DATE: 07 December 2011 CONTEXT AND POLICY ISSUES

Transcription

1 TITLE: versus Non-selective Non-steroidal Anti-Inflammatory Drugs and Proton Pump Inhibitors: Clinical Effectiveness, Safety, and Cost-Effectiveness DATE: 07 December 2011 CONTEXT AND POLICY ISSUES Non-steroidal anti-inflammatory drugs (NSAIDs) are widely used in controlling the symptoms of arthritis, including osteoarthritis (OA) and rheumatoid arthritis (RA), low back pain and soft tissue pain, by blocking cyclooxygenase (COX) enzymes. 1-3 OA is a chronic condition that involves the degeneration of cartilage between the joints. It is the most common form of arthritis and is associated with pain, substantial disability and a poor quality of life. 3 RA is a systemic autoimmune disorder that involves persistent joint inflammation. 4 Two COX isoenzymes (COX-1 and COX-2) play important roles in the effects of NSAIDs. COX-1 mediates the mucosal protection of the gastrointestinal (GI) mucosa, while COX-2 is found throughout the body, including joint and muscle, and mediates effects on pain and inflammation. Generally, NSAIDs are classified into two categories. One is non-cox-2 selective NSAIDs (nsnsaids, or conventional NSAIDs, such as ibuprofen, naproxen and dichofenac), which blocks both COX-1 and COX-2 enzymes and can cause GI bleeding. The other is COX-2 selective NSAIDs (mainly celecoxib) that targets only the COX-2 enzyme, which are deemed safer with regard to GI bleeding than nsnsaids. 3,5 Studies, however, found that they increase the risk of cardiovascular (CV) adverse events. 6-8 has been approved in Canada since and is the main COX-2 inhibitor currently available in Canada. nsnsaids, such as naproxen, ibuprofen and diclofenac have been available in Canada since Proton pump inhibitors (PPIs, for example, omeprazole, esomeprazole and lansoprazole) have been available in Canada since In recent years, more patients with OA or RA who require anti-inflammatory treatment were taking celecoxib than nsnsaids. 11 nsnsaids may cause serious gastric adverse events with continuous usage. PPIs are reported to reduce the risk of NSAID-induced gastric and duodenal ulcers. 12 is associated with a lower GI side effect profile compared with nsnsaid, but it may be more costly. 13 Currently, both celecoxib and co-therapy of nsnsaids and PPIs, such as omeprazole, esomeprazole and lansoprazole, are recommended for NSAIDs users with upper gastrointestinal (UGI) and CV risk factors. 1,2,14,15 Disclaimer: The Rapid Response Service is an information service for those involved in planning and providing health care in Canada. Rapid responses are based on a limited literature search and are not comprehensive, systematic reviews. The intent is to provide a list of sources and a summary of the best evidence on the topic that CADTH could identify using all reasonable efforts within the time allowed. Rapid responses should be considered along with other types of information and health care considerations. The information included in this response is not intended to replace professional medical advice, nor should it be construed as a recommendation for or against the use of a particular health technology. Readers are also cautioned that a lack of good quality evidence does not necessarily mean a lack of effectiveness particularly in the case of new and emerging health technologies, for which little information can be found, but which may in future prove to be effective. While CADTH has taken care in the preparation of the report to ensure that its contents are accurate, complete and up to date, CADTH does not make any guarantee to that effect. CADTH is not liable for any loss or damages resulting from use of the information in the report. Copyright: This report contains CADTH copyright material. It may be copied and used for non-commercial purposes, provided that attribution is given to CADTH. Links: This report may contain links to other information available on the websites of third parties on the Internet. CADTH does not have control over the content of such sites. Use of third party sites is governed by the owners own terms and conditions.

2 In the United States, there are an estimated 16,500 annual deaths due to NSAID-induced GI complications. 8 Considering the higher cost for celecoxib versus the lower cost of PPIs to treat GI complications, there is an increasing need to understand the cost-effectiveness of celecoxib comparing with nsnsaids plus PPIs in the management of RA and OA. The purpose of this review is to evaluate the clinical effectiveness, safety and cost-effectiveness of celecoxib relative to combination therapy of nsnsaids and PPIs in adult patients with OA or RA, or other diseases that require pain management. This report is an upgrade of a previous summary of abstracts, versus Non-selective Non-Steroidal Anti-Inflammatory Drugs and Proton Pump Inhibitors: Clinical Effectiveness, Safety, and Cost Effectiveness. RESEARCH QUESTIONS 1. What is the comparative clinical effectiveness of celecoxib versus combination non-cox-2 selective non-steroidal anti-inflammatory drugs and proton pumps inhibitors for pain management in adult patients? 2. What is the evidence for the safety of celecoxib compared with combination non-cox-2 selective non-steroidal anti-inflammatory drugs and proton pumps inhibitors for pain management in adult patients? 3. What is the cost-effectiveness of celecoxib compared with combination non-cox-2 selective non-steroidal anti-inflammatory drugs and proton pumps inhibitors for pain management in adult patients? KEY MESSAGE Limited evidence showed that celecoxib versus non-cox-2 selective non-steroidal antiinflammatory drugs (nsnsaids) and proton pumps inhibitors (PPIs) had a similar clinical efficacy and safety profile. nsnsaids plus PPIs may be marginally more cost-effective versus celecoxib in the management of adult patients with rheumatoid arthritis or osteoarthritis. METHODS Literature Search Strategy This report makes use of a literature search conducted for a previous CADTH report. The original literature search was conducted in September 2011on key resources including PubMed, The Cochrane Library (2011, Issue 9), University of York Centre for Reviews and Dissemination (CRD) databases, Canadian and major international health technology agencies, as well as a focused Internet search. No filters were applied to limit retrieval by study type. Where possible, retrieval was limited to the human population. The initial search was also limited to English language documents published between January 1, 2001 and September 6, For the current report, the PubMed database search was rerun on November 9 to capture any articles published since the initial search date. The search of major health technology agencies was also updated to include documents published since September Rapid Response reports are organized so that the evidence for each research question is presented separately. versus NSAIDs with PPIs 2

3 Selection Criteria and Methods One reviewer screened the titles and abstracts of the retrieved publications, and evaluated the full-text publications for the final article selection, according to the selection criteria present in Table 1. Table 1: Selection Criteria Population Intervention Comparator Outcomes Study Designs Adult patients with osteoarthritis, rheumatoid arthritis or other conditions requiring pain management nsnsaids in combination with PPIs Clinical benefit (for example, pain reduction) Adverse events (mainly interested in gastrointestinal events, but would like to include any adverse events) Economic evaluation HTA/SR/MA, RCT for clinical efficacy, safety Non-RCT for safety only Economic evaluations HTA=health technology assessment, SR=systematic review; RCT= randomized controlled trial; nsnsaids= non selective nonsteroidal anti-inflammatory drugs; PPIs= proton pump inhibitors. Exclusion Criteria Studies were excluded if they did not meet the selection criteria, were duplicate publications, were abstracts or conference proceedings, or were published prior to Critical Appraisal of Individual Studies SIGN 50 check lists 16,17 and Drummond s checklist 18,19 were used to assess the methodological quality for RCTs, non-randomized studies and economic evaluation reports, respectively. SUMMARY OF EVIDENCE: Quantity of Research Available The literature search from November 9, 2011, yielded 124 citations. Upon screening titles and abstracts, 103 citations were excluded, and 21 potentially relevant articles were retrieved for fulltext review. Of the 21 potentially relevant reports, 10 did not meet the inclusion criteria. Six articles representing five RCTs, one non-rct and four economic evaluations were included in this review The study selection process is outlined in a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart (Appendix 1). versus NSAIDs with PPIs 3

4 Summary of Study Characteristics Comparative Clinical Effectiveness and Safety of versus Combination nsnsaids and PPIs The main characteristics of the included studies for clinical effectiveness and safety are summarized in Appendix 2. Five RCTs and one non-rct were identified that reported clinical effectiveness and safety. The sample sizes of RCTs ranged from 242 to 4,484 patients, and the trial duration ranged from 12 weeks to six months. All RCTs compared celecoxib with nsnsaids plus PPIs in adult patients with OA or RA. The main outcomes of interest were pain reduction, recurrent GI events risk reduction and adverse events, such as dyspepsia. Hochberg et al. 21 reported the clinical efficacy of celecoxib and fixed dose combination of entericcoated naproxen and esomeprazole for knee osteoarthritis from two 12-week RCTs. The two double-blind multi-centre phase III RCTs included patients with symptomatic knee osteoarthritis (aged 50 years or older). Following an OA flare up, patients received naproxen/esomeprazole or celecoxib for 12 weeks. In one trial (Study 307), 619 patients were randomized in either treatment arm and 614 were treated. In the other trial (Study 309), 615 patients were randomized and 610 were treated. In 2004, Lai et al. 23 compared the efficacy of celecoxib with that of naproxen and lansoprazole co-therapy in the reduction of ulcer relapse in patients with a history of NSAID-related peptic ulcers. After healing of ulcers and eradication of Helicobacter pylori, patients were randomly assigned to celecoxib 200 mg daily (n = 120) or naproxen 750 mg daily and lansoprazole 30 mg daily (n = 122) for 24 weeks. Two publications by Chan et al. 20,24 reported the findings from one RCT, which evaluated the ulcer incidence in patients on celecoxib or combination therapy of diclofenac plus omeprazole. Two-hundred and eighty-seven patients with NSAIDs-related ulcer bleeding were included. After ulcer healing, Helicobacter pylori negative patients were randomized to celecoxib 200 mg bid or combination therapy of diclofenac 75 mg bid plus omeprazole 20 mg once daily for 6 months. Endoscopy was performed if they developed recurrent bleeding or at the last follow-up visit for patients without recurrent events. In a 6-month, multi-nation, multi-centre and double-blinded RCT, Chan et al. 22 compared the risk of GI events associated with celecoxib versus diclofenac plus omeprazole. The study included 4,484 patients with OA or RA with an increased GI risk, Helicobacter pylori negative, aged 60 years and older or 18 years or older with previous gastroduodenal ulceration. They were randomized to receive celecoxib 200 mg twice a day (n=2,238) or diclofenac 75 mg bid, plus omeprazole 20 mg qd (n= 2,246). The primary endpoint was clinically significant upper or lower GI events defined as GI bleeding, obstruction and perforation. Analysis was done by using intention to treat population. Targownik et al. 26 performed a population-based matched case-control analysis based on Manitoba Population Health Research Data Repository. There were 1,382 nsnsaids or COX-2 inhibitor users with upper GI complications admitted to the hospital with a primary diagnosis for an upper GI complication. These patients were matched with 33,957 age- and sex-matched controls in the community. Conditional logistic regression analysis was used to determine the versus NSAIDs with PPIs 4

5 relative efficacy of different gastroprotective strategies, such as PPIs, celecoxib, either alone or in combination and to adjust for multiple pertinent covariates. Cost-Effectiveness of versus Combination nsnsaids and PPIs The detailed characteristics of the included economic evaluations are presented in Appendix 3. Four articles 27-29,31 that reported seven economic evaluations 27-29,32-35 were identified. One study was conducted in Hungary in 2009, 27 one in The Netherlands in 2008, 29 and one in the UK in One HTA report 31 presented four individual economic evaluations studies. Among these studies, one was published in Canada (2001), 33 one in USA(2002), 35 one in Switzerland (2001), 32 and one in Hong Kong, China (2002). 34 Time horizon ranged from six months to five years. 28 These reports compared the cost effectiveness of celecoxib with nsnsaids plus PPIs, directly or indirectly. Inotai et al. 27 evaluated the cost-effectiveness of celecoxib and nsnsaids plus PPIs. A decision tree model was developed, with a one year time horizon, to simulate cohorts within celecoxib, nsnsaids and nsnsaids plus PPIs. The effectiveness of the different active agents of NSAIDs in therapeutically relevant doses was assumed to be the same. On the other hand, differences were observed in the side effect profile of the drugs, such as GI ulcers and bleeding. The direct medical costs and costs related to adverse events were calculated to measure the cost of an increase in health gain or quality-adjusted life-year (QALY), and the incremental costeffectiveness ratio (ICER) was reported. In the 2008 HTA, Chen et al. 28 reviewed the cost-effectiveness of celecoxib and nsnsaids (ibuprofen or diclofenac) plus PPI for OA and RA. Systematic reviews of RCTs and a modelbased economic evaluation were undertaken. The model was designed to run in two forms: the full Assessment Group Model (AGM), which included an initial drug switching cycle, and the simpler AGM, where an initial cycle and opportunities for the patient to switch NSAIDs were not included. In 2008, Al et al. 29 assessed the costs and upper GI side effects of celecoxib and nsnsaids plus PPIs for patients with OR in The Netherlands. A model was used to convene data from various sources on the probability of GI side effects and resource use. The probabilities of GI side effects for celecoxib and NSAIDs alone were derived from trial data. The calculations were based on six months of treatment and were from a societal perspective. An extensive probabilistic sensitivity analysis was performed to address uncertainty. In a 2006 HTA, Brown et al. 30 assessed the costs of celecoxib and nsnsaids plus PPIs for the prevention of NSAIDs-induced GI events. Four economic evaluations that examined the cost-effectiveness of celecoxib and nsnsaids plus PPIs were included. The data were reported individually. A probabilistic decision-analytic model was designed and populated with data to carry out incremental economic analysis. No ICER was reported. Summary of Critical Appraisal The strengths and limitations of all included studies in the report are summarized in Appendix 4. versus NSAIDs with PPIs 5

6 Overall, the research questions were adequately addressed in all RCTs. Randomization and double blind process, dropout and Intention to treat (ITT) analysis were well reported except one, 23 in which the double blind was not reported. The non-rct 26 adjusted the results for potential confounding factors. The economic evaluations were also well conducted in terms of study design, data collection, data analysis and interpretation. The ICERs of nsnsaids plus PPIs compared with celecoxib, however, was not reported in the majority of included studies. 29,32-35 Summary of Findings Five RCTs published in six articles and one non-rct 26 study was identified evaluating the comparative clinical effectiveness and safety compared celecoxib with nsnsaids plus PPIs. The main findings were summarized below and in Appendix 5. Comparative Clinical Effectiveness of versus nsnsaids plus PPIs Pain Reduction In 2011, Hochberg et al. 21 evaluated the efficacy of celecoxib in pain reduction compared with fixed dose combination of naproxen plus omeprazole. The primary outcomes measured were mean change from baseline to week 12 using Western Ontario and McMaster Osteoarthritis Index (WOMAC) pain and function subscales, and Patient Global Assessment of osteoarthritis using a visual analog scale (PGA-VAS). It was reported that, after 12 weeks, both naproxen/esomeprazole and celecoxib were associated with improvements (least squares mean change from baseline to week 12) in WOMAC pain (Study 307: and -41.8, respectively; Study 309: and -42.9, respectively), WOMAC function (Study 307: and -36.3, respectively; Study 309: and -36.8, respectively), and PGA-VAS (Study 307: 21.2 and 21.6, respectively; Study 309: 29.0 and 25.6, respectively). A pre-specified non-inferiority margin of 10 mm between naproxen/esomeprazole and celecoxib was satisfied for each co-primary endpoint at week 12 in both studies. The authors concluded that naproxen/esomeprazole showed comparable efficacy to celecoxib for the management of pain associated with osteoarthritis of the knee over 12 weeks. Prevention of GI Ulcer or Bleeding Relapse In 2004, Lai et al. 23 compared the efficacy of celecoxib with that of naproxen and lansoprazole combination therapy in the reduction of ulcer relapse in patients with a history of NSAID-related peptic ulcers. The primary endpoint was recurrent ulcer complications. During a median follow-up of 24 weeks, it was found that four patients [3.7%, (95% CI: 0.0% to 7.3%)] in the celecoxib group, compared with seven patients [6.3%, (95% CI 1.6% to 11.1%)] in the naproxen plus lansoprazole group, developed recurrent ulcer complications [absolute difference -2.6%; (95% CI: -9.1% to 3.7%)]. The margin for the non-inferiority was defined as less than 7% for the upper limit of the 95%CI for the difference between the two treatment groups. The results suggested that celecoxib was statistically non-inferior to naproxen-lansoprazole co-therapy in the prevention of recurrent ulcer complications. The author concluded that celecoxib was as effective as the naproxen-lansoprazole co-therapy in the prevention of recurrences of ulcer complications in patients with a history of NSAID-related complicated peptic ulcers. Chan et al. 20,24 reported the findings from one RCT, which evaluated the ulcer incidence in patients on celecoxib or combination therapy of diclofenac plus omeprazole. No statistically versus NSAIDs with PPIs 6

7 significant difference was found in the rate of recurrent ulcers in six months [celecoxib: 18.7% and diclofenac plus omeprazole: 25.6%; difference between groups: -6.7%; (95% CI: -17.8% to 3.9%; P = 0.21)]. Recurrent ulcer bleeding occurred in seven patients [4.9 % (95% CI: 3.1 to 6.7) receiving celecoxib and in nine patients (6.4 % (95% CI, 4.3 to 8.4)) receiving diclofenac plus omeprazole [difference, 1.5 % (95 %CI: 6.8 to 3.8)]. The authors concluded that celecoxib was as effective as diclofenac plus omeprazole in terms of prevention of recurrent bleeding in patients with a recent history of ulcer bleeding. In a 6-month, multi-nation, multi-centre and double-blind RCT, Chan et al. 22 compared the risk of GI events associated with celecoxib versus diclofenac plus omeprazole. It was found that 20 patients (0.9%) receiving celecoxib and 81 patients (3.8%) receiving diclofenac plus omeprazole reported upper or lower GI events [hazard ratio 4.3, (95% CI: 2.6 to 7.0); p<0.0001)]. Onehundred and fourteen patients (6%) treated with celecoxib versus 167 patients (8%) receiving diclofenac plus omeprazole withdrew before the trial ended because of GI adverse events (p=0.0006). It was indicated that the risk of GI events was statistically significant lower in patients treated with celecoxib than in those receiving a diclofenac plus a PPI. In the case-control study, Targownik et al. 26 found that, compared with nsnsaids alone, the adjusted odds ratio for overall upper GI complications was 0.67 (95% CI 0.48 to 0.95) (P<0.01) in nsnsaids plus PPIs group and 0.39 (0.32 to 0.48) P<0.01) in celecoxib group respectively. The findings suggested that celecoxib may be superior to the combination of nsnsaids with a PPI in the prevention of overall GI complications in patients with a history of GI bleeding, perforation or systematic ulcers. Comparative Safety of versus nsnsaids plus PPIs Dyspepsia and Other Adverse Events Cryer et al. 25 compared celecoxib with fixed dose combination of naproxen plus omeprazole in the treatment of patients with OR. It was found that, at the end of 12-week treatment, Modified Severity of Dyspepsia Assessment (msoda) scores improved in each group, with no significant treatment differences between celecoxib and naproxen/esomeprazole. UGI AE incidence (Study 307: 17.3%; Study 309: 20.3%) was also similar between treatment groups. However, significantly more heartburn-free days was reported in naproxen/esomeprazole treated patients compared with celecoxib. In 2004, Lai et al. 23 reported that, during a median follow-up of 24 weeks, more patients in celecoxib group [15.0%, (95% CI 9.7 to 22.5)] experienced dyspepsia than those receiving naproxen- lansoprazole (5.7%, 95% CI 2.8 to11.4, P =.02). Chan et al. 20,24 also reported that more patients experienced dyspepsia in patients on celecoxib compared with diclofenac plus omeprazole [hazard ratio 5.3%, (95% CI: 2.6 to 10.8) ]. Cost-Effectiveness of versus nsnsaids plus PPIs Seven studies 27-29,32-35 presented in four publications were found. The findings were summarized below and Appendix 6. versus NSAIDs with PPIs 7

8 Inotai et al. 27 evaluated the cost-effectiveness of and nsnsaids plus PPIs. It was reported that nsnsaids plus PPIs dominated celecoxib since celecoxib had higher costs and was less effective. In one 2008 HTA, Chen et al. 28 reviewed the cost-effectiveness of celecoxib and nsnsaids (ibuprofen or diclofenac) plus PPIs for OA and RA. It was found that celecoxib was dominated by ibuprofen or diclofenac combined with PPIs in most cases. In other words, nsnsaids plus PPIs is more cost effective compared with celecoxib. This result applies both to standard (defined as patients without previous history of GI events) and high-risk arthritis patients (defined as patients with previous GI ulcers). The author concluded that the economic models showed a wide range of possible costs per QALY gained in patients with OA and RA. In 2008, Al et al. 29 assessed the costs and upper GI side effects of celecoxib and nsnsaids plus PPIs for patients with OR in The Netherlands. It was found that the total cost per six months of therapy was: 255 Euros per patient for celecoxib and 243 Euros per patient for nsnsaids plus PPIs. Treatment with celecoxib was associated with the lowest number of GI side effects and related deaths. ICERs for celecoxib versus nsnsaids plus PPIs were not reported although the author reported that celecoxib is preferred treatment strategy in patients with more than 1.5% probability to have at least one serious GI event in one year. In a 2006 HTA, Brown et al. 30 presented the data from four economic evaluations ICERs were not presented in the HTA for any of the four economic evaluations. The authors of the HTA concluded that the economic models suggested that nsnsaids plus PPIs was more costeffective compared with other treatment strategies, including celecoxib, to avoid endoscopic ulcers in patients, who required long-term NSAIDs therapy. Limitations Firstly, the clinical studies were conducted in different countries between 2001 and Over time, the clinic management of patients with RA and OA has changed. 36,37 Secondly, the comparator (nsnsaids and PPIs) agents, dosage, and reporting outcomes varied from study to study, so it is difficult to compare their findings. Thirdly, the study durations in the primary clinical trials were less than or equal to six months. As a result, the lack of long-term clinical outcomes, such as mortality and serious GI and CV adverse events, may lead to an inaccurate estimation of the clinical benefits and harms. Finally, several economic evaluations included in the HTA by Brown 30,39 were published in 2001 or 2002, Therefore, their study findings must be interpreted with caution. Moreover, the analyses were conducted in different countries using various economic models, so their findings may not be transferable to the Canadian setting. CONCLUSIONS AND IMPLICATIONS FOR DECISION OR POLICY MAKING: Limited evidence showed that celecoxib and nsnsaids plus PPIs provided a similar clinical effectiveness and safety profile. It also indicated that nsnsaids plus PPIs may be marginally more cost-effective compared with celecoxib in the management of adult patients with RA or OA. The findings must be interpreted with caution. An economic evaluation that compares the costeffectiveness of celecoxib with nsnsaids plus PPIs in the management of OA or RA in the Canadian context is required to determine the cost-effectiveness in Canada. versus NSAIDs with PPIs 8

9 PREPARED BY: Canadian Agency for Drugs and Technologies in Health Tel: REFERENCES 1. American College of Rheumatology Ad Hoc Group on Use of Selective and Nonselective Nonsteroidal Antiinflammatory Drugs. Recommendations for use of selective and nonselective nonsteroidal antiinflammatory drugs: an American College of Rheumatology white paper. Arthritis Rheum Aug 15;59(8): National Collaborating Centre for Chronic Conditions. Osteoarthritis: national clinical guideline for care and management in adults [Internet]. In: London: Royal College of Physicians; 2008 [cited 2011 Nov 29]. (NICE clinical guideline 59). Available from: 3. Oregon Evidence-based Practice Center. Analgesics for osteoarthritis: an update of the 2006 Comparative Effectiveness Review [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (AHRQ); 2011 Oct. [cited 2011 Dec 5]. (Comparative Effectiveness Review no. 38). Available from: 38_ pdf 4. Garner SE, Fidan D, Frankish RR, Judd M, Shea B, Towheed T, et al. for rheumatoid arthritis. Cochrane Database Sys Rev [cited 2011 Dec 6];(4):CD Available from: Republished in Issue 1,2009; subscription required. 5. Gøtzsche PC. NSAIDs. Clin Evid (Online) [Internet] Jun 28 [cited 2011 Dec 5];1108. Available from: Subscription required. 6 Cardiovascular safety of Cox-2 inhibitors and non-selective NSAIDs [Internet]. London: Medicines and Healthcare products Regulatory Agency (MHRA); 2010 Jul 1. [cited 2011 Dec 6]. Available from: amiclistquery=&dynamiclistsortby=xcreationdate&dynamiclistsortorder=desc&dynam iclisttitle=&pagenumber=1&title=cardiovascular%20safety%20of%20cox- 2%20inhibitors%20and%20non-selective%20NSAIDs 7. Chen LC, Ashcroft DM. Risk of myocardial infarction associated with selective COX-2 inhibitors: meta-analysis of randomised controlled trials. Pharmacoepidemiol Drug Saf Jul;16(7): Chou R, Helfand M, Peterson K, Dana T, Roberts C. Comparative effectiveness and safety of analgesics for osteoarthritis [executive summary] [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (AHRQ); 2006 Sep. [cited 2011 Nov 30]. (Comparative versus NSAIDs with PPIs 9

10 Effectiveness Review No. 4). Available from: 9. Health Canada. Notice of Compliance (NOC) [Internet]. Ottawa: Health Canada; ; 2011 [cited 2011 Dec 1]. Available from: Evidence for PPI use in gastroesophageal reflux disease, dyspepsia and peptic ulcer disease: scientific report [Internet]. Ottawa: Canadian Agency for Drugs and Technologies in Health (CADTH); 2007 Mar. [cited 2011 Dec 5]. (Optimal therapy report; vol. 1 no. 2). Available from: Cryer B, Luo X, Assaf AR, Sands G, Mardekian J. Persistence with non-selective NSAIDs and celecoxib among patients with gastroesophageal reflux disease and osteoarthritis or rheumatoid arthritis. Curr Med Res Opin Feb;27(2): Rostom A, Dubé C, Jolicoeur E, Boucher M, Joyce J. Gastroduodenal ulcers associated with the use of non-steroidal anti-inflammatory drugs: a systematic review of preventive pharmacological interventions [Internet]. Ottawa: Canadian Coordinating Office for Health Technology Assessment (CCOHTA); [cited 2011 Nov 29]. (Technology Overview No. 12). Available from: Drug Product Database [Internet]. Ottawa: Health Canada. Drug Product Database online query; 2010 [cited 2010 May 3]. Available from: Bhatt DL, Scheiman J, Abraham NS, Antman EM, Chan FK, Furberg CD, et al. ACCF/ACG/AHA 2008 expert consensus document on reducing the gastrointestinal risks of antiplatelet therapy and NSAID use: a report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents. Circulation Oct 28;118(18): Lanza FL, Chan FK, Quigley EM, Practice Parameters Committee of the American College of Gastroenterology. Guidelines for prevention of NSAID-related ulcer complications. Am J Gastroenterol Mar;104(3): Methodology checklist 2: randomized controlled trials. In: SIGN 50: a guideline developer's handbook. Edinburgh: Scottish Intercollegiate Guidelines Network; SIGN 50: a guideline developer's handbook. Edinburgh: Scottish Intercollegiate Guidelines Network; Drummond MF, Jefferson TO. Guidelines for authors and peer reviewers of economic submissions to the BMJ. The BMJ Economic Evaluation Working Party. BMJ [Internet] Aug 3 [cited 2011 Nov 28];313(7052): Available from: Drummond MF, Richardson WS, O'Brien BJ, Levine M, Heyland D. Users' guides to the medical literature. XIII. How to use an article on economic analysis of clinical practice. A. versus NSAIDs with PPIs 10

11 Are the results of the study valid? Evidence-Based Medicine Working Group. JAMA May 21;277(19): Chan FK, Hung LC, Suen BY, Wu JC, Lee KC, Leung VK, et al. versus diclofenac and omeprazole in reducing the risk of recurrent ulcer bleeding in patients with arthritis. N Engl J Med Dec 26;347(26): Hochberg MC, Fort JG, Svensson O, Hwang C, Sostek M. Fixed-dose combination of enteric-coated naproxen and immediate-release esomeprazole has comparable efficacy to celecoxib for knee osteoarthritis: two randomized trials. Curr Med Res Opin Jun;27(6): Chan FK, Lanas A, Scheiman J, Berger MF, Nguyen H, Goldstein JL. versus omeprazole and diclofenac in patients with osteoarthritis and rheumatoid arthritis (CONDOR): a randomised trial. Lancet Jul 17;376(9736): Lai KC, Chu KM, Hui WM, Wong BC, Hu WH, Wong WM, et al. compared with lansoprazole and naproxen to prevent gastrointestinal ulcer complications. Am J Med Nov;118(11): Chan FK, Hung LC, Suen BY, Wong VW, Hui AJ, Wu JC, et al. versus diclofenac plus omeprazole in high-risk arthritis patients: results of a randomized doubleblind trial. Gastroenterology Oct;127(4): Cryer BL, Sostek MB, Fort JG, Svensson O, Hwang C, Hochberg MC. A fixed-dose combination of naproxen and esomeprazole magnesium has comparable upper gastrointestinal tolerability to celecoxib in patients with osteoarthritis of the knee: Results from two randomized, parallel-group, placebo-controlled trials. Ann Med Dec;43(8): Targownik LE, Metge CJ, Leung S, Chateau DG. The relative efficacies of gastroprotective strategies in chronic users of nonsteroidal anti-inflammatory drugs. Gastroenterology Apr;134(4): Inotai A, Mészáros A. Economic evaluation of nonsteroidal anti-inflammatory drug strategies in rheumatoid arthritis. Int J Technol Assess Health Care Apr;25(2): Chen YF, Jobanputra P, Barton P, Bryan S, Fry-Smith A, Harris G, et al. Cyclooxygenase-2 selective non-steroidal anti-inflammatory drugs (etodolac, meloxicam, celecoxib, rofecoxib, etoricoxib, valdecoxib and lumiracoxib) for osteoarthritis and rheumatoid arthritis: a systematic review and economic evaluation [Internet]. Health Technol Assess Apr [cited 2011 Nov 10];12(11):1-278, iii. Available from: Al MJ, Maniadakis N, Grijseels EW, Janssen M. Costs and effects of various analgesic treatments for patients with rheumatoid arthritis and osteoarthritis in the Netherlands. Value Health Jul;11(4): Brown TJ, Hooper L, Elliott RA, Payne K, Webb R, Roberts C, et al. A comparison of the cost-effectiveness of five strategies for the prevention of non-steroidal anti-inflammatory versus NSAIDs with PPIs 11

12 drug-induced gastrointestinal toxicity: a systematic review with economic modelling. Health Technol Assess Oct;10(38):iii-iv, xi-xiii, Engström A, Jacob J, Lundin D. Sharp drop in prices after the introduction of generic substitution [Internet]. Solna (SE): LFN: Pharmaceutical Benefits Board; 2006 Jun. [cited 2011 Dec 2]. Available from: Chancellor JV, Hunsche E, de Cruz E, Sarasin FP. Economic evaluation of celecoxib, a new cyclo-oxygenase 2 specific inhibitor, in Switzerland. Pharmacoeconomics. 2001;19(Suppl 1): Zabinski RA, Burke TA, Johnson J, Lavoie F, Fitzsimon C, Tretiak R, et al. An economic model for determining the costs and consequences of using various treatment alternatives for the management of arthritis in Canada. Pharmacoeconomics. 2001;19(Suppl 1): You JH, Lee KK, Chan TY, Lau WH, Chan FK. Arthritis treatment in Hong Kong--cost analysis of celecoxib versus conventional NSAIDS, with or without gastroprotective agents. Aliment Pharmacol Ther Dec;16(12): El-Serag HB, Graham DY, Richardson P, Inadomi JM. Prevention of complicated ulcer disease among chronic users of nonsteroidal anti-inflammatory drugs: the use of a nomogram in cost-effectiveness analysis. Arch Intern Med Oct 14;162(18): Kanis JA, Burlet N, Cooper C, Delmas PD, Reginster JY, Borgstrom F, et al. European guidance for the diagnosis and management of osteoporosis in postmenopausal women. Osteoporos Int [Internet] Apr [cited 2011 Dec 2];19(4): Available from: Doan QV, Chiou CF, Dubois RW. Review of eight pharmacoeconomic studies of the value of biologic DMARDs (adalimumab, etanercept, and infliximab) in the management of rheumatoid arthritis. J Manag Care Pharm Sep;12(7): Maetzel A, Krahn M, Naglie G. The cost effectiveness of rofecoxib and celecoxib in patients with osteoarthritis or rheumatoid arthritis. Arthritis Rheum. 2003;49(3): Brown TJ, Hooper L, Elliott RA, Payne K, Webb R, Roberts C, et al. A comparison of the cost-effectiveness of five strategies for the prevention of non-steroidal anti-inflammatory drug-induced gastrointestinal toxicity: a systematic review with economic modelling [appendices]. Health Technol Assess [Internet] [cited 2011 Nov 28];10(38): Available from: versus NSAIDs with PPIs 12

13 APPENDIX 1: Selection of Included Studies 124 citations identified from electronic literature search and screened 103 citations excluded 21 potentially relevant articles retrieved for scrutiny (full text, if available) 0 potentially relevant reports retrieved from other sources (grey literature, hand search) 21 potentially relevant reports 10 reports excluded: -irrelevant population (1) -irrelevant intervention (1) -irrelevant comparator (1) -duplicate publications (3) -other (e.g., review articles, editorials, etc.)(4) 11 reports included in review versus NSAIDs with PPIs 13

14 Appendix 2: Characteristics of Included Studies on Clinical Effectiveness and Safety First Author, Publication Year, Country Hochberg, USA Study Design, Length of Follow-up Two RCTs (Study 307 and Study 309) reported in this article 12 weeks Patient Characteristics, Sample Size Patients with knee osteoarthritis (Study 307: n=619 Study 309: n=615) Intervention Comparators Main Outcomes (200mg qd) Naproxen plus esomeprazole (fixed dose combination 500m/20mg, bid Primary efficacy endpoints (mean change from baseline at end ) WOMAC pain and function subscales; PGA-VAS). Cryer, USA Chan, Hong Kong, China From the same RCT as Hochberg, RCT 6 months Patients with knee osteoarthritis (Study 307: n=619 Study 309: n=615) Patients with OA or RA at increased GI event risk (n=4484) (200mg bid) Naproxen plus esomeprazole (fixed dose combination) Diclofenac plus omeprazole (75 mg bid plus omeprazole 20 mg qd) Tolerability end-points msoda); heartburn severity; UGI adverse events (AEs). Risk of GI events Lai, RCT weeks Chan, Hong Kong, China Chan, Hong Kong, China RCT 6 months From the same RCT as Chan, Patients with OA or RA and a history of NSAID-related peptic ulcers (n=242) Patients with arthritis and NSAIDsassociated ulcer bleeding. Randomized after healing of ulcers and eradication of Helicobacter pylori (N=287) Patients with arthritis and NSAIDsassociated ulcer bleeding; Randomized after healing of ulcers and eradication of Helicobacter pylori (n=290) 200mg qd (200mg bid) Celecox 200mg bid Naproxen plus lansoprazole (750mg/30mg qd) Diclofenac plus omeprazole (75 mg bid / 20 mg qd) Diclofenac plus omeprazole (75mg bid/20mg qd) Ulcer relapse Ulcer relapse Recurrent ulcer bleeding Targownik, Canada Populationbased matched case-control analysis NSAID users admitted to the hospital for an upper GI complication (N=1382) matched to NSAID-using controls in the community( n=33,957) nsnsaids plus PPIs Upper GI complications AE=adverse events; bid=twice a day; qd.=once daily; GI=gastrointestinal; NSAID=non-steroidal anti-inflammatory drug; nsnsaids= non-cox-2-selective non-steroidal anti-inflammatory drugs; OA=osteoarthritis; PGA-VAS=patient global assessment-visual analogue scale; PPIs=proton pump inhibitors ; msoda= Modified Severity of Dyspepsia Assessment; RA=rheumatoid arthritis; RCT=randomized controlled trials; WDAE=withdraw due to adverse events; WOMAC=Western Ontario and McMaster Osteoarthritis Index; UGI=upper gastrointestinal. versus NSAIDs with PPIs 14

15 APPENDIX 3: Characteristics of Included Economic Evaluations First Author, Study Patient Publication Design, Interventio Comparator Characteristics, Year, Time n s Sample Size Country horizon Inotai, Hungary Economic evaluations, One year RA patients; drug costs (euros) based on the average costs formulary named Pharmindex; All costs were originally calculated in HUFs (Hungarian forint), then converted to euro (2007) nsnsaids plus PPIs Assumptions Adopted a QALY value Utility values to be similar to QALY values for a 3- month period from Maetzel et al.(2003) 38 The efficacy of the different active agents of NSAIDs in therapeutically relevant dose was assumed to be the same. Mortality rate for serious GI outcomes: 0.08 AMI mortality rate (dying because of AMI in 30 days): Chen, UK HTA, Economic evaluations, 5 years OA or RA patients nsnsaids plus PPIs nsnsaids do not protect against the risk of MI Dyspepsia requiring medical consultation (base case value: 10.2%) Hospitalized if complicated UGI event (base case value: 62.7%) Symptomatic ulcer (3 months) diclofenac (0.137%) Al, The Netherlands Brown, ,39 UK Economic evaluations, 6 months HTA, Economic evaluations, 6 months to one year OA or RA patients nsnsaids plus PPIs 4 reports were nsnsaids included plus PPIs Serious GI events : probability: , Relative risk: 0.5 Anemia: Probability: RR: 0.33 All NSAIDs (excluding aspirin) have greater efficacy than simple analgesics (such as paracetamol and codeine) in reducing pain and improving physical function. All NSAIDs (excluding aspirin and simple analgesics) are of equal efficacy based on equivalent dose. AE=adverse events; AMI=acute myocardial infarction; GI=gastrointestinal; MI=myocardial infarction; msoda= Modified Severity of Dyspepsia Assessment; NSAIDs=non-steroidal anti-inflammatory drugs; nsnsaids= non-cox-2 selective non-steroidal antiinflammatory drugs; OA=osteoarthritis; PGA-VAS=patient global assessment-visual analogue scale; PPI=proton pump inhibitor; QALY=quality adjusted life year; RA=rheumatoid arthritis; WOMAC=Western Ontario and McMaster Osteoarthritis Index; UGI=upper gastrointestinal. versus NSAIDs with PPIs 15

16 APPENDIX 4: Critical Appraisal of Included Studies First Author, Publication Year Strengths Limitations Randomized Controlled Trials Hochberg, Research questions, randomization, allocation concealment, double blinding process, ITT were reported; Baseline characteristics were reported and were comparable between Treatment under investigation is the only difference between Dropout rate was approximately 9.5% and comparable between Pain reduction scale was based on visual analogue scale; Whether the results for all sites were comparable: not reported; Cryer, Research questions, randomization, allocation concealment, double blinding process, ITT were reported; Baseline characteristics were reported and comparable; between Treatment under investigation is the only difference between Dropout rate was approximately 9.5% and comparable between Chan, Research questions, randomization, allocation concealment, double blinding process, ITT were reported; Baseline characteristics were reported and comparable between Treatment under investigation is the only difference between Main outcomes were assessed using validated measurements, such as GI bleeding and endoscopic ulcers; Lai, Research questions, randomization, allocation concealment, ITT were reported; Baseline characteristics were reported and comparable between Treatment under investigation is Pain reduction scale was based on visual analogue scale; Whether the results for all sites were comparable: not reported; Dropout rate greater than 20% in both Whether the results for all sites were comparable: not reported; Open label versus NSAIDs with PPIs 16

17 First Author, Publication Year Strengths the only difference between Dropout was approximately 2% and comparable between Main outcomes were assessed using validated measurements, such as GI bleeding and endoscopic ulcers; Limitations Chan, Research questions, randomization, allocation concealment, double blinding process, ITT were reported Baseline characteristics were reported and comparable between Dropout was approximately 6% and comparable between groups Main outcomes were assessed using validated measurements, such as GI bleeding and endoscopic ulcers Chan, Research questions, randomization, allocation concealment, double blinding process, ITT were reported Baseline characteristics were reported and comparable between Dropout was approximately 6% and comparable between groups Main outcomes were assessed using validated measurements, such as GI bleeding and endoscopic ulcers Non-randomized Study Economic Evaluations Inotai, Research questions and selection criteria were reported; Sources of effectiveness estimates used were stated; Choice of model used and the key parameters on which it was based were justified; Chen, Research questions and selection criteria were reported; Sources of effectiveness Numeric imbalance in the number of patients receiving concomitant aspirin (5.2% in celecoxib arm and 12.3% in diclofenac plus omeprazole) Numeric imbalance in the number of patients receiving concomitant aspirin (5.2% in celecoxib arm and 12.3% in diclofenac plus omeprazole) Decision tree models cannot handle the long-term effect of chronic diseases, such as RA; It was also unable to consider the increased risk after the recrudescence of peptic ulcer and acute The same hospitalization rate was used for peptic ulcer for all treatments arms, which may differ in the different therapeutic strategies; Conflict of interests declared; Lack of consistency in the reporting safety events; The choice and dose of nsnsaids versus NSAIDs with PPIs 17

18 First Author, Publication Year Strengths estimates used were stated; Choice of model used and the key parameters on which it was based were justified; Conflict of interests declared; Al, Research questions and selection criteria were reported; Sources of effectiveness estimates used were stated Choice of model used and the key parameters on which it was based were justified. Conflict of interests declared; Brown, Research questions and ,39 selection criteria were reported; Sources of effectiveness estimates used were stated; Choice of model used and the key parameters on which it was based were justified; Conflict of interests declared; Limitations varies from country to country; age restrictions and other exclusion criteria also limit generalizability and potential publication bias; Unrealistic 100% patient compliance was assumed, though it might be expected that the number of GI events would increase when adherence decreases; ICER not reported for the interested comparison; Drug cost data from four included studies based on the source of ; Therefore, the conclusion from each individual study must be interpreted with caution. Very little information on the relative effectiveness between strategies owing to lack of head-to-head studies; Direct healthcare costs were available only as reported estimates from clinicians. There is little or no patient-based information about the resource use consequences of NSAID-related GI events; AMI= myocardial infarction; GI=gastrointestinal; ICER=incremental cost effectiveness ratio; ITT=intention to treat analysis; NSAID=non-steroidal anti-inflammatory drug; nsnsaids= non-cox-2 selective non-steroidal anti-inflammatory drugs; PPIs=proton pump inhibitors. versus NSAIDs with PPIs 18

19 Appendix 5: Main Study Findings and Authors Conclusions Clinical Effectiveness and Safety First Main results Author, Publication Year nsnsaid plus PPI Author s Conclusions Hochberg, WOMAC pain score change from baseline: Study 307: -41.8; Study 309: -42.9; WOMAC function change from baseline Study 307: -36.3; Study 309: -36.8; PGA-VAS change from baseline Study 307: 21.6, Study 309: 25.6, Cryer, (Companion publication of Hochberg ) (data expressed as mean change from baseline) msoda Study 307: -4.8 Study 309: -3.3 Heartburn free days Study 307: 75.8% Study 309: 72.1% UGI AE Study 307: 16.9% Study 309: 21.6% Discontinuation due to UGI WOMAC pain score: change from baseline Study 307: Study 309: WOMAC function change from baseline Study 307: Study 309: -38.9; PGA-VAS change from baseline Study 307: 21.2 Study 309: 29.0 (data expressed as mean change from baseline) msoda Study 307: -3.5 Study 309: -4.5 Heartburn free days Study 307: 83.9% Study 309: 78.4% UGI AE Study 307: 16.6% Study 309: 18.9% Discontinuation due to UGI Naproxen/esomepra zole has comparable efficacy to celecoxib in the management of pain associated with osteoarthritis of the knee over 12 weeks msoda: ( vs. naproxen/esomepra zole) difference of change from baseline) were not significant; UGI AE incidence was also similar between treatment Significantly more heartburn-free days was reported in naproxen/esomepra zole treated patients compared with celecoxib. Study 307: 1.6% Study 309: 3.7% Study 307: 1.2% Study 309: 0.8% Discontinuation due to any AE Study 307: 6.6% Study 309: 9.0% Chan, GI risk: # of patients (%) (0.9%) (hazard ratio 4.3, 95% CI 2.6 to 7.0; p< vs. nsnsaid plus PPIs) Withdrawal due to GI AE: 114 (6%) (p= vs. nsnsaids plus PPI) Discontinuation due to any AE Study 307: 7.3% Study 309: 6.6% GI risk: # of patients (%) 81 (3.8%) Withdrawal due to GI AE: 167 (8%) GI risk was lower with celecoxib than a diclofenac plus a PPI Targownik, UGI complications UGI complications In terms of UGI versus NSAIDs with PPIs 19

20 First Author, Publication Year (adjusted odds ratio (OR, 95%CI) compared with nsnsaids alone): 0.39 (0.32 to 0.48) P<0.01) Main results nsnsaid plus PPI (adjusted odds ratio (OR, 95%CI) compared with nsnsaids alone): 0.67 (0.48 to 0.95) (P<0.01) Author s Conclusions complication risk reduction, may be superior to nonselective NSAIDs plus a PPI. Lai, Ulcer relapse at 24 weeks follow up: # of patients (%) 4 (3.7%, 95% CI: 0.0% to 7.3% Absolute difference betwee groups: -2.6%; 95% CI for the difference: - 9.1% to 3.7%); GI events 22.5% Dyspepsia 15.0%, 95% CI 9.7to 22.5) (P =.02 vs. nsnsaid plus PPI CV AE: 0.8% Withdrawal: 5% Ulcer relapse at 24 weeks follow up: # of patients (%) 7 (6.3%, 95% CI 1.6% to 11.1%) GI events 15.6% Dyspepsia (5.7%, 95% CI 2.8 to11.4) CV AE: 0.8% Withdrawal: 4.1% Efficacy was comparable in the prevention of ulcer relapse; AE: more patients experienced dyspepsia in celecoxib arm Chan, Rate of ulcers relapse in 6 months %(95CI: 11.3 to 26.1) Difference between groups: -6.7%; 95% CI: -17.8% to 3.9%) (P = 0.21). Dyspepsia (celecoxib vs. nsnsaids plus PPI) Hazard ratio: %CI 2.6 to 10.8 (P<0.001) Overall withdrawal rate 5.1% WDAE 3.4% Rate of ulcers relapse in 6 months 25.6% (95%CI) ) Overall withdrawal rate Similar to celecoxib group WDAE 6.6% Recurrent ulcer bleeding 9 patients (6.4 %, 95% CI, 4.3 to 8.4) Efficacy: similar Dyspepsia: more in celecoxib arm Chan, Recurrent ulcer bleeding 7 patients (4.9 %, 95% CI: 3.1 to 6.7) (difference between groups: 1.5 %( 95 %CI 6.8% to 3.8%). Efficacy: comparable: AE: similar Dyspepsia 15.3% CV events 1.4% Renal adverse events 24.3 % WDAE 10.4% Dyspepsia 8.4% CV events 1.4% Renal adverse events: 30.8% WDAE 9.8% AE=adverse events; CV=cardiovascular; GI=gastrointestinal; UGI=upper gastrointestinal; nsnsaids= non-selective non-steroidal anti-inflammatory drugs; PPIs=proton pump inhibitors; WDAE=withdraw due to adverse events; WOMAC=Western Ontario and versus NSAIDs with PPIs 20