid or anticoagulant must be continued; after discharge from hospital for gastrointestinal bleeding.

Manitoba RESh – Patients must have tried and failed a course of Pariet 10 mg and omeprazole 20 mg before they will be considered for benefit coverage with other PPI medications. 1. Treatment of GERD symptoms: Note: Patients with non-erosive GERD could potentially be reduced to step-down therapy with an H2RA depending on symptom resolution. Patients may also be trialed by dosing on alternate days or by using a lower dose PPI. 2. For gastro-protection of NSAIDs: For use in the prevention of NSAID induced ulcers in patients who continue on a non-selective NSAID therapy and who have any one of the following risk factors: a. History of peptic or duodenal ulcer b. Age>65 years c. Concomitant warfarin use d. Concomitant corticosteroid use 3. PUD treatment (approved for eight weeks of therapy) 4. H. pylori eradication (approved for seven to14 days treatment course)[except for Apo-omeprazole 20 mg] 5. Zollinger-Ellison syndrome and Barrett’s esophagus (approved for a three-year treatment course)

RESi – Coverage as above but only for pediatric patients unable to take 20 mg dosage.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 6 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 7 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Table 3: continued

Ontario RESj – 1. Gastroesophageal Reflux Disease (GERD): For the treatment of erosive GERD or upper GI malignancy; OR For the treatment of non-erosive GERD after failure of H2-receptor antagonist therapy. Patients with GERD should be reassessed within six months after initial treatment with a PPI. The reassessment could include confirmation of need for PPI with endoscopy, a trial of PPI withdrawal, or step-down therapy to H2-receptor antagonist therapy. Note: There is a lack of published evidence to support double-dose PPI therapy in this setting. LU Authorization Period: one-year. 2. H. pylori-positive Peptic Ulcers: For the treatment of H. pylori-positive peptic ulcers where H. pylori is documented, by serology, urea breath test or endoscopy, for a one-week course in combination with antimicrobial therapy. Retreatment of H. pylori-positive peptic ulcers must be documented by persistent H. pylori infection on urea breath test or endoscopy. Maximum duration: 7 days (for retreatment, a four-week period must elapse since the end of the previous treatment). LU Authorization Period: one-year. 3. Confirmed Peptic Ulcers or NSAID-induced Ulcer Prophylaxis: For the treatment of confirmed peptic ulcers and NSAID-induced ulcers; OR For the prophylaxis of NSAID-induced ulcers for patients at increased risk of GI bleeding. Note: There is a lack of published evidence to support double-dose PPI therapy in this setting. LU Authorization Period: one-year. 4. Other Gastrointestinal Disorders: For the treatment of gastroduodenal Crohns disease, short-gut syndrome, scleroderma, or pancreatitis. Note: There is a lack of published evidence to support double-dose PPI therapy in these settings. LU Authorization Period: one-year. 5. Severe Conditions: For the treatment of severe esophagitis, Zollinger-Ellison syndrome, esophageal stricture, persistent symptoms of GERD or persistent erosive esophagitis, or upon hospital discharge following a gastrointestinal bleed. For patients receiving double-dose therapy, the need to continue treatment at higher doses should be reassessed after eight weeks. For retreatment at higher doses, a four-week period should have elapsed from the end of the previous treatment. Reassessment could include a procedural assessment of the condition or step-down therapy to lower-dose proton pump inhibitor (PPI) therapy. LU Authorization Period: one-year.

New Brunswick RESk - GERD with endoscopic evidence of esophagitis or Barrett's esophagus; Severe GERD after H2RA therapy; Severe GERD after step-down failure; Zollinger-Ellison Syndrome; Hp treatment or retreatment; Non-Hp PUD which is unresponsive to H2RAs; GIs exempted.

RESl - GERD with endoscopic evidence of esophagitis or Barrett's esophagus; Severe GERD after H2RA therapy; Severe GERD after step-down failure; Zollinger-Ellison Syndrome; Non-Hp PUD which is unresponsive to H2RAs; GIs exempted.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 8 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Table 3: continued

Nova Scotia RESm – Patients must have failed initial therapy with Apo-omeprazole or Pariet 10 mg before coverage for another PPI can be considered. 1. Moderate Symptoms of GERD and/or Mild to Moderate Esophagitis (refractory to lifestyle adjustments, antacids and eight weeks of prescription strength H2RA), severe symptoms of GERD, severe esophagitis, Barrett’s esophagus, other reflux-associated indications on the basis of a specialist recommendation. 2. Complicated PUD, NSAID induced ulcers, gastric or duodenal ulcers (H. pylori-positive), uncomplicated ulcer (after failure of eight to12 weeks of H2RA therapy), Zollinger-Ellison Syndrome.

Prince Edward Island RESn - Gastric or Duodenal Ulcer resistant to H2RAs and confirmed by endoscopy or radiology. Severe GERD, or GERD after H2RA failure, must be confirmed. ZE resistant to H2RAs and confirmed. Hp eradication in confirmed Gastric or Duodenal Ulcer cases who have failed with H2RAs and have not used NSAIDs.

Newfoundland RESo – Patients must have tried and failed a course of Pariet 10 mg and omeprazole 20 mg before they will be considered for benefit coverage with other PPI medications. 1. For the treatment of GERD, confirmed by UGI investigation (radiography or endoscopy), following failure of an adequate trial (four to eight weeks minimum, at therapeutic dosing) of a H2 antagonist. Approval in this case will be for a one-year period and is subject to an annual review with request for renewal needed if the medication is to be continued. 2. For the treatment of GERD, based on clinical diagnosis and in the absence of UGI confirmation, for an eight week period only, and following failure of an adequate trial (four to eight weeks minimum, at therapeutic dosing) of a H2 antagonist. 3. For H. Pylori eradication, in conjunction with clarithromycin and metronidazole/amoxicillin, at BID dosing for a seven day period. Requests for a second eradication treatment will be honored however additional treatments will require diagnostic confirmation of the continued presence of H. pylori. 4. For cytoprotection against NSAID-induced gastropathy in high risk patients (defined as patients requiring chronic nsaid treatment AND who are over 60 years of age, have a previous history of peptic ulcer disease, or have concomitant oral steroid use) who have failed or been intolerant to a trial of misoprostol. 5. For the treatment of active gastritis, duodenal or gastric ulceration, confirmed by UGI investigation (radiography or endoscopy), following failure of a H2 antagonist. Approval will be for four weeks for a duodenal ulcer, and 12 weeks for gastritis or gastric ulceration.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 9 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Table 3: continued

Yukon RESp – Chronic Disease Program – For treatment of PUD or chronic GERD secondary to an approved chronic disease, and where H2RAs are contraindicated or ineffective. For treatment of patients with high risk when NSAID therapy cannot be avoided for approved chronic disease. Pharmacare and Extended Benefits - For treatment of PUD or chronic GERD, and where H2RAs are contraindicated or ineffective. For treatment of H. pylori. For treatment of patients with high risk when NSAID therapy cannot be avoided.

Northwest Territories, NIHB and Nunavut RESq – As part of a multi-drug regimen for H. pylori eradication Coverage also provided if the following prerequisites are met: patient has tried at least 60 days of Apo- omeprazole and patient has tried at least 60 days of rabeprazole sodium. Total trial of 120 days will be confirmed against medication history:

Hp therapy with PUD; confirmed gastric or duodenal ulcers; mild to moderate GERD after failure or not tolerate four weeks on H2RAs; severe GERD; NSAID ulcers where NSAID must be continued; prevention of NSAID ulcers in high-risk patients; Zollinger-Ellison Syndrome; Barrett's Esophagus; esophagitis associated with connective tissue disease; others on individual basis.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 10 Consultation Document: Canadian Agency for Drugs and Technologies in Health

3.2 Conditions Related to the Use of Proton Pump Inhibitors

3.2.1 Gastroesophageal reflux disease (GERD) Gastroesophageal reflux is the movement of gastric contents from the stomach back into the esophagus. The most common symptoms of patients with GERD are heartburn and regurgitation, although many patients present with multiple symptoms.16 Endoscopic examination of the esophagus may or may not show erosive esophagitis (i.e., erosion of the esophagus by acid).17,18

In Canada, GERD is the most prevalent acid-related disease, and a significant contributor to impairment of health-related quality of life.17 A population-based study showed that up to 10% of Canadians have isolated heartburn.19

Although not very common, complications associated with GERD may include deep ulcer, strictures, hemorrhage, anemia, and Barrett’s esophagus. 1,20 Observational studies have also demonstrated that GERD is a risk factor for esophageal adenocarcinoma.21-23 However, mortality associated with GERD is very low (1/100,000).24

3.2.2 Reflux Esophagitis Reflux esophagitis is one of the complications of GERD and is characterized by inflammation and ulceration of the esophagus as seen upon endoscopic examination. It is caused by reflux of acidic gastric contents into the esophagus. Heartburn is a characteristic symptom of reflux esophagitis and may be associated with regurgitation or a feeling of warm fluid climbing up the throat.25 Approximately 30% of GERD patients in primary care and general practice showed endoscopic evidence of esophagitis in one study.17 A recent Canadian study observed that 43% of uninvestigated dyspepsia patients referred for endoscopy from family practice centres had erosive esophagitis16.

3.2.3 Barrett’s Esophagus In Barrett’s esophagus, the stratified squamous epithelium that normally lines the distal esophagus is replaced by an abnormal columnar epithelium that has intestinal features. This process is called intestinal metaplasia and it is one of the complications of GERD. It is usually detected during endoscopic examination of the esophagus when symptoms of GERD are being evaluated. This condition does not cause noticeable symptoms, however, a very small number of patients with Barrett’s esophagus develop esophageal adenocarcinoma.26

3.2.4 Dyspepsia Dyspepsia describes a heterogeneous group of symptoms (such as pain or discomfort centred in the upper abdomen) with many underlying causes (e.g. peptic ulcer disease, GERD, gastritis, gastric cancer, drug-induced dyspepsia, etc.). Dyspepsia is a common condition in Canada (prevalence of 29%)19 that significantly diminishes the quality of life of patients. An estimated 7% of the average Canadian family physician’s practice is devoted to the management of dyspepsia.27 Although the Rome II definition of dyspepsia does not include reflux disease, the Canadian Dyspepsia Working Group considers that reflux disease is an integral part of uninvestigated dyspepsia.28 In this report, only symptoms that are not predominated by heartburn are considered to be dyspepsia. Heartburn-dominant symptoms are considered under GERD.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 11 Consultation Document: Canadian Agency for Drugs and Technologies in Health

3.2.5 Peptic Ulcer Disease (PUD) Peptic ulcer disease is a condition of the upper gastrointestinal tract characterized by erosions and/or ulcerations of the gastric or duodenal walls.29 The mucosal ulceration seen in PUD is due to alterations in the normal defense mechanisms of the mucosa, and acid hypersecretion in some cases. The lifetime incidence of PUD is estimated at 10% in men and 4% in women.30 PUD may cause abdominal pain and other gastrointestinal symptoms such as heartburn, nausea, vomiting, belching and bloating. However, some patients with PUD are asymptomatic. Upper gastrointestinal bleeding, perforation, and obstruction are serious, life-threatening complications that occur in a small percentage of PUD sufferers.29 The two major etiological factors in PUD are infection with H. pylori and the use of non- steroidal anti-inflammatory drugs (NSAIDs), both of which produce mucosal damage through various mechanisms.29

A small fraction of ulcers are associated with neither H. pylori infection nor NSAID use. Some of these ulcers are due to other medical conditions such as Zollinger-Ellison syndrome or Crohn’s disease, or the ingestion of drugs other than NSAIDs.31

3.2.6 NSAID-associated Ulcers NSAID use is a well-established risk factor for the development of ulcers. The prevalence of endoscopically-proven gastric ulcers is 12% to 30% in individuals with arthritis taking NSAIDs over several months, while that of duodenal ulcers is 2% to19%. In comparison, the prevalence estimates of gastric and duodenal ulcers in the general population are 0.3% and 1.4%, respectively.29 NSAID users are also at higher risk for bleeding ulcers and other complications.31 Furthermore, the risk of ulceration in NSAID-users increases with age in comparison to NSAID non-users from 1.5 fold greater risk in those age 60 or less, to five times greater risk in those older than age 60.32 Other risk factors for NSAID-associated ulcers are: concomitant comorbidity, past history of PUD, concomitant use of steroids, combined NSAID use and the type and dose of NSAID.

3.2.7 H. pylori Infection H. pylori is a bacterium that infects the gastric mucosa and is associated with chronic gastritis, peptic ulcer, and gastric cancer. The prevalence of the infection is 20% to 40% in Canada. Although the majority of patients infected with the organism do not develop clinically apparent disease, the risk of peptic ulcer in those infected is about twice that of uninfected individuals.33 More than 90% of patients with duodenal ulcers, and 70% to 84% of those with gastric ulcer, are infected with the organism. Successful eradication of the infection leads to healing and a reduction in ulcer recurrence rates to less than 5% per year.30 H. pylori may also have a role in other upper gastrointestinal diseases such as non- ulcer dyspepsia.31

3.2.8 Zollinger-Ellison syndrome (ZES) ZES is a syndrome of single or multiple gastrointestinal ulcerations that result from benign or malignant islet cell tumours of the pancreas (gastrinomas) that secrete high levels of gastrin. The symptoms of ZES include signs of peptic ulcers: burning pain in the abdomen, diarrhea, nausea, vomiting, fatigue, weakness, weight loss, and bleeding. It is a rare disease, with an estimated prevalence of between 0.1 to 3 per million in the US.34

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 12 Consultation Document: Canadian Agency for Drugs and Technologies in Health

4 METHODS FOR CLINICAL EVALUATION

To promote optimal prescribing and use of PPIs for indications approved in Canada, COMPUS gathered evidence-based statements and recommendations from clinical practice guidelines (CPGs) and consensus documents (CDs). We evaluated the quality of the evidence supporting these statements and synthesized the information to summarize the results. An Expert Review Panel (ERP) was then convened by COMPUS to identify Evidence-Based Statements on the use of PPIs, and validate the available evidence underlying these Statements.

4.1 Literature Search Strategy for CPGs and CDs

Recently published, up-to-date CPGs and CDs related to the use and prescribing of PPIs for approved indications were obtained from published and unpublished sources and from consultation with stakeholders.

The following databases on the DIALOG® system were searched using a combination of controlled vocabulary descriptors and keywords: MEDLINE®, BIOSIS Previews®, EMBASE® and PASCAL (Appendix 2). Drug registry numbers were also used but were excluded from EMBASE® to avoid large numbers of false hits. Parallel searches were run on PubMed, the Cochrane Library and CINAHL (Ovid). Searches were not limited by date or language.

Internet-based guideline repositories were searched as well as the web sites of gastroenterology associations and organizations concerned with the creation and regulation of health information. A more general search for source documents was performed using Google™ and Yahoo! ®

4.2 Selection Criteria and Method

4.2.1 Selection Criteria

The most recent versions of CPGs and CDs were selected for evaluation if they were prepared by professional bodies in Canada, the United States, Australia, New Zealand, the UK, or Western Europe and if they contained recommendations on using pharmacotherapy for the management of GERD, esophagitis, Barrett’s esophagus, dyspepsia, PUD, NSAID-associated ulcer, H. pylori infection, or ZES in adults or children.

4.2.2 Selection Method

Two independent reviewers examined the titles and abstracts of preliminary search results to identify citations that conformed to the selection criteria. They obtained the full text of those that did and sorted them into three groups, covering the eight indications approved in Canada. One group, referred to as GERD, was used to categorize information relating to GERD, endoscopic-negative reflux disease (ENRD), erosive esophagitis and Barrett's esophagitis; another, PUD, for PUD, NSAID-associated ulcer, and H. pylori infection; and a third, dyspepsia, formed a category of its own. Once categorized, the documents were examined for statements and recommendations related to the use of PPIs for these indications in adults or children. Consensus between reviewers was necessary at each stage and a third reviewer was consulted if there was disagreement.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 13 Consultation Document: Canadian Agency for Drugs and Technologies in Health

A list of documents that met the selection criteria and contained relevant statements and recommendations was posted on the COMPUS web site on July 27, 2005 so that stakeholders could supply references to documents that may have been missed during the search procedure. Any references received by August 12, 2005 that were not already in the list of selected documents and that met the inclusion criteria were added to the final list.

4.3 Selection and Synthesis of Statements from CPGs and CDs

Two reviewers identified and extracted PPI-related recommendations and statements from the selected CPGs and CDs, and categorized them as being related to either GERD, dyspepsia or PUD. Related items were compiled into a single PPI-related Statement (PRS) that represented a synopsis of the existing statements and recommendations. These were reviewed by a gastroenterologist before the process of evidence identification and appraisal was initiated.

4.4 Identification, Evaluation and Presentation of Evidence Cited in CPGs and CDs

References cited in support of the statements and recommendations in the original source documents were identified and selected for further evaluation through the procedures described below. Information relating the selected evidence to the PRSs was presented in tables.

4.4.1 Identification Procedure

Two reviewers independently identified evidence cited in support of the statements or recommendations from CPGs/CDs that comprised each PRS. The full text of cited references was obtained and classified by study design into separate groups containing SRs, RCTs, observational studies, and all other types of evidence. Two reviewers checked and agreed on the relevance of the evidence to the PRS for which it was cited. Documents judged to be irrelevant (e.g., references clearly cited in error) were omitted from further evaluation.

4.4.2 Selection Procedure

When more than one type of evidence existed for a given PRS, sources with the highest level of evidence were selected for further evaluation, i.e., assessment of study quality and data extraction. Selection was based on a hierarchy of study design in which SRs were considered the highest level of evidence, followed by RCTs, observational studies, and information obtained from narrative reviews and expert opinion. For each PRS, selection of supporting documents for further evaluation was decided using the following rules:

• All SRs were evaluated. • RCTs were evaluated if: • No good quality SRs pertaining to the PRS existed, or; • One or more good quality SRs did exist but a given RCT was not included in these, and if the main results of the RCT differed qualitatively from those of the SRs. • Observational studies were evaluated if: • There were no SRs or RCTs, or; • When SRs and/or RCTs existed but none were of good quality, and the main results of a given observational study differed qualitatively from the results of the SRs/RCTs.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 14 Consultation Document: Canadian Agency for Drugs and Technologies in Health

4.5 Quality Assessment of the Evidence

Study quality for SRs was assessed using the 11 point AMSTAR instrument (Appendix 3). Studies were rated as good quality if they had a score ≥ 6 and poor otherwise. RCTs, cohort and case control studies were assessed using the adapted SIGN 50 methodology checklists (Appendix 4a to 4c). These studies were assigned ratings of ‘very good’, ‘good’, or ‘poor’. All assessments were conducted by two reviewers and disagreements were resolved by consensus or intervention by a third reviewer.

4.6 Data Extraction

One reviewer extracted information from the selected evidence to create tables listing study type, population, sample size, intervention, comparator, outcome measure, length of follow-up, results, and sources of funding. Additional information about study objectives, numbers and types of included studies, databases searched, data synthesis techniques, and statistical heterogeneity was collected from SRs. A second reviewer checked all tables against the original studies for accuracy.

4.7 Stakeholder Feedback

A document containing the PRSs and corresponding tables of evidence was posted on the COMPUS web site to elicit stakeholder feedback. A web-based electronic form was used to collect this feedback. Stakeholders were given the opportunity to provide comments regarding the PRSs and provide additional evidence. New evidence from stakeholder-cited documents that met our inclusion criteria were extracted and incorporated into the report provided to the COMPUS ERP prior to its deliberations.

4.8 Updating Of Evidence

A literature search from 2003 onwards was conducted in January 2006 to update the evidence obtained from CPGs and CDs (Appendix 2). The search strategy was similar to the initial literature search, except that the objective was to identify relevant SRs and RCTs not already captured in the selected evidence, and language was restricted to English. A limited grey literature search was also conducted on a selection of health technology assessment, economic, and research web sites.

First- and second-level selection procedures for the evidence update were similar to those used in the selection of CPGs and CDs. All heretofore unidentified SRs pertaining to PPI use for the listed indications were selected for quality assessment and data extraction. RCTs were included according to similar rules as for RCTs from CPGs and CDs. Evidence extracted from selected studies was included in the report provided to ERP prior to its deliberations.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 15 Consultation Document: Canadian Agency for Drugs and Technologies in Health

4.9 Expert Review Panel

The COMPUS Expert Review Panel (ERP) for PPIs consisted of 12 members: four gastroenterologists, one clinical pharmacologist, two family physicians, one geriatrician, two pharmacists, one methodologist, and one health economist. The Panel was asked to identify Evidence-Based Statements relating to the use of PPIs for the three major clinical areas for which COMPUS staff had assembled PRSs and evidence. For the purpose of finalizing Evidence-Based Statements related to PPIs, the Panel convened either in-person or by teleconference on five separate occasions in mid to late 2006 for sessions lasting one half to one-and-a-half days.

4.9.1 ERP Process for Identifying Evidence Statements

Prior to the first Panel session, COMPUS provided members with a report containing all PRSs, the associated evidence (from CPGs/CDs, stakeholders, and the evidence update) presented in table form, and pertinent comments from stakeholders. In advance of the meetings, members were asked to review the material and submit a ‘pre-vote’, comments, and any additional evidence that was felt to be relevant, for each PRS. A sample of the voting form used is shown in Appendix 5. In brief, panel members were asked to vote on two aspects of each PRS. First, a Yes/No question was asked on whether there was evidence for the PRS. Members would answer ‘yes’ only if they felt that at least one of the studies detailed was relevant to the PRS. Members voting ‘yes’ would then be asked to rate their level of acceptance with the PRS (based on the strength of the evidence) on the following five-point scale: A) ‘Accept completely’, B) ‘Accept with reservation’, C) ‘Neutral’, D) ‘Reject with reservation’, and E) ‘Reject completely’. These ‘pre-votes’ were tabulated by COMPUS staff and presented at the ERP meetings to facilitate the Panel’s deliberation and final vote on the PRS.

During the Panel meetings, each PRS and the associated evidence was discussed at length. To craft the final Evidence Statement, the Panel could alter the focus, wording, or context of a PRS to accurately reflect the available evidence. Panel members also carefully assessed the evidence presented and determined the final list of studies that should be presented as being relevant to the Evidence Statement. Studies thought to be irrelevant due to characteristics such as study population or outcome could be removed, or relevant studies not originally presented in support of a given PRS could be added. Once the final wording of the proposed Evidence Statement and the corresponding list of relevant studies were finalized, the Panel voted using the same two-part scheme used in the ‘pre-vote’. The Panel also identified any contextual information it felt was relevant to the Evidence Statement. In a few cases, the Panel felt that a Statement required further literature search or analysis of the evidence. These Statements were ‘parked’ and deferred to a subsequent meeting to allow COMPUS staff to perform the necessary tasks.

An Evidence Statement was considered endorsed by the Panel if a majority of the Panel members voted to either ‘Accept completely’ or ‘Accept with reservation’. If a majority voted to ‘Reject with reservation’ or ‘Reject completely’, the Statement was inverted so that the final wording was acceptable. For example, if the Panel voted to reject the hypothetical statement ‘PPIs are more efficacious than H2RAs for condition X’, the Statement was re-worded to state ‘PPIs are no more efficacious than H2RAs for condition X’.

A Statement was considered to be a Research Gap if any of the following occurred if: 1) A majority of the Panel voted ‘No’ to the first question regarding whether there was evidence for the Statement; 2) A

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 16 Consultation Document: Canadian Agency for Drugs and Technologies in Health majority voted ‘Neutral’ to the second question; or 3) The votes were evenly divided (i.e., there was a difference of less than 2 votes) between ‘Accept’ and ‘Reject’.

This report contains the approved Evidence-Based Statements, each with a table of evidence containing summaries of the studies identified as being of relevance to the Statement. For selected results, numbers-needed-to-treat (NNTs) were calculated by COMPUS staff and are presented in the evidence tables. Results of the Panel’s votes are presented as a percentage breakdown and as a visual analog scale (green indicating agreement). Contextual information identified by ERP as being relevant to the Statement is also reported.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 17 Consultation Document: Canadian Agency for Drugs and Technologies in Health

5. RESULTS Legend Level Interpretation A Accept completely B Accept with reservation C Cannot accept or reject (neutral) D Reject with reservation E Reject completely

5.1. GERD

5.1.1 List of Evidence Statements and Research Gaps

G1: Uninvestigated GERD G1.1: Initial therapy G1.1.1: PPIs vs. H2RAs Evidence Statements

G1.1.1A: Standard-dose PPIs, as initial therapy for up to 4 weeks, are more efficacious than H2RAs for improvement of reflux symptoms in uninvestigated GERD.

G1.1.1B: Standard-dose PPIs are more efficacious than continued H2RAs in patients with uninvestigated GERD who have incomplete response to a previous trial of H2RAs.

G1.1.2: Prognosis after initial therapy Evidence Statements

G1.1.2A: Approximately 20% of patients with uninvestigated GERD will remain asymptomatic off therapy for up to 6 months after a successful course of initial therapy (for 4-8 weeks) with a PPI or H2RA.

G1.2: Maintenance therapy G1.2.1: Long-term therapy with PPIs vs. H2RAs Evidence Statements

G1.2.1A: PPI therapy in uninvestigated GERD is more efficacious than H2RAs for control of symptoms for up to 6 months.

G1.2.1B In patients with uninvestigated GERD who have completed their initial course of PPIs, continued PPI therapy is more efficacious than step-down to H2RAs for providing symptom relief.

G1.2.2: On-demand PPIs after response to initial PPI therapy

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 18 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statements

G1.2.2A: For patients with uninvestigated GERD who respond to initial PPI therapy, subsequent ‘on– demand’ PPI therapy is more efficacious than placebo.

G1.2.2B: For patients with uninvestigated GERD who respond to initial PPI therapy, subsequent ‘on– demand’ PPI therapy is more efficacious than continuous standard-dose H2RAs.

G1.2.2C: For patients with uninvestigated GERD who respond to initial PPI therapy, subsequent ‘on– demand’ PPI therapy is less efficacious than continuous standard regular dose PPI therapy for heartburn resolution.

G2: Endoscopic Negative Reflux Disease (ENRD) G2.1: Initial therapy G2.1.1: PPIs vs. H2RAs Evidence Statements

G2.1.1A: PPIs are more efficacious than H2RAs, as initial therapy, for improvement of heartburn symptoms at four weeks in patients with ENRD.

G2.1.1B: PPIs are no more efficacious than H2RAs for improving quality of life in patients with ENRD.

G2.1.2: Half-dose vs. standard-dose PPIs Evidence Statements

G2.1.2A: For patients with ENRD, half-dose PPI as initial therapy is less efficacious than standard-dose PPI therapy for heartburn resolution.

G2.2: Maintenance therapy G2.2.1: Long-term therapy with PPIs Evidence Statements

G2.2.1A: Half-dose, continuous PPI therapy in ENRD is more efficacious than placebo for control of symptoms for up to 6 months.

G2.2.2: On-demand PPIs after response to initial PPI therapy Evidence Statements

G2.2.2A: For patients with ENRD who respond to initial PPI therapy, subsequent ‘on-demand’ PPI therapy is more efficacious than placebo.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 19 Consultation Document: Canadian Agency for Drugs and Technologies in Health

G3: Erosive Esophagitis G3.1: Initial therapy G3.1.1: PPIs vs. H2RAs Evidence Statements

G3.1.1A: PPIs are more efficacious than H2RAs for improving symptoms and improving the healing of erosive esophagitis.

G3.1.1B: The speed of improvement of heartburn and healing is faster with PPIs than H2RAs in patients with erosive esophagitis.

G3.1.2 : Double-dose vs. standard-dose PPIs Evidence Statements

G3.1.2A: Doubling the standard daily doses of PPIs, as initial therapy, is no better than standard daily dose PPI therapy for improvement of erosive esophagitis.

G3.2: Maintenance therapy G3.2.1: Long-term therapy with PPIs vs. H2RAs or placebo Evidence Statements

G3.2.1A: Long-term maintenance PPI therapy (i.e., up to 12 months) in erosive esophagitis is more efficacious than placebo for prevention of symptomatic and endoscopic relapse.

G3.2.1B: Long-term maintenance PPI therapy (i.e., up to 12 months) in erosive esophagitis is more efficacious than H2RAs for prevention of endoscopic relapse.

G3.2.1C: Long-term PPI therapy is more efficacious than H2RAs for erosive esophagitis complicated by strictures.

G3.2.1D: In patients with erosive esophagitis who have completed an initial course of PPIs, continued PPI therapy is more efficacious than step-down to H2RAs for preventing relapse and providing symptom relief.

G3.2.2: Half-dose vs. standard-dose PPIs Evidence Statements

G3.2.2A: Half-dose PPIs are less efficacious than standard-dose PPIs as maintenance treatment for the prevention of symptomatic and endoscopic relapse in patients with erosive esophagitis.

G3.2.3: On-demand PPIs after response to initial PPI therapy Evidence Statements

G3.2.3A: For patients with erosive esophagitis who respond to initial PPI therapy, subsequent ‘on- demand’ PPI therapy is less efficacious than continuous standard-dose PPI therapy.

G3.2.4: Intermittent PPI strategy

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 20 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statements

G3.2.4A: In patients with erosive esophagitis, intermittent PPI therapy (e.g., 3 days per week) was less efficacious than continuous PPI therapy.

G4: Asthma, laryngeal symptoms, and cough associated with GERD Evidence Statements

G4.1: PPIs are not efficacious in asthma patients with concomitant GERD.

G4.2: PPIs are not efficacious in improving laryngeal symptoms (i.e., cough, throat clearing, globus, hoarseness, sore throat) associated with reflux.

G4.3: PPIs are not efficacious in improving chronic cough with or without GERD.

G5: Differences among PPIs Evidence Statements

G5.1: There are no clinically important differences among standard-doses of PPIs (omeprazole 20 mg, lansoprazole 30 mg, pantoprazole 40 mg, rabeprazole 20 mg, esomeprazole 20 mg) in treatment of symptomatic GERD, ENRD and esophagitis.

G6: PPI Adverse drug reactions Evidence Statements

G6.1: PPIs have a similar adverse event rate (generally minor) as H2RAs in GERD randomized controlled trials of up to one year duration.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 21 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Research Gaps

1) Double dose PPIs are more efficacious than standard-dose PPIs for patients with uninvestigated GERD symptoms who have severe symptoms.

2) In patients with uninvestigated GERD, initial therapy with half-dose PPIs is less efficacious than standard-dose PPIs for relief of symptoms.

3) In patients with ENRD who have completed their initial course of PPIs, continued PPI therapy is more efficacious than stepping down to H2RAs.

4) In patients with uninvestigated GERD, ENRD or erosive esophagitis who have received initial PPI therapy, step-down therapy to H2RAs is more efficacious than placebo.

5) Double dose PPIs are more efficacious than continued standard-dose PPIs in patients with uninvestigated GERD, erosive esophagitis or ENRD, who remain symptomatic with regular dose PPIs.

6) For patients with ENRD who respond to initial PPI therapy, subsequent ‘on-demand’ PPI therapy is less efficacious than continuous standard-dose PPI therapy.

7) In patients with erosive esophagitis, surgical anti-reflux therapy is not more efficacious than maintenance PPI therapy (when dose adjustment based on symptoms is allowed).

8) The required duration of therapy for erosive esophagitis complicated by strictures was identified as an area where further research is required.

9) The use of PPIs in Barrett’s Esophagus was identified as an area where further research is required.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 22 Consultation Document: Canadian Agency for Drugs and Technologies in Health

5.1.2 Evidence Statements, ERP voting results, and Summaries of evidence

G1: Uninvestigated GERD G1.1: Initial therapy G1.1.1:PPIs vs. H2RAs Evidence Statement G1.1.1A: Standard-dose PPIs, as initial therapy for up to 4 weeks, are more efficacious than H2RAs for improvement of reflux symptoms in uninvestigated GERD.

Voting Results: A 75%, B 25%, C 0%, D 0%, E 0%

Summary: Five RCTs were identified for this statement. In a good quality RCT, Armstrong et al. 35 reported that initial treatment with omeprazole 20 mg daily was more efficacious than ranitidine 150 mg twice daily for the relief of heartburn at 4 weeks in patients with heartburn-dominant uninvestigated dyspepsia. However, there were no significant differences in terms of other GERD symptoms. Data from Kaplan-Machlis et al.36, Talley et al. 37, van Zyl et al. 38 and Howden et al. 39, all RCTs of good quality, suggest that initial treatment with standard-doses of PPIs was more efficacious than standard- doses of H2RAs for the relief of heartburn and other GERD symptoms in patients with uninvestigated GERD at various time points between 1-20 weeks after the start of treatment. Study Population Intervention Comparator Outcome measure Results Type(QA) Armstrong 1. 390 patients with ome 20 mg ran 150 mg bid Heartburn relief ome 20 mg start vs. ran 150 et al. 2005 heartburn dominant qd for 4-8 for 4-8 weeks at 4 weeks mg bid start: 35 uninvestigated weeks, Stepped to ome Heartburn relief from daily dyspepsia Stepped to 20 mg qd for 4- diary: 55% (95% CI : (48.1- RCT (good) 2. Score > 4 on 7- ome 40 mg 8 weeks if 62.1%) vs. 27% (21-34%) point Likert scale for qd for 4-8 heartburn (p<0.001). heartburn severity† weeks if persists NNT (95% CI) = 4 (3, 5) heartburn Sustained heartburn persists resolution: 36% (26-43%) vs. 13% (8-18%) (p<0.001). NNT (95% CI) = 4 (3, 7)

Median time to heartburn relief : 3 (2-5) days vs. 8 (5- 15) days. Median time for sustained heartburn resolution : 5 (2-8) vs. 29 (15-35) days.

Global assessment of upper GIT symptoms over 7 days before week 4 visit : Heartburn relief : 76% (70- 82%) vs. 58% (51-65%) (p = 0.0002). NNT (95% CI) = 6 (4, 11)

Regurgitation: No difference (data not shown) Epigastric pain: No difference (data not shown) Nausea or vomiting: No

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 23 Consultation Document: Canadian Agency for Drugs and Technologies in Health

difference (data not shown). van Zyl 1. 338 GERD patients pant 20 mg ran 300 mg qd Relief of key Proportion with relief of key 200438 aged 18-75 with qd for 4 for 4 weeks GERD symptoms GERD symptoms heartburn, acid weeks (retrosternal 28 days: 68.3% with pant vs. RCT (good) regurgitation or burning pain, acid 43.3% with ran, 95% CI of dysphagia symptoms regurgitation and OR: 1.84 – 4.51 (for ≥2 days prior to pain on presentation and ≥ 1 swallowing) Proportion with relief of key episode/month for GERD symptoms prior 3 months); 14 days: 56.6% with pant vs. endoscopy not 31.1% with ran, 95% CI of performed OR: 1.79-4.73 2. GERD (heartburn) symptoms were 2 points higher on Likert scale† than any other GI symptom (e.g., epigastric pain) 3. Overall average heartburn score was 5 in both groups (estimated from figure)

Talley et al. n=307 pant 20 mg ran 150 mg bid Primary: Pant vs. ran: 200237 patients ≥18 years of qd for 12 for 12 months Symptom control Proportion with ‘complete’ age with symptomatic months at 6 and 12 symptom control: RCT (good) GERD presenting to months Week 4: 67% vs. 48% (p< GP (heartburn at least (‘Complete’ = 0.001). (Data extrapolated twice a week as absence of from figure). predominant upper GI heartburn during NNT (95% CI) = 5 (3, 12) complaint) 7 days before f/u; ‘Sufficient’ = Week 8: 75% vs. 58% % in each category of mild heartburn on (p<0.001) (Data extrapolated heartburn severity (7- not more than 1 from figure). point Likert scale†): day during 7 days NNT (95% CI) = 6 (4, 15) 1: 0% before f/u) 2: 1% Proportion with ‘sufficient’ 3: 11-15% Secondary: GSRS symptom control: 4: 33-42% scores, rates of Week 4: 64% vs. 48%, 5: 25-35% symptomatic p=0.008 6: 15-18% relapse (heartburn NNT (95% CI) = 6 (4, 20) 7 2-3% of at least mild intensity on 3 of GSRS Scores: 7 days before f/u Significantly lower scores in patients who for abdominal pain, reflux, achieved indigestion, and diarrhea in remission at week pant arm but absolute 8) differences were small (data not shown) Howden et 1. 593 patients with lans 30 mg R= ran, 150 Heartburn relief Lans 30 mg qd vs. ran 150 al. 2001*39 symptomatic GERD; qd for 20 mg, bid for 20 at 20 weeks mg bid: endoscopy not weeks weeks Median Heartburn severity RCT (good) performed Pretreatment period: 1.75 vs. 1.88 2. Median Heartburn Week 1-8: 0.29 vs. 0.57

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 24 Consultation Document: Canadian Agency for Drugs and Technologies in Health

severity during (p<0.001) pretreatment period: Week 9-20: 0.17 vs. 0.36 1.75 vs. 1.88 (lans (p<0.001). vs. ran) (0 = none, 1 = mild, 2 = moderate, 3 = severe)

Kaplan- 1. 268 GERD patients ome 20 mg ran 150 mg bid GSRS at 12 and GSRS: Machlis et (clinically diagnosed) qd for 6 for 6 months 24 weeks, At 2- and 4- week: ome al. 2000*36 2. GSRS reflux months PGWS at 24 groups showed lower subscale score 3.9 – weeks adjusted reflux scores RCT (good) 4.3 (adjusted 1-month mean score 2.53) vs. ran (2.89) (p=0.005) At 3 months, GSRS reflux scores, showed overall treatment difference favoring ome (adjusted 3-month mean score 2.67) vs. ran (2.95) (p<0.05) At 6 months: Adjusted mean reflux scores 2.68 with ome vs. 2.85 with ran (p = 0.2).

PGWS: No difference in the improvement PGWB score was found between ome and ran from 1 to 6 months. †Likert scale for heartburn severity: 1 = no symptoms; 2 = minimal; 3 = mild; 4 = moderate; 5 =moderate severe; 6 = severe; 7 = very severe GSRS: Gastrointestinal Symptoms Rating Scale; PGWB: Psychological General Well Being scale; lans: lansoprazole; ome: omeprazole; pant: pantoprazole; ran: ranitidine; * indicates industry involvement

Context: • A majority of subjects in trials reported heartburn of moderate severity at baseline. • Most studies35-38 used a 7-point Likert scale to measure heartburn severity, in which ‘1’ corresponds to no heartburn, and ‘2’ to mild heartburn. One study39 used a 4-point scale, in which ‘0’ corresponds to no heartburn, and ‘1’ to mild heartburn. • Symptom relief rates ranged from 55-75% with PPIs versus 27-58% with H2RAs at 4-8 weeks in the above studies. • The ERP recognized that Quality of life (QOL) is an important issue but there was uncertainty as to which outcomes and measures of QOL are of importance. There was also uncertainty as to whether the Psychological Well-being Scale (PGWS) results should be considered.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 25 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statement G1.1.1B: Standard-dose PPIs are more efficacious than continued H2RAs in patients with uninvestigated GERD who have incomplete response to a previous trial of H2RAs.

Voting Results: A 42%, B 58%, C 0%, D 0%, E 0%

Summary: PPIs were more effective than H2RAs for the relief of heartburn that was resistant to H2RAs in one good quality RCT40. Study Outcome Population Intervention Comparator Results Type (QA) measure Maton et al. 533 patients ome 20 mg ran 150 mg Complete Patients with complete resolution of heartburn: 1999*40 with qd for 8 bid for 8 heartburn wk 4: 31% with ome vs. 11% with ran heartburn, weeks weeks resolution; (p<0.0001) RCT (good) poorly Total NNT (95% CI) = 5 (3, 9) responsive to heartburn wk 8: 46% with ome vs. 16% with ran 6 weeks of relief; (p<0.0001). treatment Heartburn- NNT (95% CI) = 3 (3, 5) with H2RAs free days at 4 and 8 Total heartburn relief (patients having no or weeks mild heartburn): wk 4: 66% with ome vs. 40% with ran (p<0.0001) NNT (95% CI) = 4 (3, 6) wk 8: 70% with ome vs. 49% with ran (p = 0.0004), NNT (95% CI) = 5 (3, 10)

The percentage of heartburn-free days: wk 4: 69% with ome vs. 48% with ran (p<0.0001) wk 8: 76% with ome vs. 56% with ran (p<0.0001)) ome: omeprazole; ran: ranitidine; * indicates industry involvement

Context: • The ERP noted that the evidence for this Statement was limited since there was only one RCT. • Less than 50% of patients taking either H2RA or PPI had complete resolution of symptoms. By week 8, almost half had ‘no’ or ‘mild’ heartburn with H2RAs. Treatment with H2RAs resulted in 56% of heartburn-free days by 8 weeks, which increased to 76% with PPIs.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 26 Consultation Document: Canadian Agency for Drugs and Technologies in Health

G1: Uninvestigated GERD G1.1: Initial therapy G1.1.2: Prognosis after initial therapy Evidence Statement G1.1.2A: Approximately 20% of patients with uninvestigated GERD will remain asymptomatic off therapy for up to 6 months after a successful course of initial therapy (for 4-8 weeks) with a PPI or H2RA.

Voting Results: A 25%, B 75%, C 0%, D 0%, E 0%

Summary: The CADET-HR study by Armstrong et al.35 was a good quality RCT in which uninvestigated heartburn-dominant dyspepsia patients were randomized to a ‘PPI-start’ or ‘H2RA-start’ strategy. If heartburn persisted, step-up to a high dose of PPI was allowed in the PPI-start arm, and to standard-dose PPI in the H2RA-start arm. Responders in the initial treatment phase were followed up in a 6-month treatment-free phase. Of the initial cohort, 81% and 77% of PPI-start and H2RA-start subjects were eligible for the treatment-free follow-up. At 6 months, 79% and 76% of PPI-start and H2RA-start patients relapsed, with no significant difference between groups. Study Population Intervention Comparator Outcome measure Results Type(QA) Armstrong 390 patients ome 20 mg qd ran 150 mg bid Heartburn relief (Treatment phase data not et al. 2005 with heartburn for 4-8 weeks for 4-8 weeks defined as none or shown) 35 dominant mild heartburn on no uninvestigated Stepped to ome Stepped to ome more than 1 of the Proportion eligible for RCT dyspepsia 40 mg qd for 4- 20 mg qd for 4- previous 7 days treatment-free f/u phase (PPI- (good) 8 weeks if 8 weeks if (maximum score of start vs. H2RA-start): 81% heartburn heartburn 3 on a 7-point Likert (n=159) vs. 77% (n=149) persisted. persisted. scale†). Measured If no relief at If no relief at by patient daily Proportion with heartburn- 12-16 weeks, 12-16 weeks, diary. relapse at 6 months (95% CI) patients patients (PPI-start vs. H2RA-start): withdrawn from withdrawn from Heartburn relapse 79% (72-85%) vs. 76% (69- study and study and during treatment- 83%), p=0.72 (NS) performed offered ome 40 free follow-up endoscopy. mg qd. defined as heartburn Median time to relapse (95% on ≥2 of the CI) (PPI-start vs. H2RA- Subjects Subjects previous 7 days with start): 8 (7-9) vs. 9 (8-11) experiencing experiencing severity ≥4 on 7- days (NS difference) heartburn relief heartburn relief point Likert scale†. at 4, 8, 12, or 16 at 4, 8, 12, or 16 Measured by patient Median # of days to moderate weeks were weeks were daily diary. to severe heartburn (≥4 on entered into entered into Likert scale) (95% CI): 6 (5- treatment-free treatment-free 7), similar for both groups follow-up phase follow-up phase of 6 months. of 6 months. Mean proportion of days spent heartburn-free from the time b/w heartburn relief and relapse (95% CI) (PPI-start vs. H2RA-start): 22% (18- 26%) vs. 23% (18-27%) (NS difference) †Likert scale for heartburn severity: 1 = no symptoms; 2 = minimal; 3 = mild; 4 = moderate; 5 =moderate severe; 6 = severe; 7 = very severe ome: omeprazole; ran: ranitidine

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 27 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Context: • The majority of patients had relapsed 6 months after treatment discontinuation. Median time to relapse was only 8-9 days.

G1: Uninvestigated GERD G1.2: Maintenance therapy G1.2.1: Long-term therapy with PPIs vs. H2RAs Evidence Statement G1.2.1A: PPI therapy in uninvestigated GERD is more efficacious than H2RAs for control of symptoms for up to 6 months.

Voting Results: A 17%, B 75%, C 8%, D 0%, E 0%

Summary: Three RCTs were identified as evidence for this statement, two of good quality37,39 and one of poor quality 41,42. The data from these trials suggest that long-term treatment with PPIs is more efficacious than H2RAs for the relief of heartburn, control of GERD symptoms, and patient satisfaction up to 6 months in patients with uninvestigated GERD. Study Outcome Population Intervention Comparator Results Type(QA) measure Talley et al. n=307 pant 20 mg ran 150 mg Primary: Pant vs. ran: 200237 patients ≥18 years of once daily twice daily for Symptom Proportion with ‘complete’ age with symptomatic for 12 12 months control at 6 symptom control: RCT GERD presenting to months and 12 At 6 months: 71% vs. 56%, (good) GP (heartburn at least months p=0.007 twice a week as (‘Complete’ NNT (95% CI) = 7 (4, 23) predominant upper GI = absence of complaint) heartburn At 12 months (pant vs. ran): 77% during 7 vs. 59%, p=0.001 days before NNT (95% CI) = 6 (4, 13) f/u; ‘Sufficient’ Proportion with ‘sufficient’ = mild symptom control: heartburn on At 12 months: 86% vs. 79%, not more p>0.05 (NS) than 1 day during 7 Relapse rates in patients who days before achieved complete symptom f/u) control at 8 weeks: At 12 months (pant vs. ran): 11% Secondary: vs. 12%, p>0.05 (NS) GSRS scores, rates of symptomatic relapse (heartburn of at least mild intensity on 3 of 7 days before f/u in patients who

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 28 Consultation Document: Canadian Agency for Drugs and Technologies in Health

achieved remission at week 8) Hansen et 2,156 patients with esome 20 mg esome 20 mg Relapse rate During 6 months: al. symptoms of GERD qd qd on demand esome 20 mg continuous vs. 2005 (no endoscopy continuously for 6 months Patient esome 20 mg on demand vs. ran 41,42 performed) for 6 months satisfaction 150 mg bid continuous: ran 150 mg bid with Patients received continuously treatment Proportion of patients with at RCT esome 40 mg qd for 4 for 6 months No least one relapse: 7.0% vs. 10.9% (poor) weeks; 1,902 heartburn, no vs. 34.4%, p<0.0001 for both symptom-free patients acid esome groups vs. ran were enrolled regurgitation. Patient satisfaction: proportion of Time spent patients completely/very satisfied: on assigned 82.2% vs. 75.4% vs. 33.5%, maintenance p<0.0001 for both esome groups therapy. vs. ran, p<0.01 for esome continous vs. esome on demand

Time on maintenance therapy (days): 170 vs. 171 vs.125, p<0.0001 for both esome groups vs. ran

Proportion without heartburn: 72.2% vs. 45.1% (p<0.0001) vs. 32.5%, p<0.0001

Proportion without heartburn acid regurgitation: 78% vs. 62% vs. 45.7%, no p value given Howden et 593 patients with lans 30 mg ran, 150 mg, Heartburn lans 30 mg qd vs. ran 150 mg bid: al. 2001*39 symptomatic GERD; qd bid for 20 relief up to Median Heartburn severity (scale endoscopy not weeks 20 weeks of 0-3): RCT performed Pretreatment period: 1.75 vs. 1.88 (good) Week 1-8: 0.29 vs. 0.57 (p<0.001) Week 9-20: 0.17 vs. 0.36 (p<0.001).

Esome: esomeprazole; lans: lansoprazole; pant: pantoprazole; ran: ranitidine.

Context: • In the only 12-month study, nearly 60% of patients on H2RA had ‘complete symptom control’ at 6 and 12 months, and there was no difference between PPI and H2RA in terms of the proportion with ‘sufficient symptom control’ at 12 months37. • Only the Talley study assessed (as a secondary outcome) relapse rates among patients who achieved complete symptom control with initial therapy. There was no significant difference between PPI and H2RA for this outcome37. • It was noted that the strength of the evidence for this Statement was limited by the fact that only studies of long-term therapy were considered by the Panel.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 29 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statement G1.2.1B: In patients with uninvestigated GERD who have completed their initial course of PPIs, continued PPI therapy is more efficacious than step-down to H2RAs for providing symptom relief.

Voting Results: A 17%, B 67%, C 8%, D 8%, E 0%

Summary: The results of one good quality RCT39 indicate that PPI treatment provides more consistent heartburn relief than step-down to H2Ras in patients with GERD. Study Outcome Population Intervention Comparator Results Type(QA) measure Howden et 593 patients with L= lans 30 R= ran, 150 Heartburn L had significantly less heartburn al. 2001*39 symptomatic GERD mg qd mg, bid for 20 relief at 20 than R (heartburn on at weeks weeks L (82%) had significantly higher 24- RCT least 50% of past 7- RL = ran 150 h heartburn relief than R (66%), RL (good) 10 days, at least one mg bid for 8 (74%), and LR (67%), p<0.001 moderate-severe weeks and then NNT (95% CI) = 6 (4, 16) for L episode) switch to lans vs.R; 30 mg qd for NNT (95% CI) = 13 (6, ∞) for L vs. Median heartburn the following RL; severity ranged 12 weeks NNT (95% CI) = 7 (4, 20) for L vs. from 1.75 to 1.88 LR = lans 30 LR; (on scale of 0-3 mg qd for 8 described in weeks and then Median heartburn severity during ‘Results’) in the 4 switch to ran study on scale 0-3 (0=none, 1=mild, arms 150 mg bid for 2=moderate, 3=severe): the following Pre-treatment period: 1.88 for R, 12 weeks 1.75 for L, 1.75 for RL, 1.70 for LR Week 1-8: 0.57 for R, 0.29 for L, 0.56 for RL, 0.34 for LR; L vs. R p<0.001, L vs. RL p<0.001, LR vs. R p<0.001, RL vs. LR p<0.001 Week 9-20: 0.36 for R, 0.17 for L, 0.19 for RL, 0.49 for LR ; L vs. R p<0.001, L vs. LR p<0.001, LR vs. RL p<0.001, RL vs. R p<0.05 Week 1-20: 0.46 for R, 0.25 for L, 0.35 for RL, 0.44 for LR; L vs. R p<0.001, L vs. RL p<0.05, L vs. LR p<0.001, RL vs. R p<0.05, LR vs. R not significant, LR vs. RL not significant lans: lansoprazole; ran: ranitidine; * indicates industry involvement

Context: • The evidence is limited to one study. • The severity of GERD at baseline may be important for the degree of success with step-down therapy. • The clinical relevance of differences in heartburn scores of less than one is uncertain.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 30 Consultation Document: Canadian Agency for Drugs and Technologies in Health

G1: Uninvestigated GERD G1.2: Maintenance therapy G1.2.2: On-demand PPIs after response to initial PPI therapy Evidence Statement G1.2.2A: For patients with uninvestigated GERD who respond to initial PPI therapy, subsequent ‘on-demand’ PPI therapy is more efficacious than placebo.

Voting Results: A 8%, B 67%, C 25%, D 0%, E 0%

Summary: The evidence identified consists of one poor quality RCT43. It found that on-demand PPI was better in terms of symptom relief, patient satisfaction and compliance than placebo for patients with GERD. Study Outcome Population Intervention Comparator Results Type(QA) measure Bigard et al. 203 patients (18-80 lans 15 mg Placebo qd Proportion of At 6 months: 2005*43 yrs) with GERD (not qd on on demand patients Proportion completing study: lans endoscopically demand for 6 for 6 months completing vs. placebo: 81.0% vs. 60.8%, RCT (poor) verified) who were months study p=0.003 asymptomatic after 4 Proportion discontinued due to weeks of lans 15 Proportion of insufficient heartburn control: lans mg/day patients vs. placebo: 15.5% vs. 27.6%, discontinued p=0.046. due to insufficient Time between initiation of on- heartburn demand therapy and control. discontinuation (days): lans vs. placebo: 162.4±53.1 136.7±70.0, Overall p=0.024. assessment of treatment Treatment efficacy judged to be efficacy (by ‘excellent or good’: lans vs. investigator) placebo: 71.4% vs. 54.6%, p=0.023. lans: lansoprazole; * indicates industry involvement

Context: • The evidence for this Statement is limited to one poor quality RCT in 203 patients in whom treatment efficacy was judged to be ‘good’ or ‘excellent’ in 55% of placebo-treated patients versus 71% of PPI-treated patients. • The clinical relevance of the outcomes studied is a potential limitation of the study. • There is no evidence on long-term use of on-demand PPI therapy.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 31 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statement G1.2.2B: For patients with uninvestigated GERD who respond to initial PPI therapy, subsequent ‘on-demand’ PPI therapy is more efficacious than continuous standard-dose H2RAs.

Voting Results: A 0%, B 58%, C 42%, D 0%, E 0%

Summary: One poor quality RCT (reported in two separate publications)41,42 was identified for this statement. On demand esomeprazole 20 mg was more effective than continuous H2RA therapy at 6 months in terms of relapse rate, patient satisfaction, and time spent on assigned therapy in subjects with GERD that were symptom-free after 4 weeks of esomeprazole 40 mg daily41,42. Study Outcome Population Intervention Comparator Results Type(QA) measure Hansen et 2,156 patients with esome 20 mg esome 20 mg Relapse rate During 6 months: al. heartburn as qd qd on esome 20 mg continuous vs. 2005*42 predominant continuously demand for 6 Patient esome 20 mg on demand vs. ran symptom for 3 of for 6 months months satisfaction 150 mg bid continuous: Hansen et past 7 days (no with treatment al. endoscopic grading ran 150 mg Proportion of patients with at least 2005*41 reported) bid No heartburn, one relapse: 7.0% vs. 10.9% vs. continuously no acid 34.4%, p<0.0001 for both esome Severity of heartburn for 6 months regurgitation. groups vs. ran at baseline: Mild RCT (poor) (11.4%), Moderate Time spent on Patient satisfaction: proportion of (70.0%), Severe assigned patients completely/very satisfied: (17.9%) maintenance 82.2% vs. 75.4% vs. 33.5%, therapy. p<0.0001 for both esome groups Patients received vs. ran, p<0.01 for esome esome 40 mg qd for continous vs. esome on demand 4 weeks; 1,902 symptom-free Time on maintenance therapy patients (no more (days): 170 vs. 171 vs.125, than 1 day with mild p<0.0001 for both esome groups heartburn in past 7 vs. ran days) were enrolled Proportion without heartburn: 72.2% vs. 45.1% (p<0.0001) vs. 32.5%, p<0.0001 NNT (95% CI) = 3 (2, 3) for esome continuous vs. ran; NNT (95% CI) = 8 (6, 14) for esome on-demand vs. ran; NNT (95% CI) = 4 (3, 5) for esome continuous vs. on-demand

Proportion without heartburn acid regurgitation: 78% vs. 62% vs. 45.7%, no p value given esome: esomeprazole; ran: ranitidine; * indicates industry involvement

Context: • The evidence for this Statement is limited to one poor quality RCT that was conducted in an open-label fashion.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 32 Consultation Document: Canadian Agency for Drugs and Technologies in Health

• The mean daily dose consumed in each treatment arm was not reported in the study. However, based on the cost analysis, it may be inferred that mean consumption of study medication was 35% lower than in the continuous esomeprazole arm42.

Evidence Statement G1.2.2C: For patients with uninvestigated GERD who respond to initial PPI therapy, subsequent ‘on-demand’ PPI therapy is less efficacious than continuous standard regular dose PPI therapy for heartburn resolution.

Voting Results: A 0%, B 67%, C 33%, D 0%, E 0%

Summary: One poor quality RCT (reported in two separate publications)41,42 identified found that continuous PPI therapy is more effective than on-demand therapy in terms of patient satisfaction and heartburn relief at 6 months. Study Outcome Population Intervention Comparator Results Type(QA) measure Hansen et 2,156 patients with esome 20 mg esome 20 mg Relapse rate During 6 months al. heartburn as qd qd on esome 20 mg continuous vs. 2005*42 predominant continuously demand for 6 Patient esome 20 mg on demand vs. ran symptom for 3 of for 6 months months satisfaction 150 mg bid continuous: Hansen et past 7 days (no with treatment al. endoscopic grading ran 150 mg Proportion of patients with at least 2005*41 reported) bid No heartburn, one relapse: 7.0% vs. 10.9% vs. continuously no acid 34.4%, p<0.0001 for both esome RCT (poor) Severity of heartburn for 6 months regurgitation groups vs. ran at baseline: Mild (11.4%), Moderate Time spent on Patient satisfaction: proportion of (70.0%), Severe assigned patients completely/very satisfied: (17.9%) maintenance 82.2% vs. 75.4% vs. 33.5%, therapy p<0.0001 for both esome groups Patients received vs. ran, p<0.01 for esome esome 40 mg qd for continous vs. esome on demand 4 weeks; 1,902 symptom-free Time on maintenance therapy patients (no more (days): 170 vs. 171 vs.125, than 1 day with mild p<0.0001 for both esome groups heartburn in past 7 vs. ran days) were enrolled Proportion without heartburn: 72.2% vs. 45.1% (p<0.0001) vs. 32.5%, p<0.0001 NNT (95% CI) = 3 (2, 3) for esome continuous vs. ran; NNT (95% CI) = 8 (6, 14) for esome on-demand vs. ran; NNT (95% CI) = 4 (3, 5) for esome continuous vs. on-demand

Proportion without heartburn acid regurgitation: 78% vs. 62% vs. 45.7%, no p value given esome: esomeprazole; ran: ranitidine; * indicates industry involvement

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 33 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Context: • The evidence for this Statement is limited to one poor quality RCT that was conducted in an open-label fashion. • The mean daily dose consumed in each treatment arm was not reported in the study. However, based on the cost analysis, it may be inferred that mean consumption of study medication was 35% lower than in the continuous esomeprazole arm42. • Although a p-value for the comparison between continuous and on-demand esomeprazole in terms of the proportion of patients with at least one relapse was not reported, the difference was numerically small (i.e., 7 vs. 11%)41. • While statistically significant, the clinical relevance of the difference between continuous and on-demand esomeprazole in terms of the proportion of patients ‘completely’ or ‘very’ satisfied is uncertain. • A significant proportion of patients did well with on-demand therapy.

G2: Endoscopic Negative Reflux Disease (ENRD) G2.1: Initial therapy G2.1.1: PPIs vs. H2RAs Evidence Statement G2.1.1A: PPIs are more efficacious than H2RAs, as initial therapy, for improvement of heartburn symptoms at four weeks in patients with ENRD.

Voting Results: A 10%, B 60%, C 10%, D 20%, E 0%

Summary: van Pinxteren et al. recently published an update44 of their good quality SR from 200445. In the update, one additional RCT (of standard-dose PPI vs. H2RA) was available to the authors for the analysis44. The combined analysis of half- and standard-dose PPIs vs. H2RAs showed that PPIs were significantly more effective than H2RAs for the relief of heartburn at 4 weeks in ENRD patients. (In the 2004 SR, the estimate of RR was similar but the confidence interval crossed unity.45) In previous meetings ERP members had discussed the need for a comparison restricted to standard-dose PPIs (i.e., without inclusion of half-dose data) vs. H2RAs. Because the paper did not contain this analysis, COMPUS staff performed their own using Easy MA, based on trial-level data presented in the paper. Standard-dose PPIs alone were not found to be significantly different from H2RAs.

For overall symptoms improvement, PPIs were significantly better than H2RAs (combined half- and standard-dose analysis)44. This result was similar to that in the 2004 version of the SR. Study Outcome Population Intervention Comparator Results Type(QA) measure van 4 RCTs ome 10, 20 or 40 cim 300 or 400 Heartburn Heartburn persistence Pinxteren (n=960), mg, qd, pant 20 mg qid, fam 20 remission at PPIs (half or standard-dose) vs. et al. patients with or 40 mg qd, mg bid or 40 4 weeks H2RAs: 200644 ENRD) for esome 20 or 40 mg qd, niz 150 RR = 0.78 (95% CI 0.62, 0.97) heartburn mg, qd, rab 10 mg bid, ran 150 SR (good) remission and 20 mg qd) for mg bid for 4 PPIs (standard-dose only) vs. 4 weeks weeks H2RAs: RR = 0.79 (95% CI 0.61, (ome 10 mg 1.02) (not presented in paper; group: n = 465; calculated by COMPUS using Easy ome 20 mg MA software) group: n = 138;

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 34 Consultation Document: Canadian Agency for Drugs and Technologies in Health

H2RAs standard-dose group: n=357)

2 RCTs (n = lans 15, 30 mg, ran 150 mg bid, Overall Overall symptom improvement: PPIs 937), patients ome 20 mg for 4 fam 20 mg bid symptom vs. H2RAs: RR= 0.82 (95% CI 0.73, with ENRD) weeks for 4 weeks improvement 0.93) for overall at 4 weeks symptom improvement lans: lansoprazole; ome: omeprazole; esome: esomeprazole; rab: rabeprazole; cim: cimetidine; ran; ranitidine; fam: famotidine; niz: nizatidine; pant: pantoprazole

Context: • Symptom improvement occurred in a significantly greater proportion of patients with PPIs vs. H2RAs. In terms of heartburn remission, PPIs were significantly more efficacious than H2RAs when both half-dose and standard-dose trials were pooled, but the result crossed unity for the subgroup of standard-dose trials alone. • The proportion achieving heartburn relief, calculated by COMPUS staff by pooling response rates across the 4 trials included in the van Pinxteren SR, were 58% vs. 40% in the comparison of standard-dose PPIs vs. H2RAs, and 51% vs. 44% in the comparison of half-dose PPIs vs. H2RAs.

Statement G2.1.1B: PPIs are no more efficacious than H2RAs for improving quality of life in patients with ENRD.

Voting Results: A 9%, B 55%, C 18%, D 18%, E 0%

Summary: One good quality SR44 found that there was no difference in improvement in the reflux dimension of the GSRS, nor in SF-36 scores, between PPIs or H2RAs. These data were based on three RCTs. Two poor quality RCTs46,47 were also identified. Fujiwara et al.46 found that there was no difference in improvement of quality of life between omeprazole 20 mg qd and famotidine 20 mg bid for 4 weeks in patients with ENRD. Similarly, Wada et al.47 reported that omeprazole 20 mg qd and famotidine 20 mg bid given for 8 weeks had similar efficacy in terms of improvement in the quality of life of patients with GERD. Study Population Intervention Comparator Outcome measure Results Type(QA) Van 2 RCTs ome 20 mg qd fam 20 mg GSRS and SF-36 No difference in improvement in Pinxteren et (n=181 with , pant 40 mg bid, niz 150 scales at 1-4 weeks reflux dimension of the GSRS (2 al. 200644 ENRD) for qd mg bid RCTs) or in SF-36 (1 RCT) were GSRS; 1 RCT found for PPIs vs. H2RAs MA (good) (n=98 with ENRD) for SF- 36

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 35 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Fujiwara et 106 patients ome 20 mg qd fam 20 mg HR-QOL HR-QOL al. 200546 with non- for 4 weeks bid for 4 GSRS (15 items GSRS: Both ome and fam erosive GERD weeks combined into 5 improved the GSRS scores from RCT ( no mucosal symptom clusters: baseline. (poor) break by LA reflux, abdominal classification pain, indigestion, ome vs. fam: upon diarrhea and reflux score: 2.36±1.42 vs. endoscopy; constipation) 2.37±1.29 (NS); appearance abdominal pain: 2.02±1.21 vs. classified as SF-36 (8 scales: 2.38±1.22 (NS); normal or physical indigestion: 2.01±1.05 vs. minimally functioning, role- 2.13±0.85 (NS); changed) physical, bodily diarrhea: 1.80±1.12 vs. 1.76±0.95 pain, general health, (NS); constipation: 2.29± 1.21 vs. vitality, social 2.19±1.06 (NS); functioning, role- total score: 2.11±0.91 vs. emotional, mental 2.12±0.73 (NS) health) and physical and mental SF-36: Both ome and fam summary scores at improved the SF-36 scores from 4 weeks baseline.

Ome vs. fam: physical functioning: 80.8±21.7 vs. 78.4±20.9 (NS); role-physical: 69.6±39.4 vs. 69.9±41.6 (NS); bodily pain: 65.8±23.5 vs. 66.0±22.6 (NS); general health: 53.0±19.6 vs. 52.8±17.8 (NS); vitality: 57.4±22.3 vs. 54.8±22.0 (NS); social functioning: 80.7±20.3 vs. 78.9±21.8 (NS); role-emotional: 66.4±43.1 vs. 73.5±39.1 (NS); mental health: 65.2±23.1 vs. 66.5±23.6 (NS) Wada et al. 54 GERD ome 20 mg qd fam 20 mg GSRS (total, reflux, PP analysis at 8 weeks: 200547 patients with for 8 weeks bid for 8 abdominal pain, Ome vs. fam: score >4 on weeks indigestion, RCT QUEST, diarrhea and GSRS: (poor) classified constipation) at 8 total: 1.80±0.54 vs. 2.13±0.78 endoscopically weeks (NS); according to reflux: 1.90±0.71 vs. 2.35±1.54 LA grade SF-36 (8 scales: (NS); physical abdominal pain: 1.82±1.04 vs. Note: 51 functioning, role- 1.86±0.81 (NS); indigestion: patients physical, bodily 1.90±0.60 vs. 2.00±0.94 (NS); completed pain, general health, diarrhea: 1.38±0.70 vs. 2.10± 1.77 study, of these, vitality, social (NS); constipation: 2.02±1.12 vs. 39 were functioning, role- 2.33± 1.41 (NS) endoscopically emotional, mental non-erosive health) at 8 weeks Change from baseline: GERD (grades ome 20 mg improved reflux score A-D) and 12 (p=0.05);

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 36 Consultation Document: Canadian Agency for Drugs and Technologies in Health

with erosive fam 20 mg significantly improved GERD (grade total, reflux, abdominal pain and N) indigestion scores from baseline, (all p<0.05) . SF-36 at 8 weeks: fam vs. ome: physical function: 83.70±24.0 vs. 84.5±16.0 (NS); role physical: 76.7±40.6 vs. 79.4±33.3 (NS); bodily pain: 62.8±27.7 vs. 64.5±29.5 (NS); general health: 56.7±14.0 vs. 47.6±24.5 (NS); vitality: 67.3±23.4 vs. 57.5±21.7 (NS); social functioning: 84.2±30.8 vs. 72.8± 23.9 (NS); role emotional: 75.6±40.8 vs. 68.6±39.9 (NS); mental health: 81.1±16.2. vs. 66.5±21.8 (NS)

Change from baseline: ome 20 mg improved general health, vitality and mental health (all p<0.05); fam 20 mg significantly improved only mental health from baseline (p<0.05). fam: famotidine; GSRS: Gastrointestinal Symptoms Rating Scale; niz: nizatidine; ome: omeprazole; pant: pantoprazole; SF- 36: Short Form- 36;

Context: • The assembled evidence does not show any improvement on general QOL scores. However, the ERP noted that the identified trials did not use GI-specific QOL scores, which may be more discriminatory to differences between treatments than general QOL scores like the SF-36. • There were conflicting opinions in both the literature, and among members of the ERP, as to whether GSRS is a quality of life scale or a symptom scale.

G2: Endoscopic Negative Reflux Disease (ENRD) G2.1: Initial therapy G2.1.2: Half-dose vs. standard-dose PPIs Evidence Statement G2.1.2A: For patients with ENRD, half-dose PPI as initial therapy is less efficacious than standard-dose PPI therapy for heartburn resolution.

Voting Results: A 0%, B 67%, C 25%, D 8%, E 0%

Summary: Six RCTs were identified for this statement. Four RCTs 48-51, three of good quality48-50 and one of poor quality51, only reported p-values for comparisons of standard and half-doses of PPIs versus placebo and not for dose comparisons for most outcomes. However, Carlsson et al.51 did find that the

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 37 Consultation Document: Canadian Agency for Drugs and Technologies in Health

GSRS score was improved to a significantly greater degree in ENRD patients treated with omeprazole 20 mg versus 10 mg. Two RCTs 52,53 of very good and poor quality respectively, reported significant improvement of heartburn symptoms after 4 weeks of therapy with omeprazole 20 mg daily versus omeprazole 10 mg daily in patients with ENRD. However, Richter et al.53 reported no statistically significant differences in the improvement of regurgitation, epigastric pain, dysphagia or nausea with omeprazole 20 mg as compared to 10 mg daily. Study Population Intervention Comparator Outcome measure Results Type(QA) Miner et al. 203 patients with rab 10 mg qd rab 20 mg qd, Heartburn relief All p values are for both 200248 nonerosive GERD as initial or placebo as Satisfactory rab groups vs. placebo, no (grade 0,1 Hetzel- therapy initial therapy heartburn relief: p-values for rab 10 mg RCT (good) Dent scale) (no more than vs. rab 20 mg. one episode of moderately- Rab 10 mg and rab 20 mg severe heartburn vs. placebo: during the At week 4: preceeding 7-day Percentage of heartburn interval) periods: 53% and 47% vs. 23% (p≤0.001). Complete heartburn relief: 29% and 28% vs. 3.4% (p≤0.001) NNT (95% CI) rab 10 mg vs. placebo = 4 (3, 7) NNT (95% CI) rab 20 mg vs. placebo = 4 (3, 8)

Satisfactory heartburn relief: 57% and 57% vs. 32% (p≤ 0.004 and p≤0.001) NNT (95% CI) rab 10 mg vs. placebo = 4 (2, 12) NNT (95% CI) rab 20 mg vs. placebo = 4 (2, 11)

Percentage of time antacid- free or night periods: 76% and 73% vs. 51% (p≤0.001). Mean antacid tablets per day over weeks 1-4: 0.9 ± 0.2 and 1.0 ± 0.2 vs. 2.3 ± 0.2 (p≤0.001). Richter et 214 patients with non- lans 15 mg lans 30 mg qd Heartburn relief lans 15 mg vs. lans 30 mg al. 1999 erosive GERD mg qd for 8 or placebo for and antacids used vs. placebo: 49 (moderate to severe weeks as 8 weeks as Overall 24 h heartburn relief daytime and or initial initial therapy Daytime heartburn relief: RCT (good) nighttime heartburn) therapy on Day 1: 45% and 39% vs. 19% (p<0.01 for both lans groups vs. placebo) NNT (95% CI) = 4 (2, 11) for lans 15mg vs. placebo; NNT (95% CI) = 5 (3, 31) for lans 30 mg vs. placebo

Week 0-4: 82% and 76% vs. 12% (p<0.001 for both

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 38 Consultation Document: Canadian Agency for Drugs and Technologies in Health

lans groups vs. placebo) NNT (95% CI) = 1 (1, 2) for lans 15mg vs. placebo; NNT (95% CI) = 2 (1, 2) for lans 30 mg vs. placebo

Week 4-8: 92% and 94% vs. 16% (p<0.001 for both lans groups vs. placebo). NNT (95% CI) = 1 (1, 2) for lans 15mg vs. placebo; NNT (95% CI) = 1 (1, 1) for lans 30 mg vs. placebo

Nighttime heartburn relief: Day 1: 61% and 51% vs. 39% % (p<0.01 and p< 0.05 for both lans groups vs. placebo) NNT (95% CI) = 5 (2, 25) for lans 15mg vs. placebo; NNT (95% CI) = 8 (3, ∞) for lans 30 mg vs. placebo

Week 0-4: 89% and 73% (p<0.05) vs. 25% (p<0.001 for both lans groups vs. placebo). NNT (95% CI) = 2 (1, 2) for lans 15mg vs. placebo; NNT (95% CI) = 2 (2, 3) for lans 30 mg vs. placebo

Week 4-8: 96% and 92% vs. 50% (p<0.001 for both lans groups vs. placebo). NNT (95% CI) = 2 (2, 3) for lans 15mg vs. placebo; NNT (95% CI) = 2 (2, 3) for lans 30 mg vs. placebo

Antacid used: % day used: Week 0-4: 16% and 27% vs. 79% (p<0.001 for both lans groups vs. placebo). Week 4-8: 11% and 9% vs. 64% (p<0.001 for both lans groups vs. placebo). Average number antacid used per day: Week 0-4: 0.32 tablets and 0.50 tablets vs. 2.54 tablets (p<0.001 for both lans groups vs. placebo) Week 4-8: 0.22 tablets and 0.17 tablets vs. 1.70 tablets

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 39 Consultation Document: Canadian Agency for Drugs and Technologies in Health

(p<0.001 for both lans groups vs. placebo) Katz et al. Patients (18-75 years) esome 20 mg esome 40 mg Proportion of At 4 weeks: 2003*50 with at least 6 months qd x 4 weeks qd x 4 weeks patients with Proportion with complete history of heartburn as initial or complete heartburn resolution: esome RCT (good) and no evidence of therapy Placebo qd x 4 heartburn 20 mg vs. esome 40 mg erosive esophagitis weeks resolution. vs. placebo: upon endoscopy As initial referred to as non- therapy Percentage of Study 1: 41.8% vs. 35% vs. erosive reflux disease heartburn-free 13.7%; days and nights. NNT (95% CI) esome 20 mg Study 1: n=368 and vs. placebo = 4 (3, 6) Study 2: n=349) NNT (95% CI) esome 40 mg (Study 1 and 2 were vs. placebo = 5 (3, 9) of identical design) Study 2: 32% vs. 35% vs. 11.9% NNT (95% CI) esome 20 mg vs. placebo = 5 (3, 10) NNT (95% CI) esome 40 mg vs. placebo = 4 (3, 8)

(p<0.001 for both esome groups vs. placebo, no p- value for esome 20 mg vs. 40 mg)

Mean % of heartburn-free days: esome 20 mg vs. esome 40 mg vs. placebo Study 1: 63% vs. 63% vs. 46% Study 2: 68% vs. 66% vs. 36% (p<0.001 for both esome groups vs. placebo, no p- value for esome 20 mg vs. 40 mg)

Mean % of heartburn-free nights: esome 20 mg vs. esome 40 mg vs. placebo: Study 1: 87.7% vs. 87.7 vs. 78.7; Study 2: 89.9% vs. 88.5% vs. 79.5; (p<0.006 for both esome groups vs. placebo, no p- value for esome 20 mg vs. 40 mg) Lind et al. 509 patients with ome 10 mg ome 20 mg qd Heartburn At 4 weeks APT and per 1997*52 heartburn (all grade, qd for 4 wks or placebo for resolution (no protocol analysis: mild, moderate, as initial 4 wks as initial more than 1 day Complete absence of RCT (very severe) without therapy therapy with mild episode heartburn: 31% (ome 10 mg) good) endoscopic of heartburn vs. 46% (ome 20 mg)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 40 Consultation Document: Canadian Agency for Drugs and Technologies in Health

esophagitis during the past 7 (p<0.002) vs. 13% (placebo) days) and (p<0.0001 both doses of ome complete absence vs. placebo) of heartburn (no NNT (95% CI) ome 10 mg episodes of vs. placebo = 6 (4, 13) heartburn during NNT (95% CI) ome 20 mg the 7 days before vs. placebo = 3 (2, 4) the clinic visit) at NNT (95% CI) ome 20 mg 4 weeks vs. ome 10 mg = 7 (4, 18)

Heartburn resolution: 49% (ome 10 mg) vs. 61% (ome 20 mg) (p<0.002) vs. 24% (placebo) (p<0.0001 both does of ome vs. placebo) NNT (95% CI) ome 10 mg vs. placebo = 4 (3, 7) NNT (95% CI) ome 20 mg vs. placebo = 3 (2, 4) NNT (95% CI) ome 20 mg vs. ome 10 mg = 8 (5, 42)

Richter et 359 patients with ome 10 mg ome 20 mg qd, Complete Complete resolution of al. 2000*53 heartburn (moderate qd for 4 wks or placebo for resolution of heartburn: to severe) without as initial 4 wks as initial heartburn 2-4 Week 2: 20% (ome 10 mg) RCT (poor) endoscopic therapy therapy week and relief of vs. 41% (ome 20 mg) esophagitis other symptoms (p<0.002) and 5% (placebo) of GERD (p<0.001). NNT (95% CI) = 7 (4, 14) for ome 10 mg vs. placebo; NNT (95% CI) = 3 (2, 4) for ome 20 mg vs. placebo; NNT (95% CI) = 5 (3, 10) for ome 20 mg vs. ome 10 mg

Weeks 4: 27% (ome 10 mg) vs. 48% (ome 20 mg) (p<0.002) and 5% (placebo) (p<0.001). NNT (95% CI) = 5 (3, 8) for ome 10 mg vs. placebo; NNT (95% CI) = 2 (2, 3) for ome 20 mg vs. placebo; NNT (95% CI) = 5 (3, 11) for ome 20 mg vs. ome 10 mg

All p values below are for both ome groups vs. placebo, no p-values for ome 10 mg vs. ome 20 mg.

4 weeks: proportion of patients reporting no symptoms of acid

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 41 Consultation Document: Canadian Agency for Drugs and Technologies in Health

regurgitation: 60% (ome 10 mg) and 68% (ome 20 mg) vs. 43% (placebo) (p ≤ 0.003 and p<0.001). Epigastric pain: 58% (ome 10 mg) and 73% (ome 20 mg) vs. 43% (placebo) (p ≤ 0.003 and p<0.001). Dysphagia: 89% (ome 10 mg) and 91% (ome 20 mg) vs. 72% (placebo) (p ≤ 0.003 and p<0.001). Nausea: 84% (ome 10 mg) and 86% (ome 20 mg) vs. 66% (placebo) (p ≤ 0.003 and p<0.001). Carlsson et 261 patients with ome 10 mg ome 20 mg qd Symptoms relief Proportion with symptom al. 1998 51 ENRD qd for 4 or placebo for at 4 weeks relief at 4 weeks : weeks as 4 weeks as (no p values provided for RCT (poor) initial initial therapy any outcome) therapy Ome 10 mg vs. ome 20 mg vs. placebo: Relief of heartburn: : 58% vs. 56% vs. 23% Relief of regurgitation: 53% vs. 64% vs. 34% Relief of epigastric pain: 43% vs. 46% vs. 35% Complete relief: 31% vs. 29% vs. 19% Sufficient control of symptoms: 37% vs. 48% vs. 35% GSRS score: Ome 10 mg vs. ome 20 mg vs. placebo: Baseline: 2.80 for all 4 weeks: 1.75 and 1.80 vs. 2.25 (p< 0.01 for both ome 10 vs. placebo and ome 20 vs. placebo, no p value reported for ome 10 vs. ome 20) Esome: esomeprazole; lans: lansoprazole; ome: omeprazole; rab: rabeprazole; * indicates industry involvement

Context: • There is relatively scant evidence that half-dose PPIs are inferior to full-dose over the long-term. Only the Richter et al.53 and Lind et al.52 RCTs, both of which studied omeprazole, reported p- values for the comparison of half-dose and standard-dose PPIs. • Some differences between doses are not likely to be clinically significant. In other cases, the difference in heartburn resolution rates between standard-dose and half-dose PPIs is more substantial (e.g., ~20% in Richter et al.53). • There appeared to be less of a difference between standard and half-doses of rabeprazole and esomeprazole.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 42 Consultation Document: Canadian Agency for Drugs and Technologies in Health

G2: Endoscopic Negative Reflux Disease (ENRD) G2.2: Maintenance therapy G2.2.1: Long-term therapy with PPIs Evidence Statement G2.2.1A: Half-dose, continuous PPI therapy in ENRD is more efficacious than placebo for control of symptoms for up to 6 months.

Voting Results: A 0%, B 50%, C 25%, D 25%, E 0%

Summary: Venables et al.54, a poor quality RCT, assessed continuous half-dose omeprazole versus placebo and found that active treatment was more effective than placebo for heartburn relief at 8 and 16 weeks but not at 24 weeks, quality of life was better in the omeprazole group at 6 months, and acid regurgitation was less frequent in the omeprazole group at 8 weeks but not at 16 and 24 weeks. Study Population Intervention Comparator Outcome measure Results Type(QA) Venables et 495 patients with ome 10 mg Matching % of patients who % of patients who al. (1997) 54 GERD, but without daily for 6 placebo qd for discontinue discontinue therapy before RCT (poor) erosive esophagitis at mo 6 mo therapy on the 6 months because of endoscopy, who had basis of life table inadequate relief of obtained successful estimates for heartburn or any other control of heartburn cumulative group reason (including side after 4-8 weeks of relapse rate effects) initial therapy Symptom relapse Ome vs. placebo

Quality of life 27% vs. 52% (P<0.0001) (GSRS score) Symptom relapse

1) % patients with heartburn symptom

Ome vs. placebo

at 8 weeks 47% vs. 60%, p<0.01 at 16 weeks 37% vs. 56%, p<0.001 at 24 weeks 32% vs. 34 %, p>0.05 (NS)

2). % patients with acid regurgitation symptom Ome vs. placebo at 8 weeks 22% vs. 38% ( P<0.001) at 16 weeks 19% vs. 21% (NS, no p value provided) at 24 weeks 14% vs. 19% (NS, no p value provided)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 43 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Quality of life (GSRS score)

mean GSRS score reduction ( from randomization to completion of the study)

Ome vs. placebo

0.85 vs. 0.95 (P<0.05)

Esome: esomeprazole; lans: lansoprazole; ome: omeprazole; rab: rabeprazole

Context: • At 6 months, PPI was superior to placebo only in terms of QOL. There were no significant differences in terms of heartburn or regurgitation. Furthermore, although the differences were statistically significant at 8 and 16 weeks, their clinical relevance is uncertain. • In light of the chronic nature of ENRD, the ERP expressed reservation about the utility of the evidence given the 6-month duration of the trial. There was also concern regarding the ‘poor’ quality rating of the study.

G2: Endoscopic Negative Reflux Disease (ENRD) G2.2: Maintenance therapy G2.2.2: On-demand PPIs after response to initial PPI therapy Evidence Statement G2.2.2A: For patients with ENRD who respond to initial PPI therapy, subsequent ‘on-demand’ PPI therapy is more efficacious than placebo.

Voting Results: A 25%, B 75%, C 0%, D 0%, E 0%

Summary: The evidence for this Statement consisted of one poor quality SR55 and three poor quality RCTs56-58. The SR55 showed that on-demand PPIs were more effective than placebo in terms of patients’ willingness to continue and number of antacids consumed in patients with ENRD over 6 months. Similarly, in their RCT Bytzer et al. 58 found that on-demand PPIs were more effective than placebo for controlling symptoms, decreasing consumption of antacids and improving quality of life (PGWB) in patients with ENRD over 6 months. Two further RCTs56,57 had similar findings, and also reported greater patient satisfaction with treatment.

Study Outcome Population Intervention Comparator Results Type(QA) measure Zacny et al. 5 studies esome 20 mg Placebo for 6 Willingness to Willingness to continue PPI: 83% 200555 (n=2,322) in and 40 mg, months continue, with ome 20 mg, 70% with ome 10 patients with ome 10 mg number of mg vs. 56% with placebo (p< 0.01); SR (poor) ENRD and 20 mg, antacids 85% with esome vs. 48% with rab 10 mg, consumed placebo (p<0.0001); 93% and 91% lans 30 mg, with esome 40 gm and 20 mg vs. all ‘on- 60% with placebo (p<0.05) (meta demand’ analysis not performed). for 6 months Number of antacids consumed per

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 44 Consultation Document: Canadian Agency for Drugs and Technologies in Health

day: 0.77 to 0.91 with ome 20 mg and 10 mg vs. 1.15 with placebo (NS); 0.39 with esome 20 mg vs. 1.06 with placebo (NS); 0.39 with esome vs. 1.06 with placebo; 0.48 with esome 40 mg, 0.44 with esome 20 mg vs. 1.07 with placebo (p values not provided). Kaspari et 536 patients (18- pant 20 mg Placebo qd Perceived At 6 months: al. 75 yrs) with qd on on demand average daily PADSL: pant vs. placebo: 1.5 vs. 2005*56 endoscopically demand for 6 for 6 months symptom load 2.2, p<0.05. Difference = 0.16 (95% confirmed non- months (PADSL) CI: -0.34, -0.01). RCT (poor) erosive GERD including (grades 0-I, heartburn, acid Proportion unwilling to continue Savary Miller) regurgitation because of: and pain upon a) Insufficient symptom control of swallowing symptoms (pant vs. placebo): 2.9% was assessed vs. 10.1%, Difference = 7.2% (95% by patient CI: -11.1%, -3.3%) diary; patient b) Unsatisfactory treatment (pant vs. unwillingness placebo): 3.4% vs. 11.4%, to continue Difference = -8.0% (95% CI: - 12.0%, -3.9%). c) Any reason (pant vs. placebo): 5.3% vs. 14.4%, Difference = -9.1% (95% CI: -13.8%, -4.5%) Scholten et 548 patients with pant 40 mg placebo on- Perceived Perceived average daily symptom al. 2005*57 non-erosive or and 20 mg demand for average daily load: pant 40 mg 2.71 vs. placebo mild GERD and qd on- 24 wks symptom load; 3.93, p<0.0001; RCT (poor) heartburn, relieved demand for unwillingness pant 20 mg 2.91 vs. placebo 3.93, of heartburn after 24 wks to continue due p<0.0001; 4 weeks of PPI to inadequate pant 40 mg vs. pant 20 mg, NS treatment heartburn control; Proportion unwilling to continue: discontinuation pant 40 mg 0.9% vs. placebo 10.9%, of tablet intake; p<0.05; time since last pant 20 mg 2.8% vs. placebo 10.9%, intake of p<0.05; medication pant 40 mg vs. pant 20 mg, NS at 24 weeks Proportion discontinuing tablet intake: pant 40 mg 6.4% vs. pant 20 mg 10.1% vs. placebo 24.1%, p value not provided

Tablet intake average/day: Antacids (Mean ± SD): pant 40 mg (0.33 ± 0.52) vs. pant 20 mg (0.45 ± 0.79) vs. placebo (0.68 ± 0.77), p value not provided

Mean duration (±SD) of treatment (days) (PP analysis): pant 40 mg (149.5 ± 38.9) vs. pant 20 mg (140.4 ± 47.4) vs. placebo (127.8 ± 52.4), p=0.0005

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 45 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Bytzer et al. 535 patients with rab 10 mg qd Placebo qd Proportion of At 6 months: 2004*58 history of reflux on demand x on demand x patients Proportion of patients d/c therapy symptoms and 6 months. 6 months. discontinuing due to inadequate heartburn control: RCT (poor) negative treatment due rab 6% vs. placebo 20%, p<0.00001. endoscopy to inadequate NNT (95% CI) = 7 (5, 13) heartburn control. Mean change in symptom severity from baseline: Antacids rab 0.7 vs. placebo 1.0, p<0.05. consumed per week. Proportion with sufficient heartburn control: rab 86.4% vs. placebo QOL by 67.6%, p=0.00002. PGWBI. NNT (95% CI) = 5 (4, 9)

Maximum duration of symptoms (days): rab 6.7 vs. Placebo 7.5, p=0.0256.

Proportion with maximum symptom episode duration of < 2 days: rab 30% vs. Placebo 18%, p=0.0106 and < 4 days: rab 59% vs. Placebo 45%, p=0.0096.

Mean Antacid consumption per week: rab 2.0 vs. placebo 3.9,p = 0.0009 PGWBI: In patients completing 6 months, transformed scores = rab 72.8 vs. placebo 73.2, (p<0.0001). In patients who discontinued in the placebo group, decrease in score (- 12.9) was significant (p<0.0001) but not in the rab group (-7.4); NS. esome: esomeprazole; lans: lansoprazole; ome: omeprazole; pant: pantoprazole; rab: rabeprazole; * indicates industry involvement

Context: • The included studies used both half-dose and standard-dose PPI regimens. • A number of the trials in the Zacny et al. SR55 were single-blinded. • The Bytzer et al. trial was included in the Zacny et al. SR, however, it is presented in the evidence table because several trial outcomes other than those captured by Zacny et al. were reported58. • The value of some clinical outcomes reported in the included studies is uncertain. • A large proportion of subjects, from about 50-90%, were willing to continue on on-demand placebo for up to 6 months. The difference in satisfaction between PPI and placebo appeared to be small.

G3: Erosive Esophagitis G3.1: Initial therapy G3.1.1: PPIs vs. H2RAs

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 46 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statement G3.1.1A: PPIs are more efficacious than H2RAs for improving symptoms and improving the healing of erosive esophagitis.

Voting Results: A 92%, B 8%, C 0%, D 0%, E 0%

Summary: The evidence for this statement consisted of 2 poor quality SRs60,61, and 5 RCTs62-66, one of which was of very good quality66 and one of good quality62. All of these studies reported superior heartburn relief and endoscopic remission rates with PPIs as compared to H2RAs at various time points from 4 weeks to 6 months. The ERP also considered an as yet unpublished Cochrane review in its deliberations. The findings of this study were supportive of the Evidence Statement. Study Outcome Population Intervention Comparator Results Type(QA) measure Carlsson et 4 trials, 1154 ome 10 and 20 ran 150 mg bid Endoscopic Endoscopic remission : al. 1997*60 patients with mg qd and placebo remission at 6 82.4% (95%CI: 78.2%, 86.6%) erosive months with ome 20 mg vs. 72% SR (poor) esophagitis (95%CI: 65.5%, 78.3%) with ome 10 mg vs. 52.3% (95%CI: 44.4%, 60.1%) with ran 150 mg vs. 10.6% (95 CI: 5.0%, 16.3%) with placebo.

Time to remission: ome 20 mg was superior to ome 10 mg (p=0.04) and ran (p<0.0001) and placebo (p<0.0001). Ome 10 mg is superior to ran (p<0.0001). Chiba et al. 2,123 patients lans 30, 60 mg cim 800-1600 Heartburn relief Proportion heartburn-free (±SD): 199761 with erosive qd, ome 20, 40 mg/day, fam 40 ≤12 week 77±10.4% with PPIs vs. esophagitis mg qd, pant 40 or 80 mg/day, 48±15.5% with H2RAs SR (poor) (grade II-IV), mg qd niz 300-600 Healing (p<0.0001). 16 studies mg/day, ran proportion 1- NNT = 4 300-1200 12 weeks mg/day Healing proportion (±SD): 84±11.4% with PPIs vs. 52±17.1% with H2RAs (p< 0.0005) NNT (95% CI) = 4 (3, 4) Koop et al. 249 patients pant 40 mg qd for ran 150 mg bid Symptom relief Symptom relief: 1995*62 with acute 4-8 weeks for 4-8 weeks at 2 and 4 wk 2: 46 % (95% CI: 38%,54%) symptomatic weeks. with pant vs. 37 % (95% CI: RCT (good) reflux 65%,80%) with ran (NS) esophagitis Healing rate at (grade II and 4 and 8 weeks wk 4: 72% (95% CI: 65%,80% III, Savary- with pant vs. 52% (95% CI: Miller 40%,63%) with ran (p<0.01). classifiication) Healing rates (pant vs. ran): wk 4 : 62% (95%CI: 55%, 70%) (pant 40 mg) vs. 47% (95%CI: 36%, 58%) (ran), (p<0.05)

wk 8 : 74% (95%CI: 67%, 80%)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 47 Consultation Document: Canadian Agency for Drugs and Technologies in Health

(pant 40 mg) vs. 55% (95%CI: 45%, 66%) (ran), (p<0.01) . Robinson et 184 patients ome 20 mg qd for ran 150 mg bid Heartburn and Improvement in night time al. 1993*63 with erosive 8 weeks + metoclopram- symptom relief heartburn wk 4: 64% with ome esophagitis ide 10 mg qid 1-8 weeks 20 mg vs. 25.1% in the RCT (poor) (grade II-IV) combination for Regurgitation combination group (p<0.01). 8 weeks NNT (95% CI) = 3 (2, 4)

wk 8: 73% with ome vs. 52% with the combination group (p<0.01). NNT (95% CI) = 5 (3, 14)

Night-time and day-time heartburn relief Over the 8 week, the night-time and day-time heartburn relief were better with ome than with ran and metoclopramide combination (p<0.01).

Complete relief of acid regurgitation wk 4: 68% with ome vs. 35% in combination (p<0.01). wk 8: 78% with ome vs. 57% with combination (p<0.01). Frame 172 patients ome 20 mg qd for ran 150 mg bid Healing rates at wk 4: 72% with ome vs. 54% 1991*64 with erosive or 4-8 weeks for 4-8 weeks 4 and 8 weeks with ran, Diff (95%CI): 18% ulcerative (1%, 34%), p=0.042. RCT (poor) esophagitis (grade 2-3) wk 8: 90% with ome vs. 76% with ran, Diff (95%CI): 14% (2%, 27%), p=0.03. Bate et al. 272 patients ome 20 mg qd for cim 400 mg qid Symptom relief Symptom relief: 1990*65 with reflux 4 weeks for 4 weeks at 4 weeks. at 4 weeks: esophagitis 46% with ome vs. 22% with cim RCT (poor) (grade I-IV). Healing (p<0.001). (endoscopic and/or Endoscopic healing: histological) wk 4: 56% with ome vs. 26% proportions at 4 with cim (p<0.001) weeks and 8 wk 8: 71% with ome vs. 35% weeks with cim (p<0.001) Patients both asymptomatic Both asymptomatic and healed of and healed of esophagitis: 38% of patients with esophagitis ome vs. 12% with cim were assessed by (p<0.001). endoscopy and/or Abnormal histology: histology. About 60% of the patients had at entry. At 8 weeks, 33% of patients taking ome vs. 52% with cim continued to have abnormal histology (p<0.001).

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 48 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Endoscopically/histologically healed and symptoms relieved: 45% with ome vs. 22% with cim (p<0.01) Zeitoun et 156 patients ome 20 mg for 4- ran 150 mg bid Healing rates at At 29+/- 6 days: 74% with ome al. 1989*66 with erosive or 8 weeks for 4-8 weeks 4 and 8 weeks vs. 41% with ran (p<0.001) ulcerative RCT (very esophagitis At 57+/- 6 days: 87% with ome good) (grade 2-4) vs. 56% with ran (p<0.001) cim: cimetidine; fam: famotidine; lans: lansoprazole; pant: pantoprazole; niz: nizatidine; rab: rabeprazole; ran: ranitidine, ome; omeprazole; * indicates industry involvement

Context: • The evidence for this statement consisted of a number of short-term (i.e., 4-12 week) studies, and a single SR that assessed healing at 6 months60. • In the short-term studies, symptomatic response rates to H2RAs ranged from approximately 22- 52%, while those of PPIs ranged from 46-78%. In terms of healing rates in short term studies, 56-90% of PPI treated subjects experienced healing of esophagitis, as compared to 26-76% with H2RAs. The SR on long-term therapy reported healing in 82% of those given standard-dose PPI versus 52% with standard-dose H2RA at 6 months.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 49 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statement G3.1.1B: The speed of improvement of heartburn and healing is faster with PPIs than H2RAs in patients with erosive esophagitis.

Voting Results: A 33%, B 77%, C 0%, D 0%, E 0%

Summary: PPIs produce faster relief of heartburn and improvement of healing than H2RAs in patients with erosive or reflux esophagitis. The data are based on a poor quality SR61 and a good quality RCT67. Study Outcome Population Intervention Comparator Results Type (QA) measure Chiba et al. 2,123 patients lans 30-60 mg H2RAs, cim The rate of Rate of heartburn improvement 199761 with erosive qd, ome 20-40 800-1600 heartburn (±SD):: (12±1%/week) with PPIs vs. esophagitis mg qd, pant 40 mg/day, fam 40 improvement. (6±1%/week) with H2RAs (p<0.05). SR (poor) (grade II-IV), mg qd for 12 or 80 mg/day, 16 studies. weeks niz 300-600 Speed of Healing rates (±SD):: mg/day, ran healing/week wk 2: 32±3%/week with PPIs vs. 300-1200 mg 1- 12 weeks; 15±6%/week with H2RAs /day for 12 overall healing wk 4: 17±3%/week with PPIs vs. weeks rates 9±3%/week with H2RAs wk 6: 15±1%/week with PPIs vs. 6±2%/week with H2RAs wk 8: 11±1%/week with PPIs vs. 7±2%/week with H2RAs. wk 12: 8±0.1%/week with PPIs vs. 5±1%/week with H2RAs

Overall healing rates: 12±1%/week with PPIs vs. 6±0.2%/week with H2RAs The r2 for regression slopes (0.91 for PPIs vs. 0.89 for H2RAs) (p<0.0001). Green et al. 198 patients ome 20 mg qd, ran 150 mg bid Median time to 7 days with ome vs. 19 days with ran 1995*67 with reflux (or 40 mg qd if report no (p<0.001). esophagitis needed) heartburn and RCT (good) (1-IV) to consume no antacid cim: cimetidine; lans: lansoprazole; niz: nizatidine; ome: omeprazole, pant; pantoprazole, fam; famotidine; ran: ranitidine; * indicates industry involvement

Context: • Subjects with various degrees of esophagitis severity were included in trials. The included studies did not stratify results based on grade of esophagitis, and the ERP noted that results may vary in lower versus higher grades of esophagitis. The Panel therefore expressed reservation regarding the applicability of this Statement to a given patient with esophagitis.

G3: Erosive Esophagitis G3.1: Initial therapy G3.1.2 : Double-dose vs. standard-dose PPIs

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 50 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statement G3.1.2A: Doubling the standard daily doses of PPIs, as initial therapy, is no better than standard daily dose PPI therapy for improvement of erosive esophagitis.

Voting Results: A 50%, B 42%, C 8%, D 0%, E 0%

Summary: The evidence consists of 6 RCTs68-73, two of which were of very good quality68,69, and one of good quality70. Three good quality RCTs68-70and two poor quality RCTs71,72 reported no benefit of using double dose over standard-dose PPIs for healing in patients with reflux or erosive esophagitis. However, one poor quality RCT73 reported better healing with the double dose than the standard-dose of PPI in patients with reflux esophagitis. The ERP also considered an as yet unpublished Cochrane review in its deliberations. The findings of this study were supportive of the Evidence Statement. Study Outcome Population Intervention Comparator Results Type(QA) measure van 192 patients pant 40 mg pant 80 mg Healing at 4 wk 4: 69% (pant 40 mg) vs. 66% (pant 80 Resenburg with reflux qd for 8 qd for 8 and 8 weeks mg) et al. esophagitis weeks weeks wk 8 : 85% (pant 40 mg) vs. 86% (pant 80 1996*68 mg)

RCT (very good) Bardhan et 229 patients lans 30 mg ran 150 mg Healing rate at wk 4: 84% (lans 30 mg) vs. 72% (lans 60 al. 1995*69 with qd or lans 60 bid for 4-8 4 and 8 weeks mg) endoscopically mg qd for 4- weeks wk 8: 92% (lans 30 mg) vs. 90% (lans 60 RCT (very confirmed 8 weeks mg) good) reflux The difference between lans 30 mg and lans esophagitis 60 mg rates was NS. (grade 1-3) Sontag et 230 patients ome 20 mg Placebo for Healing rates Healing rates: al. 1992*70 with qd or ome 40 4-8 weeks at 4 and 8 wk 4: 44% (ome 40 mg) vs. 38% (ome 20 symptomatic mg qd for 4- weeks. mg) RCT erosive 8 weeks Symptom wk 8: 72.5% (ome 40 mg) vs. 73% (ome 20 (good) esophagitis relief. mg) (grade ≥ 2) There were no significant differences in healing rates between the two dose groups when adjusting for baseline severity of esophagitis.

Heartburn relief: 82.1% (ome 40 mg) vs. 80% (ome 20 mg). Hetzel et 132 patients of ome 20 mg ome 40 mg Healing rates Grade II: al. 1988*71 severe peptic qd for 4 qd for 4 at 4 weeks 87% (ome 20 mg) vs. 97% (ome 40 mg) esophagitis weeks weeks Grade III: RCT (grade 2-4) 67% (ome 20 mg) vs. 88% (ome 40 mg) (poor) Grade IV: 48% (ome 20 mg) vs. 44% (ome 40 mg) Earnest et 292 patients lans 15mg Placebo for Healing rates wk 4: 73% (lans 30 mg) vs. 76% (lans 60 al. 1998*72 with reflux qd, 30 mg qd 4-8 weeks at 4, 6 and 8 mg) esophagitis or 60 mg qd weeks wk 6: 87% (lans 30 mg) vs. 86% (lans 60 RCT (grade ≥ 2) for 4-8 mg) (poor) weeks wk 8: 87% (lans 30 mg) vs. 89% (lans 60 mg) Bate et al. 313 patients ome 20 mg ome 40 mg Healing rates 45% (ome 20 mg) vs. 64% (ome 40 mg) 1993*73 with reflux qd for 8 qd for 8 at 8 weeks (p<0.02) esophagitis weeks weeks

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 51 Consultation Document: Canadian Agency for Drugs and Technologies in Health

RCT (grade 2-4) (poor) Esome: esomeprazole; ome: omeprazole; pant: pantoprazole; lans: lansoprazole; ran: ranitidine; * indicates industry involvement

Context: • The ERP noted there is some evidence that double dose esomeprazole may be superior to standard doses of esomeprazole and other PPIs for esophagitis healing74,75, although the observed differences may not be clinically important. The Panel recognized that the recommended dose of esomeprazole for esophagitis in Canada is 40 mg/day, and that the 20 mg and 40 mg strengths are of equal cost. • The Panel expressed some reservation regarding the relatively small number of RCTs that specifically addressed esophagitis of Grades 2-4, and the ‘poor’ quality ratings of these trials.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 52 Consultation Document: Canadian Agency for Drugs and Technologies in Health

G3: Erosive Esophagitis G3.2: Maintenance therapy G3.2.1: Long-term therapy with PPIs vs. H2RAs or placebo Evidence Statement G3.2.1A: Long-term maintenance PPI therapy (i.e., up to 12 months) in erosive esophagitis is more efficacious than placebo for prevention of symptomatic and endoscopic relapse.

Voting Results: A 92%, B 8%, C 0%, D 0%, E 0%

Summary: The evidence identified consists of one good quality SR76 and two poor quality RCTs77,78. Donnellan et al.76 reported that a maintenance dose (half-dose) of PPI was more effective in maintaining healed esophagitis and symptom relief of GERD as compared to placebo over 26-52 weeks of follow-up. Pilotto et al.77 demonstrated that pantoprazole 20 mg for 6 months was better than placebo for healing and symptom relief in patients >65 yrs with erosive esophagitis. Sontag et al. 78 reported that omeprazole 20 mg daily was more effective than placebo for maintaining remission in patients with healed erosive esophagitis of at least grade 2. Study Population Intervention Comparator Outcome measure Results Donnellan Patients with PPIs (ome 10 Placebo for 26- Relapse of healed Proportion with relapse of et al. esophagitis; mg, lans 15 52 weeks esophagitis esophagitis (PPI vs. 2004*76 10 RCTs (n=1465 mg, pant 20 and GERD placebo): 36.1% vs. 75.4% for maintenance of mg, rab 10 mg, symptoms (no p-value given) SR (good) healing, n=1426 for esome 10 mg) between 26-52 symptom relief) qd for 26-52 weeks Proportion with significant weeks symptoms (PPI vs. placebo): 44.4% vs. 73.4%

RR of maintaining symptomatic remission (maintenance dose vs. placebo) = 0.54 (95% CI 0.42-0.69), NNT (95% CI) = 3 (2.2, 4.8) Pilotto et al. 164 patients (> 65 pant 20 mg qd Placebo qd for Healing and Healing rates with pant: 2003 77 yrs) with acute for 6 months 6 months symptoms up to At 8 weeks: 81.1% (95% CI: oesophagitis 12 months 75.1, 87.1) RCT (grades I-III by At 6 months: 82% (95% CI: (poor) Savary-Miller) 75.4-88.5) Pant vs. placebo Patients treated At 12 months,: with pant 40 mg for Healing: 79.6% (95% CI: 8 weeks; if healed 68.3, 90.9%) vs. 30.4% (95% of erosive CI: 18.3, 42.4%), p=0.0001 oesophagitis, treated with pant 20 Reflux symptoms: 22.4% vs. mg qd for 6 67.7% p<0.0001; months. Cured Heartburn: 8.3% vs. 54.8%, patients were p<0.0001; randomized to pant NNT (95% CI) = 2 (2, 3) 20 mg qd or placebo for 6 Acid regurgitation 12.5% vs. months 41.9%, p=0.006 Note: 105 (ITT) for double-blind phase

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 53 Consultation Document: Canadian Agency for Drugs and Technologies in Health

(12 month results) Sontag et al. Phase I: 472 ome 20 mg qd Placebo for 6 Endoscopic Proportion with remission: 1997* 78 patients with active for 3 months remission at 6 70% with ome 20 mg qd vs. erosive esophagitis consecutive months 34% with ome 3 days/week RCT (poor) (grade ≥2) healed days/week for ome 20 mg qd vs. 11% with placebo, with ome 40 mg qd 6 months for 6 months p<0.001 for ome arms vs. x 4-8 weeks (data placebo, p<0.001 for ome not presented here) daily vs. 3 days/week

Phase II: 406 patients healed in phase I were enrolled in phase II (maintenance study) Esome: esomeprazole; lans: lansoprazole; ome: omeprazole; pant: pantoprazole; rab: rabeprazole; * indicates industry involvement.

Context: • All except one study78 used half-dose PPIs.

Evidence Statement G3.2.1B: Long-term maintenance PPI therapy (i.e., up to 12 months) in erosive esophagitis is more efficacious than H2RAs for prevention of endoscopic relapse.

Voting Results: A 22%, B 67%, C 11%, D 0%, E 0%

Summary: The evidence identified for this Statement consists of one good quality RCT79. Lundell et al. compared omeprazole 40 mg daily with ranitidine 300 mg twice daily and found that the former was more effective in producing remission at 12 months in erosive or ulcerative eophagitis of at least grade 2 that was resistant to H2RAs. Study Population Intervention Comparator Outcome measure Results Lundell et 98 patients with ome 40 mg qd ran 300 mg bid Endoscopic Proportion with remission: al. 1991*79 erosive and /or for 12 months for 12 months remission of 67% of patients taking ome ulcerative esophagitis after vs. 10% taking ran at 12 RCT (good) esophagitis (grade 12 months months (p<0.0001) ≥ 2) resistant to H2RAs Ome: omeprazole; ran: ranitidine; * indicates industry involvement

Context: • The ERP noted the following limitations regarding the evidence in support of this Statement: single trial with small sample size (n=92), study population that was resistant to H2RAs, use of double-dose H2RA, and lack of data on symptom control.

Evidence Statement G3.2.1C: Long-term PPI therapy is more efficacious than H2RAs for erosive esophagitis complicated by strictures.

Voting Results:

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 54 Consultation Document: Canadian Agency for Drugs and Technologies in Health

A 8%, B 67%, C 25%, D 0%, E 0%

Summary: The evidence consists of 3 RCTs, one of good quality80, and two of poor quality81,82. Smith et al.81 and Marks et al.82 showed that long term treatment (6-12 months) with PPIs (or H2RAs) was effective in reducing the need for redilatation and symptom relief. They also reported that long-term treatment with PPIs was superior to ranitidine for relieving dysphagia and reducing the need for redilatation in patients with esophageal strictures. However, Marks et al.82 found no significant difference between omeprazole and ranitidine at 3 months in such patients. In addition, Swarbrick et al.80 did not find a significant difference between PPIs and double-dose ranitidine in patients with esophageal stricture. Study Outcome Population Intervention Comparator Results Type(QA) measure Smith et 366 patients with ome 20 mg ran 150 mg Stricture Patients requiring redilatation: al. esophageal qd for 1 year bid for 1 recurrence 30% in ome arm vs. 46% in ran arm, 1994*81 strictures (grade year (redilat- (p<0.01) 1-2), esophagitis ation), RCT (grade 0-4), symptom Patients requiring redilatation: (poor) mean age 71 occurrence, 0.48 in ome arm vs. 1.08 in ran arm, years adverse (p<0.01) events at 1 year Symptoms relief: No dysphagia: 76% in ome arm vs. 64% in ran arm (p<0.05)

Able to eat normal diet: 83% in ome arm vs. 69% in ran arm (p< 0.01)

Asymptomatic: 65% in ome arm vs. 43% in ran arm (p<0.001)). Marks et 37 patients with ome 20 mg ran 150 mg Esophagitis Healing: al. esophageal qd for 6 mos bid or fam healing; At 3 months, 61% in ome arm and 47% in 1994*82 strictures 20 mg bid dysphagic ran arm (NS). (stricture 1, 2), for 6 mos relief; At 6 months, 100% of patients in ome arm RCT esophagitis Need for and 53% in ran arm were healed (p<0.01). (poor) (grade ≥ 2) age dilatation at 57-64 years. 3 and 6 Patients requiring dilatation: months At 6 months: 41% with ome vs. 73% with ran (p=0.07)

Dysphagia relief: At 3 months: 50% with ome vs.33% with ran (p = 0.34). At 6 months: 94% with ome vs.40% with ran (p<0.01). Swarbrick 158 patients of lans 30 mg ran 300 mg Time to Time to redilatation and probability of no et al. esophageal qd for 1 year bid for 1 redilatation; redilatations: These were higher in lans 1996*80 strictures (grade year proportion of group than ran group (P = 0.053); 1-4), esophagitis pts needing RCT (present or redilatation; Redilatation: (good) absent, but not number of At 12 months, fewer redilatations 30.8% in graded, present redilatations. lans group vs. 43.8% in ran group (NS). in 80% of dysphagia

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 55 Consultation Document: Canadian Agency for Drugs and Technologies in Health

patients) (mean relief; # of redilatation: age 68 years) reduction in No significant difference between groups in stricture # of redilatations. grade presence of Dysphagia grade: esophagitis Significantly lower dysphagia grades in lans at 1 year group at 6 months (p = 0.009) but not at 12 months (p = 0.074).

Reduction in stricture grade: There is more reduction for lans group than ran at both 6 months and 12 months. (p=0.11 and 0.33, respectively).

Esophagitis present at 12 months: 30% with lan vs. 52% with ran groups. fam: famotidine; lans: lansoprazole; ome: omeprazole; ran: ranitidine; * indicates industry involvement

Context: • Although the evidence was considered poor and sometimes contradictory, PPIs were thought likely to be superior to H2RAs for esophagitis with strictures. • The trial that found fewer significant differences between treatments used double-dose H2RA80. • All studies assessed initial therapy for 6 months to one year.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 56 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statement G3.2.1D: In patients with erosive esophagitis who have completed an initial course of PPIs, continued PPI therapy is more efficacious than step-down to H2RAs for preventing relapse and providing symptom relief.

Voting Results: A 42%, B 58%, C 0%, D 0%, E 0%

Summary: One good quality SR76was identified for this statement. It demonstrated that maintenance doses of PPIs are better than H2RAs in maintaining remission of symptoms and healing of esophagitis at one year in patients with reflux esophagitis. Study Population Intervention Comparator Outcome Results Type (QA) measure Donnellan et Maintenance PPI H2RAs (cim Esophagitis PPI maintenance dose vs. H2RA : al. 2004*76 of healed maintenance 800 mg/day, relapse and esophagitis: dose (ome 10 ran 150 bid, presence of Proportion with relapse of esophagitis: SR (good) 6 RCTs, mg qd, lans fam 20 mg symptoms at 39% vs. 66%; RR = 0.57; (95% CI n=1156 15 mg qd, bid) for 24-52 24-52 weeks 0.47, 0.69), NNT (95% CI) = 3.3 (2.5, pant 20 mg weeks 5.0) Maintenance qd, rab 10 mg of symptom qd, esome 10 Proportion with relapse of symptoms: relief: mg qd) for 31% vs. 44%; RR = 0.55; (95% CI 4 RCTs, 24-52 weeks 0.47, 0.65), NNT (95% CI) = 4.0 (3.1, n=831 5.3)

Cim: cimetidine; esome: esomeprazole; fam: famotidine; lans: lansoprazole; ome: omeprazole; pant: pantoprazole; rab: rabeprazole; ran: ranitidine; * indicates industry involvement

Context: • A large proportion of patients experience relapse of esophagitis on both PPIs (44%) and H2RAs (66%).

G3: Erosive Esophagitis G3.2: Maintenance therapy G3.2.2: Half-dose vs. standard-dose PPIs Evidence Statement G3.2.2A: Half-dose PPIs are less efficacious than standard-dose PPIs as maintenance treatment for the prevention of symptomatic and endoscopic relapse in patients with erosive esophagitis.

Voting Results: A 50%, B 50%, C 0%, D 0%, E 0%

Summary: The evidence for this Statement consists of 1 good quality SR76 and one poor quality SR83. Donnellan et al.76 showed that standard-dose PPIs were more effective than half-dose PPIs for the healing and symptom relief of esophagitis at 24-52 weeks. Edwards et al. 83 reported that standard-dose lansoprazole was more effective at maintaining healing at 6 and 12 months than half-dose lansoprazole. . Study Outcome Population Intervention Comparator Results Type(QA) measure Donnellan PPI healing PPI PPI healing Healing and PPI healing dose vs. maintenance dose:

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 57 Consultation Document: Canadian Agency for Drugs and Technologies in Health et al. vs.maintenance maintenance dose/ standard- symptom 2004*76 dose (22 RCTs, dose / half dose (ome 20 relief Proportion with relapse of esophagitis: n=5,964) doses (ome mg, lans 30 between 24 18% with healing dose vs. 29% with SR (good) 10 mg, lans mg, rab 20 mg, to 52 weeks maintenance dose, p<0.00001; RR of Maintenance of 15 mg, rab pant 40 mg, relapse = 0.63 (95% CI 0.55, 0.73), NNT symptom relief: 10 mg, pant esome 20 mg, (95% CI) = 9.1 (6.7, 14.3) (significant (18 RCTs, 20 mg, 40 mg), heterogeneity) n=5116 ) esome 10 H2RAs (fam mg), for 24 20 mg bid, ran Proportion with significant symptoms: to 52 weeks 150 mg, 300 31% vs. 36%, p<0.009; RR of symptoms mg bid), relapse = 0.78 (95% CI 0.68, 0.88), NNT placebo for 24 (95% CI) = 12.5 (8.3, 25) to 52 weeks)

Edwards et Maintaining of lans 15 mg lans 30 mg for Healing at 6 Maintainenance of healed reflux al. 2002*83 healed reflux for 6-12 6-12 months and 12 esophagitis (lans 30 mg vs.lans 15 mg): esophagitis at 6 months months SR (poor) months: 6 At 6 months: RR = 1.12 (95% CI 1.06 - RCTs, (n=1160) 1.18) in patients with reflux At 12 months = RR 1.12 (95% CI 1.05 - esophagitis 1.18)

At 12 months: 7 RCTs (n=1505) in patients with reflux esophagitis (5 RCTs are common between the two durations) Esome: esomeprazole; fam: famotidine; lans: lansoprazole; ome: omeprazole; pant: pantoprazole; rab: rabeprazole; ran: ranitidine; * indicates industry involvement

Context: • Full dose appears to be superior to half-dose but the difference is small and may be of limited clinical significance. For example, the proportions with significant symptoms at 6-12 months were 31% with half-dose and 36% with standard-dose PPIs in the Donnellan et al. SR76.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 58 Consultation Document: Canadian Agency for Drugs and Technologies in Health

G3: Erosive Esophagitis G3.2: Maintenance therapy G3.2.3: On-demand PPIs after response to initial PPI therapy. Evidence Statement G3.2.3A: For patients with erosive esophagitis who respond to initial PPI therapy, subsequent ‘on-demand’ PPI therapy is less efficacious than continuous standard-dose PPI therapy.

Voting Results: A 8%, B 58%, C 33%, D 0%, E 0%

Summary: The evidence for this statement consisted of one poor quality RCT84, in which continuous esomeprazole 20 mg was more effective in erosive esophagitis patients than the same dose on-demand in terms of proportion in remission, and severity of symptoms (acid regurgitation, dysphagia, and epigastric pain) at 6 months. However, there were no differences in terms of the proportion with symptomatic relapse or patient satisfaction. Study Population Intervention Comparator Outcome Results Type(QA) measure Sjostedt et al. 539 patients with esome 20 mg esome 20 mg Endoscopic Cumulative proportion in 2005*84 erosive reflux on demand qd remission at 6 remission at 6 months: 81% esophagitis (LA for 6 months maintenance months with once daily vs. 58% with RCT (poor) classification grades therapy for 6 on demand, p<0.0001 A-D) months Symptomatic relapse over 6 Endoscopic remission at 6 Patients were treated months months by grade of esophagitis with esome 40 mg qd (once daily vs. on demand): for 4-8 weeks; 477 grade A: 93% vs. 78% healed patients were (p=0.03); enrolled grade B: 90% vs. 65% (p<0.0001); grade C: 90% vs. 51% (p=0.0002); grade D: 80% vs. 44% (NS)

Symptomatic relapse at 6 months: once daily 5.0% vs. on demand 5.7% (NS)

No. with GERD symptoms at 6 months: normal/mild/moderate/severe (once daily vs. on demand): acid regurgitation: 194/9/0/1 vs. 143/15/8/0 (p=0.003); Dysphagia: 202/2/0/0/ vs. 158/5/2/1 (p=0.02); Epigastric pain: 188/13/2/1 vs. 140/23/2/1 (p=0.02)

Patient satisfaction (once daily vs. on demand): At 3 months: 96% vs. 94% (NS); At 6 months: 96% and 99% (NS)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 59 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Study drug taken: 95% of treatment days with once daily vs. 57% with on demand Esome: esomeprazole; * indicates industry involvement

Context: • The evidence for this Statement is limited to one poor quality RCT. • On-demand and continuous esomeprazole were similar in terms of the proportion with symptomatic relapse at 6 months. Most patients were normal or had mild symptoms in both groups at 6 months. Patient satisfaction was also similar. • There are no studies assessing the effect of on-demand therapy in terms of long-term complications of erosive esophagitis.

G3: Erosive Esophagitis G3.2: Maintenance therapy G3.2.4: Intermittent PPI strategy Evidence Statement G3.2.4A: In patients with erosive esophagitis, intermittent PPI therapy (e.g., 3 days per week) was less efficacious than continuous PPI therapy.

Voting Results: A 25%, B 75%, C 0%, D 0%, E 0%

Summary: The evidence identified consisted of one poor quality SR55. It demonstrated that, in patients with erosive esophagitis, continuous treatment with omeprazole 20 mg qd was better than intermittent treatment with omeprazole 20 mg 3 days per week or continuous treatment with ranitidine 150 mg bid. Study Outcome Population Intervention Comparator Results Type(QA) measure Zacny et 2 RCTs (n=565) RCT 1: RCT 1: Healing at 12 RCT 1 (n=159): ome 20 mg daily vs. 3 al. 200555 in patients with ome 20 mg ome 20 mg 3 months days/week vs. ran 150 mg: erosive qd for 6-12 days/week SR (poor) esophagitis months Proportion with healing maintained (grade 2-4) ran 150 mg after 12 months: 89% (95% CI: 80%, bid for 6-12 98%) vs. 32% (95% CI: 19%, 46%) vs. months 25% (95%CI 13%, 38%) (p<0.001 for each comparison) RCT 2: RCT 2: ome 20 mg ome 20 mg Proportion experiencing relapse of qd for 6 intermittent moderate to severe symptoms: 15% vs. months for 6 months 67% vs. 57%, p<0.001 for each comparison Placebo for 6 months RCT 2 (n=406): ome 20 mg qd vs. ome 20 mg intermittent vs. placebo:

Proportion free of esophagitis: 70% vs. 34% vs.11% (p< 0.001 for both ome arms vs. placebo, no p-value for ome 20 mg qd vs. intermittent)

Symptoms recurrence: 74% vs. 37%

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 60 Consultation Document: Canadian Agency for Drugs and Technologies in Health

vs. 10% (p< 0.001 for both ome arms vs. placebo, no p-value for ome 20 mg qd vs. intermittent)

Ome: omeprazole; ran: ranitidine.

Context: • The single source of evidence was a poor quality SR of only two RCTs. • There are no data on intermittent therapy in esophagitis of less than Grade 2.

G4: Asthma, laryngeal symptoms, and cough associated with GERD Evidence Statement G4.1: PPIs are not efficacious in asthma patients with concomitant GERD.

Voting Results: A 67%, B 27%, C 9%, D 0%, E 0%

cim: cimetidine; ome: omeprazole; FEV1: Forced expiratory volume in one second 8583 SummaryNote : Only data: One related good to qualityPPIs were SR extracted. on PPI treatment for asthma associated with GERD was identified. This study reported that anti-reflux treatment with a PPI (omeprazole 20-80 mg per day) or H2RA (ranitidine 300 mg, cimetidine 1000 to 1200 mg per day) did not improve FEV1, morning peak expiratory flow, airway responsiveness or use of inhalers in patients with GERD-associated asthma. However, nocturnal symptoms and esophageal pH improved with omeprazole 40 mg and 80 mg respectively. Study Outcome Population Intervention Comparator Results Type(QA) measure Gibson et 12 RCTs PPI (ome Placebo for Peak FEV1: No improvement with ome (4 al. 20038583 (n=432) in 20 mg, 40 4-26 weeks expiratory RCTs) or H2RAs (4 RCTs) vs. placebo patients with mg, 80 mg flow, airway Morning and evening Peak expiratory SR (good) asthma per day for hyper- flow: No significant effect with ome or (confirmed by 4-12 weeks) responsiveness, H2RAs vs.placebo (p = 0.8 for both vs. history, H2RAs (ran asthma placebo) diagnosis or 300 mg, symptoms, Airway responsiveness: Not affected by evidence of cim 1000 esophageal pH ome or H2RAs vs. placebo (p value not respiratory mg, 1200 given) airflow mg per day Nocturnal symptoms score: 1 RCT obstruction) and for 4- 26 reported improvement with ome 40 mg GERD weeks) vs. placebo (p<0.04). 2 RCTs reported (confirmed by Surgery improvement with H2RAs vs. placebo (p symptoms, <0.02 and 0.05). esophageoscopy Use of inhalers per day : No improvement with or without with ome or H2RAs vs. placebo (p=0.4 biopsy) and 0.6 respectively)

Asthma symptoms: Improvement of wheezing in patients with asthma associated with GERD: H2RA vs. placebo: RR = 2.07 (95% CI 1.20, 3.58), p<0.009 Esophageal pH: Improved with ome 80 mg per day but not with H2RA or low dose omeprazole vs.placebo (p value not given)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 61 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Context: • Although 12 RCTs were included in the Gibson et al. SR, the total sample size was relatively small (N=432)85.

Evidence Statement G4.2: PPIs are not efficacious in improving laryngeal symptoms (i.e., cough, throat clearing, globus, hoarseness, sore throat) associated with reflux.

Voting Results: A 70%, B 20%, C 10%, D 0%, E 0%

Summary: The evidence identified consisted of 3 RCTs, one of good quality86 and 2 poor quality87,88. Vaezi et al.86 and Eherer et al.88, both placebo-controlled studies, found no significant effect of PPIs on the improvement of symptoms of laryngitis in patients with chronic posterior laryngitis or reflux associated laryngitis. Steward et al.87 focused on PPI treatment for chronic laryngo-pharyngeal reflux. This RCT revealed that rabeprazole 20 mg for 2 months did not improve reflux symptoms, laryngeal symptoms, health status or laryngeal appearance as compared to placebo. Study Population Intervention Comparator Outcome measure Results Type(QA) Vaezi et al. 146 adult patients esome 40 bid Placebo for 16 chronic posterior 1. patients with CPL 2006 86 with 1 or more for 16 weeks weeks laryngitis (CPL) symptom resolved at final RCT (good) symptoms (throat symptom (throat visit (16 weeks) clearing, cough, clearance, esome vs.placebo: (%) globus, sore throat or hoarseness, 14.7 % vs.16.0% ( p= 0.799) hoarseness) for ≥3 cough, globus, Throat clearing: 10% months sore throat) relief vs. 13% Hoarseness: 11% vs. 9% All subjects had CPLI ( chronic Cough: 9% vs. 25% laryngoscopic signs posterior Globus: 25% vs. 20% consistent with reflux laryngitis index) Sore throat: 44% vs. 33% based on CPL index score changes based on 2. CPLI change from Pharyngeal acid reflux laryngoscpoic baseline to final visit documented in 13% examination esome vs. placebo: (mean ± and 16% of esome and SD) placebo treated -1.66±2.13 vs. -2.0±2.55 patients respectively (p=0.446) by ambulatory pH- monitoring (NB: pH readings for 47 patients excluded due to technical error)

Patients with moderate to severe heartburn on 3 or more days per week were excluded Steward et 42 adult patients with rab 20 mg placebo bid for Primary: Changes from baseline: al. 2004*87 history of hoarseness bid for 2 2 months and Change in total Total reflux score: rab throat clearing, cough, months and lifestyle reflux symptom 9.7±11.1, 95% CI: 4.2%, RCT (poor) globus and sore throat lifestyle modification scores, proportion 15.2%), p=0.002 vs. placebo

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 62 Consultation Document: Canadian Agency for Drugs and Technologies in Health

for > 4 weeks and modification instructions noting significant 6.6±12.5, 95% CI: 0.6%, with diagnosis of instructions global 12.7%), p=0.03. (NS) larngo-pharyngeal improvement Significant global reflux (edema, (reported as improvement: rab 53.3% erythema, and/or worse, (95% CI: 30%, 75%) pachydermia) unchanged, vs. placebo 50% (95% CI: slightly better, 29%, 71%); (p=1.0) (NS) 40% and 50% in much better, and Reflux severity: rab 4.7±5.2, placebo and PPI group gone) 95% CI: 2.1%, 7.3%), respectively had p=0.001 vs. placebo 3.0±6.1, abnormal pH at Secondary: 95% CI: 0.1%, 6.0%), baseline on 24-hr Change in p=0.04 (NS) esophageal pH components of Reflux frequency: rab monitoring reflux symptom 5.0±7.0, 95% CI: 1.5%, scores (freq 8.5%), p=0.008 vs. placebo Smokers: 31.6 % severity, ‘typical’ 3.6±7.1, 95% CI: 0.2%, patients in control (heartburn and 7.0%), p=0.04 (NS) group and 57.9 % acid ‘Typical’ symptoms: rab patients in PPI group regurgitation), 1.0±5.3, 95% CI: -1.6%, were smokers ‘laryngeal’, 3.6%), p=0.43 vs. placebo (p=0.19) ‘pharyngeal’, and 0.6±3.9, 95% CI: -1.3%, individual 2.4%), p=0.55 (NS) symptoms) ‘Laryngeal’ symptoms: rab 4.4±4.6, 95% CI: 2.1%, 6.6%), p=0.0008 vs. placebo 2.4±5.6, 95% CI: -0.3%, 5.1%), p=0.08 (NS) ‘Pharyngeal’ symptoms: rab 3.9±4.5, 95% CI: 1.7%, 6.2%), p=0.002 vs. placebo 2.8±5.1, 95% CI: 0.3%, 5.2%), p=0.03 (NS) SF36 MCS: rab 0.03±11.0, 95% CI: -5.4%, 5.5%), p=0.99 vs. placebo -0.9±8.2, 95% CI: -4.8%, 3.0%), p=0.64 (NS) SF36 PCS: rab -2.1±10.3, 95% CI: -7.3%, 3.0%), p=0.64 vs. placebo (NS)1.2±7.8, 95% CI: -2.5%, 4.9%), p=0.51 Videostobe grade: rab 0.6±1.8, 95% CI: -0.3%, 1.6%), p=0.19 vs. placebo 0.5±2.3, 95% CI: -0.9%, 1.9%), p=0.44 (NS)

Mean difference in treatment effects between groups (placebo vs. rab): total reflux score: -3.06±3.9 (95% CI: -10.99%, 4.86%), p=0.44 Sig global improvement: - 3.3±17.5% (95% CI: -37.5%, 30.9%), p=1.0

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 63 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Reflux severity: -1.67±1.87 (95% CI: -5.46%, 2.12%), p=0.38 Reflux frequency: - 1.39±2.33 (95% CI: -6.12%, 3.33%), p=0.55 ‘Typical’ symptoms: - 0.45±1.52 (95% CI: -3.54%, 2.64%), p=0.77 ‘Laryngeal” symptoms: - 1.97±1.69 (95% CI: -5.4%, 1.47%), p=0.3 ‘Pharyngeal’ symptoms: - 1.18±1.57 (95% CI: -4.37%, 2.01%), p=0.46 SF36 MCS: -0.92±3.17 (95% CI: -7.36%, 5.51%), p=0.77 SF36 PCS: 3.34±2.98 (95% CI: -2.71%, 9.4%), p=0.27 Videostrobe grade: - 0.11±0.75 (95% CI: -1.65%, 1.43%), p=0.69 Eherer et al. 62 non-smoker pant 40 mg placebo bid for Laryngitis scores, % time with pH <4: no 200388 patients with bid for 3 3 months symptoms after 3 significant difference hoarseness for ≥2 months months between 2 treatment groups. RCT (poor) months and other 24-h pH-metry: before and laryngeal symptoms after 3 months of pant or (diagnosed by placebo treatment, no telescopic video- significant diff laryngo-stroboscopy ) Mean esophageal symptom without obvious cause score at baseline (±SD): of laryngitis (24 of 62 pant/placebo group 3.3±1.2 had GERD were significantly lower vs. demonstrated by 24-h placebo/pant group 11.0 ±2.6 pH-metry) (p value not reported) Change in scores after first treatment not significant between the second treatment groups: 5.4±2.8 vs. 2.2±1.4 Laryngeal symptoms score: before treatment: pant vs. placebo: 14.6±3.1 vs. 17.4±3.1 Change in scores after treatment: 8.3±3.6 vs. 10.3 ±3.9 (NS) Decrease of laryngeal score: pant vs. placebo: decrease of 8.0±1.4 decrease of 5.6±2.6 (NS) Esome: esomeprazole; pant: pantoprazole; rab: rabeprazole; * indicates industry involvement

Context: • There was limited evidence for this Statement. The three trials that were identified were of small sample size. Furthermore, the validity of some outcomes was uncertain, and results were

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 64 Consultation Document: Canadian Agency for Drugs and Technologies in Health

sometimes contradictory. • All studies enrolled subjects based on laryngopharyngeal signs and symptoms. Only about 15- 50% of subjects had reflux as measured by 24-hour pH-metry. The ERP noted that chronic posterior laryngitis (CPL) is a non-specific sign of GERD, and cannot be relied upon to confirm the presence of concomitant GERD. • The study population differed across studies. For example, in Vaezi et al.86, subjects had laryngoscopic signs consistent with reflux but those with overt GERD were excluded. However, exclusion of patients with overt GERD symptoms was not described in the other two RCTs87,88. In the absence of GERD symptoms, there is no evidence that PPIs are useful.

Evidence Statement G4.3: PPIs are not efficacious in improving chronic cough with or without GERD.

Voting Results: A 33%, B 67%, C 0%, D 0%, E 0%

Summary: One good quality SR89 was identified for this statement. It showed that PPIs are not efficacious in improving cough associated with GERD. Study Outcome Population Intervention Comparator Results Type(QA) measure Chang et al. Patients with chronic PPIs (ome Placebo for 8 Cough score at Patients still coughing at the 200589 (≥3 weeks) non 40 mg, 80 to 12 weeks various time end of trial: No benefit of specific cough mg, pant 40 points using PPI vs. placebo; SR (good) without any mg) for 8 to Pooled OR = 0.24 (95% CI respiratory symptoms, 12 weeks 0.04, 1.27), p=0.09 (NS) sign or systemic illness Mean cough score at the end of trial: PPI vs. placebo; Patients still coughing SMD = -0.51 (95% CI -1.02, at the end of trial: 3 0.01), p=0.05 RCTs (n= 49) (smoking status was Change in cough score at not reported) end of intervention (vs. baseline): No benefit of Mean cough score at using PPI vs. placebo; SMD the end of trial: 3 = -0.40 (95% CI -0.86, RCTs (n=61) 0.06), p=0.09 (NS) (smokers were excluded in 2 RCTs, Change in cough score at 4 smoking status not weeks of treatment (vs. reported in 1 RCT) baseline): No benefit of using PPI vs. placebo; SMD Difference in cough = -0.51 (95% CI -1.08, score at trial end vs. 0.06), p=0.09 (NS) baseline: 4 RCTs (n = 77) ((smokers were Difference in cough score at excluded in 2 RCTs, week 8 vs. weeks 4: No smoking status not benefit of using PPI vs. reported in 2 RCTs) placebo; SMD = -0.44 (95% CI -0.04, 0.16), p=0.1 (NS) Change of cough score at 4 weeks of treatment: 2 RCTs (n

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 65 Consultation Document: Canadian Agency for Drugs and Technologies in Health

= 51) (smokers were excluded in both RCTs)

Difference in cough score at week 8 vs. week 4: 2 RCTs (n=51) (smokers were excluded in both RCTs)

Ome: omeprazole; pant: pantoprazole

Context: • The evidence for this Statement is limited to trials with small sample sizes that were likely underpowered. • It was evident that there was heterogeneity in study populations across the individual RCTs that comprised the Chang et al. SR89. Two trials included patients with laryngeal symptoms (including cough) and GERD defined by pH-metry, two trials enrolled patients with chronic cough and GERD defined by pH-metry, while the remaining two RCTs included patients with laryngitis symptoms (including cough) with or without GERD defined by pH-metry.

G5: Differences among PPIs Evidence Statement G5.1: There are no clinically important differences among standard-doses of PPIs (omeprazole 20 mg, lansoprazole 30 mg, pantoprazole 40 mg, rabeprazole 20 mg, esomeprazole 20 mg) in treatment of symptomatic GERD, ENRD and esophagitis.

Voting Results: A 50%, B 50%, C 0%, D 0%, E 0%

Summary: The evidence identified for this Statement consisted of 6 good quality SRs76,90-94, 3 poor quality SRs83,95,96, 1 good quality RCT97, and 7 poor quality RCTs98-104.

Systematic Reviews Edwards et al.91 found that esomeprazole 40 mg/day was significantly more likely to produce healing of esophagitis at 4 and 8 weeks as compared to omeprazole 20 mg/day, however there were no significant differences between omeprazole and other PPIs. McDonagh et al.90 also found no significant differences in most comparisons of one PPI with another, except that esomeprazole was significantly better than comparator PPIs for some outcomes. Donnellan et al.76 reported that omeprazole was more effective for symptom relief but not for healing of esophagitis as compared to lansoprazole. Edwards et al. 83 found no difference between omeprazole and lansoprazole in the proportion of healed reflux esophagitis patients that maintained healing at 6 or 12 months. McQuaid et al. 94 reported that omeprazole was more effective than lansprazole and esomeprazole in heartburn relief on day 1, although there was no significant difference between lansoprazole and esomeprazole. Wang et al.93 reported that there was no significant difference in esophagitis healing rates between omeprazole and other PPIs. Vakil et al.92reported a significantly higher rate of symptom relief in GERD patients treated with lansoprazole versus omeprazole on days 1-3 of treatment, as well as with esomeprazole versus omeprazole at 8

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 66 Consultation Document: Canadian Agency for Drugs and Technologies in Health weeks. However, there was no significant difference between omeoprazole and pantoprazole at 4 and 8 weeks, nor between rabeprazole and omeprazole at 4 weeks. Vakil92 et al. also reported that esomeprazole was more effective than lansoprazole in the healing of the GERD at 8 weeks, but that there were no significant differences between other PPIs in the healing of GERD. Edward et al.91 showed that esomeprazole was more effective than omeprazole in the healing of GERD at 4 and 8 weeks, but no difference was noted between omeprazole and the other PPIs for healing of esophagitis. The remaining two SRs95,96 showed no significant difference among lansoprazole, omeprazole, pantoprazole, and rabeprazole on the healing of reflux esophagitis in patients with erosive or ulcerative esophagitis and GERD. The ERP also considered an as yet unpublished Cochrane review in its deliberations. The findings of this study were supportive of the Evidence Statement.

RCTs Chey et al. 102 showed no significant difference between lansoprazole 30 mg and esomeprazole 40 mg when both were given for 2 weeks for GERD symptoms. However, Lauritsen et al.103reported that esomeprazole 20 mg for 6 months was more effective than lansoprazole 15 mg in maintaining remission at 6 months in patients with healed reflux esophagitis. For GERD of grade BC or ≥ 2, omeprazole 20 mg and rabeprazole 20 mg were similar in terms of healing and symptoms improvement97. Similarly, in erosive or reflux esophagitis (grade 1-4), omeprazole and pantoprazole were not different in terms of healing and symptoms104. Chen et al. 98 reported that esomeprazole was as effective as omeprazole for healing of erosive esophagitis. Finally, of the RCTs conducted in patients with non-erosive GERD99-101, none revealed significant differences among the various PPIs tested in terms of symptom relief.

Study Outcome Population Intervention Comparator Results Type(QA) measure Donnellan Maintenance of ome 20 mg qd lans 30 mg or Healing, Proportion with esophagitis relapse: et al. healed for 48 to 52 rab 20 mg qd symptom ome vs. lans or rab: 10.3% vs. 2004*76 esophagitis: weeks for 48 to 52 relief at 48 11.1%, p>0.05 (NS); RR = 0.93 3 RCTs (n=1020) weeks to 52 weeks (95%CI: 0.65, 1.32) SR Maintenance of Proportion wth relapse of (good) symptom relief: 3 significant symptoms: ome vs. lans RCTs (n=1001) or rab: 36.6% vs. 30.8%, p< 0.03; RR = 1.19; (95%CI 1.02, 1.38), NNT (95% CI) = 16.7 (9.1 to 100) McDonagh 12 RCTs ome 20 mg lans 30 mg, Symptom At 4 weeks: No difference between et al. (n=11,923) in qd for 4-8 pant 40 mg relief and ome, lans, pant, and rab for relief of 200590 patients with weeks rab 20 mg, healing symptoms and healing of esophagitis qd for 4-8 esophagitis. SR (good) weeks 2 RCTs in patients By baseline severity: with moderate to ome 20 mg vs. lans 30 mg: severe esophagitis equivalent healing rates in patients for 4 weeks, with moderate to severe n=372, esophagitis. 8 week, n=358 Proportion with symptom relief at 4 esome 40 mg vs. esome 40 mg ome 20 mg for weeks: esome 40 mg superior to ome 20 mg: qd for 4-8 4-8 weeks ome 20 mg, pooled risk difference 3 RCTs in patients weeks = 10% (95% CI 6%, 14%); esome with esophagitis, 40 mg vs. lans 30 mg or pant 40 grade C-D mg, no significant difference (p (n=1343) value not given)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 67 Consultation Document: Canadian Agency for Drugs and Technologies in Health

esome 40 mg vs. esome 40 mg, lans 30 mg qd Healing lans 30 mg: qd for 4-8 for 4-8 weeks esome 40 mg had higher 4-8 weeks 2 RCTs in patients weeks healing rates than ome 20 mg; with moderate to esome 40 mg vs. lans 30 mg: 5% severe esophagitis higher healing (NNT =20) at 4 (n= 2285), 1RCT weeks; at 8 weeks pooled estimates (n=109) in not significantly different patients with moderate to severe Moderate to severe esophagitis esophagitis lans 30 mg vs. ome 20 mg: pooled risk difference at 4 weeks 1%; esome 40 mg vs. esome 40 mg, pant 40 mg qd (95%CI -13%, 16%) (NS); at 8 pant 40 mg, qd for 4-8 for 4-8 weeks weeks: 3%; (95%CI -4%, 10%). patients with weeks (NS) moderate esophagitis esome 40 mg vs. ome 20 mg: (n=769) Pooled risk difference at 4 weeks: 16%; (95%CI 11%, 22%); at 8 weeks: 13%; (95%CI 9%, 17%)

esome 40 mg vs. lans 30 mg: pooled risk difference at 4 weeks: 8%; (95%CI 4%, 12%) (NNT = 13); 8 weeks: 9%; (95%CI 5%, 12%) (NNT =11).

Another RCT at 8 weeks: lans 30 mg vs. esome 40 mg (difference = 10%; 95%CI -2%, 22%) (NS)

esome 40 mg vs. pant 40 mg: healing rates 45% vs. 67%; of patients with grade C esophagitis at baseline, 91% improved to grade A or B with esome 40 mg vs. 100% with pant 40 mg at 10 weeks Edwards et 2 RCTs (n=859) lans 30 mg qd ome 20 mg qd Healing at 6 Maintaintenance of healed reflux al. 2002*83 in patients with for 12 mos for 12 mos and 12 esophagitis: lans vs. ome healed reflux months At 6 months: RR = 0.99 (95%CI: SR (poor) esophagitis 0.95, 1.03)

At 12 months: RR = 1.01 (95%CI 0.96, 1.06) McQuaid et 6 RCTs ome 20 mg, qd lans 30 mg, Heartburn 24-hour heartburn relief on day 1: al. 200594 (n=13,400) of PPI (duration not esome 20 mg, relief at day ome 20 mg vs. all other PPIs (4 intraclass mentioned) 40 mg, rab 20 1 RCTs, n=7,875): RR = 0.83 (95% SR (good) comparisons in mg, qd CI 0.75, 0.92); patients with (duration not NNH = 16 (11, 34) erosive mentioned) esophagitis ome 20 mg vs. lans 30 mg (2 RCTs, n=3,579): RR = 0.82 (95%CI: 0.75, 0.88); NNH = 18 (13, 26)

ome 20 mg vs. esome 40 mg (2 RCTs, n=4,296): RR = 0.78 (95% CI: 0.71, 0.85);

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 68 Consultation Document: Canadian Agency for Drugs and Technologies in Health

NNH = 11 (9, 16)

lans 30 mg vs. esome 40 mg (2 RCTs, n=5,522): RR = 1.03 (95% CI 0.87, 1.22)

Daytime heartburn relief: lans 30 mg vs. esome 40 mg: (2 RCTs, n=3,213): RR 1.03 (95% CI 0.94, 1.13)

Nightime relief: lans 30 mg vs. esome 40 mg: (2 RCTs, n=3,213): RR 1.03 (95% CI 0.96, 1.10)

Day 1 sustained heartburn relief: ome 20 mg vs. other PPIs (3 RCTs, n = 7,790): RR = 0.80 (95% CI 0.75, 0.86); NNH = 20 (16, 29)

ome 20 mg vs. esome 40 mg (2 RCTs, n = 4,296): RR = 0.78 (95% CI 0.71, 0.86) NNH = 17 (13, 26) Wang et al. 23 RCTs Standard-dose ome 20 mg qd Healing rate Healing with standard-dose PPIs vs. 200593 (n=6534) in PPIs (lans 30 for 8 weeks at 2, 4, 8 ome 20 mg: patients with mg, pant 40 weeks 2 weeks: (RR = 1.05; 95% CI 0.95, SR (good) esophagitis mg, esome 40 1.17) mg, rab 20 mg) 4 weeks: (RR = 1.04; 95% CI 0.98, qd for 8 weeks 1.11). 8 weeks: (RR = 1.1; 95% CI 0.98, 1.06) Vakil et al. 10 RCTs (n not rab 20 mg, lans ome 20 mg, Healing rates Proportion with symptom relief: 2003*92 given) in patients 30 mg, qd, for pant 40 mg and with esophagitis 8 weeks qd, for 8 weeks symptom At day 4: lans 30 mg vs. ome 20 mg SR (good) relief at 8 (1 RCT): 56% vs. 49% (P<0.0001) and 52 At 4 weeks: ome 20 mg vs. pant 40 weeks mg (1 RCT): 78% vs. 84%, p>0.05 (NS); ome 20 mg vs. rab 20 mg (1 RCT): 84% vs. 83%, p>0.05 (NS) At 8 weeks: ome 20 mg vs. pant 40 mg (1 RCT): 81% vs. 89% (NS); esome 40 mg vs. ome 20 mg (2 RCTs): 94% and 94% vs. 87% and 84% (p< 0.05 and p< 0.001 respectively in the 2 RCTs)

Healing rates (%): At 4 weeks : ome 20 mg vs. pant 40 mg (1 RCT): 78% vs. 74%, p>0.05 (NS) At 8 weeks: rab 20 mg vs. ome 20 mg (2 RCTs): 91.5 % vs. 94%, p>0.05 (NS); ome 20 mg vs. pant 40 mg (1 RCT): 94% vs. 90%,

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 69 Consultation Document: Canadian Agency for Drugs and Technologies in Health

p>0.05 (NS); esome 40 mg vs. ome 20 mg (1RCT): 93% vs. 89%, p<0.0001; lans 30 mg vs. pant 40 mg (1 RCT): 86% vs. 90% (NS)

At 52 weeks: rab 20 mg vs. ome 20 mg (2 RCTs): 95% vs. 96%, p>0.05 (NS)

1 RCT (n not ome 10 mg qd ome 20 mg qd Symptom Proportion with symptom relief at 4 given) in patients for 4 weeks for 4 weeks relief at 4 weeks : with ENRD weeks ome 10 mg vs. ome 20 mg : 35% vs. 41%, no p-value given

1 RCT (n not ome 10 mg qd ome 20 mg, qd Symptom Proportion with symptom relief: given) in patients for 4 weeks for 4 weeks relief at 4 61% with ome 20 mg vs. 49% with of GERD weeks ome 10 mg , p<0.02

1 RCT in rab 10 mg/day ome 20 mg/ Healing rates Proportion with maintained healing patients with or 20 mg/day day for 52 and of GERD: rab 10 mg vs. rab 20 mg esophagitis ( no for 52 weeks weeks symptom vs. ome 20 mg: grade info (no relief at 52 95% vs. 96% vs. 95%, p-values not provided) administration (no weeks given info provided) administration info provided)

Klok et al. pant vs. ome 4, ome 20 mg qd pant 40 mg qd, Endoscopic wk 4: other PPIs vs. ome: 200395 studies (n=604), lans 30 mg qd, healed 0.97 (95%CI: 0.88, 1.06) for pant, lans vs. ome 6, rab 20 mg qd, GERD 1.02 (95%CI: 0.96, 1.08) for lans, SR (poor) studies (n=1881), esome 40 mg 0.98 (95%CI: 0.91, 1.06) for rab. rab vs. ome, 2 qd studies (n=409), esome 40 mg vs. ome RR : esome vs. ome, 2 1.18, (95%CI: 1.14, 1.23),.NNT = studies (n=3729) 7.7. with endoscopically determined GERD. Caro et al. 8 studies for lans 30 ran 300 Healing RR healing rates (PPIs (lans, pant, 2001*96 healing involving mg/day, ome mg/day proportion rab) vs. ome 20 mg): 1298 patients with 20 mg/day, and wk 4 : SR (poor) endoscopically pant 40 symptom 1.04 (95%CI : 0.99, 1.10) for lans, confirmed GERD mg/day, rab 20 relief and 0.92 (95%CI : 0.85, 1.00) for rab, (grade 0-4) mg/day rates at 4 and 0.96 (95%CI : 0.85, 1.08) for pant 8 weeks Relapse at 1 wk 8: year 1.02 (95%CI: 0.98, 1.06) for lans; 0.93 (95%CI: 0.87, 1.00) for rab, 0.98 (95%CI : 090, 1.07) for pant.

Overall heartburn relief RR PPIs (lans, pant, rab) compared with ome 20 mg: At 4 wk: 1.02 (95%CI: 0.94, 1.11).

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 70 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Relapse: at 1 year: lans vs. rab (4.1% vs. 5%, respectively) Edwards et 12 RCTs (1-5 esome 40 mg, ome 20 mg qd Endoscopic Healing rates: al. 2001*91 RCTs in different lans 30 mg, for 8 weeks healing at 4 esome 40 mg vs. ome 20 mg SR (good) arms) pant 40 mg, and 8 weeks 4 weeks: RR 1.14 (95%CI 1.10, (n=2446) rab 20 mg qd 1.18) for 8 weeks 8 weeks : RR: 1.08 (95%CI 1.05, Patients with 1.10) esophagitis lans 30 mg vs. ome 20 mg 4 weeks: RR : 1.02 (95%CI 0.97, 1.08) 8 weeks: RR : 1.01 (95%CI 0.97, 1.06)

pant 30 mg vs. ome 20 mg 4 weeks: RR: 0.99 (95% CI 0.91, 1.07) 8 weeks: RR: 0.98 (95%CI 0.93, 1.04)

rab 20 mg vs. ome 20 mg 4 weeks: RR: 1.00 (95%CI 0.87, 1.14) 8 weeks : RR: 0.98; (95% 0.91, 1.05) Chen et al. 48 patients with esome 40 mg ome 20 mg qd EE healing EE healing rates at 8 wks: esome 2005*98 erosive reflux qd for 8 weeks for 8 weeks rates at 8 64% (95% CI: 44.3%, 83.8%) vs. esophagitis (EE) weeks, ome 45.5% (95%CI: 22.7%, RCT (poor) remission 68.3%), p=0.2481 (NS) rates for heartburn on Remission rate on day 1: esome the 1st day 77% vs. ome 65%, p>0.05 (NS) after treatment Monnikes et 529 patients with pant 20 mg qd esome 20 mg Symptom Median (68% range) time (days) al. 2005*99 endoscopy- for 28 days qd for 28 days relief to: negative GERD 1st symptom relief: pant 2 (1, 13) RCT (poor) vs. esome 2 (1, 14); Sustained symptom relief: pant 10 (1, 28) vs. esome 13 (1, 28);

Mean (SD) time to: 1st symptom relief (days): pant 5.9 (8.1) vs. esome 6.4 (9.0); Sustained symptom relief: pant 13.2 (11.6) vs. esome 13.5 (11.6);

Proportion of patients with symptom relief within 14 days: 1st symptom relief: pant 86.3% vs. esome 84.5%; Sustained symptom relief: pant 56.4% vs. esome 54.4%

Proportion of patients with

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 71 Consultation Document: Canadian Agency for Drugs and Technologies in Health

symptom relief within 28 days: 1st symptom relief: pant 92.8% vs. esome 89.7%; Sustained symptom relief: pant 80.2% vs. esome 79.4%

No difference between groups for all endpoints Fock et al. 134 patients with rab 10 mg qd esome 20 mg Symptom Time (days) to achieve 24-h 2005*100 reflux symptoms for 4wks qd for 4wks relief symptom-free interval: & normal Heartburn: rab 8.5 vs. esome 9, RCT (poor) endoscopy p=0.265 (NS); (NERD) Regurgitation: rab 6 vs. esome 7.5, p=0.405 (NS)

Proportion achieving 24-h symptom-free: Heartburn: rab 84% vs. esome 61% (NS); Regurgitation: rab 90% vs. esome 68% (NS)

Complete relief of day-time heartburn: At wk1: esome 23.4% vs. rab 26.9%, p>0.05 (NS); At wk4: esome 41.1% vs. rab 55.3.9%, p>0.05 (NS) Complete relief of night-time heartburn: At wk1: esome 20.9% vs. rab 28.8%, p>0.05 (NS); At wk4: esome 41.0% vs. rab 44.4%, p>0.05 (NS)

Symptom severity score during first 5days: no statistically significant difference between groups for day & night time heartburn or regurgitation

Armstrong Study A: 1282 Study A: Study A: Complete Complete resolution: et al. patients with esome 40 mg esome 20 mg resolution of At 2 weeks: 2004*101 ENRD qd for 4wks qd, or ome 20 heartburn at Study A: Study B: 693 Study B: mg qd for 4 2wks and 4 esome 40 mg vs. esome 20 mg RCT (poor) patients with esome 40 mg wks wks, vs. ome 20 mg: ENRD qd for 4 wks Study B: ome adequate 35% (95% CI: 30%, 39%) vs. Study C: 670 Study C: 20 mg qd for 4 control of 40% (95% CI: 35%, 45%) vs. patients with esome 20 mg wks heartburn at 38% (95% CI: 33%, 42%); ENRD qd for 4wks Study C: ome 4wks Study B: esome 40 mg 41% (95% 20 mg qd for CI: 36%, 47%) vs. ome 20 mg 43% 4wks (95% CI: 37%, 48%); Study C: esome 20 mg 41% (95% CI: 36%, 47%) vs. ome 20 mg 44% (95% CI: 39%, 50%); no significant difference between treatment groups in each study

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 72 Consultation Document: Canadian Agency for Drugs and Technologies in Health

At 4 weeks: Study A: esome 40 mg 57%(95% CI: 52%, 62%) vs. esome 20 mg 61% (95% CI: 52%, 62%) vs. ome 20 mg 58% (95% CI: 53%, 63%); Study B: esome 40 mg 70% (95% CI: 65%, 75%) vs. ome 20 mg 68% (95% CI: 63%, 73%); Study C: esome 20 mg 62% (95% CI: 57%, 67%) vs. ome 20 mg 60% (95% CI: 54%, 65%); no significant difference between treatment groups in each study

Chey et al. 3034 patients with lans 30 mg qd esome 40 mg Heartburn Heartburn relief at 2 wk: 2003*102 GERD for 2wks qd for 2wks relief, Average daytime heartburn sustained severity: lans 0.52 vs. esome 0.56, RCT (poor) resolution of p-value not given; heartburn, % of days with heartburn: lans symptom 36.2% vs. esome 38.2%, p=0.09 assessment, (NS); patient % of nights with heartburn: lans satisfaction 38.3% vs. esome 38.6%, p=0.6 with (NS); treatment Average severity of night-time heartburn: lans 0.60. vs. esome 0.61, p=0.563 (NS); Sustained resolution of heartburn: 25% had achieved sustained resolution of heartburn at 4 days in both groups, p=0.647 Lauritsen et 1236 patients with esome 20 mg lans 15 mg qd Remission Remission rate al. 2003*103 healed reflux qd for 6months for 6 months rate, life table estimates: esome 83% esophagitis symptoms (95% CI: 80%, 86%) vs. lans 74% RCT (poor) (95% CI: 70%, 78%), p<0.00001, log rank test; remission rate for grade A esophagitis: esome 87% vs. lans 84%, p<0.01; grade B: esome 83% vs. lans 72%, p<0.01; grade C: esome 75% vs. lans 61%, p<0.05; grade D: esome 77% vs. lans 50%, p<0.05;

Proportion heartburn-free: esome 78% vs. lans 71%, p<0.01; NNT (95% CI) = 14 (8, 47) Proportion free of acid regurgitation: esome 81% vs. lans 72%, p<0.001; Proportion free of epigastric pain: esome 80% vs. lans 75%, p<0.05 Dekkers et 202 patients with rab 20 mg qd ome 20 mg qd Healing rates Healing rates: al. 199997 erosive or and wk 4: 81% with rab vs. 81% with

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 73 Consultation Document: Canadian Agency for Drugs and Technologies in Health

ulcerative GERD symptom ome (95%CI: ±11%) (NS) RCT (good) (≥2) relief at 4 wk 8: 92% with rab vs. 94% with and 8 weeks. ome (95%CI: -9%, 5%) (NS)

Heartburn improvement: wk 4: 68% with rab vs. 75% with ome (p=0.36) wk 8: 73% with rab vs. 76% with ome (p=0.66) Vcev et al. 120 patients with pant 40 mg qd ome 20 mg qd Healing rates wk 4:pant 75% vs. ome 70% (NS) 1999104 reflux esophagitis at 4 and 8 wk 8: pant 90% vs. ome 87% (NS) (grade 1-4, weeks RCT (poor) Savary-Miller classification) esome: esomeprazole; lans: lansoprazole; ome: omeprazole; pant: pantoprazole; rab: rabeprazole; ran: ranitidine; * indicates industry involvement

Context: • There is some evidence that esomeprazole 40 mg produces slightly higher healing rates in erosive esophagitis than standard-doses of other PPIs, although the clinical relevance of the observed differences is uncertain. The evidence consists of one very good quality RCT75, and one good quality RCT74. Kahrilas et al.75 reported that esomeprazole 40 mg was more effective than omeprazole 20 mg for esophagitis healing and heartburn symptom relief at 8 weeks. Conversely, Gillessen et al.74 found no significant difference in healing rates between esomepazole 40 mg and pantoprazole 40 mg in terms of healing rates at up to 10 weeks or in terms of symptom relief. The ERP also considered an as yet unpublished Cochrane review that was consistent with the above findings. • The ERP recognized that the Canadian Association of Gastroenterology defines the standard- dose of esomeprazole as 40 mg/day. • In the Product Monograph the recommended initial dose of esomeprazole is 20 mg/day for all indications, with the exception of erosive esophagitis and H. pylori eradication.

G6: PPI adverse drug reactions Evidence Statement G6.1: PPIs have a similar adverse event rate (generally minor) as H2RAs in GERD randomized controlled trials of up to one year duration.

Voting Results: A 50%, B 50%, C 0%, D 0%, E 0%

Summary: One good quality SR compared the rate of adverse effects with PPIs to H2RAs76. Donnellan et al. reported that the proportion with overall adverse effects was higher with a maintenance dose (i.e., half-dose) of PPIs than H2RAs, but there was no difference between healing doses (i.e., single dose) of PPIs and H2RAs. Healing doses of PPIs were associated with headache more often than H2RAs. There was no significant difference in the risk of abdominal pain with either healing or maintenance doses. All adverse effect data in the Donnellan et al. SR were extracted from esophagitis studies. The ERP also considered an as yet unpublished Cochrane review in its deliberations. The findings of this study were supportive of the Evidence Statement. Study Outcome Population Intervention Comparator Results Type(QA) measure Donnellan Esophagitis PPI healing H2RAs (ran Adverse Esophagitis Studies: (PPI vs. H2RA)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 74 Consultation Document: Canadian Agency for Drugs and Technologies in Health et al. Studies: (PPI vs. dose (ome 150 bid, 300 effects 2004*76 H2RA) 20 mg qd, mg bid, 150 Proportion with overall adverse effects: lans 30 mg mg tid, fam Healing dose of PPI (at 26-52 weeks): 19% SR (good) Overall adverse qd, pant 40 20 mg bid), vs. 15%, p=0.2 (NS); RR = 1.26 (95% CI effects: mg qd, rab for 24-52 0.89, 1.80) Healing dose: 20 mg qd, weeks Maintenance dose of PPI: 44% vs. 31%, 3 RCTs (n=469) esome 20 p<0.003; RR = 1.38 (95% CI 1.11, 1.72) Maintenance mg qd) for dose: 3 RCTs, 24-52 weeks Proportion with Headache: (n=574) Healing dose of PPI: 19% vs. 8%, p = 0.04;

RR = 2.27 (95 CI 1.04, 4.97) PPI Headache: Maintenance dose of PPI: 9.7% vs. 8.4%, maintenance Healing dose: 1 p>0.05 (NS); RR = 1.15 (95% CI 0.46, dose (half RCT (n=189) 2.85) the healing Maintenance dose) dose: 1 RCT Proportion with Abdominal Pain: (n=188) Healing dose of PPI: 9% vs. 10%, p>0.05 (NS); RR = 0.90 (95% CI 0.43, 1.90) Abdominal Maintenance dose of PPI: 11.3% vs. 10.2%, pain: p>0.05 (NS); RR = 1.1 (95% CI 0.55, 2.24) Healing dose: 1 RCT (n=259) Maintenance dose: 1 RCT (n=261)

Diarrhea: 0 RCTs

ENRD Studies: No PPI vs. H2RA studies

Esome: esomeprazole; fam: famotidine; lans: lansoprazole; ome: omeprazole; pant: pantoprazole; rab: rabeprazole; ran: ranitidine; * indicates industry involvement

Context: • Post-marketing surveillance data were not considered for this Statement. The ERP recognizes that safety information derived from RCTs has limitations. • Observational studies have linked PPI use with C. difficile infection. Associations between PPI use and pneumonia have also been reported.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 75 Consultation Document: Canadian Agency for Drugs and Technologies in Health

5.1.3 Research Gaps

1) Double dose PPIs are more efficacious than standard-dose PPIs for patients with uninvestigated GERD symptoms who have severe symptoms. No evidence was identified for this Statement.

2) In patients with uninvestigated GERD, initial therapy with half-dose PPIs is less efficacious than standard-dose PPIs for relief of symptoms. No evidence was identified for this Statement.

3) In patients with ENRD who have completed their initial course of PPIs, continued PPI therapy is more efficacious than stepping down to H2RAs. No evidence was identified for this Statement.

4) In patients with uninvestigated GERD, ENRD or erosive esophagitis who have received initial PPI therapy, step-down therapy to H2RAs is more efficacious than placebo. No evidence was identified for this Statement.

5) Double dose PPIs are more efficacious than continued standard-dose PPIs in patients with uninvestigated GERD, erosive esophagitis or ENRD, who remain symptomatic with regular dose PPIs. No evidence was identified for this Statement.

6) For patients with ENRD who respond to initial PPI therapy, subsequent ‘on-demand’ PPI therapy is less efficacious than continuous standard-dose PPI therapy Although the ERP identified relevant studies, a majority of the panel voted ‘C’ (Neutral), therefore this Statement was considered a Research Gap. The voting results and summary of evidence are shown below.

Voting Results: A 0%, B 27%, C 55%, D 18%, E 0%

Summary: One good quality RCT105 and two poor quality RCTs were identified for this statement106,107. Pace et al.106 found that HR-QOL was better in patients treated with continuous esomeprazole 20 mg as compared to on-demand at 6 months. However, Janssen et al. reported that on-demand pantoprazole 20 mg was non-inferior to continuous administration in terms of failure rate (based on uncontrolled symptoms). Finally, Tsai et al.107 reported that on-demand therapy with esomeprazole 20 mg was better than continuous therapy with lansoprazole 15mg in patients with ENRD at 6 months in terms of time to discontinuation, unwillingness to continue, and patient satisfaction at 1 month, although patient satisfaction was similar at 3 and 6 months. Study Outcome Population Intervention Comparator Results Type(QA) measure Pace et al. 6,017 patients esome 20 mg esome 20 mg HR-QOL by At 6 months: Significant difference 2005*106 with GERD qd on qd QOLRAD (5 in QOLRAD scores was found at the (grades 0-I, demand for 6 continuously dimensions: end of maintenance phase in favour RCT (poor) Savary Miller months for 6 months emotional of continuous vs. on demand classification) distress, sleep (p<0.0001), but magnitude of

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 76 Consultation Document: Canadian Agency for Drugs and Technologies in Health

disturbance, difference was small (data in figure) Patients treated vitality, with esome 40 mg food/drink Proportion of patients very satisfied: for 4 weeks; 5,265 problems, 64.5% with continuous vs. 59.7% patients with physical/social with on demand (p value not complete functioning). reported) resolution or mild symptoms were Patient enrolled satisfaction score.

Janssen et 558 patients with pant 20 mg pant 20 mg Difference in Failure rate: On demand vs. al. endoscopically on demand qd estimated continuous: 30.7% (95% CI: 24.5%, 2005*105 confirmed GERD for 24 weeks continuously failure rates 37.0%) vs. 18.6% (95% CI: 13.4%, (grades 0-I, for 24 weeks (uncontrolled 23.8%); Difference = 12.1% (95% RCT (good) Savary Miller) heartburn, acid CI: -∞, 18.9%), thus non-inferiority regurgitation was concluded Patients treated and pain on with pant 20 mg swallowing) On demand vs. continuous: qd for 4 weeks; between Perceived mean daily symptom load: 470 patients treatments after 1.26±1.49 vs. 0.82±1.34, p<0.001 completely free of 24 weeks. GERD symptoms Unwillingness to continue: 6% vs. were enrolled Note: on 7.8% demand tested for non- Number of pantoprazole 20 mg inferiority vs. tablets used per patient: daily treatment Daily 0.51±0.31 vs. 0.93±0.17, using a p<0.001 predefined non- Total: 83.2±51.4 vs. 152.4±38.2, inferiority p<0.001 margin of 20%.

Perceived mean daily symptom load. Tsai et al. 774 patients (18- esome 20 mg lans 15 mg Time to At 6 months: 2004*107 80 yrs) with on demand x qd discontinuation Time to discontinuation due to ENRD 6 months. continuous x due to unwillingness to continue: esome > RCT (poor) 6 months. unwillingness lans; p=0.001 based on survival to continue. estimates (figure only)

Patient Proportion of patients that satisfaction discontinued due to unwillingness to with treatment. continue: esome 6% (95%CI: 2.8%, 8.8%) vs. lans 13% (95%CI: 9.2%, 16.8%), p=0.001.

At 1 month, proportion of patients ‘satisfied’ with treatment: esome 93.2% vs. lans 87.8%; p=0.02. Difference 5.5% (95%CI: 0.88%, 10.1%). At 3 and 6 months, satisfaction remained high and was similar for both groups (data in figure)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 77 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Mean number of days with heartburn symptoms in previous 7 days at 6 months: esome 1.6 vs. lans 0.9, no p value given

Note: patients on esome took treatment approximately 1/3 of days vs. 4/5 of days for lans esome: esomeprazole; lans: lansoprazole; pant: pantoprazole; * indicates industry involvement

Context: • All but one RCT were of poor quality. • The outcomes studied in the included trials were generally subjective. The differences between on-demand and continuous therapy in terms of QOL and satisfaction were usually small. The evidence was often contradictory (e.g., Tsai et al. found that on-demand was significantly better than continuous therapy in terms of willingness to continue107, but Pace et al. found that continuous was better in terms of satisfaction and QOL106). • On-demand therapy may be preferred by some patients.

7) In patients with erosive esophagitis, surgical anti-reflux therapy is not more efficacious than maintenance PPI therapy (when dose adjustment based on symptoms is allowed). Although the ERP identified relevant studies, a majority of the panel voted ‘C’ (Neutral), therefore this Statement was considered a Research Gap. The voting results and summary of evidence are shown below.

Voting Results: A 10%, B 20%, C 70%, D 0%, E 0%

Summary: The evidence identified consisted of one poor quality RCT108 and one good quality RCT109,110. Mahon et al.108 reported that, in patients with GERD, laparoscopic Nissen fundoplication led to significantly less acid exposure of the lower esophagus at 3 months and produced greater improvement in gastrointestinal symptoms and general well-being after 12 months as compared to PPIs. Lundell et al. 109,110 reported that ARS was superior to omeprazole 20 mg or 40 mg in terms of the proportion experiencing treatment failure at 3 and 5 years. However, dose increases were allowed in the omeprazole arm. When the data were analyzed such that the definition of treatment failure excluded subjects who were able to maintain remission on an increased omeprazole dose, there was no significant difference between ARS and omeprazole. There were also no differences between the two groups in terms of endoscopic outcomes, quality of life (PGWB score), and GSRS scores. Study Outcome Population Intervention Comparator Results Type(QA) measure Mahon et 340 patients Laparoscopic One of five 24-h pH Number of patients (n) indicated for each al. with ≥ 6 Nissen PPIs monitoring efficacy variable. Change (95% CI) is the 2005*108 month fundoplication previously and adjusted difference by analysis of co-variance history of (LNF) prescribed esophageal using corresponding pre-treatment baseline RCT GERD, 217 by a manometry at value as covariate: (poor) with primary 3 months pathological care Total well-being score (LNF vs. PPI): reflux based physician: QOL, At baseline: (n=104) 127.1±22.9 vs. (n=108) on Psychological 132.8±21.2;

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 78 Consultation Document: Canadian Agency for Drugs and Technologies in Health

endoscopic rab 10 mg General At 3 months: (n=97) 143.5±22.5 vs. PPI esophagitis qd, Well-being (n=101) 136.3±20.8; Change 10.1 (4.7, 15.5); and/or 24-h pant 20 mg Index p<0.001. pH studies, qd, lans 15 (PGWBI) and At 12 months: (n=99) 142.4±20.0 vs. with excess mg qd, GSRS at 3 (n=96)136.8±22.4; acid reflux ome and 12 Change: 9.5 (4.2, 14.8), p<0.001 and good 20 mg qd, months symptom or GI well-being score (LNF vs. PPI) correlation esome 20 Note: At baseline: (n=104) 31.7±6.7 vs. (n=108) mg qd individual 34.3±6.6; scores for At 3 months: (n=92) 37.3±5.9 vs. (n=100) (Daily PGWBI or 36.1±6.2; doses were GSRS were Change: 2.1 (0.5, 3.7), p=0.010; At 12 adjusted to not reported months: (n=80) 37.0±5.4 vs. (n=86) 35.0±7.3; a level that Change: 3.0 (1.1, 4.9), p=0.003 abolished all reflux General well-being score (LNF vs. PPI): symptoms At baseline (n=104) 95.4±19.0 vs. (n=108) and were 98.5±16.6; expressed At 3 months: (n=92) 106.0±18.2 vs. (n=100) as 100.2±17.0; multiples Change: 8.2 (3.8, 12.5), p<0.001; of the At 12 months: (n=79) 106.2±16.3 vs. above (n=86)100.4±18.9. Change: 7.1 (2.5, 11.7), doses (e.g. p=0.003 ome 10 mg =0.5, lans LOSP (mmHg) (LNF vs. PPI): 30 mg =2) At baseline: (n=109) 6.3±5.8 vs. (n=107) 8.1±7.6; At 3 months: (n=83) 17.2±7.0 vs. (n=87) 7.9±7.7; Change: 10.2 (8.4, 12.0), p<0.001

DeMeester score (LNF vs. PPI): At baseline: (n=102) 42.7±33.1 vs. (n=105) 36.9±26.5; At 3 months: (n=77) 8.6±16.3 vs. (n=84) 17.1±21.4, p<0.001

% of time pH<4 LNF vs. PPI At baseline: (n=97) 12.9±10.9 vs. (n=98) 9.5±7.3; At 3 months: (n=79) 1.4±3.6 vs. (n=85) 3.8±7.8, p=0.002 Lundell et 310 patients Antireflux ome (either Treatment (35 subjects had dose increased to ome 40 al. 2000* with chronic surgery 20 mg or failure*, mg/day, 1 subject had dose increased to ome 109 GERD (ARS) (total 40 mg qd) Symptomatic 60 mg/day) symptoms or partial relapse, Lundell et and fundoplication (dose esophagitis, No. without treatment failure at 3 years (ARS al. 2001*110 esophagitis through adjustment symptoms, vs. initial ome dose): 97 vs. 77, log-rank test upon laparotomy) to 40-60 quality of life p=0.0016 endoscopy mg allowed scales at up RCT if to 5 years No. without treatment failure at 3 years (ARS (good) symptoms vs. initial or adjusted ome dose): data not recurred) reported, but no significant difference

No. without treatment failure at 5 years (ARS

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 79 Consultation Document: Canadian Agency for Drugs and Technologies in Health

vs. initial ome dose): 83 vs. 65, log-rank test p<0.001

No. without treatment failure at 3 years (ARS vs. initial or adjusted ome dose): data not reported, but no significant difference (log- rank test p=0.088)

No. with Symptomatic relapse at 3 years: 17 experienced symptom relapse in ARS vs. 50 in ome group, p-value not reported

No. with Symptomatic relapse at 5 years: 20 experienced symptom relapse in ARS vs. 49 in ome group, p-value not reported

Similar proportion reported none-mild reflux symptoms at pre-defined time points throughout study

Esophagitis at 3 years: 14 had esophagitis upon endoscopy in ARS group vs. 18 in ome group, p-value not reported. No difference in prevalence of Barrett’s esophagus or strictures.

Esophagitis at 5 years: 18 had esophagitis upon endoscopy in ARS group vs. 20 in ome group, p-value not reported. No difference in prevalence of Barrett’s esophagus or strictures.

Quality of life (PGWB score) (ome vs. ARS): 12 months: 103.7 (16.7) vs. 102.1 (19.0) 24 months: 103.5 (17.0) vs. 103.1 (19.4) 36 months: 103.2 (17.8) vs. 104.7 (17.1) 48 months: 102.7 (17.6) vs. 103.2 (18.8) 60 months: 104.4 (16.7) vs. 103.5 (19.1) No difference between treatments.

GSRS score (ome vs. ARS): 12 months: 1.9 (0.7) vs. 1.9 (0.6) 24 months: 1.9 (0.7) vs. 1.9 (0.7) 36 months: 1.8 (0.7) vs. 1.7 (0.5) 48 months: 1.9 (0.7) vs. 1.9 (0.7) 60 months: 1.9 (0.7) vs. 2.0 (0.9) Comparable outcomes between treatments.

GSRS reflux dimension score (ome vs. ARS): 12 months: 1.8 (1.0) vs. 1.4 (0.7) 24 months: 1.7 (0.8) vs. 1.3 (0.7) 36 months: 1.7 (0.9) vs. 1.3 (0.6) 48 months: 1.7 (0.8) vs. 1.4 (0.8) 60 months: 1.6 (0.8) vs. 1.4 (0.9) ‘Difference in favour of ARS in reflux

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 80 Consultation Document: Canadian Agency for Drugs and Technologies in Health

dimension’; p-values not reported ARS: antireflux surgery; esome: esomeprazole; GSRS: Gastrointestinal Symptom Rating Scale; ome: omeprazole; lans: lansoprazole; LOSP = lower oesophageal sphincter pressure;ome: omeprazole; pant: pantoprazole;PGWB: Psychological General Well-Being score; rab: rabeprazole; * indicates industry involvement

Context: • The ERP expressed reservations regarding several limitations of the evidence: lack of blinding and patient selection, lack of reporting of p-values for some outcomes, clinical relevance and validity of outcomes, and difficulty in interpretation of outcomes. • Variations in techniques and expertise among studies may be important confounders. • Long-term follow-up data is needed to measure outcomes after laparscopic surgery.

8) The required duration of therapy for erosive esophagitis complicated by strictures was identified as an area where further research is required.

9) The use of PPIs in Barrett’s Esophagus was identified as an area where further research is required.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 81 Consultation Document: Canadian Agency for Drugs and Technologies in Health

5.2 Dyspepsia

5.2.1 List of Evidence Statements and Research Gaps

D1: Uninvestigated Dyspepsia D1.1: Initial therapy D1.1.1: PPIs vs. H2RAs Evidence Statements

D1.1.1A: Initial (up to 4 weeks) standard-dose PPI therapy is more efficacious than standard-dose H2RAs at reducing symptoms in patients with H. pylori-negative, otherwise uninvestigated dyspepsia. D1.1.2: PPIs vs. prompt endoscopy Evidence Statements

D1.1.2A: Management strategy based on initial (prompt) endoscopy does not produce better outcomes (e.g., improvement of symptoms, failure of treatment strategy, quality of life) than empirical PPI therapy in patients with uninvestigated dyspepsia. D1.2: Maintenance therapy D1.2.1: On-demand PPI vs. on-demand H2RAs or placebo Evidence Statements

D1.2.1A: On-demand PPI therapy is more efficacious than on-demand placebo for the maintenance treatment of dyspepsia symptoms in H. pylori-negative patients with uninvestigated dyspepsia.

D1.2.1B: On-demand PPI therapy is not more efficacious than on-demand H2RAs for the maintenance treatment of dyspepsia symptoms in H. pylori-negative patients with uninvestigated dyspepsia.

D2: Functional dyspepsia D2.1: Initial Therapy D2.1.1: PPIs vs. H2RAs or placebo Evidence Statements D2.1.1A: Standard-dose PPIs for 4-8 weeks are more efficacious than placebo for the improvement of symptoms in functional dyspepsia.

D2.1.1B: Standard-dose PPIs for 4-8 weeks are no more efficacious than standard-dose H2RAs for the improvement of symptoms in functional dyspepsia.

Research gaps

2) Standard-dose PPI therapy is more efficacious than antacids/alginates at reducing dyspeptic symptoms in patients with uninvestigated dyspepsia.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 82 Consultation Document: Canadian Agency for Drugs and Technologies in Health

3) Standard-dose PPI therapy is more efficacious than standard-dose H2RAs at reducing dyspeptic symptoms in patients with uninvestigated dyspepsia (H. pylori status unknown).

4) Continuous maintenance therapy with standard-dose PPIs is more efficacious than continuous maintenance therapy with H2RAs in uninvestigated dyspepsia.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 83 Consultation Document: Canadian Agency for Drugs and Technologies in Health

5.2.2 Evidence Statements, ERP voting results, and Summaries of evidence

D1: Uninvestigated dyspepsia D1.1: Initial therapy D1.1.1: PPIs vs. H2RAs Evidence Statement D1.1.1A: Initial (up to 4 weeks) standard-dose PPI therapy is more efficacious than standard-dose H2RAs at reducing symptoms in patients with H. pylori-negative, otherwise uninvestigated dyspepsia.

Voting Results: A 0%, B 73%, C 27%, D 0%, E 0%

Summary: In their good quality RCT, Veldhuyzen van Zanten et al. 111 demonstrated that once daily omeprazole provided superior symptom relief to ranitidine and placebo for the treatment of Hp-negative primary care dyspeptic patients at 4 weeks. Outcome Study Population Intervention Comparator Results measure Veldhuyzen n=512 ome 20 mg ran 150 mg Global overall Success rate at 4 weeks (ome vs. van Zanten et 1) Age: mean qd for 4 bid for 4 severity (GOS) ran vs. cis vs. placebo) (95% CI): al. 2005*111 (range): 42.5 y (18- wks, then weeks score; 51% (43-55%) vs. 36% (28-44%) 78 years) on-demand Treatment vs. 31% (22-39%) vs. 23% (16- RCT (good) 2) Pts with epigastric therapy for cis 20 mg success rates 31%), p=0.01 for ome vs. ran, pain or discomfort, an bid for 4 (defined as a p=0.001 for ome vs. cis, p<0.001 with or without additional 5 weeks score of either 1 for ome vs. placebo; heartburn and/or months (none) or 2 acid regurgitation placebo for (minimal) on the NNT (95% CI)=7 (4, 29) for ome (NB: excluded those 4 weeks GOS scale); vs. ran with heartburn Complete NNT (95% CI) = 4 (3, 6) for ome and/or regurgitation Each of the symptom relief vs. placebo without epigastric above was at 4 wks and 6 pain, or previous followed by months Percentage with complete diagnosis of GERD) on-demand symptom relief (GOS=1) at 4 3) Primary care therapy for weeks (ome vs. ran vs. cis vs. setting an placebo) (95% CI): 24% (17%, 4) Hp status negative additional 5 31%) vs. 11% (6%, 16%) vs. 8% 5) Endoscopy not months (3%, 13%) vs. 4% (0.5%, 7%), performed p=0.005 for ome vs. ran, p=0.001 for ome vs. cis, p=0.03 for ome vs. placebo NNT (95% CI) for ome vs. ran = 8 (5, 25) NNT (95% CI) for ome vs. placebo = 5 (4, 8)

Percent reponders (GOS

Cis: cisapride; ome: omeprazole; ran: ranitidine; * indicates industry involvement

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 84 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Context: • Patients with overt GERD (i.e., heartburn or regurgitation as the only symptoms) were excluded in the trial by van Zanten et al.111, however, patients could have had various degrees of heartburn in addition to epigastric pain/discomfort. Heartburn non-dominant dyspepsia is defined as ‘no’ or ‘minimal’ heartburn according to the scale used to measure heartburn severity in this study. In a subgroup analysis of patients with ‘no’ or ‘minimal’ heartburn/regurgitation (n=301), the response rate at 4 weeks for PPI was 48.7% (95% CI: 37.6%-59.8%) vs.39.5% (95% CI: 28.9%- 50.2%) with H2RA (p-value not reported). • After 4 weeks, therapy was on demand. Therefore the evidence for this Statement is limited to therapy of four weeks duration.

D1: Uninvestigated dyspepsia D1.1: Initial therapy D1.1.2: PPIs vs. prompt endoscopy Evidence Statement D1.1.2A: Management strategy based on initial (prompt) endoscopy does not produce better outcomes (e.g., improvement of symptoms, failure of treatment strategy, quality of life) than empirical PPI therapy in patients with uninvestigated dyspepsia.

Voting Results: A 9%, B 45%, C 45%, D 0%, E 0%

Summary: One RCT112 of poor quality compared empirical PPI therapy with a strategy of prompt endoscopy. Laheij et al. did not find a significant difference between PPI empirical therapy and prompt endoscopy in terms of the number of symptom-free days in patients with uninvestigated dyspepsia. Study Outcome Population Intervention Comparator Results Type(QA) measure Laheij et n=84 ome 20 mg Prompt Percentage Dyspeptic symptom-free days al. 1998*112 1) Mean age = 43-44 yrs qd for two endoscopy + of pts (mean): PPI vs. endoscopy: 166 in both groups wks appropriate undergoing d (95% CI: 128d, 204d) vs. 159 d RCT 2) Pts with persistent treatment (no endoscopy; (95% CI: 119d, 198d), p value not (poor) dyspeptic symptoms of details dyspeptic reported sufficient severity as reported) symptom- judged by GP, were free days; QOL score at 1 yr: PPI vs. referred for upper GI QOL (at 10 endoscopy: 15 (95% CI: 13, 17) endoscopy wks & 1 vs. 16 (95% CI: 14, 17), p value 3) Tertiary setting year) not reported 4) Hp status: not available Percentage of pts undergoing 5) Endoscopy not done endoscopy at 1 year follow-up: before study entry PPI vs. prompt endoscopy: 31% 6) No history of PPI use vs. 100%, p value not reported NNT (95% CI) = 1 (1, 2) ome: omeprazole; QOL: Quality of Life; Rx: prescription; * indicates industry involvement

Context: • With a sample size of 84 referral patients, the Laheij et al. study112 was very small.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 85 Consultation Document: Canadian Agency for Drugs and Technologies in Health

• The study refers to ‘prompt endoscopy’, meaning immediate endoscopy. The ERP considered it unlikely that such a strategy would be practical for initial management in Canada.

D1: Uninvestigated dyspepsia D1.2: Maintenance therapy D1.2.1: On-demand PPI vs. on-demand H2RAs or placebo Evidence Statement D1.2.1A: On-demand PPI therapy is more efficacious than on-demand placebo for the maintenance treatment of dyspepsia symptoms in H. pylori-negative patients with uninvestigated dyspepsia.

Voting Results: A 0%, B 78%, C 22%, D 0%, E 0%

Summary: In their good quality RCT, Veldhuyzen van Zanten et al. 111 reported that once daily omeprazole provided superior symptom relief to ranitidine and placebo for the treatment of Hp-negative primary care dyspeptic patients at 4 weeks. The investigators excluded subjects who presented with heartburn and/or regurgitation alone without epigastric pain. An additional 6-month course of omeprazole on-demand therapy did not produce better symptom relief than on-demand ranitidine or on- demand placebo. However, for the subgroup of patients who responded to initial treatment, the proportion who were still responders at 6 months was significantly higher in the omeprazole group vs. placebo, but not vs. ranitidine.

Study Population Intervention Comparator Outcome Results Type (QA) measure Veldhuyzen n=512 ome 20 mg ran 150 mg Global overall (Cisparide data omitted): van Zanten et 1) Age: mean qd for 4 bid for 4 severity (GOS) al. 2005*111 (range): 42.5 y (18- wks, then weeks score; Success rate at 4 weeks (GOS=1 78 years) on-demand Treatment or 2) (95% CI): RCT (good) 2) Pts with therapy for cis 20 mg success rates ome vs. ran: dyspepsia symptoms an bid for 4 (defined as a 51% (43-55%) vs.36% (28-44%) of at least moderate additional 5 weeks score of either 1 p=0.01, NNT (95% CI)=7 (4, 29) severity months (none) or 2 3) Excluded patients placebo for (minimal) on the ome vs. placebo: with heartburn 4 weeks GOS scale); 51% (43-55%) vs.23% (16-31%) and/or regurgitation p<0.001, NNT (95% CI) = 4 (3, 6) alone w/o epigastric Each of the Complete pain, or a previous above was symptom relief Percentage with complete diagnosis of GERD followed by at 4 wks and 6 symptom relief (GOS=1) at 4 4) No alarm on-demand months weeks (95% CI): symptoms therapy for ome vs. ran: 5) Primary care an 24% (17-31%) vs.11% (6-16%) setting additional 5 p=0.005 6) Hp status negative months NNT (95% CI) = 8 (5, 25) 7) No upper endoscopy or barium ome vs. placebo: study in previous 6 24% (17-31%) vs. 4% (0.5-7%), months p<0.001 NNT (95% CI) = 5 (4, 8)

Percent reponders (GOS

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 86 Consultation Document: Canadian Agency for Drugs and Technologies in Health

ome vs. ran: 44% (6-53%) vs. 41% (33-49%), NS

ome vs. placebo: 44% (6-53%) vs. 35% (27-43%), NS

Proportion of responders at 4 weeks who remained responders at 6 months (95% CI): ome vs. ran: 31.1% (23-39%) vs. 20.9% (14- 27%), p>0.05 (NS)

ome vs. placebo: 31.1% (23-39%) vs. 13.5% (8- 20%), p=0.001 Cis: cisapride; ome: omeprazole; ran: ranitidine; RRR: relative risk reduction; * indicates industry involvement

Context: • It should be noted that on-demand omeprazole was significantly more efficacious than on- demand placebo only in the subset of patients who responded to initial 4-week therapy.

Evidence Statement D1.2.1B: On-demand PPI therapy is not more efficacious than on-demand H2RAs for the maintenance treatment of dyspepsia symptoms in H. pylori-negative patients with uninvestigated dyspepsia.

Voting Results: A 0%, B 80%, C 10%, D 10%, E 0%

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 87 Consultation Document: Canadian Agency for Drugs and Technologies in Health

with heartburn 4 weeks GOS scale); 51% (43-55%) vs.23% (16-31%) and/or regurgitation p<0.001, NNT (95% CI) = 4 (3, 6) alone w/o epigastric Each of the Complete pain, or a previous above was symptom relief Percentage with complete diagnosis of GERD followed by at 4 wks and 6 symptom relief (GOS=1) at 4 4) No alarm on-demand months weeks (95% CI): symptoms therapy for ome vs. ran: 5) Primary care an 24% (17-31%) vs.11% (6-16%) setting additional 5 p=0.005 6) Hp status negative months NNT (95% CI) = 8 (5, 25) 7) No upper endoscopy or barium ome vs. placebo: study in previous 6 24% (17-31%) vs. 4% (0.5-7%), months p<0.001 NNT (95% CI) = 5 (4, 8)

Percent reponders (GOS

ome vs. placebo: 44% (6-53%) vs. 35% (27-43%), NS

Proportion of responders at 4 weeks who remained responders at 6 months (95% CI): ome vs. ran: 31.1% (23-39%) vs. 20.9% (14- 27%), p>0.05 (NS)

ome vs. placebo: 31.1% (23-39%) vs. 13.5% (8- 20%), p=0.001 Cis: cisapride; NS: not statistically significant; ome: omeprazole; ran: ranitidine; * indicates industry involvement

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 88 Consultation Document: Canadian Agency for Drugs and Technologies in Health

D2: Functional dyspepsia D2.1: Initial Therapy D2.1.1: PPIs vs. H2RAs or placebo Evidence Statement D2.1.1A: Standard-dose PPIs for 4-8 weeks are more efficacious than placebo for the improvement of symptoms in functional dyspepsia.

Voting Results: A 18%, B 73%, C 0%, D 9%, E 0%

Summary: Two good quality SRs113,114 and three RCTs, one of very good quality115 and two of good quality116,117 comprised the evidence for this statement. In both SRs, Moayyedi et al.113and Shiau et al. 114 found that PPIs were more effective than placebo for the relief of functional dyspepsia. van Zanten et al. compared esomeprazole 40 mg daily with placebo in patients with functional dyspepsia without heartburn- or regurgitation-predominant symptoms116. They found that the proportion with ‘no’ or ‘mild’ symptoms was significantly higher in the esomeprazole arm at 4 weeks but not at 8 weeks. A similar pattern was observed for the outcome of symptom improvement (defined as reduction in the Global Overall Symptom (GOS) score of at least 2 points from baseline). Blum et al.117 reported that omeprazole 10 mg and 20 mg was superior to placebo for the treatment of dyspepsia symptoms only in H. pylori-positive functional dyspeptics, while only the 20 mg dose produced a statistically significant benefit in terms of the proportion achieving complete symptom relief. Talley et al.115 showed that omeprazole was significantly more effective than placebo for symptom relief in functional dyspeptics, more than 40% of whom were infected with H. pylori. In subgroup analyses, those with reflux-like and ulcer-like dyspepsia demonstrated a significantly greater response with omeprazole vs. placebo, however, in the dysmotility-like subgroup there was no significant difference. Overall, data from the two RCTs indicate that PPIs are more effective than placebo in terms of symptom relief for patients with functional dyspepsia. Study Outcome Population Intervention Comparator Results Type (QA) measure Moayyedi et PPI vs. placebo: ome 10 or 20 Placebo Dyspepsia cure Dyspepsia at 2-8 weeks: al. 2006*113 8 RCTs (n=3293) mg qd x 2-4 (no more than 1) Adults presenting with weeks; lans 15 minimal 1) PPIs vs. placebo: RRR = SR (good) dyspepsia symptoms or 30 mg qd x symptoms) 14% (95% CI: 5%, 23%), with negative or 4-8 weeks NNT (95% CI) = 9 (6, 26) insignificant findings upon endoscopy or barium study 2) Hp status not reported

PPI vs. H2RA: ome 10 or 20 ran 150 mg Dyspepsia cure 2) PPI vs. H2RA: RRR = 2 RCTs (n=739) mg qd for 2 qd or bid (global symptom 7% (95% CI: 16, -2%) 1) Non-ulcer dyspepsia weeks; lans 30 relief) (NS) patients mg qd for 8 2) Hp status : not weeks provided 3) Endoscopy: normal

Shiau et al. 6 RCTs (n=2368) PPI for at least Placebo Symptom relief Symptom relief: 2002114 1) adults with functional 1 week PPIs vs. placebo: dyspepsia (no evidence (excellent outcome) SR (good) of organic disease, OR = 1.81 (95% CI: 1.49,

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 89 Consultation Document: Canadian Agency for Drugs and Technologies in Health

including at upper 2.20) endoscopy, to explain the NNT (95% CI) = 8 (6, 12) symptoms) 2) Hp status: not Combined good and available excellent response: OR = 3) endoscopy: used for 1.53 (95% CI: 1.29, 1.81), diagnosis of functional NNT (95% CI) = 10 (6.67, dyspepsia. 16.67) van Zanten et n=224 Esome 40 mg Placebo Primary: Esome vs. Placebo: al. 2006*116 1. age: 18 yr or older qd for 8 weeks Symptom relief (range 20-79) (defined as % with Symptom relief: RCT (good) 2. Rome II criteria FD having At 4 weeks: pts with at least moderate symptoms of no 50.5% (95% CI: 40.7%, dyspepsia symptoms more than 60.2%) vs. 32.2% (95% CI: (defined as a score of >4 minimal severity 23.8%, 41.5%), p=0.009 on a 7-point Global GOS ≤2) during overall Symptom (GOS) the previous 2 At 8 weeks: scale). Excluded those days) at 8 weeks 55.1% (95% CI: 40.7%, with predominant 60.2%) vs. 46.1% (95% CI: symptoms of heartburn Secondary: 36.8%, 55.6%), p=0.16 or acid regurgitation. Symptom relief (NS) 3. primary care setting at 4 weeks, 4. Hp status: 22-23% Hp- symptom % with Symptom positive improvement (Δ resolution: 5. endoscopy: GOS ≥2), At 4 weeks: endoscopically symptom 22.9% (95% CI: 15.4%, confirmed FD resolution 32.0%) vs. 13.0% (95% CI: (GOS=1) at 4 7.5%, 20.6%), p>0.05 (NS) and 8 weeks, QOL At 8 weeks: 32.1% (95% CI: 23.5%, 41.7%) vs. 22.6% (95% CI: 15.3%, 31.4%), p>0.05 (NS)

% with Symptom improvement: At 4 weeks: 60.6% (95% CI: 50.7%, 69.8%) vs. 42.6% (95% CI: 33.4%, 52.2%), p=0.005

At 8 weeks: 60.6% (95% CI: 50.7%, 69.8%) vs. 54.8% (95% CI: 45.2%, 64.1%), p>0.05 (NS)

QOL: Esome significantly better than placebo at 4 weeks, but not 8 weeks

Blum et al. n=792 ome 10 mg or Placebo ‘Primary In Hp (+) patients: 2000*117 1. Age ≥18 yrs: mean 20 mg qd for response’ at 2 RCT (good) age: 38-48 yrs among 14 days weeks (defined Difference in response different groups as lack of rate (percent of patients

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 90 Consultation Document: Canadian Agency for Drugs and Technologies in Health

2) Patients with ran 150 mg at dyspeptic with 95% CI) compared functional dyspepsia bedtime symptoms with. placebo ( (epigastric symptoms for 14 days requiring further therapeutic gain ) w/o organic disease); had management) to have severe dyspeptic 1.) complete symptom symptoms for ≥3 days Complete relief) : during 1 week treatment symptom relief ome 20 mg vs. ome 10 mg with antacids during at 2 weeks vs. ran: 19.6% (8.6%, screening period 30.7%) vs. 10.8% (0.4%, 3). Primary care setting 21.1%) vs. 7.4% (-2.2%, 4): Hp status: Hp status 17.1%), at baseline. 53.3% Hp p<0.001 for ome 20 mg vs. (+), 46.7% Hp (-) placebo, 5) Gastroscopy performed at baseline p<0.05 for ome 10 mg vs. placebo,

p>0.05 (NS) for ran vs. placebo,

p<0.05 for ome 20 mg vs. ran,

p>0.05 (NS) for ome 10 mg vs. ran

NNT (95% CI) for ome 20 mg vs. placebo = 6 (4, 12)

NNT (95% CI) for ome 10 mg vs. placebo = 10 (5, 50)

2). primary response’ (95% CI): ome 20 mg vs. ome 10 mg vs. ran: 17.6% (4.2%, 31.0%) vs. 6.8% (- 6.7%, 20.4%) vs. 8.9% (- 4.2%, 21.9%),

p<0.025 for ome 20 mg vs. placebo,

p>0.05 (NS) for ome 10 mg and ran vs. placebo,

p>0.05 (NS) for ome 10 mg vs. ran

NNT (95% CI) for ome 20 mg vs. placebo = 6 (4, 24)

In Hp (-) patients:

Difference in response rate (percent of patients with 95% CI) compared with. placebo (

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 91 Consultation Document: Canadian Agency for Drugs and Technologies in Health

therapeutic gain )

1.) complete symptom relief) :

ome 20 mg vs. ome 10 mg vs. ran: 16.7% (3.2%, 30.1%) vs. 10.5% (-2.1%, 23.1%) vs. 12.7% (-0.7%, 26.2%),

p<0.025 for ome 20 mg vs. placebo,

p>0.05 (NS) for ome 10 mg and ran vs. placebo,

p>0.05 (NS) for ome 10 mg vs. ran

NNT (95% CI) for ome 20 mg vs. placebo = 6 (4, 32)

2). primary response ome 20 mg vs. ome 10 mg vs. ran: 5.5% (-8.0%, 19.1%) vs. -1.1% (-14.4%, 12.3%) vs. 7.9% (-5.8%, 21.7%),

p>0.05 (NS) for all pairwise comparisons Talley et al. n=1,262 ome 20 mg qd Placebo Symptom relief Proportion with complete 1998*115 1) Age: mean (range): 43 for 4 weeks at 4 weeks symptom relief (all (18-80) yrs patients): ome 20 mg vs. RCT (very 2) Pts presenting with ome 10 mg qd ome 10 mg vs. placebo: good) functional dyspepsia for 4 weeks 38% vs. 36% vs. 28%, (endoscopically normal) p=0.002 for ome 20 mg vs. with persistent or placebo, recurrent epigastric pain and/or epigastric p=0.02 for ome 10 mg vs. discomfort experienced placebo; on at least one of 3 days immediately prior to RRR for ome 20 mg vs. study entry; pts also placebo = 14% (95% required to have at CI:5.3%, 21.9%), NNT minimum 1 mo history (95% CI) = 10 (6, 27) of dyspeptic symptoms, with symptoms having NNT (95% CI) for ome 20 had to occur for at least mg vs. placebo = 10 (6, 28) 25% of days during that month NNT (95% CI) for ome 10 3) Health care centre mg vs. placebo = 13 (7, 63) setting, both GP and GI specialist recruited Proportion with complete patients symptom relief (ulcer-like 4) Hp status: 38%, 42% dyspepsia) (n=708): ome and 44.6% Hp(+) in ome 20 mg vs. ome 10 mg vs.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 92 Consultation Document: Canadian Agency for Drugs and Technologies in Health

20 mg, ome 10 mg, and placebo: 40% vs. 35% vs. placebo groups 27%, p=0.006 for ome 20 respectively mg vs. placebo, 5) Endoscopy performed for diagnosis of FD p=0.08 (NS) for ome 10 6) Excluded subjects mg vs. placebo with heartburn or acid regurgitation as the only NNT (95% CI) for ome 20 symptom w/o epigastric mg vs. placebo = 8 (5, 22) component Proportion with complete symptom relief (reflux-like dyspepsia) (n=143): ome 20 mg vs. ome 10 mg vs. placebo: 54% vs. 45% vs. 23%, p=0.002 for ome 20 mg vs. placebo,

p=0.02 for ome 10 mg vs. placebo

NNT (95% CI) for ome 20 mg vs. placebo = 3 (2, 8)

ome 10 mg vs. placebo: 5 (2, 29)

Proportion with complete symptom relief (dysmotility-like symptoms) (n=291): ome 20 mg vs. ome 10 mg vs. placebo: 32% vs. 37% vs. 31%, p=0.92 (NS) for ome 20 mg vs. placebo, p=0.33 (NS) for ome 10 mg vs. placebo Esome: esomeprazole; FD: functional dyspepsia; NS: not statistically significant; ome: omeprazole; ran: ranitidine; RRR = relative risk reduction; VAS: Visual Analog Scale; * indicates industry involvement

Context: • The Talley et al. trial included 143 patients with reflux-like dyspepsia (i.e., heartburn/acid regurgitation-predominant symptoms) who should not have been included according to the current definition of functional dyspepsia. Both doses of omeprazole were significantly more efficacious than placebo in this subgroup. In the remaining subgroups, ulcer-like dyspepsia (representing 708 of the 1262 patients enrolled) demonstrated a significantly greater response with omeprazole 20 mg but not 10 mg as compared to placebo. In those with dysmotility-like dyspepsia (291/1262) or other dyspepsia (106/1262), there were no significant differences between either dose of omeprazole versus placebo115. • A limitation of the Blum et al. study was that investigators were aware of H. pylori status117.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 93 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statement D2.1.1B: Standard-dose PPIs for 4-8 weeks are no more efficacious than standard-dose H2RAs for the improvement of symptoms in functional dyspepsia.

Voting Results: A 9%, B 45%, C 45%, D 0%, E 0%

Summary: Moayyedi et al. 113, in a good quality SR, identified one RCT in which PPIs and H2RAs were compared for the treatment of functional dyspepsia that was not heartburn-dominant. The only data that was reported for this study was the RR for relief of epigastric pain, which showed that PPIs were significantly superior to placebo. In a good quality RCT by Blum et al., standard-dose PPI was significantly more efficacious than ranitidine 150 mg daily in terms of complete symptom relief at 2 weeks in H. pylori-positive functional dyspeptics. However, there was no significant difference in H. pylori-negative subjects. There was also no difference between treatments in either subgroup in terms of the primary response, i.e., dyspeptic symptoms requiring further management. Study Outcome Population Intervention Comparator Results Type (QA) measure Moayyedi et al. 1 RCT (n=151)† lans 30 mg ran 150 mg Complete RR (PPI vs. H2RA) = 0.92 2006*113 qd for 8 bid for 8 relief of (0.71, 1.18) (NS) Patients with normal weeks weeks epigastric pain SR (good) endoscopy and over last 7 predominant epigastric days Of the 2 RCTs pain/discomfort included in the analysis of PPI vs. H2RA, ERP identified one in which the population did not have heartburn- dominant dyspepsia; only this RCT is shown here Blum et al. n=792 ome 10 mg Placebo ‘Primary In Hp (+) patients: 2000*117 1. Age ≥18 yrs: mean or 20 mg qd response’ at 2 RCT (good) age: 38-48 yrs among for 14 days weeks Difference in response rate different groups (defined as (percent of patients with 2) Patients with ran 150 mg at lack of 95% CI) compared with functional dyspepsia bedtime dyspeptic placebo: (epigastric symptoms for 14 days symptoms w/o organic disease); requiring 1.) complete symptom relief: had to have severe further ome 20 mg vs. ome 10 mg dyspeptic symptoms for management) vs. ran: 19.6% (8.6%, 30.7%) ≥3 days during 1 week vs. 10.8% (0.4%, 21.1%) vs. treatment with antacids Complete 7.4% (-2.2%, 17.1%), during screening period symptom p<0.001 for ome 20 mg vs. 3). Primary care setting relief at 2 placebo, 4): Hp status: Hp status weeks at baseline. 53.3% Hp p<0.05 for ome 10 mg vs. (+), 46.7% Hp (-) placebo, 5) Gastroscopy performed at baseline p>0.05 (NS) for ran vs. placebo,

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 94 Consultation Document: Canadian Agency for Drugs and Technologies in Health

p<0.05 for ome 20 mg vs. ran,

p>0.05 (NS) for ome 10 mg vs. ran

2). primary response (95% CI): ome 20 mg vs. ome 10 mg vs. ran: 17.6% (4.2%, 31.0%) vs. 6.8% (-6.7%, 20.4%) vs. 8.9% (-4.2%, 21.9%),

p<0.025 for ome 20 mg vs. placebo,

p>0.05 (NS) for ome 10 mg and ran vs. placebo,

p>0.05 (NS) for ome 20 mg and 10 mg vs. ran

In Hp (-) patients:

Difference in response rate (percent of patients with 95% CI) compared with placebo:

1.) complete symptom relief):

ome 20 mg vs. ome 10 mg vs. ran: 16.7% (3.2%, 30.1%) vs. 10.5% (-2.1%, 23.1%) vs. 12.7% (-0.7%, 26.2%),

p<0.025 for ome 20 mg vs. placebo,

p>0.05 (NS) for ome 10 mg and ran vs. placebo,

p>0.05 (NS) for ome 20 mg and 10 mg vs. ran

2). primary response ome 20 mg vs. ome 10 mg vs. ran: 5.5% (-8.0%, 19.1%) vs. - 1.1% (-14.4%, 12.3%) vs. 7.9% (-5.8%, 21.7%),

p>0.05 (NS) for all pairwise comparisons FD: functional dyspepsia; lans: lansoprazole; NS = not statistically significant; ome: omeprazole; ran: ranitidine; RRR = relative risk reduction; VAS: Visual Analog Scale; * indicates industry involvement

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 95 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Context: • A limitation of the Blum et al. study was that investigators were aware of H. pylori status117. • Data for the unpublished trial identified by Moayyedi et al. was obtained from company sources by the authors113. Few details regarding the design or outcomes of this study were reported in the SR.

5.2.3 Research Gaps

1) Test and treat for H. pylori infection as an initial therapeutic strategy for the treatment of uninvestigated dyspepsia (H. pylori status unknown) is not more efficacious than empirical PPI therapy. Although the ERP identified relevant studies, the panel’s votes were evenly split between ‘Accept’ and ‘Reject’, therefore this Statement was considered a Research Gap. The voting results and summary of evidence are shown below.

Voting Results: A 0%, B 36%, C 36%, D 27%, E 0%

Summary: In a good quality RCT, Manes et al.118 found that there was a trend (p=0.05) towards greater improvement in dyspepsia symptoms at 4 weeks in the empirical PPI arm as compared to the test and treat arm. At 6 and 12 months follow-up, dyspepsia symptom scores were improved to a significantly greater extent in the test and treat arm than in the empirical PPI arm, and demonstrated a lower overall rate of endoscopy. However, test and treat subjects reported significantly fewer symptom-free days. Study Outcome Population Intervention Comparator Results Type(QA) measure Manes et n=219 Hp test and Empirical PPI: Symptom Test and treat vs. Empirical PPI al. 2003118 1) Age: mean (range): 38 treat: 1 wk ome 20 mg qd relief (18-45) yrs triple for 4 wks assessed by Proportion with symptom RCT 2) Pts presenting with eradication dyspepsia improvement at 4 wks: (good) uninvestigated upper (ome 20 mg, severity 71% (95% CI: 61-79%) vs. 83% abdominal symptoms clar 500 mg, score every (95% CI: 74-89%); p=0.05 3) Hospital GI unit & tini 500 2 mos; NNT (95% CI) = 8 (4, 104) 4) Hp status: assessed mg all bid) use of with C-UBT, if (+), if Hp +ve, medical Dyspepsia symptom scores at 6 eradication therapy, if (-), ome 20 mg resources; and 12 mos: Test and treat had go PPI-only therapy qd for 4 clinical significantly better scores at 6 & 5) Control group (PPI weeks if Hp outcome 12 mos vs. PPIs alone (p<0.001 empirical treatment) –ve for both 6 & 12 mos without Hp testing comparisons) (data shown in 6) Endoscopy not figure) performed at study entry Mean # of days w/o symptoms: test and treat vs. PPI: 139.3 (95% 117.9-160.7), vs. 231.5 (95% CI: 205.7, 257.5) p<0.0001

Overall endoscopy rates at 1 year follow-up:

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 96 Consultation Document: Canadian Agency for Drugs and Technologies in Health

test and treat vs. PPI

55% (95% CI, 46-65%) vs. 88% (95% CI : 80- 93%) p<0.0001 NNT (95% CI) = 3 (2, 5) clar: clarithromycin; C-UBT: C13-urea breath test; NS = not statistically significant; ome: omeprazole; tini: tinidazole; * indicates industry involvement

Context: • The rate of H. pylori infection was 61% in the H. pylori test and treat arm of the Manes et al. trial118. The ERP noted that this is higher than the overall prevalence of H. pylori infection in Canada, although certain populations may have a higher risk of infection (e.g., certain ethnic backgrounds, recent immigrants, urban populations, etc.).

2) Standard-dose PPI therapy is more efficacious than antacids/alginates at reducing dyspeptic symptoms in patients with uninvestigated dyspepsia. There was no evidence from studies in which heartburn-dominant dyspepsia was excluded.

3) Standard-dose PPI therapy is more efficacious than standard-dose H2RAs at reducing dyspeptic symptoms in patients with uninvestigated dyspepsia (H. pylori status unknown). There was no evidence from studies in which heartburn-dominant dyspepsia was excluded.

4) Continuous maintenance therapy with standard-dose PPIs is more efficacious than continuous maintenance therapy with H2RAs in uninvestigated dyspepsia. No evidence was identified for this Statement.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 97 Consultation Document: Canadian Agency for Drugs and Technologies in Health

5.3 PUD

5.3.1 List of Evidence Statements and Research Gaps

P1: H. pylori eradication P1.1: PPI-based eradication triple therapy P1.1.1: Differences among PPIs Evidence Statements

P1.1.1A: All PPIs have similar efficacy in triple therapy regimens for H. pylori eradication.

P1.1.2 : Double-dose vs. standard-dose PPIs Evidence Statements

P1.1.2A: Standard-dose PPI administered twice daily is more efficacious than standard-dose PPI administered once daily (when used in a PAC triple-therapy regimen) for H. pylori eradication.

P1.1.2B: Standard-dose PPI administered twice daily is no more efficacious than standard-dose PPI administered once daily (when used in a PMC triple-therapy regimen) for H. pylori eradication.

P1.1.3: Duration of triple-therapy Evidence Statements

P1.1.3A: H. pylori eradication therapy with triple therapy (PAC and PMC) for 7 days is less efficacious than 14 days.

P1.1.3B: H. pylori eradication therapy with triple therapy (PAC and PMC) for 7 days is as efficacious as 10 days. P1.2: PPI treatment for peptic ulcer following H. pylori eradication therapy Evidence Statements

P1.2.1A: Continued treatment with PPI after a course of H. pylori eradication therapy does not produce higher ulcer healing rates than eradication therapy alone in H. pylori-infected patients with uncomplicated duodenal ulcer. This does not apply to gastric ulcers.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 98 Consultation Document: Canadian Agency for Drugs and Technologies in Health

P2: NSAID-associated ulcer P2.1: Treatment of NSAID-associated ulcer P2.1.1: PPIs vs. H2RAs and misoprostol Evidence Statements

P2.1.1A: Standard-dose PPI therapy for 4-8 weeks produces higher healing rates of NSAID-associated ulcers than H2RAs, when NSAIDS are continued.

P2.1.1B: Standard-dose PPI therapy for 4-8 weeks produces higher healing rates of NSAID-associated ulcers than 800 mcg/day misoprostol, when NSAIDs are continued. P2.1.2: Differences among PPIs Evidence Statements

P2.1.2A: Different PPIs produce similar healing rates of NSAID-associated ulcer.

P2.2: Prevention of NSAID-associated ulcer P2.2.1: PPIs vs. H2RAs, misoprostol, and placebo Evidence Statements

P2.2.1A: Standard-dose PPIs are more efficacious than placebo for the prevention of NSAID-associated endoscopic gastric and duodenal ulcers.

P2.2.1B: Standard-dose PPIs are more efficacious than standard-dose H2RAs for the secondary prevention of NSAID-associated endoscopic gastric and duodenal ulcers.

P2.2.1C: In patients with a history of ulcer, standard-dose PPIs have similar efficacy to misoprostol 400-800 mcg/day for the prevention of NSAID associated endoscopic gastric and duodenal ulcers.

P2.2.2: PPIs and non-selective NSAIDs vs. COX-2 NSAIDs Evidence Statements

P2.2.2A: There is no difference in ulcer recurrence and bleeding rates between COX-2 selective NSAIDs and the combination of PPI and conventional NSAIDs in patients with previous NSAID- associated upper GI bleeding. P2.2.3: PPIs vs. H. pylori eradication Evidence Statements

P2.2.3A: Standard-dose PPIs are more efficacious than H. pylori eradication in reducing the risk of GI complications in Hp-positive, NSAID-associated ulcers with a history of NSAID-associated ulcer bleeding.

P2.2.4: Differences among PPIs Evidence Statements

P2.2.4A: Different PPIs reduce ulcer risk to a similar degree when given to NSAID users for ulcer prophylaxis.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 99 Consultation Document: Canadian Agency for Drugs and Technologies in Health

P3: ASA-associated ulcer Evidence Statements

P3.1: In patients in whom H. pylori has been eradicated, standard-dose PPIs are more efficacious than placebo for the secondary prevention of ulcer complications in users of low dose (≤325 mg/day) ASA.

P3.2: In H. pylori-negative patients who have a history of ulcer bleeding on low-dose ASA alone, the combination of low-dose ASA and a PPI is associated with a lower risk of recurrence of ulcer complications at 1 year as compared to clopidogrel alone.

Research Gaps

1) H. pylori eradication therapy with triple therapy (PAC and PMC) for 10 days is as efficacious as 14 days.

2) In patients with uncomplicated gastric ulcers, continued treatment with a PPI after a course of H. pylori eradication therapy produces higher ulcer healing rates than eradication therapy alone in H. pylori-infected patients.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 100 Consultation Document: Canadian Agency for Drugs and Technologies in Health

5.3.2 Evidence Statements, ERP voting results, and Summaries of evidence

P1: H. pylori eradication P1.1: PPI-based eradication triple therapy P1.1.1: Differences among PPIs Evidence Statement P1.1.1A: All PPIs have similar efficacy in triple therapy regimens for H. pylori eradication.

Voting Results: A 100%, B 0%, C 0%, D 0%, E 0%

Summary: Five good quality systemic reviews90,119-122and two poor quality systematic reviews95,123found that, in general, there were no significant differences in Hp eradication rates among the various PPIs. Two good quality RCTs by Hawkey et al.124 and Wong et al.125 also found no significant difference between omeprazole with rabeprazole in 7-day PPI-based triple therapy regimens. In another good quality RCT, Tulassay et al.126 did not find a significant difference between omeprazole and esomeprazole.

However, a few studies have noted differences in eradication rates. In one systematic review that did not pool RCT results due to heterogeneity90, one RCT found a lower eradication rate with pantoprazole 40 mg versus omeprazole 40 mg or pantoprazole 80 mg, and another study observed a lower rate with rabeprazole 20 and 40 mg versus lansoprazole 30 mg. In a good quality RCT by Spinzi et al.127, lansoprazole was found to be somewhat more effective than omeprazole in the PAC regimen. Finally, in a poor quality RCT by Hsu et al.128, esomeprazole as part of the PAC regimen was more effective than pantoprazole. Study Outcome Population Intervention Comparator Results Type (QA) measure McDonagh et 20 RCTs (n not PPI PPI Hp No diffs b/w various agents in al. 200590 provided) eradication SRs and MAs. Hp +ve adults rate RCTs were heterogeneous in SR (good) terms of populations, regimens, duration, ulcer status, methods of assessing Hp eradication and PPI dose. Hp eradication ranged from a low of 62.5% (rab 20 mg) to a high of 100% (pant 40 mg). One study found lower rate for pant 40 mg vs. ome 40 mg or pant 80 mg, and another found lower rate for rab 20 and 40 mg vs. lans 30 mg. Gisbert & 4 RCTs (n=1,292) esome- ome-based Hp Mean erad rates (95% CI): esome Pajares PUD and NUD adults based eradication eradication 85% (82-87%) vs. ome 82% 2004119 treated for Hp eradication therapy rate (79%-85%); OR (95% CI) = 1.19 infection therapy (0.81-1.74), NS SR (good) Gisbert et al. 7 RCTs (n=1,137) pant-based ome and lans- Hp Mean erad rates (95% CI): Pant 2004120 Adults with PUD or eradication based eradication 83% (80-86%) vs. other PPIs 81% NUD treated for Hp therapy eradication rate (77%-84%); OR (random effects) SR (good) infection therapy (95% CI) = 1 (0.61-1.64), NS - similar results for separate

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 101 Consultation Document: Canadian Agency for Drugs and Technologies in Health

comparisons (Pant vs. Ome and Pant vs. Lans)

Gisbert et al. 12 RCTs (n=2,226) rab-based ome and lans- Hp Mean erad rates (95% CI): rab 2003121 Adults with PUD or eradication based eradication 79% (77-82%) vs. other PPIs 77% NUD treated for Hp therapy eradication rate (75%-79%); OR (95% CI) = 1.16 SR (good) infection therapy (0.94-1.43), NS - similar results for separate comparisons (rab vs. ome and rab vs. lans)

rab 10 mg vs. other PPIs at std dose (7 RCTs): OR = 1.21 (0.75- 1.95), NS Vergara et al. 6 RCTs (n=1,085) ome in lans in triple Hp Mean erad rate: ome 74.7% vs. 2003122 Adults treated for Hp triple eradication eradication lans 76%; Peto OR (95% CI) = infection eradication therapy rate 0.91 (0.69-1.21), NS SR (good) therapy 4 RCTs (n=825) ome in rab in triple Hp Mean erad rate: ome 77.9% vs. Adults treated for Hp triple eradication eradication rab 81.2%; Peto OR (95% CI) = infection eradication therapy rate 0.81 (0.58-1.15), NS therapy 2 RCTs (n=833) ome in esome in Hp Mean erad rate: ome 87.7% vs. Adults treated for Hp triple triple eradication esome 89%; Peto OR (95% CI) = infection eradication eradication rate 0.89 (0.58-1.35), NS therapy therapy 3 RCTs (n=550) lans in rab in triple Hp Mean erad rate: lans 81% vs. rab Adults treated for Hp triple eradication eradication 85.7%; Peto OR (95% CI) = 0.77 infection eradication therapy rate (0.48-1.22), NS therapy Klok et al. 5 RCTs (n=860) lans 60 mg ome 40 mg in Hp Pooled RR (95% CI) = 1.05 200395 concerning Hp in Hp Hp eradication (0.95-1.15), NS eradication eradication eradication rate SR (poor) therapy therapy 2 RCTs (n=196) lans 30 mg ome 40 mg in Hp Pooled RR (95% CI) = 1.06 concerning Hp in Hp Hp eradication (0.92-1.23), NS eradication eradication eradication rate therapy therapy 1 RCT (n=149) lans 30 mg ome 20 mg in Hp RR (95% CI) = 1.10 (0.76-1.57) concerning Hp in triple triple eradication eradication eradication eradication rate therapy therapy 1 RCT (n=27) lans 30 mg pant 40 mg in Hp RR (95% CI) = 0.62 (0.17-2.19) in dual dual eradication eradication eradication rate therapy therapy 2 RCTs (n=354) rab 40 mg lans 60 mg in Hp Pooled RR (95% CI) = 1.04 concerning Hp in triple triple eradication (0.95-1.13), NS eradication eradication eradication rate therapy therapy 2 RCTs (n=314) rab 20 mg ome 40 mg in Hp Pooled RR (95% CI) = 0.99 concerning Hp in Hp Hp eradication (0.87-1.14), NS eradication eradication eradication rate therapy therapy 2 RCTs (n=311) rab 40 mg ome 40 mg in Hp Pooled RR (95% CI) = 1.02 concerning Hp in triple triple eradication (0.93-1.13), NS eradication eradication eradication rate

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 102 Consultation Document: Canadian Agency for Drugs and Technologies in Health

therapy therapy 2 RCTs (n=213) pant 40 mg ome 40 mg in Hp Pooled RR (95% CI) = 0.94 concerning Hp in triple triple eradication (0.86-1.03), NS eradication eradication eradication rate therapy therapy 2 RCTs (n=349) pant 80 mg ome 40 mg in Hp Pooled RR (95% CI) = 0.96 concerning Hp in triple triple eradication (0.88-1.04), NS eradication eradication eradication rate therapy therapy 2 RCTs (n=833) esome 40 ome 40 mg in Hp Pooled RR (95% CI) = 1.00 concerning Hp mg in triple triple eradication (0.90-1.11), NS eradication eradication eradication rate therapy therapy Moayyedi 10 RCTs (n=1,348) PPI-triple PPI-triple Hp 2% difference in eradication rate and Murphy Patients with Hp therapy therapy with eradication in favour of ome; p=0.35, NS 2001123 infection (ulcer status with ome lans rate not provided) MA (poor) Hsu et al. 200 Hp +ve adults EAC: esome PAC: pant 40 Hp EAC vs. PAC: 94% vs. 82%, 2005128 with endoscopically 40 mg, mg, amox 1g, eradication p=0.009 verified PUD or amox 1g & & clar 500 rate 8 weeks NNT (95% CI) = 8 (5, 32) RCT (poor) gastritis clar 500 mg, mg, & all bid after all bid for 7 for 7 days, completion of days, then then pant 40 therapy pant 40 mg/day x 3 mg/day x 3 weeks for pts weeks for with PUD or pts with antacids x 3 PUD or weeks for pts antacids x 3 with gastritis weeks for pts with gastritis Hawkey et al. 348 adults with PUD RAC (rab OAC: (ome Hp Hp eradication rate: RAC + RMC 2003*124 & Hp-positive 40 mg/d + 40 mg/d + eradication pooled rate: 77% vs. OAC + amox 2 g/d amox 2 g/d + rate at 4 wks OMC pooled rate: 75%; RCT (good) + clar 1 g/d) clar 1 g/d) for difference (95% CI) = 1.5% (- 7 days 7.4%, 10.4%), NS RMC (rab 40 mg/d + OMC: (ome met 800 40 mg/d + mg/d + clar met 800 mg + 1 g/d) for 7 clar 1 g/d) for days 7 days Wong et al. 173 adults with Hp RAC7: rab 7-day OAC: Hp Hp eradication rate and (95% CI): 2001*125 infection & no active 10 mg, ome 20 mg, eradication RAC7 vs. RAC3 vs. OAC was: bleeding amox 1 g, amox 1 g, clar at 6 wks 88% (77%, 95%) vs. 72% (59%, RCT (good) clar 500 mg, 500 mg, each 83%) vs. 82% (70%, 91%), NS each bid for bid for 7 difference b/w groups 7 days days

RAC3: rab 20 mg, amox 1 g, clar 500 mg, each bid for 3 days

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 103 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Tulassay et 433 adults with DU EAC group: OAC group: Hp Hp eradication rate and (95% CI): al. & Hp-positive esome 20 ome 20 mg, eradication EAC vs. OAC was: 86% (81%, 2001*126 mg, amox 1 amox 1 g, clar at 4 to 6 wks 90%) vs. 88% (83%, 92%); g and clar 500 mg, all p>0.05, NS RCT (good) 500 mg, all bid for 7 bid for 7 days, then days ome 20 mg qd for 3 wks Spinzi et al. 356 adults with DU LAC group: OAC group: Hp Hp eradication rate and (95% CI): 1998127 or GU & Hp-positive lans 30 mg ome 20 mg eradication LAC vs. OAC was: bid, amox 1 bid, amox 1 g at ≥ 6 wks 72% (65%, 78%) vs. 62% (54%, RCT (good) g bid, clar bid, clar 500 69%), p=0.043 500 mg bid mg bid for 7 NNT (95% CI) = 10 (5, 341) for 7 days days amox: amoxicillin; clar: clarithromycin; esome: esomeprazole; lans: lansoprazole; met: metronidazole; ome: omeprazole; pant: pantoprazole; rab: rabeprazole; * indicates industry involvement

P1: H. pylori eradication P1.1: PPI-based eradication triple therapy P1.1.2: Double-dose vs. standard-dose PPIs Evidence Statement P1.1.2A: Standard-dose PPI administered twice daily is more efficacious than standard-dose PPI administered once daily (when used in a PAC triple-therapy regimen) for H. pylori eradication.

Voting Results: A 18%, B 55%, C 27%, D 0%, E 0%

Evidence Statement P1.1.2B: Standard-dose PPI administered twice daily is no more efficacious than standard-dose PPI administered once daily (when used in a PMC triple-therapy regimen) for H. pylori eradication.

Voting Results: A 0%, B 73%, C 27%, D 0%, E 0%

Summary: A good quality meta-analysis by Vallve et al.129 showed that the H. pylori eradication rate was significantly higher with double dose PPI than single dose in the PAC regimen, but not in the PMC regimen containing 250 mg bid clarithromycin and 500 mg bid metronidazole. However, the meta- analysis for the latter consisted of only 2 studies with a total of 304 patients. Study Outcome Population Intervention Comparator Results Type (QA) measure Vallve et al. Over all: Double doses of Single doses of Hp eradication Overall Hp eradication rate 2002129 11 RCTs PPI (ome, lans, PPI (ome, lans, rate and (95% CI): double dose PPI (n=2,391) pant or rab), pant or rab), vs. single dose: 83.9% (81%, MA (good) clar (any dose) clar (any dose) 85%) vs. 77.7% (72%, 77%). Patients with bid, and either bid, and either Odds ratio was 1.51 (95% CI: Hp infection amox or met amox or met 1.23, 1.85); p<0.01 (ulcer status (any doses) bid (any doses) bid NNT (95% CI) = 16 (12, 30) not provided) for any duration for any duration Hp eradication rate (95% CI) PAC regimen in PAC regimen: double dose

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 104 Consultation Document: Canadian Agency for Drugs and Technologies in Health

(Most used PPI vs. single dose: 84.5% PPI, clar 500 (82%, 87%) vs. 75.9% (73%, mg bid and 79%). Odds ratio was 1.73 amox 1 g bid): (95% CI: 1.38, 2.18). 9 RCTs (n= 1,870) Hp eradication rate (95% CI) in PMC regimen: double dose PMC regimen PPI vs. single dose: 74.8% (PPI, clar 250 (67%, 81%) vs. 74.5% (67%, mg bid, met 81%). Odds ratio was 1.01 500 mg bid): 2 (95% CI: 0.60, 1.69); p>0.05, RCTs (n=304) NS. amox: amoxicillin; clar: clarithromycin; lans: lansoprazole; met: metronidazole; ome: omeprazole; pant: pantoprazole; rab : rabeprazole

Context: • The difference between single vs. double dose PPI in the PMC triple therapy was not statistically significant (eradication rate was 74.5% vs. 74.8%, with an odds ratio of 1.01). It was based on 2 RCTs with 304 patients in total. • The difference between single vs. double dose PPI in the PAC triple therapy was statistically significant. (eradication rate was 75.9% vs. 84.5%, with an odds ratio of 1.73). It was based on 9 RCTs with 1,870 patients in total.

P1: H. pylori eradication P1.1: PPI-based eradication triple therapy P1.1.3: Duration of triple-therapy Evidence Statement P1.1.3A: H. pylori eradication therapy with triple therapy (PAC and PMC) for 7 days is less efficacious than 14 days.

Voting Results: A 36%, B 27%, C 9%, D 27%, E 0%

Evidence Statement P1.1.3B: H. pylori eradication therapy with triple therapy (PAC and PMC) for 7 days is as efficacious as 10 days.

Voting Results: A 0%, B 46%, C 27%, D 27%, E 0%

Summary: A poor quality meta-analysis by Calvet et al.130 showed that the H. pylori eradication rate was significantly higher with a 14-day PPI triple therapy as compared to 7 days. However, there were no significant differences in eradication rates between 7 and 10-day, or 10 and 14-day regimens. Study Outcome Population Intervention Comparator Results Type (QA) measure Calvet et al. 13 RCTs PPI, clar, and PPI, clar, and Hp eradication Hp eradication rate and 2000130 (n=906) either amox or either amox or rate (95%CI): 14 days vs. 7 days met for 10 to 14 met for 7 days was: 81% (77%, 85%) vs. 72% MA (poor) Patients with days (doses not (doses not (68%, 76%). Overall OR and Hp infection specified) specified) (95% CI) = 0.62 (0.45, 0.84) NNT (95% CI) = 10 (6, 28)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 105 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Hp eradication rate and (95% CI): 10 days vs. 14 days: 82% (77%, 86%) vs. 84% (79%, 89%); p>0.05, NS

Hp eradication rate and (95% CI): 7 days vs. 10 days was: 80% (71%, 86%) vs. 83% (75%, 89%); p>0.05, NS

NNT for 10-14 days as compared to 7 days to obtain one extra cure = 6 to 23

Context: • The Calvet et al. meta-analysis130 was updated in the 2004 NICE dyspepsia guidelines31. For both PAC and PMC regimens, 7-day therapy was significantly less likely to cure H. pylori than 14-day therapy. Pooled eradication rates were 67.8% for 7-day versus 77.2% for 14-day therapy (RD:9%, 95% CI: 5% to 14%; NNT:11, 95% CI: 7 to 20). Statistically, therapy of longer duration remained significantly more effective for each of PAC and PMC when considered separately. The updated meta-analysis also showed a statistically non-significant trend towards greater eradication for 10-day as compared to 7-day therapy. It was not clear from the description in the NICE guideline whether ITT data was used in the meta-analyses. • It was observed that the vote results did not yield a clear message, suggesting that the evidence is not clear that regimens greater than 7 days produce higher eradication rates (also see Research Gap #1 regarding H. pylori eradication efficacy of 10 vs. 14 days, below).

P1: PPIs in H. pylori eradication triple therapy regimens and PUD P1.2: PPI treatment for peptic ulcer following H. pylori eradication therapy Evidence Statement P1.2.1A: Continued treatment with PPI after a course of H. pylori eradication therapy does not produce higher ulcer healing rates than eradication therapy alone in H. pylori-infected patients with uncomplicated duodenal ulcer. This does not apply to gastric ulcers.

Voting Results: A 25%, B 75%, C 0%, D 0%, E 0%

Summary: This statement is based on results from one good quality SR131. Gisbert et al. found no significant difference in GU or DU healing rates in patients treated with 7-days PPI-based Hp eradication triple therapy alone, as compared to those treated with the same eradication therapy followed by PPI alone for a further 2-4 weeks131. Study Outcome Population Intervention Comparator Results Type (QA) measure Gisbert et 6 RCTs 7-day PPI-based 7-day PPI-based Ulcer healing Mean healing rate (95% CI) al. 2005131 (n=862) triple therapy with triple therapy, then rate = 92% (89-96%) with triple DU or GU with no further PPI PPI alone for 2-4 therapy then PPI alone vs. SR (good) Hp infection; treatment weeks 91% (87-95%) with triple ulcer size or therapy alone; Pooled OR

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 106 Consultation Document: Canadian Agency for Drugs and Technologies in Health

presence of (95%CI) was 1.11 (0.71- complications 1.74), NS not specified - only one study assessed both DU and GU, others only DU (similar results when DU only studies were assessed)

Context: • It was noted that if H. pylori is eradicated then there is no further treatment required and that the rate of H. pylori-infected patients with duodenal ulcer is decreasing.

P2: NSAID-associated ulcer P2.1: Treatment of NSAID-associated ulcer P2.1.1: PPIs vs. H2RAs and misoprostol Evidence Statement P2.1.1A: Standard-dose PPI therapy for 4-8 weeks produces higher healing rates of NSAID-associated ulcers than H2RAs, when NSAIDs are continued.

Voting Results: A 37%, B 63%, C 0%, D 0%, E 0%

Summary: Two RCTs, both of good quality, compared PPI therapy with ranitidine 300 mg daily for the healing of NSAID-associated ulcers132,133. Agrawal et al. reported that the rate of gastric ulcer healing at 4 weeks was significantly higher for both lansoprazole 15 mg and 30 mg per day at both 4 and 8 weeks132. Yeomans et al. reported that the overall success rate (defined as a composite of healed ulcer, fewer than 5 erosions, and no more than mild dyspepsia) at 8 weeks was significantly higher for both omeprazole 20 mg and 40 mg per day as compared to ranitidine 300 mg/day. Both omeprazole doses also healed a greater proportion of total ulcers and gastric ulcers than ranitidine 300 mg/day. Duodenal ulcer healing rates were higher with the two omeprazole doses than ranitidine, although the difference was not statistically significant for omeprazole 40 mg/day.

Study Outcome Population Intervention Comparator Results Type (QA) measure Agrawal et al. 353 patients lans: 15 ran 300 GU healing Rate of ulcer healing at wk 4: 2000*132 with GU ≥ 5 mg/day; lans mg/day rate at 4 and 8 lans 15 mg: 47% vs. lans 30 mg: 57% mm, using an 30 mg/day wks vs. ran: 30%; RCT (good) NSAID ≥ lans 15 mg vs. ran: p < 0.01 1month; lans 30 mg vs. ran: p < 0.001 excluded lans 15 mg vs. lans 30 mg: p > 0.05 patients with (NS) active bleeding or perforated Rate of ulcer healing at wk 8: ulcers lans 15 mg: 69% vs. lans 30 mg: 73% vs. ran: 53%; NSAIDs lans 15 mg vs. ran: p = 0.01 continued lans 30 mg vs. ran: p < 0.01 throughout trial lans 15 mg vs. lans 30 mg: p > 0.05 (NS) Yeomans et al. 541 patients ome 20 ran 300 Treatment Overall success rate: 1998*133 with DU or GU mg/day; mg/day success rate at ome 20 mg: 80% vs. ome 40 mg: ≥ 3 mm or >10 ome 40 8 weeks 79% vs. ran: 63%;

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 107 Consultation Document: Canadian Agency for Drugs and Technologies in Health

RCT (good) erosions, mg/day (healing of ome 20 mg vs. ran: p < 0.001 receiving ulcer, < 5 ome 40 mg vs. ran: p = 0.001 NSAIDs; erosions, no ome 40 mg vs. ome 20 mg: p > 0.05 excluded more than mild (NS) patients with dyspepsia) NNT (95% CI) for ome 20 mg vs. ran pyloric stenosis = 6 (4, 13) or major active NNT (95% CI) for ome 40 mg vs. ran GI bleeding = 6 (4, 15) (120 patients had DU and % with DU healing: 421 patients ome 20 mg: 92% vs. ome 40 mg: had GU or both 88% vs. ran: 81%; GU and DU). ome 20 mg vs. ran: p < 0.03 ome 40 mg vs. ran: p > 0.05 (NS) NSAIDs ome 40 mg vs. ome 20 mg: p > 0.05 continued (NS) throughout trial NNT (95% CI) for ome 20 mg vs. ran = 9 (4, ∞)

% with GU healing: ome 20 mg: 84% vs. ome 40 mg: 87% vs. ran: 64%; ome 20 mg vs. ran: p < 0.001 ome 40 mg vs. ran: p < 0.001 ome 40 mg vs. ome 20 mg: p > 0.05 (NS) NNT (95% CI) for ome 20 mg vs. ran = 5 (3, 17) NNT (95% CI) for ome 40 mg vs. ran = 4 (3, 11) lans: lansoprazole; ome: omeprazole; ran: ranitidine; * indicates industry involvement

Context: • Both studies assessed the healing of endoscopic ulcers. The relevance of these findings to more clinically important outcomes such as healing of symptomatic or complicated ulcers is uncertain. • Studies showed that standard-dose PPIs were significantly better than standard-dose ranitidine in healing gastric ulcers. Double dose PPI was not superior to single-dose in these studies. • Yeomans et al.133 looked at PPIs vs. ranitidine in both gastric and duodenal ulcer patients and the overall success rate was better than that in Agrawal et al.132 study which studied gastric ulcer patients.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 108 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statement P2.1.1B: Standard-dose PPI therapy for 4-8 weeks produces higher healing rates of NSAID-associated ulcers than 800 mcg/day misoprostol, when NSAIDs are continued.

Voting Results: A 0%, B 76%, C 12%, D 12%, E 0%

Summary: Omeprazole and misoprostol had similar ‘success rates’ in a good quality RCT.134 As compared to misoprostol 800 mcg/day, DU healing occurred in a significantly greater proportion of omeprazole treated subjects at 8 weeks. GU healing occurred in significantly more subjects treated with omeprazole 20 mg as compared to misoprostol, but not in those treated with omeprazole 40 mg. Study Outcome Population Intervention Comparator Results Type (QA) measure Hawkey et 935 NSAID ome 20 mis 800 Success rate: Success rate at wk 8: al. 1998*134 users with mg/day, ome mcg/day (ulcer healed, ome 20 mg: 76% vs. ome 40 mg: DU, GU or 40 mg/day <5 erosions, no 75% vs. mis 800 mcg: 71%; RCT (good) both (≥ 3 more than mild ome 40 mg vs. mis: p > 0.05 (NS) mm) or >10 dyspepsia), DU ome 20 mg vs. mis: p > 0.05 (NS) erosions; and GU ulcer ome 40 mg vs. ome 20 mg: p > 0.05 excluded healing at 8 (NS) patients with weeks clinically Rate of DU healing at wk 8: important ome 20 mg: 93% vs. ome 40 mg: upper GI 89% vs. mis 800 mcg: 77%; bleeding or ome 40 mg vs. mis: p < 0.001 pyloric ome 20 mg vs. mis: p < 0.001 stenosis ome 40 mg vs. ome 20 mg: p-value not reported NSAIDs NNT (95% CI) for ome 40 mg vs. mis continued = 8 (4, ∞) throughout NNT (95% CI) for ome 20 mg vs. mis trial = 6 (4, 29)

Rate of GU healing at wk 8: ome 20 mg: 87% vs. ome 40 mg: 80% vs. mis 800 mcg: 73%; ome 40 mg vs. mis: p > 0.05 (NS) ome 20 mg vs. mis: p = 0.004 ome 40 mg vs. ome 20 mg: p-value not reported NNT (95% CI) for ome 20 mg vs. mis = 7 (4, 25) mis:misoprostol; ome: omeprazole; * indicates industry involvement

Context: • The single study that compared the efficacy of PPIs vs. misoprostol assessed healing rates of endoscopic ulcers, as well as a combined endpoint of endoscopic ulcer, erosions, and dyspepsia134. The relevance of these findings to more clinically important outcomes such as healing of symptomatic or complicated ulcers is uncertain.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 109 Consultation Document: Canadian Agency for Drugs and Technologies in Health

P2: NSAID-associated ulcer P2.1: Treatment of NSAID-associated ulcer P2.1.2: Differences among PPIs Evidence Statement P2.1.2A: Different PPIs produce similar healing rates of NSAID-associated ulcer.

Voting Results: A 0%, B 100%, C 0%, D 0%, E 0%

Summary: A good systematic review 90 found that there were no head-to-head comparisons of the various PPIs for the healing of NSAID-associated ulcer. However, indirect comparisons using data from trials comparing PPIs to H2RAs or misoprostol did not indicate intraclass differences in healing efficacy. Also, similar symptomatic improvement was observed for standard and double doses of omeprazole or lansoprazole. It should be noted that formal statistical methods for indirect comparisons were not employed in this review. Study Outcome Population Intervention Comparator Results Type (QA) measure McDonagh 3 RCTs PPI H2RA or Endoscopic No head-to-head studies assessing PPI et al. 200590 (n=1,829) mis healing, intraclass differences were found. NSAID users; symptoms In indirect comparisons: For GU healing SR (good) ulcer size or at 8 weeks, all CI’s for PPI vs. H2RA and presence of PPI vs. mis, comparisons overlapped; complications similar symptom improvement observed not specified; for ome 20 and 40 mg, and lans 15 mg continued and 30 mg NSAID use during trials not specified mis: misoprostol

Context: • No direct comparisons (i.e., head-to-head trials) were found. There may have been heterogeneity in the populations studied in the trials used in the indirect comparison by McDonagh et al.90

P2: NSAID-associated ulcer P2.2: Prevention of NSAID-associated ulcer P2.2.1: PPIs vs. H2RAs, misoprostol, and placebo Evidence Statement P2.2.1A: Standard-dose PPIs are more efficacious than placebo for the prevention of NSAID-associated endoscopic gastric and duodenal ulcers.

Voting Results: A 57%, B 43%, C 0%, D 0%, E 0%

Summary: According to two good quality systematic reviews, PPIs were more effective than placebo for the prevention of NSAID-associated endoscopic gastric and duodenal ulcers135,136. Study Population Intervention Comparator Outcome measure Results Type(QA)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 110 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Hooper et 6 RCTs (n=1,358) PPI placebo Primary: Serious GI Primary outcomes: insufficient al. 2004135 NSAID users; complications data for most primary outcomes: primary/secondary (hemorrhage, recurrent symptomatic ulcers: RR (95% CI) SR (good) prophylaxis not GI bleeding, was 0.09 (0,0.5), but significance specified perforation, lost on sensitivity analysis obstruction, melena, NNT (95% CI) =85 (77, 154) death from any of these), symptomatic Secondary Outcomes: endoscopic ulcers, HRQL, ulcers: RR (95% CI) was 0.37 mortality, serious CV (0.3,0.5) or renal illness NNT (95% CI) =8 (7, 10)

Secondary: totals GI symptoms, endoscopic ulcers (≥3mm), anemia, occult bleeding, total dropouts, dropouts due to GI symptoms Rostom et 5 RCTs PPI Placebo Endoscopic ulcer DU RR (95% CI) = 0.19 (0.09, al. 2002136 (n = 1,216) 0.37) Subjects requiring NNT (95% CI) = 13 (12, 16) SR (good) chronic NSAID use taking GU RR (95% CI) = 0.40 (0.32, NSAIDs > 3 0.51) weeks, w/ or w/o NNT (95% CI) = 6 (6, 8) past ulcer Total Endoscopic Ulcer: Peto OR (95% CI) = 0.23 (0.18, 0.31) NNT (95% CI) = 5 (4, 5)

Results similar for primary and secondary prophylaxis trials * indicates industry involvement; HRQL: Health-Related Quality of Life.

Context: • Data on ulcer complications, the most clinically relevant outcome, were sparse. Most studies identified in the SRs used endoscopic ulcers as an outcome, which has commonly been used as a surrogate for ulcer complications.

Evidence Statement P2.2.1B: Standard-dose PPIs are more efficacious than standard-dose H2RAs for the secondary prevention of NSAID-associated endoscopic gastric and duodenal ulcers.

Voting Results: A 14%, B 86%, C 0%, D 0%, E 0%

Summary: In their good quality SR, Rostom et al. reported the results of one RCT that showed standard-dose omeprazole was superior to standard-dose ranitidine in preventing both duodenal and gastric ulcer recurrence136. Study Population Intervention Comparator Outcome measure Results Type(QA)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 111 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Rostom et 1 RCT ome 20 ran 150 mg Endoscopic ulcer DU RR (95% CI) = 0.11 (0.01, 0.89) al. 2002136 (n = 425) mg/day bid NNT (95% CI) = 27 (25, 217) Subjects SR (good) requiring GU RR (95% CI) = 0.32 (0.17, 0.62) chronic NSAID NNT (95% CI) = 10 (8, 17) use taking NSAIDs > 3 Total endoscopic ulcers RR (95% CI) = weeks; past 0.28 (0.15, 0.51) ulcer history NNT (95% CI) = 7 (6, 10) not reported Ome: omeprazole; ran: ranitidine

Context: • Data on ulcer complications, the most clinically relevant outcome, were not identified in the SR. Studies assessed endoscopic ulcers, a surrogate for ulcer complications. • Studies comparing single-dose PPI vs. double-dose H2RAs in preventing NSAID-associated ulcers are lacking.

Evidence Statement P2.2.1C: In patients with a history of ulcer, standard-dose PPIs have similar efficacy to misoprostol 400-800 mcg/day for the prevention of NSAID associated endoscopic gastric and duodenal ulcers.

Voting Results: A 0%, B 56%, C 44%, D 0%, E 0%

Summary: One poor quality RCT137 and one good quality RCT134 were identified for this Statement. Graham et al.137 reported that misoprostol 800 mcg /day was significantly better than lansoprazole 30 mg daily for preventing gastric ulcer in H. pylori-negative adults who were long-term users of NSAIDs and had a history of endoscopically-documented gastric ulcer with or without coexisting duodenal ulcer or gastrointestinal bleeding. However, patients taking misoprostol 800 mcg/day had significantly poorer compliance (p<0.001) and more adverse effects (p<0.05 to 0.006 for all adverse events, except for ‘nights with abdominal pain’) than lansoprazole 30 mg. Hawkey et al.134 reported a higher ulcer remission rate with omeprazole 20 mg qd than misoprostol 400 mcg /day in adult patients with successfully healed NSAIDs-induced ulcers who required continuous treatment with NSAIDs. However, relapse of gastric ulcer was similar between omeprazole and misoprostol (no p-value reported). Relapse of duodenal ulcer occurred less frequently in omeprazole-treated subjects as compared to misoprostol, whereas multiple erosions at relapse tended to be more frequent in the omeprazole group (no p-values reported). Study Outcome Population Intervention Comparator Results Type(QA) measure Graham et 537 patients (≥ 18 Lans 30 mg Miso 200 1. Time free (lans 15mg data not shown) al. 2002137 years,) H. pylori- qd x 12 mcg qid x 12 from GU up to negative, long- weeks weeks 12 weeks 1. Time free from GU (life table RCT (poor) term users of methods): NSAIDs, history Lans 15mg Placebo x 12 2. Percentage of of endoscopically qd x 12 weeks patients Miso 800 mcg remained free of GU documented GU weeks remaining ulcer longer than lans 30 mg (p = 0.04) with or without free at 12 weeks coexisting DU or Lans 30 mg groups remained free of GI bleeding; GU 3. ADRs and GU significantly longer than placebo

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 112 Consultation Document: Canadian Agency for Drugs and Technologies in Health

or DU <5 mm in compliance (p<0.001) diameter 2. % of patients remaining free of GU at 12 weeks 93% (95% CI 87.2%-97.9%) for miso, 82% (95% CI 75%-89.6%) for lans 30 mg, 51% (95% CI, 41.1%- 61.3%) for placebo, (p=0.04, miso vs. lans; p<0.001 for lans and miso vs. placebo) NNT (95% CI) (mis vs. lans 30 mg) = 9 (5, 32)

3. % of patients remaining free of GU and DU at 12 weeks 88% for miso, 83% for lans 30 mg, 47% for placebo, (NS for miso vs. lans; p<0.001 for lans and miso vs. placebo)

4. % completing trial: 83% in mis arm vs. 86% in lans 30 mg arm (no p value reported).

5. % d/c therapy due to adverse effects: 10% in mis arm vs. 8% in lans 30 mg arm, no p-value reported

When withdrawn subjects were classified as treatment failures, the percentages of patients who were treatment successes were 67% for miso, 68% for lans 30 mg and 34% for placebo, no p value reported

Patients withdrawing prematurely classified as worse case (i.e., having had a GU or DU), percentages of patients free of gastroduodenal ulcer disease throughout the study: 67% for miso, 68% for lans 30 mg, 34% for placebo. (no p value reported).

5. ADR and compliance:

a) Daytime abdominal pain, % days with pain (From patients diary): 41% with miso, 31% with lans 30 mg and 35% with placebo (p<0.05 for miso vs. lans)

b). Nighttime abdominal pain, % nights with pain (From patients diary): 33% with miso, 27% with lans 30 mg and 30% with placebo. (NS for miso vs. lans)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 113 Consultation Document: Canadian Agency for Drugs and Technologies in Health

c). Antacids use, %Days used (From patients diary): 42% with miso, 25% with lans 30 mg and 37% with placebo (p<0.001 for miso vs. lans).

d).Any treatment-related ADR: 31% with Miso, 16% with lans 30 mg and 10% with placebo; (p =0.006, miso vs. lans 30 mg)

i). Diarrhea: 22% with Miso, 7% with lans 30 mg and 3% with placebo. (p<0.01 for each comparison vs. miso)

ii) Treatment related Abdominal pain and nausea: 6% and 4% with miso

e) Compliance: more than 90% with placebo and lans 30 mg compared to 73% with miso (p<0.001). Hawkey et 732 adults with Ome 20 mg Miso 200 Remission rate: Remission rate at 6 months al. successful healing qd x 6 mcg bid x 6 (absence of ome vs. mis vs. placebo: 61% vs. 1998134 of NSAID- months months ulcer, >10 48% vs. 27%, p=0.001 (ome vs. mis) induced ulcer gastric erosions, and p<0.001 (for comparisons of RCT (good) (DU, GU or ≥10 Placebo x 6 >10 duodenal placebo with ome and mis) erosions in months erosions, NNT (95% CI) (ome vs. mis) = 8 (3, stomach or moderate or ∞) duodenum) who worse dyspeptic require symptoms, GU at relapse continuous adverse events ome. vs. mis vs. placebo: 35/274 treatment of resulting in (13%) vs. 31/296 (10%) vs. 50/155 NSAID; 48% treatment (32%); p-values not reported were H. pylori- discontinuation) positive at 6 months. GU ≥5mm at relapse ome. vs. mis vs. placebo: 21/274 (8%) vs. 23/296 GU, DU and (8%) vs. 31/155 (20%); p-values not erosion reported recurrence rate DU at relapse: ome. vs. mis vs. placebo: 7/274 (3%) vs. 30/296 (10%) vs. 19/155 (12%); p-values not reported

DU ≥5mm at relapse ome. vs. mis vs. placebo: 7/274 (3%) vs. 26/296 (9%) vs. 15/155 (10%); p- values not reported

Multiple erosions at relapse ome. vs. mis vs. placebo: 34/274 (12%) vs. 21/296 (7%) vs. 21/155 (14%); p-values not reported ome: omeprazole; lans: lansoprazole; miso : misoprostol ; ADR: Adverse drug reaction

Context:

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 114 Consultation Document: Canadian Agency for Drugs and Technologies in Health

• Graham et al. 2002137 was the only study that examined ulcer remission rate as the primary outcome. Hawkey et al. 1998134 defined a composite outcome that included ulcers, erosions, and dyspepsia. • As both studies focused on the prevention of endoscopic ulcers, the generalizability to ulcer complications is limited. The data supporting PPIs for the prevention of ulcer complications is poor. • Graham et al. 2002137 reported that misoprostol 800 mcg daily was associated with significantly poorer compliance and a significantly greater incidence of treatment-related adverse effects as compared to lansoprazole 30 mg. However, the total proportion completing the study was similar in the two groups. Furthermore, even when withdrawals were considered to be treatment failures (i.e., as having an ulcer), the total proportion that was ulcer-free was similar in both groups (i.e., 67% vs. 68% for misoprostol and lansoprazole 30 mg, respectively). • Graham138 also examined the effect of H. pylori infection on the effectiveness of omeprazole for the prevention of GU or DU in a post hoc anlaysis of data from Hawkey et al. 1998134. He found that the total ulcer relapse was 14.5% in omeprazole arm vs. 19.6% in misoprostol arm vs. 42.5% in placebo (p=0.93 for mis vs. ome). There was no significant difference between omeprazole and misoprostol in terms of GU relapse in all patients or in H. pylori-positive patients, but misoprostol was superior to omeprazole in H. pylori-negative patients. In terms of DU relapse, omeprazole was superior to misoprostol in all patients and in H. pylori-positive patients, however, there was no difference between treatments in H. pylori-negative subjects.

P2: NSAID-associated ulcer P2.2: Prevention of NSAID-associated ulcer P2.2.2: PPIs and non-selective NSAIDs vs. COX-2 NSAIDs Evidence Statement P2.2.2A: There is no difference in ulcer recurrence and bleeding rates between COX-2 selective NSAIDs and the combination of PPI and conventional NSAIDs in patients with previous NSAID-associated upper GI bleeding.

Voting Results: A 14%, B 57%, C 29%, D 0%, E 0%

Summary: A good quality SR139 identified one RCT that compared the probability of recurrent ulcer bleeding after 6 months of either celecoxib or diclofenac and omeprazole in arthritis patients with previous GI hemorrhage on NSAIDs. There was no significant difference between the two treatment arms. Further outcomes from this trial were reported in a separate study140 published after the SR, in which there were also no significant differences between treatments in the combined endpoint of bleeding and endoscopic ulcer at 6 months. Similarly, another good quality RCT141 found no significant difference in the recurrence rate of ulcer complications in a study of celecoxib versus naproxen and lansoprazole in patients with NSAID-associated upper GI bleeding and endoscopically verified DU or GU. Study Population Intervention Comparator Outcome measure Results Type(QA) Rostom et For COX-2 Cele 200 mg DO: Diclo 75 6 month Cele vs. DO: al. 2003139 selective vs. Non- twice daily x mg twice probability of Risk of recurrent ulcer selective NSAID + 6 months daily, recurrent ulcer bleeding at 6 months = 4.9% SR (good) PPI: 1 RCT (n=287) omeprazole bleeding vs. 6.4%, p>0.05 (NS) in arthritis patients 20 mg daily x with NSAID 6 months

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 115 Consultation Document: Canadian Agency for Drugs and Technologies in Health

associated ulcer bleeding, healing confirmed endoscopically, Hp- negative or successful eradication of Hp. Lai et al. 242 patients using Cele 200 mg NL: Nap 250 Recurrence of Cele vs. NL (95% CI) : 3.7% 2005141 NSAIDs with upper once daily x mg thrice ulcer (0.0%, 7.3%) vs. 6.3% (1.6%, GI bleeding, 24 weeks daily, lans 30 complications at 11.1%), Difference = -2.6% (- RCT (good) endoscopically mg once daily 24 weeks 9.1, 3.7) (NS); met a priori verified DU or GU x 24 weeks criterion for non-inferiority ≥5 mm, successful ulcer healing and Time to recurrence not sig. diff Hp eradication b/w groups

Chan et al. 287 arthritis patients Cele 200 mg DO: Diclo 75 Endoscopic ulcer % with Endoscopic Ulcer 2004140 with NSAID twice daily x mg twice at 6 months in among subjects w/o recurrence associated ulcer 6 months daily, ome 20 subjects w/o of ulcer complications (Cele RCT (good) bleeding, healing mg daily x 6 recurrent GI vs. DO) (95% CI): confirmed months complications 18.7% (11.3%, 26.1%) vs. (post hoc endoscopically, Hp- 25.6% (17.1%, 34.1%), p>0.05 analysis of negative or Combined (log-rank test) (NS) the RCT successful endpoint of identified by eradication of Hp. bleeding and % with Combined Endpoint Rostom et endoscopic ulcers (Cele vs. DO) (95% CI): al., above) 24.1% (16.3%, 31.8%) vs. 32.3% (23.7%, 40.9%), p>0.05 (log-rank test) (NS)

Cele: celecoxib; diclo: diclofenac; lans: lansoprazole; nap: naproxen

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 116 Consultation Document: Canadian Agency for Drugs and Technologies in Health

P2: NSAID-associated ulcer P2.2: Prevention of NSAID-associated ulcer P2.2.3 : PPIs vs. H. pylori eradication Evidence Statement P2.2.3A: Standard-dose PPIs are more efficacious than H. pylori eradication in reducing the risk of GI complications in Hp-positive, NSAID-associated ulcers with a history of NSAID-associated ulcer bleeding.

Voting Results: A 11%, B 67%, C 22%, D 0%, E 0%

Summary: A good quality RCT by Chan et al. 2001142 showed that H. pylori eradication was inferior to maintenance therapy with omeprazole for the prevention of recurrent upper GI bleeding due to NSAIDs. Study Outcome Population Intervention Comparator Results Type(QA) measure Chan et al. 400 users of Hp ome 20 Recurrent Recurrent bleeding at 6 months (ASA 2001142 NSAIDs or ASA eradication mg/day for 6 upper GI users): for ≥ 6 months therapy (bis months bleeding at 6 Hp erad: 1.9% vs. ome: 0.9%, RCT (ASA users = 250; 480 mg/day, months difference = 1.0% (95% CI: -1.9%, (good) naproxen users = met 1.6 3.9%) 150) g/day, tet 2 g/day for 7 Recurrent bleeding at 6 months days), then (naproxen users): placebo for Hp erad: 18.8% vs. ome: 4.4%, 6 months difference = 14.4% (95% CI: 4.4%, 24.4%) bis: bismuth; met: metronidazole; ome: omeprazole; tet: tetracycline.

Context: • All studies included patients who were H. pylori-negative or had infection successfully eradicated. • Only two NSAIDs (i.e., diclofenac and naproxen) were studied in these trials.

P2: NSAID-associated ulcer P2.2: Prevention of NSAID-associated ulcer P2.2.4: Differences among PPIs Evidence Statement P2.2.4A: Different PPIs reduce ulcer risk to a similar degree when given to NSAID users for ulcer prophylaxis.

Voting Results: A 0%, B 67%, C 33%, D 0%, E 0%

Summary: A good quality systematic review by McDonagh et al.90 did not find any head-to-head studies assessing intraclass differences between the various PPIs with respect to NSAID ulcer prevention. However, there did not appear to be differences among the various PPIs based on a qualitative assessment of effect sizes from studies that compared PPIs with other ulcer prophylaxis medications such as H2RAs or misoprostol. It should be noted that formal statistical methods for indirect comparisons were not employed in this review. Study Population Intervention Comparator Outcome measure Results Type(QA)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 117 Consultation Document: Canadian Agency for Drugs and Technologies in Health

McDonagh 7 RCTs (n not PPI H2RA, placebo, Ulcer prevention No head-to-head studies et al. 200590 provided) mis found. NSAID or ASA users; No apparent differences SR (good) both primary and between various PPIs secondary prophylaxis from studies comparing trials included; PPIs vs. other ulcer included trials with prophylaxis drugs. either endoscopic ulcers or ulcer complications as outcome

Context: • The statement is based on an indirect comparison. • One RCT (Regula et al. 2006143) directly comparing one PPI with another for the prevention of NSAID-induced GI lesions was referred to COMPUS after the Expert Review Panel had completed its deliberations on the Evidence Statements for PPIs. This study found that standard- dose pantoprazole was similar to standard-dose omeprazole for the prevention of NSAID- associated GI lesions over a period of 6 months. The probability of remaining in remission (i.e., lack of therapeutic failure) after 6 months was 93% for pantoprazole versus 89% for omeprazole, and the probability of remaining in remission (i.e., lack of endoscopic failure) was 95% for pantoprazole versus 93% for omeprazole.

P3: ASA-associated ulcer Evidence Statement P3.1: In patients in whom H. pylori has been eradicated, standard-dose PPIs are more efficacious than placebo for the secondary prevention of ulcer complications in users of low dose (≤325 mg/day) ASA.

Voting Results: A 29%, B 57%, C 14%, D 0%, E 0%

Summary: A good quality RCT by Lai et al., showed that PPI therapy reduces recurrence rates of ulcer complications due to low-dose ASA at one year.144 Study Outcome Population Intervention Comparator Results Type(QA) measure Lai et al. 123 Hp-infected lans 30 placebo for 1 Ulcer Ulcer complication rate at 12 2002*144 patients with mg/day for 1 yr; ASA 100 complications months: complicated ulcer ≥5 yr; ASA 100 mg/day at 12 months lans: 1.6% vs. placebo:14.8%; RCT (good) mm, receiving ASA mg/day p = 0.008 Hazard ratio (95% ≤325mg/day for ≥ 1 CI) (placebo vs. lans): 10.6 month before ulcer (1.3, 86) complications NNT (95% CI) =8 (4, 27) occurred lans: lansoprazole; * indicates industry involvement

Context: There is no evidence regarding H. pylori-negative or H. pylori status unknown patients.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 118 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Evidence Statement P3.2: In H. pylori-negative patients who have a history of ulcer bleeding on low- dose ASA alone, the combination of low-dose ASA and a PPI is associated with a lower risk of recurrence of ulcer complications at 1 year as compared to clopidogrel alone.

Voting Results: A 44%, B 56%, C 0%, D 0%, E 0%

Summary: Two very good quality RCTs145,146 demonstrated that, after healing of ulcer bleeding associated low-dose ASA alone in H. pylori-negative patients, the combination of ASA and a PPI provided significantly better protection from recurrent ulcer complications than clopidogrel alone. Study Population Intervention Comparator Outcome Results Type (QA) measure Lai et al. 170 adults requiring C: clop 75 AE: ASA 100 Recurrence No. of ulcer complications at 52 2006145 continuous treatment mg qd x 52 mg qd & of ulcer weeks: C vs. AE : 9 vs. 0 with low-dose weeks esome 20 mg complications RCT (very aspirin, with DU qd x 52 weeks at 52 weeks Cumulative incidence of good) and/or GU ≥5mm gastroduodenal ulcers and ulcer and upper GI complications at 52 weeks: C vs. bleeding associated AE: 13.6% vs. 0%, absolute with low dose ASA; difference = 13.6% (95% CI, endoscopically 6.3%, 20.9%), p=0.0019 (log- healed and Hp –ve or rank test) Hp successfully NNT (95% CI) = 7 (5, 16) eradicated Chan et al. 320 patients using C: clop 75 AE: ASA 80 Recurrent C vs. AE (95% CI): 8.6% (4.1, 2005146 low-dose aspirin with mg qd x 12 mg qd & ulcer 13.1) vs. ulcer bleeding, months esome 20 mg bleeding at 0.7% (0, 2.0), p=0.001 RCT (very healing confirmed bid x 12 12 months NNT (95% CI) = 13 (8, 30) good) endoscopically, and months Hp –ve or successfully eradicated clop: clopidogrel, ASA: aspirin, esome: esomeprazole, ome: omeprazole

Context: • The evidence consists of relatively small trials studying only one PPI (i.e., esomeprazole 20 mg). • A recent case control study by Hallas et al. 2006147 highlighted the increased risk of upper GI bleeding with combined antithrombotic therapy. The adjusted OR for upper GI bleeding was 1.8 (95% CI: 1.5 to 2.1) for low dose asprin, 1.1 (1.3 to 2.4) for clopidogrel, and 7.4 (3.5 to 15) for a combination of asprin and clopidogrel. There is no evidence regarding the efficacy of PPIs for prevention of GI bleeding due to clopidogrel either alone or in combination with ASA.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 119 Consultation Document: Canadian Agency for Drugs and Technologies in Health

5.3.3 Research Gaps

1) H. pylori eradication therapy with triple therapy (PAC and PMC) for 10 days is as efficacious as 14 days. Although the ERP identified relevant studies, the panel’s votes were evenly split between ‘Accept’ and ‘Reject’, therefore this Statement was considered a Research Gap. The voting results and summary of evidence are shown below.

Voting Results: A 0%, B 36%, C 28%, D 36%, E 0%

Hp eradication rate and (95% CI): 10 days vs. 14 days: 82% (77%, 86%) vs. 84% (79%, 89%); p>0.05, NS

Hp eradication rate and (95% CI): 7 days vs. 10 days was: 80% (71%, 86%) vs. 83% (75%, 89%); p>0.05, NS

NNT for 10-14 days as compared to 7 days to obtain one extra cure = 6 to 23 Amox: amoxicillin; clar: clarithromycin; met: metronidazole

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 120 Consultation Document: Canadian Agency for Drugs and Technologies in Health

References

1. Prodigy Knowledge. Dyspepsia: proven gastro-oesophageal reflux disease [Prodigy guidance]. Rev. Newcastle upon Tyne: Sowerby Centre for Health Informatics at Newcastle; 2005 Jul.

2. Canadian Pharmacists Association. Compendium of pharmaceuticals and specialties [database online]. Ottawa: The Association; 2005.

3. Notice of compliance listings. [database online]. Ottawa: Therapeutic Products Directorate, Health Canada; 2005. Available: http://www.hc-sc.gc.ca/hpfb-dgpsa/tpd- dpt/index_drugs_noc_e.html (accessed 2005 Dec 15).

4. PrLosec® (omeprazole) capsules: 10, 20 and 40 mg delayed release: H+, K+-ATPase inhibitor [product monograph]. Rev. Mississauga: AstraZeneca Canada; 2004 Jun.

5. PrLosec® (omeprazole magnesium): 10 and 20 mg delayed release tablets: H+, K+-ATPase inhibitor [product monograph]. Rev. Mississauga: AstraZeneca Canada; 2003 Sep.

6. PrLosec® MUPS™ (omeprazole magnesium): 10 mg and 20 mg delayed release tablets: H+, K+-ATPase inhibitor [product monograph]. Rev. Mississauga: AstraZeneca Canada; 2003 Sep.

7. Apo-omeprazole: omeprazole capsules: 20 mg and 40 mg: H+, K+-ATPase inhibitor [product monograph]. Rev. Weston (ON): Apotex; 2004 Sep.

8. PrPrevacid®: lansoprazole delayed-release capsules (manufacturer's standard), 15 mg and 30 mg; lansoprazole granules for delayed-release oral suspension, 15 mg and 30 mg. PrPrevacid® FasTab: lansoprazole delayed-release tablets, 15 mg and 30 mg. PrPrevacid® I.V.: lansoprazole sodium for injection; lyophilized powder for reconstitution, 30 mg. H+, K+-ATPase inhibitor [product monograph]. Rev. Lake Forest: TAP Pharmaceuticals; 2005 Jun.

9. PrPantoloc®: pantoprazole sodium: enteric-coated tablets, 20 mg and 40 mg: H+, K+-ATPase inhibitor [product monograph]. Rev. Markham (ON): Solvay Pharma; 2005 May.

10. PrNexium®: esomeprazole magnesium trihydrate delayed release tablets: 20 and 40 mg esomeprazole: H+, K+-ATPase inhibitor [product monograph]. Rev. Mississauga: AstraZeneca Canada; 2005 Nov.

11. PrPariet™: rabeprazole sodium: 10 and 20 mg enteric-coated tablets: H+, K+-ATPase inhibitor [product monograph]. Rev. Toronto: Janssen-Ortho; 2005 Jan.

12. Hp-PAC®: lansoprazole 30 mg delayed-release capsules; clarithromycin 500 mg film-coated tablets, and amoxicillin 500 mg capsules: Helicobacter pylori eradication therapy [product monograph]. Rev. Lake Forest: TAP Pharmaceuticals; 2004 Jun.

13. PrNovo-ranidine tablets (ranitidine tablets, USP) 150 mg and 300 mg film coated tablets; PrNovo-ranidine oral solution (Ranitidine oral solution, USP), 15 mg/ml; PrNovo-ranidine injection (Ranitidine injection, USP), 25 mg/ml. Therapeutic classification: histamine H2 receptor antagonist [product monograph]. Control #056969 (oral solution); #041987 (injection). Toronto: Novopharm Ltd; 2004 May 25.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 121 Consultation Document: Canadian Agency for Drugs and Technologies in Health

14. Novo-famotidine (famotidine) USP 20 and 40 mg tablets: histamine H2 receptor antagonist [product monograph]. Control #072862. Toronto: Novopharm Ltd.; 2001 Dec 13.

15. Novo-nizatidine (nizatidine); 150 mg and 300 mg capsules, USP, histamine H2 receptor antagonist [product monograph]. Control #055409. Toronto: Novopharm Ltd.; 1999 Jun 25.

16. Thomson AB, Barkun AN, Armstrong D, Chiba N, White RJ, Daniels S, et al. The prevalence of clinically significant endoscopic findings in primary care patients with uninvestigated dyspepsia: the Canadian Adult Dyspepsia Empiric Treatment - Prompt Endoscopy (CADET-PE) study. Aliment Pharmacol Ther 2003;17(12):1481-91.

17. Armstrong D, Marshall JK, Chiba N, Enns R, Fallone CA, Fass R, et al. Canadian consensus conference on the management of gastroesophageal reflux disease in adults: update 2004. Can J Gastroenterol 2005;19(1):15-35.

18. DeVault KR, Castell DO. Updated guidelines for the diagnosis and treatment of gastroesophageal reflux disease. Am J Gastroenterol 2005;100(1):190-200.

19. Tougas G, Chen Y, Hwang P, Liu MM, Eggleston A. Prevalence and impact of upper gastrointestinal symptoms in the Canadian population: findings from the DIGEST study. Domestic/International Gastroenterology Surveillance Study. Am J Gastroenterol 1999;94(10):2845-54.

20. Johnson DA. Workshop consensus report on the extraesophageal complications of gastroesophageal reflux disease. J Clin Gastroenterol 2000;30(3 Suppl):S51-S53.

21. Chow WH, Finkle WD, McLaughlin JK, Frankl H, Ziel HK, Fraumeni JF. The relation of gastroesophageal reflux disease and its treatment to adenocarcinomas of the esophagus and gastric cardia. JAMA 1995;274(6):474-7.

22. Ye W, Chow WH, Lagergren J, Yin L, Nyrén O. Risk of adenocarcinomas of the esophagus and gastric cardia in patients with gastroesophageal reflux diseases and after antireflux surgery. Gastroenterology 2001;121(6):1286-93.

23. Farrow DC, Vaughan TL, Sweeney C, Gammon MD, Chow WH, Risch HA, et al. Gastroesophageal reflux disease, use of H2 receptor antagonists, and risk of esophageal and gastric cancer. Cancer Causes Control 2000;11(3):231-8.

24. Williams DB. Gastroesophageal reflux disease. In: DiPiro JT, Talbert RL, Yee GC, Matzke GR, Wells BG, Posey LM, editors. Pharmacotherapy: a pathophysiologic approach. 5th ed. Toronto: McGraw-Hill, Medical Pub. Division; 2002. p.585-601.

25. Goyal RK. Diseases of the esophagus. In: Kasper DL, Braunwald E, Fauci AS, Hauser SL, Longo DL, Jameson JL, editors. Harrison's principles of internal medicine. 16th ed. Toronto: McGraw-Hill, Medical Pub. Division; 2005. p.1739-46.

26. Spechler SJ. Clinical practice: Barrett's esophagus. N Engl J Med 2002;346(11):836-42.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 122 Consultation Document: Canadian Agency for Drugs and Technologies in Health

27. Chiba N, Bernard L, O'Brien BJ, Goeree R, Hunt RH. A Canadian physician survey of dyspepsia management. Can J Gastroenterol 1998;12(1):83-90.

28. Van Zanten VSJ, Flook N, Chiba N, Armstrong D, Barkun A, Bradette M, et al. An evidence- based approach to the management of uninvestigated dyspepsia in the era of Helicobacter pylori. Canadian Dyspepsia Working Group. CMAJ 2000;162(12 Suppl):S3-23.

29. Berardi RR. Peptic ulcer disease. In: DiPiro JT, Talbert RL, Yee GC, Matzke GR, Wells BG, Posey LM, editors. Pharmacotherapy: a pathophysiologic approach. 5th ed. Toronto: McGraw- Hill, Medical Pub. Division; 2002. p.603-24.

30. Ontario Program for Optimal Therapeutics. Ontario guidelines for peptic ulcer disease and gastroesophageal reflux. 1st ed. Toronto: Queen's Printer of Ontario; 2000. Available: http://www.thecem.net/Downloads/gerd.pdf (accessed 2005 Jul 5).

31. North of England Dyspepsia Guideline Development Group. Dyspepsia: management of dyspepsia in adults in primary care [Evidence-based clinical practice guideline]. London: National Institute for Clinical Excellence; 2004 Aug. Available: http://www.nice.org.uk/pdf/CG017fullguideline.pdf (accessed 2005 Jul 27).

32. Lanza FL. A guideline for the treatment and prevention of NSAID-induced ulcers. Members of the Ad Hoc Committee on Practice Parameters of the American College of Gastroenterology. Am J Gastroenterol 1998;93(11):2037-46.

33. Hunt R, Thomson AB. Canadian Helicobacter pylori consensus conference. Canadian Association of Gastroenterology. Can J Gastroenterol 1998;12(1):31-41.

34. Hirschowitz BI. Zollinger-Ellison syndrome: pathogenesis, diagnosis, and management. Am J Gastroenterol 1997;92(4 Suppl):44S-8S.

35. Armstrong D, Veldhuyzen van Zanten SJ, Barkun AN, Chiba N, Thomson AB, Smyth S, et al. Heartburn-dominant, uninvestigated dyspepsia: a comparison of 'PPI-start' and 'H2-RA-start' management strategies in primary care: the CADET-HR study. Aliment Pharmacol Ther 2005;21(10):1189-202.

36. Kaplan Machlis B, Spiegler GE, Zodet MW, Revicki DA. Effectiveness and costs of omeprazole vs ranitidine for treatment of symptomatic gastroesophageal reflux disease in primary care clinics in West Virginia. Arch Fam Med 2000;9(7):624-30.

37. Talley NJ, Moore MG, Sprogis A, Katelaris P. Randomised controlled trial of pantoprazole versus ranitidine for the treatment of uninvestigated heartburn in primary care. Med J Aust 2002;177:415-9.

38. van Zyl J, Van Rensburg C, Vieweg W, Fischer R. Efficacy and safety of pantoprazole versus ranitidine in the treatment of patients with symptomatic gastroesophageal reflux disease. Digestion 2004;70(1):61-9.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 123 Consultation Document: Canadian Agency for Drugs and Technologies in Health

39. Howden CW, Henning JM, Huang B, Lukasik N, Freston JW. Management of heartburn in a large, randomized, community-based study: comparison of four therapeutic strategies. Am J Gastroenterol 2001;96(6):1704-10.

40. Maton PN, Orlando R, Joelsson B. Efficacy of omeprazole versus ranitidine for symptomatic treatment of poorly responsive acid reflux disease: a prospective, controlled trial. Aliment Pharmacol Ther 1999;13(6):819-26.

41. Norman HA, Bergheim R, Fagertun H, Lund H, Moum B. A randomised prospective study comparing the effectiveness of esomeprazole treatment strategies in clinical practice for 6 months in the management of patients with symptoms of gastroesophageal reflux disease. Int J Clin Pract 2005;59(6):665-71.

42. Hansen AN, Wahlqvist P, Jørgensen E, Bergheim R, Fagertun H, Lund H, et al. Six-month management of patients following treatment for gastroesophageal reflux disease symptoms: a Norwegian randomized, prospective study comparing the costs and effectiveness of esomeprazole and ranitidine treatment strategies in a general medical practitioners setting. Int J Clin Pract 2005;59(6):655-64.

43. Bigard MA, Genestin E. Treatment of patients with heartburn without endoscopic evaluation: on- demand treatment after effective continuous administration of lansoprazole 15 mg. Aliment Pharmacol Ther 2005;22(7):635-43.

44. van Pinxteren B, Numans ME, Bonis PA, Lau J. Short-term treatment with proton pump inhibitors, H2-receptor antagonists and prokinetics for gastro-oesophageal reflux disease-like symptoms and endoscopy negative reflux disease. Cochrane Database Syst Rev 2006;(3):CD002095.

45. van Pinxteren B, Numans ME, Bonis PA, Lau J. Short-term treatment with proton pump inhibitors, H2-receptor antagonists and prokinetics for gastro-oesophageal reflux disease-like symptoms and endoscopy negative reflux disease. Cochrane Database Syst Rev 2004;(3):CD002095.

46. Fujiwara Y, Higuchi K, Nebiki H, Chono S, Uno H, Kitada K, et al. Famotidine vs. omeprazole : a prospective randomized multicentre trial to determine efficacy in non-erosive gastro- oesophageal reflux disease. Aliment Pharmacol Ther 2005;21 Suppl 2:10-8.

47. Wada T, Sasaki M, Kataoka H, Tanida S, Itoh K, Ogasawara N, et al. Efficacy of famotidine and omeprazole in healing symptoms of non-erosive gastro-oesophageal reflux disease: randomized- controlled study of gastro-oesophageal reflux disease. Aliment Pharmacol Ther 2005;21 Suppl 2:2-9.

48. Miner P, Orr W, Filippone J, Jokubaitis L, Sloan S. Rabeprazole in nonerosive gastroesophageal reflux disease: a randomized placebo-controlled trial. Am J Gastroenterol 2002;97(6):1332-9.

49. Richter JE, Kovacs TO, Greski-Rose PA, Huang section, Fisher R. Lansoprazole in the treatment of heartburn in patients without erosive oesophagitis. Aliment Pharmacol Ther 1999;13(6):795- 804.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 124 Consultation Document: Canadian Agency for Drugs and Technologies in Health

50. Katz PO, Castell DO, Levine D. Esomeprazole resolves chronic heartburn in patients without erosive oesophagitis. Aliment Pharmacol Ther 2003;18(9):875-82.

51. Carlsson R, Dent J, Watts R, Riley S, Sheikh R, Hatlebakk J, et al. Gastro-oesophageal reflux disease in primary care: an international study of different treatment strategies with omeprazole. International GORD Study Group. Eur J Gastroenterol Hepatol 1998;10(2):119-24.

52. Lind T, Havelund T, Carlsson R, Anker Hansen O, Glise H, Hernqvist H, et al. Heartburn without oesophagitis: efficacy of omeprazole therapy and features determining therapeutic response. Scand J Gastroenterol 1997;32(10):974-9.

53. Richter JE, Peura D, Benjamin SB, Joelsson B, Whipple J. Efficacy of omeprazole for the treatment of symptomatic acid reflux disease without esophagitis. Arch Intern Med 2000;160(12):1810-6.

54. Venables TL, Newland RD, Patel AC, Hole J, Copeman MB, Turbitt ML. Maintenance treatment for gastro-oesophageal reflux disease: a placebo-controlled evaluation of 10 milligrams omeprazole once daily in general practice. Scand J Gastroenterol 1997;32(7):627-32.

55. Zacny J, Zamakhshary M, Sketris I, Veldhuyzen VZ. Systematic review: the efficacy of intermittent and on-demand therapy with histamine H2-receptor antagonists or proton pump inhibitors for gastro-oesophageal reflux disease patients. Aliment Pharmacol Ther 2005;21(11):1299-312.

56. Kaspari S, Kupcinskas L, Heinze H, Berghöfer P. Pantoprazole 20 mg on demand is effective in the long-term management of patients with mild gastro-oesophageal reflux disease. Eur J Gastroenterol Hepatol 2005;17(9):935-41.

57. Scholten T, Dekkers CP, Schutze K, Korner T, Bohuschke M, Gatz G. On-demand therapy with pantoprazole 20 mg as effective long-term management of reflux disease in patients with mild GERD: the ORION trial. Digestion 2005;72(2-3):76-85.

58. Bytzer P, Blum A, de Herdt D, Dubois D. Six-month trial of on-demand rabeprazole 10 mg maintains symptom relief in patients with non-erosive reflux disease. Aliment Pharmacol Ther 2004;20(2):181-8.

60. Carlsson R, Galmiche JP, Dent J, Lundell L, Frison L. Prognostic factors influencing relapse of oesophagitis during maintenance therapy with antisecretory drugs: a meta-analysis of long-term omeprazole trials. Aliment Pharmacol Ther 1997;11(3):473-82.

61. Chiba N, De Gara CJ, Wilkinson JM, Hunt RH. Speed of healing and symptom relief in grade II to IV gastroesophageal reflux disease: a meta-analysis. Gastroenterology 1997;112(6):1798-810.

62. Koop H, Schepp W, Dammann HG, Schneider A, Lühmann R, Classen M. Comparative trial of pantoprazole and ranitidine in the treatment of reflux esophagitis: results of a German multicenter study. J Clin Gastroenterol 1995;20(3):192-5.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 125 Consultation Document: Canadian Agency for Drugs and Technologies in Health

63. Robinson M, Decktor DL, Maton PN, Sabesin S, Roufail W, Kogut D, et al. Omeprazole is superior to ranitidine plus metoclopramide in the short-term treatment of erosive oesophagitis. Aliment Pharmacol Ther 1993;7(1):67-73.

64. Frame MH. Omeprazole produces significantly greater healing of erosive or ulcerative reflux oesophagitis than ranitidine. The Italian Reflux Oesophagitis Study Group. Eur J Gastroenterol Hepatol 1991;3(7):511-7.

65. Bate CM, Keeling PW, O'Morain C, Wilkinson SP, Foster DN, Mountford RA, et al. Comparison of omeprazole and cimetidine in reflux oesophagitis: symptomatic, endoscopic, and histological evaluations. Gut 1990;31(9):968-72.

66. Zeitoun P, Rampal P, Barbier P, Isal JP, Eriksson S, Carlsson R. [Omeprazole (20 mg daily) compared to ranitidine (150 mg twice daily) in the treatment of esophagitis caused by reflux: results of a double-blind randomized multicenter trial in France and Belgium]. Gastroenterol Clin Biol 1989;13(5):457-62.

67. Green JRB, Tildesley G, Theodossi A, Bate CM, Bradby GV, Axon ATR, et al. Omeprazole 20 mg to 40 mg once daily is more effective in providing complete symptom relief and endoscopic healing in patients with reflux oesophagitis. Br J Clin Res 1995;6:63-76.

68. van Rensburg CJ, Honiball PJ, Grundling HD, van Zyl JH, Spies SK, Eloff FP, et al. Efficacy and tolerability of pantoprazole 40 mg versus 80 mg in patients with reflux oesophagitis. Aliment Pharmacol Ther 1996;10(3):397-401.

69. Bardhan KD, Hawkey CJ, Long RG, Morgan AG, Wormsley KG, Moules IK, et al. Lansoprazole versus ranitidine for the treatment of reflux oesophagitis. UK Lansoprazole Clinical Research Group. Aliment Pharmacol Ther 1995;9(2):145-51.

70. Sontag SJ, Hirschowitz BI, Holt S, Robinson MG, Behar J, Berenson MM, et al. Two doses of omeprazole versus placebo in symptomatic erosive esophagitis: the U.S. Multicenter Study. Gastroenterology 1992;102(1):109-18.

71. Hetzel DJ, Dent J, Reed WD, Narielvala FM, Mackinnon M, McCarthy JH, et al. Healing and relapse of severe peptic esophagitis after treatment with omeprazole. Gastroenterology 1988;95(4):903-12.

72. Earnest DL, Dorsch E, Jones J, Jennings DE, Greski Rose PA. A placebo-controlled dose- ranging study of lansoprazole in the management of reflux esophagitis. Am J Gastroenterol 1998;93(2):238-43.

73. Bate CM, Booth SN, Crowe JP, Hepworth Jones B, Taylor MD, Richardson PD. Does 40 mg omeprazole daily offer additional benefit over 20 mg daily in patients requiring more than 4 weeks of treatment for symptomatic reflux oesophagitis? Aliment Pharmacol Ther 1993;7(5):501-7.

74. Gillessen A, Beil W, Modlin IM, Gatz G, Hole U. 40 mg pantoprazole and 40 mg esomeprazole are equivalent in the healing of esophageal lesions and relief from gastroesophageal reflux disease-related symptoms. J Clin Gastroenterol 2004;38(4):332-40.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 126 Consultation Document: Canadian Agency for Drugs and Technologies in Health

75. Kahrilas PJ, Falk GW, Johnson DA, Schmitt C, Collins DW, Whipple J, et al. Esomeprazole improves healing and symptom resolution as compared with omeprazole in reflux oesophagitis patients: a randomized controlled trial. The Esomeprazole Study Investigators. Aliment Pharmacol Ther 2000;14(10):1249-58.

76. Donnellan C, Sharma N, Preston C, Moayyedi P. Medical treatments for the maintenance therapy of reflux oesophagitis and endoscopic negative reflux disease [Cochrane review]. In: Cochrane Database of Systematic Reviews 2004 Issue 4. Chichester (UK): John Wiley & Sons, Ltd; 2004. DOI: 10.1002/14651858.CD003245.pub2.

77. Pilotto A, Leandro G, Franceschi M. Short- and long-term therapy for reflux oesophagitis in the elderly: a multi-centre, placebo-controlled study with pantoprazole. Aliment Pharmacol Ther 2003;17(11):1399-406.

78. Sontag SJ, Robinson M, Roufail W, Hirschowitz BI, Sabesin SM, Wu WC, et al. Daily omeprazole surpasses intermittent dosing in preventing relapse of oesophagitis: a US multicentre double-blind study. Aliment Pharmacol Ther 1997;11(2):373-80.

79. Lundell L, Backman L, Ekström P, Enander LK, Falkmer S, Fausa O, et al. Prevention of relapse of reflux esophagitis after endoscopic healing: the efficacy and safety of omeprazole compared with ranitidine. Scand J Gastroenterol 1991;26(3):248-56.

80. Swarbrick ET, Gough AL, Foster CS, Christian J, Garrett AD, Langworthy CH. Prevention of recurrence of oesophageal stricture: a comparison of lansoprazole and high-dose ranitidine. Eur J Gastroenterol Hepatol 1996;8(5):431-8.

81. Smith PM, Kerr GD, Cockel R, Ross BA, Bate CM, Brown P, et al. A comparison of omeprazole and ranitidine in the prevention of recurrence of benign esophageal stricture. Gastroenterology 1994;107(5):1312-8.

82. Marks RD, Richter JE, Rizzo J, Koehler RE, Spenney JG, Mills TP, et al. Omeprazole versus H2-receptor antagonists in treating patients with peptic stricture and esophagitis. Gastroenterology 1994;106(4):907-15.

83. Edwards SJ, Lind T, Lundell L. Systematic review of proton pump inhibitors for the maintenance of healed reflux oesophagitis. J Drug Assess 2002;5:165-78.

84. Sjöstedt S, Befrits R, Sylvan A, Harthon C, Jörgensen L, Carling L, et al. Daily treatment with esomeprazole is superior to that taken on-demand for maintenance of healed erosive oesophagitis. Aliment Pharmacol Ther 2005;22(3):183-91.

85. Gibson PG, Henry RL, Coughlan JL. Gastro-oesophageal reflux treatment for asthma in adults and children [Cochrane review]. In: Cochrane Database of Systematic Reviews 2003 Issue 1. Chichester (UK): John Wiley & Sons, Ltd; 2003. DOI: 10.1002/14651858.CD001496.

86. Vaezi MF, Richter JE, Stasney CR, Spiegel JR, Iannuzzi RA, Crawley JA, et al. Treatment of chronic posterior laryngitis with esomeprazole. Laryngoscope 2006;116(2):254-60.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 127 Consultation Document: Canadian Agency for Drugs and Technologies in Health

87. Steward DL, Wilson KM, Kelly DH, Patil MS, Schwartzbauer HR, Long JD, et al. Proton pump inhibitor therapy for chronic laryngo-pharyngitis: a randomized placebo-control trial. Otolaryngol Head Neck Surg 2004;131(4):342-50.

88. Eherer AJ, Habermann W, Hammer HF, Kiesler K, Friedrich G, Krejs GJ. Effect of pantoprazole on the course of reflux-associated laryngitis: a placebo-controlled double-blind crossover study. Scand J Gastroenterol 2003;38(5):462-7.

89. Chang AB, Lasserson TJ, Gaffney J, Connor FL, Garske LA. Gastro-oesophageal reflux treatment for prolonged non-specific cough in children and adults [Cochrane review]. In: Cochrane Database of Systematic Reviews 2005 Issue 1. Chichester (UK): John Wiley & Sons, Ltd; 2005. DOI: 10.1002/14651858.CD004823.pub2.

90. McDonagh MS, Carson S. Drug class review on proton pump inhibitors: final report update 3. Portland (OR): Oregon Health & Science University; 2005.

91. Edwards SJ, Lind T, Lundell L. Systematic review of proton pump inhibitors for the acute treatment of reflux oesophagitis. Aliment Pharmacol Ther 2001;15(11):1729-36.

92. Vakil N, Fennerty MB. Direct comparative trials of the efficacy of proton pump inhibitors in the management of gastro-oesophageal reflux disease and peptic ulcer disease. Aliment Pharmacol Ther 2003;18(6):559-68.

93. Wang W, Huang J, Zheng G, Xia HHX, Wong W, Lam S, et al. Head-to-head comparison of H2- receptor antagonists and proton pump inhibitors in the treatment of erosive esophagitis: a meta- analysis. World J Gastroenterol 2005;11(26):4067-77. Available: http://www.wjgnet.com/1007- 9327/abstract_en.asp?f=4067&v=11 (accessed 2006 Nov 28).

94. McQuaid KR, Laine L. Early heartburn relief with proton pump inhibitors : a systematic review and meta - analysis of clinical trials. Clin Gastroenterol Hepatol 2005;3(6):553-63.

95. Klok RM, Postma MJ, van Hout BA, Brouwers JR. Meta-analysis: comparing the efficacy of proton pump inhibitors in short-term use. Aliment Pharmacol Ther 2003;17(10):1237-45.

96. Caro JJ, Salas M, Ward A. Healing and relapse rates in gastroesophageal reflux disease treated with the newer proton-pump inhibitors lansoprazole, rabeprazole, and pantoprazole compared with omeprazole, ranitidine, and placebo: evidence from randomized clinical trials. Clin Ther 2001;23(7):998-1017.

97. Dekkers CP, Beker JA, Thjodleifsson B, Gabryelewicz A, Bell NE, Humphries TJ. Double-blind comparison [correction of double-blind, placebo-controlled comparison] of rabeprazole 20 mg vs. omeprazole 20 mg in the treatment of erosive or ulcerative gastro-oesophageal reflux disease. The European Rabeprazole Study Group. Aliment Pharmacol Ther 1999;13(1):49-57.

98. Chen C, Lu C, Luo J, Chang F, Lee S, Lai Y. Esomeprazole tablet vs omeprazole capsule in treating erosive esophagitis. World J Gastroenterol 2005;11(20):3112-7. Available: http://www.wjgnet.com/1007-9327/abstract_en.asp?f=3112&v=11 (accessed 2006 Nov 28).

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 128 Consultation Document: Canadian Agency for Drugs and Technologies in Health

99. Mönnikes H, Pfaffenberger B, Gatz G, Hein J, Bardhan KD. Novel measurement of rapid treatment success with ReQuest™: first and sustained symptom relief as outcome parameters in patients with endoscopy-negative GERD receiving 20 mg pantoprazole or 20 mg esomeprazole. Digestion 2005;71(3):152-8.

100. Fock KM, Teo EK, Ang TL, Chua TS, Ng TM, Tan YL. Rabeprazole vs esomeprazole in non- erosive gastro-esophageal reflux disease: a randomized, double-blind study in urban Asia. World J Gastroenterol 2005;11(20):3091-8. Available: http://www.wjgnet.com/1007- 9327/abstract_en.asp?f=3091&v=11 (accessed 2006 Nov 28).

101. Armstrong D, Talley NJ, Lauritsen K, Moum B, Lind T, Tunturi-Hihnala H, et al. The role of acid suppression in patients with endoscopy-negative reflux disease: the effect of treatment with esomeprazole or omeprazole. Aliment Pharmacol Ther 2004;20(4):413-21.

102. Chey W, Huang B, Jackson RL. Lansoprazole and esomeprazole in symptomatic GERD: a double-blind, randomised, multicentre trial in 3000 patients confirms comparable symptom relief. Clin Drug Invest 2003;23(2):69-84.

103. Lauritsen K, Devière J, Bigard MA, Bayerdörffer E, Mózsik G, Murray F, et al. Esomeprazole 20 mg and lansoprazole 15 mg in maintaining healed reflux oesophagitis: Metropole study results. Aliment Pharmacol Ther 2003;17(3):333-41.

104. Vcev A, Stimac D, Vceva A, Takac B, Ivandic A, Pezerovic D, et al. Pantoprazole versus omeprazole in the treatment of reflux esophagitis. Acta Med Croatica 1999;53(2):79-82.

105. Janssen W, Meier E, Gudrun GATZ, Pfaffenberger B. Effects of pantoprazole 20 mg in mild gastroesophageal reflux disease: once-daily treatment in the acute phase, and comparison of on- demand versus continuous treatment in the long term. Curr Ther Res 2005;66(4):345-63.

106. Pace F, Negrini C, Wiklund I, Rossi C, Savarino V. Quality of life in acute and maintenance treatment of non-erosive and mild erosive gastro-oesophageal reflux disease. Aliment Pharmacol Ther 2005;22(4):349-56.

107. Tsai HH, Chapman R, Shepherd A, McKeith D, Anderson M, Vearer D, et al. Esomeprazole 20 mg on-demand is more acceptable to patients than continuous lansoprazole 15 mg in the long- term maintenance of endoscopy-negative gastro-oesophageal reflux patients: the COMMAND Study. Aliment Pharmacol Ther 2004;20(6):657-65.

108. Mahon D, Rhodes M, Decadt B, Hindmarsh A, Lowndes R, Beckingham I, et al. Randomized clinical trial of laparoscopic Nissen fundoplication compared with proton-pump inhibitors for treatment of chronic gastro-oesophageal reflux. Br J Surg 2005;92(6):695-9.

109. Lundell L, Miettinen P, Myrvold HE, Pedersen SA, Thor K, Lamm M, et al. Long-term management of gastro-oesophageal reflux disease with omeprazole or open antireflux surgery: results of a prospective, randomized clinical trial. The Nordic GORD Study Group. Eur J Gastroenterol Hepatol 2000;12(8):879-87.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 129 Consultation Document: Canadian Agency for Drugs and Technologies in Health

110. Lundell L, Miettinen P, Myrvold HE, Pedersen SA, Liedman B, Hatlebakk JG, et al. Continued (5-year) followup of a randomized clinical study comparing antireflux surgery and omeprazole in gastroesophageal reflux disease. J Am Coll Surg 2001;192(2):172-9.

111. van Zanten SJOV, Chiba N, Armstrong D, Barkun A, Thomson A, Smyth S, et al. A randomized trial comparing omeprazole, ranitidine, cisapride, or placebo in helicobacter pylori negative, primary care patients with dyspepsia: the CADET-HN study. Am J Gastroenterol 2005;100(7):1477-88.

112. Laheij RJ, Severens JL, Van de Lisdonk EH, Verbeek AL, Jansen JB. Randomized controlled trial of omeprazole or endoscopy in patients with persistent dyspepsia: a cost-effectiveness analysis. Aliment Pharmacol Ther 1998;12(12):1249-56.

113. Moayyedi P, Soo S, Deeks J, Delaney B, Innes M, Forman D. Pharmacological interventions for non-ulcer dyspepsia [Cochrance review]. Cochrane Database Syst Rev 2006;(2):CD001960.

114. Shiau JY, Shukla VK, Dubé C. The efficacy of proton pump inhibitors in adults with functional dyspepsia. Ottawa: Canadian Coordinating Office for Health Technology Assessment; 2002.

115. Talley NJ, Meineche Schmidt V, Pare P, Duckworth M, Raisanen P, Pap A, et al. Efficacy of omeprazole in functional dyspepsia: double-blind, randomized, placebo-controlled trials (the Bond and Opera studies). Aliment Pharmacol Ther 1998;12(11):1055-65.

116. van Zanten SV, Armstrong D, Chiba N, Flook N, White RJ, Chakraborty B, et al. Esomeprazole 40 mg once a day in patients with functional dyspepsia: the randomized, placebo-controlled "ENTER" trial. Am J Gastroenterol 2006;101(9):2096-106.

117. Blum AL, Arnold R, Stolte M, Fischer M, Koelz HR. Short course acid suppressive treatment for patients with functional dyspepsia: results depend on Helicobacter pylori status. The Frosch Study Group. Gut 2000;47(4):473-80. Available: http://gut.bmj.com/cgi/content/abstract/47/4/473 (accessed 2006 Nov 28).

118. Manes G, Menchise A, de Nucci C, Balzano A. Empirical prescribing for dyspepsia: randomised controlled trial of test and treat versus omeprazole treatment. BMJ 2003;326(7399):1118-21. Available: http://bmj.bmjjournals.com/cgi/reprint_abr/326/7399/1118 (accessed 2006 Mar 2).

119. Gisbert JP, Pajares JM. Esomeprazole-based therapy in Helicobacter pylori eradication: a meta- analysis. Dig Liver Dis 2004;36(4):253-9.

120. Gisbert JP, Khorrami S, Calvet X, Pajares JM. Pantoprazole based therapies in Helicobacter pylori eradication: a systematic review and meta-analysis. Eur J Gastroenterol Hepatol 2004;16(1):89-99.

121. Gisbert JP, Khorrami S, Calvet X, Pajares JM. Systematic review: rabeprazole-based therapies in Helicobacter pylori eradication. Aliment Pharmacol Ther 2003;17(6):751-64.

122. Vergara M, Vallve M, Gisbert JP, Calvet X. Meta-analysis: comparative efficacy of different proton-pump inhibitors in triple therapy for Helicobacter pylori eradication. Aliment Pharmacol Ther 2003;18(6):647-54.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 130 Consultation Document: Canadian Agency for Drugs and Technologies in Health

123. Moayyedi P, Murphy B. Helicobacter pylori: a clinical update. J Appl Microbiol 2001;(30):126S-33S.

124. Hawkey CJ, Atherton JC, Treichel HC, Thjodleifsson B, Ravic M. Safety and efficacy of 7-day rabeprazole- and omeprazole-based triple therapy regimens for the eradication of Helicobacter pylori in patients with documented peptic ulcer disease. Aliment Pharmacol Ther 2003;17(8):1065-74.

125. Wong BC, Wong WM, Yee YK, Hung WK, Yip AW, Szeto ML, et al. Rabeprazole-based 3-day and 7-day triple therapy vs. omeprazole-based 7-day triple therapy for the treatment of Helicobacter pylori infection. Aliment Pharmacol Ther 2001;15(12):1959-65.

126. Tulassay Z, Kryszewski A, Dite P, Kleczkowski D, Rudzinski J, Bartuzi Z, et al. One week of treatment with esomeprazole-based triple therapy eradicates Helicobacter pylori and heals patients with duodenal ulcer disease. Eur J Gastroenterol Hepatol 2001;13(12):1457-65.

127. Spinzi GC, Bierti L, Bortoli A, Colombo E, Fertitta AM, Lanzi GL, et al. Comparison of omeprazole and lansoprazole in short-term triple therapy for Helicobacter pylori infection. Aliment Pharmacol Ther 1998;12(5):433-8.

128. Hsu PI, Lai KH, Lin CK, Chen WC, Yu HC, Cheng JS, et al. A prospective randomized trial of esomeprazole-versus pantoprazole-based triple therapy for Helicobacter pylori eradication. Am J Gastroenterol 2005;100(11):2387-92.

129. Vallve M, Vergara M, Gisbert JP, Calvet X. Single vs. double dose of a proton pump inhibitor in triple therapy for Helicobacter pylori eradication: a meta-analysis. Aliment Pharmacol Ther 2002;16(6):1149-56.

130. Calvet X, García N, López T, Gisbert JP, Gené E, Roque M. A meta-analysis of short versus long therapy with a proton pump inhibitor, clarithromycin and either metronidazole or amoxycillin for treating Helicobacter pylori infection. Aliment Pharmacol Ther 2000;14(5):603- 9.

131. Gisbert JP, Pajares JM. Systematic review and meta-analysis: is 1-week proton pump inhibitor- based triple therapy sufficient to heal peptic ulcer? Aliment Pharmacol Ther 2005;21(7):795-804.

132. Agrawal NM, Campbell DR, Safdi MA, Lukasik NL, Huang B, Haber MM. Superiority of lansoprazole vs ranitidine in healing nonsteroidal anti-inflammatory drug-associated gastric ulcers: results of a double-blind, randomized, multicenter study. NSAID-Associated Gastric Ulcer Study Group. Arch Intern Med 2000;160(10):1455-61.

133. Yeomans ND, Tulassay Z, Juhász L, Rácz I, Howard JM, van Rensburg CJ, et al. A comparison of omeprazole with ranitidine for ulcers associated with nonsteroidal antiinflammatory drugs. Acid Suppression Trial: Ranitidine versus Omeprazole for NSAID-associated Ulcer Treatment (ASTRONAUT) Study Group. N Engl J Med 1998;338(11):719-26.

134. Hawkey CJ, Karrasch JA, Szczepanski L, Walker DG, Barkun A, Swannell AJ, et al. Omeprazole compared with misoprostol for ulcers associated with nonsteroidal antiinflammatory

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 131 Consultation Document: Canadian Agency for Drugs and Technologies in Health

drugs. Omeprazole versus Misoprostol for NSAID-induced Ulcer Management (OMNIUM) Study Group. N Engl J Med 1998;338(11):727-34.

135. Hooper L, Brown TJ, Elliott R, Payne K, Roberts C, Symmons D. The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti- inflammatory drugs: systematic review. BMJ 2004;329(7472). Available: http://www.bmj.com/cgi/content/full/329/7472/948 (accessed 2006 Nov 28).

136. Rostom A, Dube C, Wells G, Tugwell P, Welch V, Jolicoeur E, et al. Prevention of NSAID- induced gastroduodenal ulcers. Cochrane Database Syst Rev 2002;(4):CD002296.

137. Graham DY, Agrawal NM, Campbell DR, Haber MM, Collis C, Lukasik NL, et al. Ulcer prevention in long-term users of nonsteroidal anti-inflammatory drugs: results of a double-blind, randomized, multicenter, active- and placebo-controlled study of misoprostol vs lansoprazole. Arch Intern Med 2002;162(2):169-75.

138. Graham DY. Critical effect of Helicobacter pylori infection on the effectiveness of omeprazole for prevention of gastric or duodenal ulcers among chronic NSAID users. Helicobacter 2002;7(1):1-8.

139. Rostom A, Dubé C, Jolicoeur E, Boucher M, Joyce J. Gastro-duodenal ulcers associated with the use of non-steroidal anti-inflammatory drugs: a systematic review of preventive pharmacological interventions [Technology report no 38]. Ottawa: Canadian Coordinating Office for Health Technology Assessment; 2003. Available: http://www.cadth.ca/index.php/en/hta/reports-publications/search/publication/430 (accessed 2006 Nov 28).

140. Chan FKL, Hung LT, Suen BY, Wong VWS, Hui AJ, Wu JCY, et al. Celecoxib versus diclofenac plus omeprazole in high-risk arthritis patients: results of a randomized double-blind trial. Gastroenterology 2004;127(4):1038-43.

141. Lai K, Chu K, Hui W, Wong BCY, Hu WHC, Wong W, et al. Celecoxib compared with lansoprazole and naproxen to prevent gastrointestinal ulcer complications. Am J Med 2005;118(11):1271-8.

142. Chan FK, Chung SC, Suen BY, Lee YT, Leung WK, Leung VK, et al. Preventing recurrent upper gastrointestinal bleeding in patients with Helicobacter pylori infection who are taking low- dose aspirin or naproxen. N Engl J Med 2001;344(13):967-73.

143. Regula J, Butruk E, Dekkers CP, de Boer SY, Raps D, Simon L, et al. Prevention of NSAID- associated gastrointestinal lesions: a comparison study pantoprazole versus omeprazole. Am J Gastroenterol 2006;101(8):1747-55.

144. Lai KC, Lam SK, Chu KM, Wong BC, Hui WM, Hu WH, et al. Lansoprazole for the prevention of recurrences of ulcer complications from long-term low-dose aspirin use. N Engl J Med 2002;346(26):2033-8.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 132 Consultation Document: Canadian Agency for Drugs and Technologies in Health

145. Lai KC, Chu KM, Hui WM, Wong BC, Hung WK, Loo CK, et al. Esomeprazole with aspirin versus clopidogrel for prevention of recurrent gastrointestinal ulcer complications. Clin Gastroenterol Hepatol 2006;4(7):860-5.

146. Chan FKL, Ching JYL, Hung LCT, Wong VWS, Leung VKS, Kung NNS, et al. Clopidogrel versus aspirin and esomeprazole to prevent recurrent ulcer bleeding. N Engl J Med 2005;352(3):238-44.

147. Hallas J, Dall M, Andries A, Andersen BS, Aalykke C, Hansen JM, et al. Use of single and combined antithrombotic therapy and risk of serious upper gastrointestinal bleeding: population based case-control study. BMJ 2006;333(7571):726. Available: http://www.bmj.com/cgi/content/full/333/7571/726 (accessed 2006 Nov 28).

148. PrPantoloc M® (pantoprazole magnesium enteric-coated tablet); 40mg pantoprazole tablet; H+, K+ - ATPase inhibitor [product monograph]. Oakville: ALTANA Pharma Inc.; 2005 Apr 22.

149. Scottish Intercollegiate Guidelines Network. Methodology checklist 2: randomised controlled trials. In: SIGN 50: a guideline developers' handbook. Edinburgh: The Network; 2004. Available: http://www.sign.ac.uk/guidelines/fulltext/50/checklist2.html (accessed 2005 Nov 17).

150. Scottish Intercollegiate Guidelines Network. Methodology checklist 3: cohort studies. In: SIGN 50: a guideline developers' handbook. Edinburgh: The Network; 2004. Available: http://www.sign.ac.uk/guidelines/fulltext/50/checklist3.html (accessed 2005 Nov 17).

151. Scottish Intercollegiate Guidelines Network. Methodology checklist 4: case-control studies. In: SIGN 50: a guideline developers' handbook. Edinburgh: The Network; 2004. Available: http://www.sign.ac.uk/guidelines/fulltext/50/checklist4.html (accessed 2005 Nov 17).

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 133 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 1: Indications for Proton Pump Inhibitors in Canada

A. Apo-Omeprazole omeprazole – 20 mg capsules Information from product monograph revised September 3, 2004 from Apotex Inc.7

Apo-Omeprazole is indicated in the treatment of conditions where a reduction of gastric acid secretion is required, such as: 1. duodenal ulcer; 2. gastric ulcer; 3. reflux esophagitis; 4. symptomatic gastroesophageal reflux disease (GERD); 5. Zollinger-Ellison Syndrome (pathological hypersecretory conditions); 6. NSAID-associated gastric and duodenal ulcers. Use in Children: The safety and effectiveness of omeprazole in children has not yet been established.

B. Losec and Losec MUPS omeprazole magnesium – 10 mg and 20 mg delayed released tablets Information from product monograph revised September 23, 2003 from AstraZeneca Canada Inc.5 omeprazole magnesium – 10 mg and 20 mg delayed release tablets [MUPS formulation] Information from product monograph revised September 23, 2003 from AstraZeneca Canada Inc.6 omeprazole – 10 mg, 20 mg and 40 mg delayed release capsules 4 Information from product monograph revised June 22, 2004 from AstraZeneca Canada Inc.

Losec tablets/capsules are indicated in the treatment of conditions where a reduction of gastric acid secretion is required, such as: Duodenal ulcer; Gastric ulcer; NSAID-associated gastric and duodenal ulcers; Reflux esophagitis; Symptomatic gastroesophageal reflux disease (GERD) i.e., heartburn and regurgitation; Dyspepsia: a complex of symptoms which may be caused by any of the organic diseases listed above, or upon investigation no identifiable organic cause is found (i.e., functional dyspepsia) [omeprazole capsules 10 mg, 20 mg and 40 mg are not indicated for dyspepsia]; Zollinger-Ellison syndrome (pathological hypersecretory condition); Eradication of H. pylori.

Losec, in combination with clarithromycin and either amoxicillin or metronidazole, is indicated for the treatment of patients with peptic ulcer disease associated with Helicobacter infection. The optimal timing for eradication therapy in patients whose ulcer is not clinically active (i.e. asymptomatic) remains to be determined. Use in Children: The safety and effectiveness of Losec tablets in children have not yet been established.

C. Nexium esomeprazole magnesium trihydrate – 20 mg and 40 mg delayed release tablets. Information from product monograph revised November 23, 2005 from AstraZeneca Canada Inc.10

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 134 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Nexium is indicated for treatment of conditions where a reduction in gastric acid secretion is required such as: Reflux esophagitis Maintenance treatment of patients with reflux esophagitis Symptomatic gastroesophageal reflux disease (i.e. heartburn and regurgitation) Healing of NSAID-associated gastric ulcers Reduction of risk of NSAID-associated gastric ulcers Helicobacter pylori (H. pylori) eradication

Nexium, in combination with clarithromycin and amoxicillin, is indicated for the treatment of patients with duodenal ulcer disease associated with Helicobacter pylori infection to eradicate the H. pylori and heal ulcers. Eradication of H. pylori has been shown to reduce the risk of duodenal ulcer recurrence. Pediatrics: The safety and effectiveness of Nexium tablets in children have not yet been established.

D. Pantoloc pantoprazole sodium – 20 mg and 40 mg enteric-coated tablets. Information from product monograph revised May 17, 2005 from Solvay Pharma Inc.9

Pantoloc is indicated for the treatment of conditions where a reduction of gastric acid secretion is required, such as the following: Duodenal ulcer Gastric ulcer Reflux esophagitis Symptomatic gastro-esophageal reflux disease (such as, acid regurgitation and heartburn). Prevention of gastrointestinal lesions induced by non-steroidal anti-inflammatory drugs (NSAIDs) in patients with a need for continuous NSAID treatment, who have increased risk to develop NSAID-associated upper gastrointestinal lesions. Helicobacter pylori associated duodenal ulcer Pantoprazole, in combination with clarithromycin and either amoxicillin or metronidazole, is indicated for the treatment of patients with an active duodenal ulcer who are H. pylori-positive. Clinical trials using combinations of pantoprazole with appropriate antibiotics have indicated that such combinations are successful in eradicating H. pylori

For the maintenance treatment of patients with reflux esophagitis and the rapid resolution of symptoms associated with reflux esophagitis, such as heartburn, regurgitation and dyspepsia, 20 mg pantoprazole once daily in the morning has been used for up to 12 months in controlled clinical trials, and in continuous maintenance treatment, in a limited number of patients for up to eight years. Pediatrics: The safety and effectiveness of pantoprazole in children have not yet been established.

D. Pantoloc M pantoprazole magnesium – 40 mg enteric-coated tablets. Information from product monograph prepared April 22, 2005 from ALTANA Pharma Inc.148 Pantoloc M is indicated for the treatment of conditions where a reduction of gastric acid secretion is required, such as the following: Duodenal ulcer Gastric ulcer Reflux esophagitis Symptomatic gastro-esophageal reflux disease (such as, acid regurgitation and heartburn).

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 135 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Helicobacter pylori associated duodenal ulcer Pantoprazole, in combination with clarithromycin and either amoxicillin or metronidazole, is indicated for the treatment of patients with an active duodenal ulcer who are H. pylori-positive. Clinical trials using combinations of pantoprazole with appropriate antibiotics have indicated that such combinations are successful in eradicating H. pylori

Pediatrics: The safety and effectiveness of pantoprazole in children have not yet been established.

Geriatrics (>65 years of age): No dosage adjustment is recommended based on age. The daily dose used in elderly patients, as a rule, should not exceed the recommended dosage regimens.

E. Pariet rabeprazole sodium – 10 mg and 20 mg enteric-coated tablets Information from product monograph revised January 26 2005 from Janssen-Ortho Inc.11

Pariet is indicated for: Treatment of conditions where a reduction of gastric acid secretion is required, such as: 1. Symptomatic relief and healing of erosive or ulcerative gastroesophageal reflux disease (GERD). 2. Long-term maintenance of healing of erosive or ulcerative gastroesophageal reflux disease (GERD). 3. Treatment of symptoms (i.e. heartburn and regurgitation) in symptomatic gastroesophageal reflux disease (GERD), also called non-erosive reflux disease (NERD). 4. Symptomatic relief and healing of duodenal ulcers. 5. Symptomatic relief and healing of gastric ulcers. 6. Long-term treatment of pathological hypersecretory conditions, including Zollinger-Ellison syndrome. 7. Eradication of H. pylori associated with duodenal ulcer disease (active or history within the past 5 years). Eradication of H. pylori has been shown to reduce the risk of duodenal ulcer recurrence. Clinical trials using combinations of rabeprazole with appropriate antibiotics have indicated that such combinations are successful in eradicating H. pylori.

Pediatrics (< 18 years of age): The safety and efficacy of rabeprazole have not been established in children under the age of 18 years.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 136 Consultation Document: Canadian Agency for Drugs and Technologies in Health

F. Prevacid lansoprazole – 15mg and 30 mg delayed release capsules Information from product monograph revised June 15, 2005 from TAP Pharmaceuticals Inc. (Distributed by Abbott Laboratories, Limited)8

Prevacid is indicated in the treatment of conditions where a reduction of gastric acid secretion is required, such as: 1. Duodenal ulcer. 2. Gastric ulcer. 3. Reflux esophagitis including patients with Barrett’s esophagus, and patients poorly responsive to an adequate course of therapy with histamine H2-receptor antagonists. 4. Healing of NSAID-Associated Gastric Ulcer; treatment of NSAID-associated gastric ulcer in patients who continue NSAID use. (Controlled studies did not extend beyond 8 weeks). 5. Reduction of Risk of NSAID-Associated Gastric Ulcers in patients with a history of gastric ulcers who require to continue taking a NSAID. (A controlled study did not extend beyond 12 weeks). 6. Symptomatic Gastroesophageal reflux disease (GERD); treatment of heartburn and other symptoms associated with GERD. 7. Pathological hypersecretory conditions including Zollinger-Ellison Syndrome. 8. Eradication of Helicobacter pylori (H. pylori).

Triple Therapy: Lansoprazole, in combination with clarithromycin plus amoxicillin as triple therapy, is indicated for the treatment of patients with H. pylori infection and active duodenal ulcer disease. Eradication of H. pylori has been shown to reduce the risk of duodenal ulcer recurrence.

Pediatric GERD (erosive and non-erosive esophagitis) (1 to 17 years of age): Prevacid is indicated for treatment of erosive and non-erosive GERD in children, aged 1 to 17 years. The clinical trial treatment period did not extend beyond 12 weeks. Dose safety and effectiveness have not been established in patients <1 year.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 137 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 2: Literature Search Strategies

Guide to DIALOG® Search Syntax ? Truncation symbol. Retrieves plural and variant endings. n Proximity operator. Words can be in any order. w Proximity operator. Words must be adjacent, in given order. l Proximity operator. Subject heading must be linked to subject subheading. ti Title. Search in article titles. ab Abstract. Search in article abstracts. de Descriptor (i.e. subject heading). Search in subject headings. ! Explode descriptor (i.e. retrieve the search concept plus all narrower terms). dt Publication type. rn Registry number.

GUIDELINES SEARCH Search Logic

#1 Indications for the use of proton pump inhibitors #2 Proton pump inhibitors #3 Guidelines and/or consensus statements #4 (#1 OR #2) AND #3 #5 Apply human limit DATABASES LIMITS SUBJECT HEADINGS/KEYWORDS DIALOG One Search® #1 INDICATIONS FOR THE USE OF PROTON PUMP INHIBITORS (May 18, 2005) Human (gastrointestinal hemorrhage OR peptic ulcer hemorrhage)/de from MEDLINE MEDLINE® OR (1955-present) (gastrointestinal hemorrhage OR peptic ulcer bleeding OR upper gastrointestinal BIOSIS bleeding)/de from EMBASE Previews® (1969- OR present) (gastrointestinal hemorrhage OR upper gastrointestinal bleeding)/de from BIOSIS EMBASE® Previews (1974-present) OR PASCAL ((gastrointestinal OR gastro(w)intestinal OR gi)(2n)(hemorrhag? OR haemorrhag? OR perforat? OR bleed? OR rebleed?))/ti,ab OR (ulcer?(2n)((hemorrhag? OR haemorrhag? OR perforat? OR bleed? OR rebleed?) OR (gastrointestinal OR gastro(w)intestinal OR gi)))/ti,ab OR gastric mucosa(l)in from MEDLINE OR stomach mucosa injury/de from EMBASE OR gastric mucosal injury/de from BIOSIS Previews OR mucosa?(2n)injur?/ti,ab OR peptic ulcer!/de from MEDLINE, EMBASE, BIOSIS Previews OR

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 138 Consultation Document: Canadian Agency for Drugs and Technologies in Health

peptic ulcer disease/de from BIOSIS Previews OR ((peptic OR stomach OR duoden? OR gastroduoden? OR gastric)(2n)ulcer?)/ti,ab OR gastroesophageal reflux/de from MEDLINE,EMBASE,BIOSIS Previews OR barrett esophagus/de from MEDLINE,EMBASE OR barrett's esophagus/de from BIOSIS Previews OR (gastro-esophageal reflux OR gastro-esophageal reflux disease OR gastroesophageal reflux disease)/de from BIOSIS Previews OR (esophageal(w)reflux OR gastro(w)oesophageal(w)reflux OR gastroesophageal(w)reflux OR gerd OR gord OR gastric(w)regurgitation OR acid(w)reflux OR barrett?(w)esophagus OR barrett?(w)oesophagus)/ti,ab OR (dyspepsia OR heartburn)/de from MEDLINE, EMBASE, BIOSIS Previews OR (dyspepsia? OR indigestion OR heartburn)/ti,ab OR helicobacter infections/de from MEDLINE OR helicobacter infection/de from EMBASE OR (helicobacter pylori gastritis OR helicobacter pylori infection)/de from BIOSIS Previews OR (helicobacter OR h(w)pylori OR campylobacter)(n4)(infection OR infections) OR gastric acid(l)se from MEDLINE OR stomach acid secretion/de from EMBASE OR gastric acid secretion/de from BIOSIS Previews OR (gastric(2n)hypersecret?) OR idiopathic(w)hypersecretion/ti,ab OR zollinger-ellison syndrome/de from BIOSIS Previews OR (zollinger(w)ellison OR ellison(w)zollinger OR zes)/ti,ab OR esophagitis!/de from MEDLINE, EMBASE, BIOSIS Previews OR (esophagitis OR esophagitides OR oesophagitis OR oesophagitides)/ti,ab

#2 PROTON PUMP INHIBITORS

proton pumps(l)ai/de from MEDLINE OR proton pump inhibitor!/maj from EMBASE OR

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 139 Consultation Document: Canadian Agency for Drugs and Technologies in Health

proton pump inhibitors/de from BIOSIS Previews OR (proton(w)pump(w)inhibitor? OR ppi OR ppis)/ti,ab OR omeprazole/de from MEDLINE, BIOSIS Previews OR (omeprazole OR Antra OR Audazol OR Aulcer OR Belmazol OR CCRIS(w)7099 OR Ceprandal OR Danlox OR Demeprazol OR Desec OR Dizprazol OR Dudencer OR Elgam OR Emeproton OR Epirazole OR Erbolin OR Exter OR Gasec OR Gastrimut OR Gastroloc OR Gibancer OR H(w)168(w)68 OR HSDB(w)3575 OR Indurgan OR Inhibitron OR Inhipump OR Lensor OR Logastric)/ti,ab OR (Lomac OR Losec OR Mepral OR Miol OR Miracid OR Mopral OR Morecon OR Nilsec OR Nopramin OR OMEP OR OMP OR OMZ OR Ocid OR Olexin OR Omapren OR Omebeta(w)20 OR Omed OR Omegast OR Omepral OR Omeprazol OR Omeprazole OR Omeprazolum OR Omeprazon OR Omeprol OR Omesek OR Omezol OR Omezolan OR Omid OR Omisec)/ti,ab OR (Omizac OR Ompanyt OR Ortanol OR Osiren OR Ozoken OR Paprazol OR Parizac OR Pepticum OR Pepticus OR Peptilcer OR Prazentol OR Prazidec OR Prazolit OR Prilosec OR Procelac OR Proclor OR Prysma OR Ramezol OR Regulacid OR Sanamidol OR Secrepina OR Tedec Ulceral OR Ulceral OR Ulcesep OR Ulcometion OR Ulcozol OR Ulcsep OR Ulsen OR Ultop OR Ulzol)/ti,ab OR (Victrix OR Zefxon OR Zegerid OR Zepral OR Zimor OR Zoltum OR Zanprol OR Ufiprazole OR Ufiprazol OR Ufiprazolum OR Andra)/ti,ab OR s rn=(73590-58-6 OR 73590-85-9 OR 88546-55-8 OR 95382-33-5 OR 95510-70- 6 OR 102332-89-8 OR 120003-84-1) from MEDLINE,BIOSIS Previews,PASCAL OR esomeprazole/de from BIOSIS Previews OR (esomeprazole OR Nexium OR Perprazole OR Nexiam OR Inexium OR Sompraz OR Axagon OR Esopral OR Lucen OR Axiago)/ti,ab OR rn=(119141-88-7 OR 161796-78-7 OR 161973-10-0 OR 217087-09-7) from MEDLINE, BIOSIS Previews, PASCAL OR lansoprazole/de from BIOSIS Previews OR (lansoprazole OR A(w)65006 OR AG(w)1749 OR Agopton OR Alexin OR Amarin OR Aprazol OR BRN(w)4333393 OR Bamalite OR Blason OR Compraz OR Dakar OR Estomil OR Fudermex OR Gastrex OR Gastride OR Gastroliber OR HSDB(w)7204 OR Ilsatec OR Ketian OR Keval)/ti,ab OR (Lancid OR Lanfast OR Lanproton OR Lansopep OR Lansoprazol OR Lansoprazole OR Lansoprazolum OR Lansox OR Lanston OR Lanz OR Lanzo OR Lanzogastro OR Lanzol OR Lanzol(w)30 OR Lanzopral OR Lanzor OR Lasoprol OR Limpidex OR Lizul OR Mesactol)/ti,ab

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 140 Consultation Document: Canadian Agency for Drugs and Technologies in Health

OR (Monolitum OR Ogast OR Ogasto OR Ogastro OR Opiren OR Pampe OR Peptomil OR Prevacid OR Prezal OR Pro(w)Ulco OR Promp OR Prosogan OR Suprecid OR Takepron OR Ulcertec OR Uldapril OR Ulpax OR Unival OR Zoprol OR Zoton)/ti,ab OR rn=(103577-45-3) from MEDLINE, BIOSIS Previews, PASCAL OR pantoprazole/de from BIOSIS Previews OR (pantoprazole OR BY(w)1023 OR Pantoprazol OR Pantoprazole OR Pantoprazolum OR SK&F(w)96022 OR Controloc OR Pantoloc OR Protonix OR Angastra OR Apton OR Eupantol OR Inipomp OR Gastromax OR Noprop OR Pamgest OR Pantecta OR Pant OR Pantoc)/ti,ab OR (Pantocal OR Pantocarm OR Pantodac OR Pantop OR Pantopan OR Pantopaz OR Pantorc OR Pantozol OR Pantozol(w)Rifun OR Pantus OR Peptazol OR Protium OR Rifun OR Singastril OR Somac OR Supracam OR Ulcemex OR Ulcotenal OR Ulserch OR Ziprol OR Zurcal OR Zurcale OR Zurcazol)/ti,ab OR rn=(102625-70-7 OR 138786-67-1 OR 164579-32-2) from MEDLINE, BIOSIS Previews, PASCAL OR rabeprazole/de from BIOSIS Previews OR (rabeprazole OR Aciphex OR E(w)3810 OR Gastrodine OR LY(w)307640(w)sodium OR Pariet OR Rabec OR Rabeloc)/ti,ab OR rn=(117976-90-6) from MEDLINE, BIOSIS Previews, PASCAL

#3 GUIDELINES AND/OR CONSENSUS STATEMENTS

guidelines!/de from MEDLINE, BIOSIS Previews OR (clinical guidelines OR clinical practice guidelines)/de from BIOSIS Previews OR (critical pathways OR health planning guidelines)/de from MEDLINE OR consensus development conferences!/de from MEDLINE OR practice guideline!/de from EMBASE OR dt=(practice guideline OR guideline OR consensus development conference OR consensus development conference, nih) OR (cpg OR cpgs OR (critical OR clinical OR practice)(w)(path OR paths OR pathway OR pathways OR protocol OR protocols OR guideline OR guidelines) OR care(w)(path OR paths OR pathway OR pathways OR map OR maps OR plan OR plans) OR consensus)/ti,ab

The Cochrane Same search logic, MeSH descriptors and keywords as DIALOG® MEDLINE

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 141 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Library 2005, search; adapted search commands for Wiley InterScience® search interface. issue 2 (May 19, 2005) PubMed Human Same search logic, MeSH descriptors and keywords as DIALOG® MEDLINE® (May 17, 2005) search; adapted search commands for PubMed search interface. CINAHL Same search logic and keywords as DIALOG® MEDLINE® search; converted (May 19, 2005) MeSH descriptors for CINAHL thesaurus; adapted search commands for Ovid search interface. Searched online guidelines collections (including CMA Infobase, AHRQ's National Guidelines Clearinghouse, the NHS National Electronic Library of Health Guidelines Finder, Guidelines International Network) as well as the web sites of guideline producing bodies, relevant professional associations and other online databases and web sites.

EVIDENCE UPDATE SEARCH Search Logic

(Indications AND PPIs) AND (Study Types: RCTs OR systematic reviews) Limits: 2003-2006, Human, English language DATABASES LIMITS SUBJECT HEADINGS/KEYWORDS DIALOG One Search® 2003-2006 INDICATIONS FOR THE USE OF PROTON PUMP INHIBITORS (January 18, 2006) Human ((gastrointestinal OR gastro()intestinal OR gi)(2n)(hemorrhag? OR haemorrhag? OR perforat? OR bleed? OR rebleed?))/ti,ab ® MEDLINE English OR (1990-present) s (ulcer?(2n)((hemorrhag? OR haemorrhag? OR perforat? OR bleed? OR BIOSIS rebleed?) OR (gastrointestinal OR gastro()intestinal OR gi)))/ti,ab Previews® (1993- OR present) (gastrointestinal hemorrhage OR Peptic Ulcer Hemorrhage)/de from EMBASE® MEDLINE (1993-present) OR PASCAL (gastrointestinal hemorrhage OR peptic ulcer bleeding OR upper gastrointestinal bleeding)/de from EMBASE OR (gastrointestinal hemorrhage OR upper gastrointestinal bleeding)/de from BIOSIS OR gastric mucosa(l)in from MEDLINE OR stomach mucosa injury/de from EMBASE OR gastric mucosal injury/de from BIOSIS OR mucosa?(2n)injur?/ti,ab OR Peptic ulcer!/de from MEDLINE,EMBASE,BIOSIS OR Peptic ulcer disease/de from BIOSIS OR

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 142 Consultation Document: Canadian Agency for Drugs and Technologies in Health

((Peptic OR stomach OR duoden? OR gastroduoden? OR gastric)(2n)ulcer?)/ti,ab OR Gastroesophageal Reflux/de from MEDLINE,EMBASE,BIOSIS OR Barrett Esophagus/de from MEDLINE,EMBASE OR Barrett's esophagus/de from BIOSIS OR (gastro-esophageal reflux OR gastro-esophageal reflux disease OR gastroesophageal reflux disease)/de from BIOSIS OR (Esophageal()Reflux OR Gastro()oesophageal()Reflux OR Gastroesophageal()Reflux OR GERD OR GORD OR Gastric()Regurgitation OR acid()reflux OR barrett?()esophagus OR barrett?()oesophagus)/ti,ab OR (Dyspepsia OR Heartburn)/de from MEDLINE,EMBASE,BIOSIS OR (Dyspepsia? OR indigestion OR heartburn)/ti,ab OR Helicobacter infections/de from MEDLINE OR Helicobacter infection/de from EMBASE OR (Helicobacter pylori gastritis OR helicobacter pylori infection)/de from BIOSIS OR (Helicobacter OR h()pylori OR campylobacter)(n4)(infection OR infections)/ti,ab OR gastric acid(l)se from MEDLINE OR stomach acid secretion/de from EMBASE OR gastric acid secretion/de from BIOSIS OR (gastric(2n)hypersecret?) OR idiopathic()hypersecretion/ti,ab OR Zollinger-Ellison syndrome/de from BIOSIS OR (Zollinger()Ellison OR Ellison()Zollinger OR ZES)/ti,ab OR Esophagitis!/de from MEDLINE,EMBASE,BIOSIS OR (esophagitis OR Esophagitides OR oesophagitis OR oesophagitides)/ti,ab

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 143 Consultation Document: Canadian Agency for Drugs and Technologies in Health

PROTON PUMP INHIBITORS

Proton pumps(l)ai/de from MEDLINE OR Proton pump inhibitor!/maj from EMBASE OR Proton pump inhibitors/de from BIOSIS OR (Proton()pump()inhibitor? OR PPI OR PPIs)/ti,ab OR omeprazole/de from MEDLINE,BIOSIS OR (omeprazole OR Antra OR Audazol OR Aulcer OR Belmazol OR CCRIS()7099 OR Ceprandal OR Danlox OR Demeprazol OR Desec OR Dizprazol OR Dudencer OR Elgam OR Emeproton OR Epirazole OR Erbolin)/ti,ab OR (Exter OR Gasec OR Gastrimut OR Gastroloc OR Gibancer OR H()168()68 OR HSDB()3575 OR Indurgan OR Inhibitron OR Inhipump OR Lensor OR Logastric)/ti,ab OR (Lomac OR Losec OR Mepral OR Miol OR Miracid OR Mopral OR Morecon OR Nilsec OR Nopramin OR OMEP OR OMP OR OMZ OR Ocid OR Olexin OR Omapren OR Omebeta()20 OR Omed OR Omegast OR Omepral OR Omeprazol OR Omeprazole)/ti,ab OR (Omeprazolum OR Omeprazon OR Omeprol OR Omesek OR Omezol OR Omezolan OR Omid OR Omisec OR Omizac OR Ompanyt OR Ortanol OR Osiren OR Ozoken OR Paprazol OR Parizac OR Pepticum OR Pepticus OR Peptilcer)/ti,ab OR (Prazentol OR Prazidec OR Prazolit OR Prilosec OR Procelac OR Proclor OR Prysma OR Ramezol OR Regulacid OR Sanamidol OR Secrepina OR Tedec Ulceral OR Ulceral OR Ulcesep OR Ulcometion OR Ulcozol OR Ulcsep OR Ulsen)/ti,ab OR (Ultop OR Ulzol OR Victrix OR Zefxon OR Zegerid OR Zepral OR Zimor OR Zoltum OR Zanprol OR Ufiprazole OR Ufiprazol OR Ufiprazolum OR Andra)/ti,ab OR rn=(73590-58-6 OR 73590-85-9 OR 88546-55-8 OR 95382-33-5 OR 95510-70-6 OR 102332-89-8 OR 120003-84-1) from MEDLINE,BIOSIS,PASCAL OR esomeprazole/de from BIOSIS OR (esomeprazole OR Nexium OR Perprazole OR Nexiam OR Inexium OR

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 144 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Sompraz OR Axagon OR Esopral OR Lucen OR Axiago)/ti,ab OR rn=(119141-88-7 OR 161796-78-7 OR 161973-10-0 OR 217087-09-7) from MEDLINE,BIOSIS,PASCAL OR lansoprazole/de from BIOSIS OR (lansoprazole OR A()65006 OR AG()1749 OR Agopton OR Alexin OR Amarin OR Aprazol OR BRN()4333393 OR Bamalite OR Blason OR Compraz OR Dakar OR Estomil OR Fudermex OR Gastrex OR Gastride OR Gastroliber)/ti,ab OR (HSDB()7204 OR Ilsatec OR Ketian OR Keval OR Lancid OR Lanfast OR Lanproton OR Lansopep OR Lansoprazol OR Lansoprazole OR Lansoprazolum OR Lansox OR Lanston OR Lanz OR Lanzo OR Lanzogastro OR Lanzol)/ti,ab OR (Lanzol()30 OR Lanzopral OR Lanzor OR Lasoprol OR Limpidex OR Lizul OR Mesactol OR Monolitum OR Ogast OR Ogasto OR Ogastro OR Opiren OR Pampe OR Peptomil OR Prevacid OR Prezal OR Pro()Ulco OR Promp)/ti,ab OR (Prosogan OR Suprecid OR Takepron OR Ulcertec OR Uldapril OR Ulpax OR Unival OR Zoprol OR Zoton)/ti,ab OR rn=(103577-45-3) from MEDLINE,BIOSIS,PASCAL OR pantoprazole/de from BIOSIS OR (pantoprazole OR BY()1023 OR Pantoprazol OR Pantoprazole OR Pantoprazolum OR SK&F()96022 OR Controloc OR Pantoloc OR Protonix OR Angastra OR Apton OR Eupantol OR Inipomp OR Gastromax)/ti,ab OR (Noprop OR Pamgest OR Pantecta OR Pant OR Pantoc OR Pantocal OR Pantocarm OR Pantodac OR Pantop OR Pantopan OR Pantopaz OR Pantorc OR Pantozol OR Pantozol()Rifun OR Pantus)/ti,ab OR (Peptazol OR Protium OR Rifun OR Singastril OR Somac OR Supracam OR Ulcemex OR Ulcotenal OR Ulserch OR Ziprol OR Zurcal OR Zurcale OR Zurcazol)/ti,ab OR rn=(102625-70-7 OR 138786-67-1 OR 164579-32-2) from MEDLINE,BIOSIS,PASCAL OR rabeprazole/de from BIOSIS OR

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 145 Consultation Document: Canadian Agency for Drugs and Technologies in Health

(rabeprazole OR Aciphex OR E()3810 OR Gastrodine OR LY()307640()sodium OR Pariet OR Rabec OR Rabeloc)/ti,ab OR rn=(117976-90-6) from MEDLINE,BIOSIS,PASCAL OR (benatoprazole OR leminoprazole OR picoprazole OR timoprazole)/ti,ab

STUDY TYPES: RCTs OR SYSTEMATIC REVIEWS

random? OR RCT?? OR dt=meta-analysis from MEDLINE OR meta-analysis/de from MEDLINE,BIOSIS OR (meta analysis OR systematic review)/de from EMBASE OR (meta()analy? OR metaanaly? OR met()analy? OR metanaly?)/ti,ab OR (systematic?()((literature()review?) OR review? OR overview?))/ti,ab OR (methodologic?()((literature()review?) OR review? OR overview?))/ti,ab OR (quantitative()(review? OR overview? OR synthes?))/ti,ab OR (research()(integration? OR overview?))/ti,ab OR Technology Assessment, Biomedical/de from MEDLINE OR biomedical technology assessment/de from EMBASE OR (HTA OR HTAs OR health()technology()assessment? OR (biomedical OR bio()medical)()technology()assessment?)/ti,ab

The Cochrane Same search logic, MeSH descriptors and keywords as DIALOG® MEDLINE Library 2006, search; adapted search commands for Wiley InterScience® search interface. issue 1 (January 19, 2006) Additional searching was conducted on key health technology assessment, systematic reviewing, and health economic resources, including University of York Centre for Reviews and Dissemination (CRD) databases, HEED, ECRI, EuroScan, and international HTA agencies.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 146 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 3: AMSTAR Instrument for Systematic Reviews

A MeaSurement Tool to Assess Reviews (AMSTAR), 2005

1. Was an ‘a priori’ design provided? Yes The research question and inclusion criteria should be established before the conduct of the review. No Can’t answer Not applicable 2. Was there duplicate study selection and data extraction? Yes There should be at least two independent data extractors and the consensus procedure for No disagreements should be reported. Can’t answer Not applicable 3. Was a comprehensive literature search performed? Yes At least two electronic sources should be searched. The report must include years and databases (e.g. No Central, EPOC, and MEDLINE). Key words and/or MESH terms must be stated and where feasible Can’t answer the search strategy should be provided. All searches should be supplemented by consulting current Not applicable contents, reviews, textbooks, specialized registers, or experts in the particular field of study, and by reviewing the references in the studies found. 4. Was the status of publication (i.e. grey literature) used as an exclusion criterion? Yes The authors should state that they searched for reports regardless of their publication type. The No authors should state whether or not they excluded any reports (from the systematic review), based on Can’t answer their publication status. Not applicable 5. Was a list of studies (included and excluded) provided? Yes A list of included and excluded studies should be provided. No Can’t answer Not applicable 6. Were the characteristics of the included studies provided? Yes In an aggregated form such as a table, data from the original studies should be provided on the No participants, interventions and outcomes. The ranges of characteristics in all the studies analyzed e.g. Can’t answer age, race, sex, relevant socioeconomic data, disease status, duration, severity, or other diseases Not applicable should be reported. 7. Was the scientific quality of the included studies assessed and reported? Yes ‘A priori’ methods of assessment should be reported (e.g., for effectiveness studies if the author(s) No chose to include only randomized, double-blind, placebo controlled studies, or allocation Can’t answer concealment as inclusion criteria); for other types of studies alternative items will be relevant. Not applicable 8. Was the scientific quality of the included studies used appropriately in formulating Yes conclusions? No The results of the methodological rigor and scientific quality should be considered in the analysis Can’t answer and the conclusions of the review, and explicitly stated in formulating recommendations. Not applicable 9. Were the methods used to combine the findings of studies appropriate? Yes For the pooled results, a test should be done to ensure the studies were combinable, to assess the No homogeneity (i.e. Chi-squared test for homogeneity, I²). If heterogeneity exists a random effects Can’t answer model should be used and/or the clinical appropriateness of combining should be taken into Not applicable consideration (i.e. is it sensible to combine?). 10. Was the likelihood of publication bias assessed? Yes An assessment of publication bias should include a combination of graphical aids (e.g., funnel plot) No and statistical tests (e.g., Egger regression test). Can’t answer Not applicable 11. Was the conflict of interest stated? Yes Potential sources of support should be clearly acknowledged in both the systematic review and the No included studies. Can’t answer Not applicable AMSTAR 2005 (Beverley Shea, CIET, Institute of Population Health, Ottawa: personal communication, 2005 Oct)

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 147 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 4a: Adapted SIGN 50 Checklist for Randomized Controlled Trials149

Indication: Recommendation #: Lead Author: Title: Reviewer: Date: RefMan #:

Methodology Checklist: Randomized Controlled Trials S I G N Section 1: Internal validity

In a well conducted RCT study….. In this study this criterion is: Well covered Poorly addressed Not applicable 1.1 The study addresses an appropriate and

clearly focused question. Adequately addressed Not reported Not addressed 1.2 Well covered Poorly addressed Not applicable The assignment of subjects to treatment groups is randomised Adequately addressed Not reported Not addressed 1.3 Well covered Poorly addressed Not applicable An adequate concealment method is used Adequately addressed Not reported Not addressed 1.4 Well covered Poorly addressed Not applicable Subjects and investigators are kept ‘blind’ about treatment allocation Adequately addressed Not reported Not addressed 1.5 The treatment and control groups are similar Well covered Poorly addressed Not applicable at the start of the trial Adequately addressed Not reported Not addressed 1.6 Well covered Poorly addressed Not applicable The only difference between groups is the treatment under investigation Adequately addressed Not reported Not addressed 1.7 Well covered Poorly addressed Not applicable All relevant outcomes are measured in a standard, valid and reliable way Adequately addressed Not reported Not addressed 1.8 What percentage of the individuals or clusters recruited into each treatment

arm of the study dropped out before the study was completed?

1.9 All the subjects are analysed in the groups Well covered Poorly addressed Not applicable to which they were randomly allocated (often referred to as intention to treat Adequately addressed Not reported Not addressed analysis) Well covered Poorly addressed Not applicable 1.10 Where the study is carried out at more than

one site, results are comparable for all sites Adequately addressed Not reported Not addressed Section 2: Overall Assessment Of The Study

2.1 How well was the study done to minimise bias? Code ++, +, or − Section 3: Others How was this study funded? 3.1 List all sources of funding quoted in the article, whether Government, voluntary sector, or industry.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 148 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 4b: Adapted SIGN 50 Checklist for Cohort Studies150

Indication: Recommendation #: Lead Author: Title: Reviewer: Date: RefMan #:

Methodology Checklist: Cohort studies S I G N Section 1: Internal validity In a well conducted cohort study: In this study the criterion is: Well covered Poorly addressed Not applicable The study addresses an appropriate and clearly focused 1.1 question. Adequately addressed Not reported Not addressed SELECTION OF SUBJECTS 1.2 The two groups being studied are selected from source Well covered Poorly addressed Not applicable populations that are comparable in all respects other than the factor under investigation. Adequately addressed Not reported Not addressed Well covered Poorly addressed Not applicable The study indicates how many of the people asked to 1.3 take part did so, in each of the groups being studied. Adequately addressed Not reported Not addressed 1.4 The likelihood that some eligible subjects might have Well covered Poorly addressed Not applicable the outcome at the time of enrolment is assessed and taken into account in the analysis. Adequately addressed Not reported Not addressed

1.5 What percentage of individuals or clusters recruited into each arm of the study dropped out before the study was completed.

Well covered Poorly addressed Not applicable 1.6 Comparison is made between full participants and those

lost to follow up, by exposure status . Adequately addressed Not reported Not addressed ASSESSMENT Well covered Poorly addressed Not applicable 1.7 The outcomes are clearly defined. Adequately addressed Not reported Not addressed Well covered Poorly addressed Not applicable The assessment of outcome is made blind to exposure 1.8 status. Adequately addressed Not reported Not addressed 1.9 Where blinding was not possible, there is some Well covered Poorly addressed Not applicable recognition that knowledge of exposure status could have influenced the assessment of outcome. Adequately addressed Not reported Not addressed Well covered Poorly addressed Not applicable 1.10 The measure of assessment of exposure is reliable. Adequately addressed Not reported Not addressed Well covered Poorly addressed Not applicable Evidence from other sources is used to demonstrate that 1.11 the method of outcome assessment is valid and reliable. Adequately addressed Not reported Not addressed Well covered Poorly addressed Not applicable Exposure level or prognostic factor is assessed more 1.12 than once. Adequately addressed Not reported Not addressed CONFOUNDING

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 149 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Well covered Poorly addressed Not applicable The main potential confounders are identified and taken 1.13 into account in the design and analysis. Adequately addressed Not reported Not addressed STATISTICAL ANALYSIS

1.14 Have confidence intervals been provided?

Section 2: Overall Assessment Of The Study

2.1 How well was the study done to minimise the risk of bias or confounding, and to establish a causal relationship between exposure and effect? Code ++, +, or −

Section 3: Others

3.1 How was this study funded? List all sources of funding quoted in the article, whether Government, voluntary sector, or industry.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 150 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 4c: Adapted SIGN 50 Checklist for Case Control Studies151

Indication: Recommendation #: Lead Author: Title: Reviewer: Date: RefMan #:

Methodology Checklist 4: Case-control studies S I G N

Section 1: Internal validity In an well conducted case control study: In this study the criterion is: Well covered Poorly addressed Not applicable 1.1 The study addresses an appropriate and clearly focused question Adequately addressed Not reported Not addressed SELECTION OF SUBJECTS Well covered Poorly addressed Not applicable 1.2 The cases and controls are taken from comparable populations Adequately addressed Not reported Not addressed Well covered Poorly addressed Not applicable 1.3 The same exclusion criteria are used for both cases and controls Adequately addressed Not reported Not addressed Cases: 1.4 What percentage of each group (cases and controls) participated in the study? Controls: 1.5 Comparison is made between participants and non- Well covered Poorly addressed Not applicable participants to establish their similarities or differences Adequately addressed Not reported Not addressed Well covered Poorly addressed Not applicable 1.6 Cases are clearly defined and differentiated from controls Adequately addressed Not reported Not addressed Well covered Poorly addressed Not applicable 1.7 It is clearly established that controls are non-cases Adequately addressed Not reported Not addressed ASSESSMENT Well covered Poorly addressed Not applicable 1.8 Measures will have been taken to prevent knowledge of primary exposure influencing case ascertainment Adequately addressed Not reported Not addressed Well covered Poorly addressed Not applicable 1.9 Exposure status is measured in a standard, valid and reliable way Adequately addressed Not reported Not addressed CONFOUNDING Well covered Poorly addressed Not applicable 1.10 The main potential confounders are identified and taken into account in the design and analysis Adequately addressed Not reported Not addressed STATISTICAL ANALYSIS 1.11 Confidence intervals are provided

Section 2: Overall Assessment Of The Study How well was the study done to minimise the risk of bias or confounding? 2.1 Code ++, +, or −

Section 3: Others

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 151 Consultation Document: Canadian Agency for Drugs and Technologies in Health

How was this study funded? 3.1 List all sources of funding quoted in the article, whether Government, voluntary sector, or industry.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 152 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 5: Sample Voting Form

Question :

Statement :

Evidence Summary:

There is evidence that addresses this statement YES NO

If YES, please vote below. If NO, please proceed to next statement

On the basis of cited evidence (systematic reviews or RCTs) this statement is:

Accept Accept with Neutral (cannot be Reject with Reject completely Completely reservation accepted or reservation rejected) (A) (B) (C) (D) (E)

Please include any contextual information about this statement:

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 153 Consultation Document: Canadian Agency for Drugs and Technologies in Health Appendix 6a: Quality Assessment Summary - Systematic Reviews - GERD

question studies studies and conclusions the findings (score) + provided provided reported of studies inclusion criteria) Carlsson et al. 199760 Yes No No No Yes No No No No No No Poor (2/11) Caro et al. 200196 Yes Yes No No Yes No No No Yes Yes No Poor (5/11) Chang et al. 200589 Yes Yes Yes Yes Yes Yes Yes Yes Yes No No Good (9/11) Chiba et al. 199761 Yes Yes No No No No Yes No No No No Poor (3/11) Donnellan et al. 2004*76 Yes Yes Yes No Yes Yes Yes Yes Yes No No Good (8/11) Edwards et al. 200283 Yes No No No Yes No No No No No No Poor (2/11) Edwards et al. 200191 Yes No No No Yes Yes Yes Yes Yes No No Good (6/11) Gibson et al. 200385 Yes Yes No No Yes Yes Yes No Yes No No Good (6/11) Klok et al. 200395 Yes No No No No Yes No No No No No Poor (2/11) McDonagh et al. 200590 Yes Yes Yes Yes No Yes Yes Yes Yes No No Good (8/11) McQuaid et al. 200594 Yes Yes No No Yes Yes Yes Yes Yes No No Good (7/11) Vakil et al. 200392 Yes Yes Yes No Yes No Yes Yes No No No Good (6/11) Van Pinxteren et al. 200445 Yes Yes Yes No Yes Yes Yes Yes Yes No No Good (8/11) Van Pinxteren et al. 200644 Yes Yes Yes No Yes Yes Yes Yes Yes No No Good (8/11) Wang et al. 200593 Yes Yes Yes No Yes No Yes No Yes Yes No Good (7/11) Zacny et al. 200555 Yes Yes Yes No Yes Yes No No No No NO Poor (5/11) Note: Equal weight to each domain Poor quality: overall score less than 6 out of 11 Good quality: overall score is 6 or more out of 11

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 154 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 6b: Quality Assessment Summary - Systematic Reviews - Dyspepsia

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Priori Duplicate Compre Grey list of Quality OA used Assessment Conflict design selection -hensive literature included Character of appropriately Appropriat of of provided & data literatur performe & -istics of studies in e method of publication interest Study Overall QA (research extraction e search d exclude included assessed formulating combining bias stated (score) question d studies and conclusions the findings + studies provided reported of studies inclusion provide criteria) d Moayyedi et al. 2006113 Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Good (10/11) Can’t Shiau et al. 2002114 Yes Yes Yes Yes Yes Yes Yes Yes Yes No Good (9/11) answer Note: Equal weight to each domain Poor quality: overall score less than 6 out of 11 Good quality: overall score is 6 or more out of 11

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 155 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 6c: Quality Assessment Summary - Systematic Reviews - PUD, H. pylori Infection and NSAID Associated Ulcer

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Priori Duplicate Compre- Grey list of Quality OA used Assessment Conflict design selection hensive literature included Character- of appropriately Appropriate of of provided & data literature performed & istics of studies in method of publication interest Study Overall QA (research extraction search excluded included assessed formulating combining bias stated (score) question studies studies and conclusions the findings + provided provided reported of studies inclusion criteria) Calvet et al. 2000130 Yes No No Yes Yes Yes No No Yes No No Poor (5/11) Gisbert et al. 2003121 Yes Yes Yes Yes Yes Yes Yes Yes Yes No No Good (9/11) Gisbert et al. 2005131 Yes No Yes Yes Yes Yes Yes Yes Yes No No Good (8/11) Gisbert et al. 2004119 Yes No Yes Yes No Yes Yes Yes Yes No No Good (8/11) Gisbert et al. 2004120 Yes Yes Yes Yes Yes Yes Yes Yes Yes No No Good (9/11) Hooper et al. 2004135 Yes Yes Yes Yes No No Yes No Yes Yes No Good (7/11) Klok et al. 200395 Yes No No No No Yes No No No No No Poor (2/11) McDonagh et al. 200590 Yes Yes Yes Yes No Yes Yes Yes Yes No No Good (8/11) Moayyedi and Murphy 2001123 Yes Yes No Yes No No No No Yes No No Poor (4/11) Rostom et al. 2003139 Yes Yes Yes Yes No Yes Yes Yes Yes Yes No Good (9/11) Rostom et al.2002136 Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Good (10/11) Vallve et al. 2002129 Yes Yes No Yes Yes No Yes No Yes No No Good (6/11) Vergara et al. 2003122 Yes Yes No Yes Yes Yes Yes No Yes No No Good (7/11) Note: Equal weight to each domain Poor quality: overall score less than 6 out of 11 Good quality: overall score is 6 or more out of 11

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 156 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 7a: Quality Assessment Summary – RCTs - GERD

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Appropriate Randomized Adequate Blinding of Groups are The only Standard, valid Percentage ITT analysis Comparable and clearly assignment concealment Subjects and similar at difference and reliable of Performed** results for Overall Study focused (Fundamental ** investigators baseline between groups measurement drop out multi study QA question factor)*** is treatment of outcome(s) sites under investigation** NR Armstrong et Allocation 26% vs. AA AA AA AA AA AA AA NAd good al. 200535 method not 27% reported NR NAd Armstrong et Randomization Allocation AA AA AA AA AA NR WC NAd poor al. 2004101 method not concealment reported not mentioned NR NAd Bate et al. Randomization Allocation NR AA AA AA AA 4% AA NAd poor 199373 method not concealment reported not mentioned Bardhan et al. 1% vs. 13% AA WC AA WC AA WC AA WC WC very good 199569 vs. 9% NR NAd Bate et al. Randomization Allocation AA NR AA PA WC 37% vs. 31% AA NAd poor 199065 method not concealment reported not mentioned NR NAd Bigard et al. Randomization Allocation AA PA AA AA AA 30% AA NAd poor 200543 method not concealment reported not mentioned Bytzer et al. AA PA AA AA AA AA AA 16.9% AA NAp poor 200458 NAd PA Carlsson et al. Allocation 11% vs. 4% ITT analysis detail AA NR NR PA AA AA NAd poor 199851 concealment vs. 5% not reported, not not mentioned clearly addressed PA Chen et al. Allocation is AA AA WC AA AA AA 9% AA NAd poor 200598 not clearly addressed NAd Chey et al. Allocation AA PA AA AA AA AA 3.2% AA AA poor 2003102 concealment not mentioned NR NAd Dekkers et al. Randomization Allocation AA AA AA WC AA 5% WC NAd good 199997 method not concealment reported not mentioned

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 157 Consultation Document: Canadian Agency for Drugs and Technologies in Health

reported not mentioned NR NAd Fujiwara et al. Randomization Allocation PA WC NR AA AA AA <10% NAp 200546 method not concealment Not ITT reported not mentioned NR NR Allocation Green et al. Randomization NAd AA concealment AA AA AA 15% vs. 33% AA NAd good 199567 method not Open label method is not reported reported NAd Hansen et al. Allocation AA AA NAp AA AA AA 12% WC NAd poor 200541,42 concealment not mentioned NR NAd Hetzel et al. Randomization Allocation AA AA AA AA PA NR AA NAd poor 198871 method not concealment reported not mentioned Howden et al. 17% vs. 14% AA WC AA WC AA AA AA NA NAd good 200139 vs. 11% Janssen et al. 2005105 AA AA AA PA AA AA AA 7% AA NAp good

Kaplan- Machlis et al. AA AA WC NAd AA AA AA 5.8%, 6.9% WC NAd Good 200036

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 158 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Lundell et al. 14% vs. WC AA AA AA NR AA WC WC NAp poor 199179 24%

Lundell et al. PA 2000109 AA AA AA NAd AA AA AA 10%, 16% NAd good Not clear if ITT

Lundell et al. PA 2001110 AA AA AA NAd AA AA AA 14%, 21% NAd good Not clear if ITT

NR NAd Maton et al. Randomization Allocation AA PA AA AA AA 5% v. 9% WC NAd good 199940 method not concealment reported not mentioned NAd Mahon et al. Allocation PA 2005108 AA PA PA PA AA AA NR NAd poor concealment Not clear if ITT

not mentioned NAd NR NAd Only Marks et al. Randomization Allocation physicians WC AA AA AA 14% NAd NAd poor 199482 method not concealment assessing reported not mentioned outcomes were blinded NR NAd Monnikes et Randomization Allocation AA NR AA AA AA 10% AA NAd poor al. 200599 method not concealment reported not mentioned Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 159 Consultation Document: Canadian Agency for Drugs and Technologies in Health

reported not mentioned NAd Scholten et al. Allocation 2%, 3%, AA PA AA AA AA AA AA PA poor 200557 concealment 14% not mentioned Sjostedt et al. PA 200584 AA AA Not PA AA AA AA 24% AA NAp poor mentioned NR NAd PA Smith et al. Randomization Allocation NR Most outcomes PA AA AA AA 23% vs. 34% NAd poor 199481 method not concealment reported as PP reported not mentioned analysis PA NR NR Sontag et al. randomization Allocation Blinding AA AA AA AA 5% AA NAp poor 199778 is not clearly method not method not addressed mentioned mentioned NR NR Sontag et al. Randomization Allocation 11% vs. 4% AA AA AA AA AA WC NAd good 199270 method not method not vs. 7% reported mentioned NAd Steward et al. AA AA AA AA AA AA AA 10% Excluded lost to NAp poor 200487 f/u in analysis

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 160 Consultation Document: Canadian Agency for Drugs and Technologies in Health

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Appropriate Randomized Adequate Blinding of Groups are The only Standard, valid Percentage ITT analysis Comparable and clearly assignment concealment Subjects and similar at difference and reliable of Performed** results for Overall Study focused (Fundamental ** investigators baseline between groups measurement drop out multi study QA question factor)*** is treatment of outcome(s) sites under investigation** PA Swarbrick et Allocation is 26% vs. WC AA WC AA AA AA WC NAd good al. 199680 not clearly 33.8% addressed NR Talley et al. Allocation AA WC AA WC AA AA 46% vs. 34% AA NAd good 200237 method not reported PA Tsai et al. AA NR Not clear if PA AA AA AA 13% AA PAd poor 2004107 concealed NAd Vaezi et al. Allocation AA NR AA AA AA AA NR AA NAd good 200686 concealment not mentioned van Resenburg AA WC WC WC WC AA AA 11% vs. 8% WC NAd very good et al. 199668 NAd van Zyl et al. Allocation AA AA AA WC AA AA 3% AA NAd good 200438 concealment not mentioned NR NAd Vcev et al. Randomization Allocation NAd AA AA AA AA 5% vs. 8% AA NAd poor 1999104 method not concealment Single blinded reported not mentioned NR NAd PA Venebles 1997 Randomization Allocation ITT analysis detail AA AA AA AA AA 9% vs. 16% NAp poor 54 method not concealment not reported, not reported not addressed clearly addressed NR NAd Wada et al. Randomization Allocation AA PA AA AA AA <10% AA NAp poor 200547 method not concealment reported not addressed Zeitoun et al. AA WC AA WC WC WC AA 20% vs. 24% WC NAd very good 198966

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 161 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Overall assessment Very good quality • All killer factors (3, 6 and 9) are well covered (WC) or adequately addressed (AA), the overall assessment might get very good. Poor quality • Any of the three killer factors (3, 6 or 9) is regarded as poor (NAd/NAp), the overall assessment might get poor. Good quality • If all of the three killer factors (3, 6 and 9) are well covered (WC) or adequately addressed (AA) but their related items (4, 5 or 8) were ranked NAd/NAp, the overall quality could move down to good. • If any of the three killer factors (3, 6 or 9) is regarded as poor (NAd/NAp) but the related factors (4, 5 or 8) was ranked WC/AA, plus other items were WC/AA/PA/NR, the overall quality could move up to good.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 162 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 7b: Quality Assessment Summary – RCTs - Dyspepsia

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Randomized Adequate Blinding of Groups are The only Standard, valid Percentage of ITT analysis Comparabl Appropriate assignment allocation Subjects and similar at difference and reliable drop out in Performed e results Overall Study and clearly (Fundamenta concealment** investigators baseline between groups measurement of each treatment ** for multi QA focused l factor)*** is treatment outcome(s) arm study sites question under investigation** PA Blum et al. 6.9%, 8.2%, PA NR Method not AA AA AA AA AA NAd Good 2000117 5.9%, 9.3% described NAd Laheij et al. NAd AA AA NAd AA AA AA 0%, 9.5% Not ITT NR Poor 1998112 Not mentioned analysis Manes et al. AA AA AA NAd AA AA AA 0% WC NAp Good 2003118 Talley et al. 2.4%, 2.2%, AA WC WC WC AA AA AA AA NAp Very good 1998115 3.6% van Zanten et AA NR NR AA AA AA AA 6.4%, 10.4% WC NAd Good al. 2006116 van Zanten et 6.0%, 8.6%, AA WC AA AA AA AA AA AA PAd Good al. 2005111 8.6%, 14.3% * WC: Well Covered; AA: Adequately Addressed; PA: Poorly Addressed; NR: Not Reported; NAd: Not Addressed; NAp: Not applicable ** Note: Shaded cells are killer factors. *** Fundamental factor: If randomization (2) was not addressed (NAd) or not applicable (NAp), the study was considered to be a cohort study. † Total dropouts, losses to follow-up, and subjects withdrawn; breakdown by treatment arm not reported.

Overall assessment Very good quality • All killer factors (3, 6 and 9) are well covered (WC) or adequately addressed (AA), the overall assessment might get very good. Poor quality • Any of the three killer factors (3, 6 or 9) is regarded as poor (NAd/NAp), the overall assessment might get poor. Good quality • If all of the three killer factors (3, 6 and 9) are well covered (WC) or adequately addressed (AA) but their related items (4, 5 or 8) were ranked NAd/NAp, the overall quality could move down to good. If any of the three killer factors (3, 6 or 9) is regarded as poor (NAd/NAp) but the related factors (4, 5 or 8) was ranked WC/AA, plus other items were WC/AA/PA/NR, the overall quality could move up to good.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 163 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Appendix 7c: Quality Assessment Summary – RCTs - PUD, H .pylori Infection and NSAID Associated Ulcer

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Randomized Adequate Blinding of Groups The only Standard, valid Percentage of ITT Comparabl Appropriat assignment allocation Subjects and are difference and reliable drop out in analysis e results for Overall Study e and (Fundamental concealment* investigators similar at between groups is measurement each treatment Performed multi study QA clearly factor)*** * baseline treatment under of outcome(s) arm ** sites focused investigation** question Chan et al. PA Modified AA AA AA AA PA AA AA 13.6%† NAp Good 2004140 ITT analysis NAd 9%, 11% in Chan et al. (Only NSAID group AA WC AA AA AA WC AA NAd Good 2001142 endoscopist 3%, 1% in ASA blinded) group Chan et al. AA WC AA AA AA AA AA 8.8%, 11.8% AA NAp Very good 2005146 Graham et NAd AA NR NR AA AA AA 15% AA NAd Poor al. 2002137 Not mentioned

Hawkey et NAd AA NR NR AA WC WC 0% WC NAp Good al. 2003124 Not mentioned

Hawkey et NAd AA PA NR AA WC WC <0.4%† AA NAd Good al. 1998134 Not mentioned Hsu et al. NAd AA NR NAd AA AA AA 1%, 2% AA NAp Poor 2005128 Not mentioned Lai et al. AA AA AA AA AA AA AA 2.3%, 3.6% AA NAp Very good 2006145 Lai et AA AA AA NAd AA AA AA 2.5%, 1.6% AA NAp Good al.2005141 NR – Lai et al. concealed, but WC AA AA AA AA WC 6.4%, 6.6% AA NAp Good 2002144 method not described Spinzi et al. AA AA AA PA NAd AA WC 9.4%, 16.4% WC NAp Good 1998127 Tulassay et NAd al. AA NR AA AA AA WC 0.7%† WC NAp Good Not mentioned 2001126 Wong et al. WC AA AA PA AA AA WC 2.9%† WC NAp Good 2001125 Yeomans et NAd AA NR NR AA AA WC 0% AA NAd Good al. 1998133 Not mentioned * WC: Well Covered; AA: Adequately Addressed; PA: Poorly Addressed; NR: Not Reported; NAd: Not Addressed; NAp: Not applicable ** Note: Shaded cells are killer factors. *** Fundamental factor: If randomization (2) was not addressed (NAd) or not applicable (NAp), the study was considered to be a cohort study. † Total dropouts, losses to follow-up, and subjects withdrawn; breakdown by treatment arm not reported.

Overall assessment Very good quality Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 164 Consultation Document: Canadian Agency for Drugs and Technologies in Health

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Randomized Adequate Blinding of Groups The only Standard, valid Percentage of ITT Comparabl Appropriat assignment allocation Subjects and are difference and reliable drop out in analysis e results for Overall Study e and (Fundamental concealment* investigators similar at between groups is measurement each treatment Performed multi study QA clearly factor)*** * baseline treatment under of outcome(s) arm ** sites focused investigation** question • All killer factors (3, 6 and 9) are well covered (WC) or adequately addressed (AA), the overall assessment might get very good. Poor quality • Any of the three killer factors (3, 6 or 9) is regarded as poor (NAd/NAp), the overall assessment might get poor. Good quality • If all of the three killer factors (3, 6 and 9) are well covered (WC) or adequately addressed (AA) but their related items (4, 5 or 8) were ranked NAd/NAp, the overall quality could move down to good. • If any of the three killer factors (3, 6 or 9) is regarded as poor (NAd/NAp) but the related factors (4, 5 or 8) was ranked WC/AA, plus other items were WC/AA/PA/NR, the overall quality could move up to good.

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 165 Consultation Document: Canadian Agency for Drugs and Technologies in Health

Draft Scientific Report: Evidence for PPI Use in GERD, Dyspepsia and PUD 166