Extended report Ann Rheum Dis: first published as 10.1136/ard.2008.099861 on 3 December 2008. Downloaded from Efficacy and toxicity of methotrexate (MTX) monotherapy versus MTX combination therapy with non-biological disease-modifying antirheumatic drugs in rheumatoid arthritis: a systematic review and meta-analysis W Katchamart,1,2 J Trudeau,3 V Phumethum,1 C Bombardier4,5 c Additional figures and ABSTRACT combination DMARD therapy be used: initially or appendixes are published online Objective: To evaluate the efficacy and toxicity of only after a trial of MTX monotherapy? Finally, only at http://ard.bmj.com/ which is the preferred combination DMARD content/vol68/issue7 methotrexate (MTX) monotherapy compared with MTX combination with non-biological disease-modifying anti- strategy? These questions are particularly salient 1 Rheumatology Division, rheumatic drugs (DMARDs) in adults with rheumatoid as formularies in many countries require the use of Department of Medicine, University of Toronto, Toronto, arthritis. MTX mono and MTX combo therapies before Ontario, Canada; Method: A systematic review of randomised trials reimbursing for the more expensive biological 2 Rheumatology Division, comparing MTX alone and in combination with other non- drugs. The objective of this paper was to system- Department of Medicine, Siriraj biological DMARDs was carried out. Trials were identified atically review randomised trials that compared Hospital, Mahidol University, MTX monotherapy with MTX in combination Bangkok, Thailand; 3 Hoˆspital in Medline, EMBASE, the Cochrane Library and ACR/ Notre-Dame, Department of EULAR meeting abstracts. Primary outcomes were with other non-biological DMARDs. This manu- Rheumatology, Montreal, withdrawals for adverse events or lack of efficacy. script is part of the 3E (Evidence, Expertise and 4 Canada; Division of Results: A total of 19 trials (2025 patients) from 6938 Exchange) Initiative described in more details in Rheumatology and Department the same issue of this journal.7 of Health Policy, Management, citations were grouped by the type of patients and Evaluation, University of randomised. Trials in DMARD naive patients showed no Toronto, Toronto, Ontario, significant advantage of the MTX combination versus MATERIALS AND METHODS 5 Canada; Division of Clinical monotherapy; withdrawals for lack of efficacy or toxicity Literature search Decision Making and Health Care, Toronto General Research were similar in both groups (relative risk (RR) = 1.16; We performed a search of electronic bibliographic Institute, University Health 95% CI 0.70 to 1.93). Trials in MTX or non-MTX DMARD databases including Medline (1950 to June week 3 Network, Toronto, Ontario, inadequate responder patients also showed no difference 2007), EMBASE (1980 to 2007 week 25) and the Canada in withdrawal rates between the MTX combo versus Cochrane Central Register of Controlled Trials mono groups (RR = 0.86; 95% CI 0.49 to 1.51 and http://ard.bmj.com/ Correspondence to: (2nd quarter 2007) using a search strategy that Dr C Bombardier, Institute for RR = 0.75; 95% CI 0.41 to 1.35), but in one study the combined terms for ‘‘rheumatoid arthritis’’, Work and Health, 481 University specific combination of MTX with sulfasalazine and ‘‘methotrexate’’ and ‘‘randomised controlled trials’’ Avenue, Suite 800, Toronto, hydroxychloroquine showed a better efficacy/toxicity ratio (full search strategy available online at http:// Ontario M5G 2E9, Canada; [email protected] than MTX alone with RR = 0.3 (95% CI 0.14 to 0.65). www.annrheumdis.com/supplemental). We also Adding leflunomide to MTX non-responders improved searched the abstracts of the Annual scientific Accepted 17 November 2008 efficacy but increased the risk of gastrointestinal side meetings of the American College of Published Online First effects and liver toxicity. Withdrawals for toxicity were Rheumatology (ACR) and European League on September 30, 2021 by guest. Protected copyright. 3 December 2008 most significant with ciclosporin and azathioprine Against Rheumatism (EULAR) from 2005 to combinations. 2007, the references lists of all relevant studies, Conclusion: In DMARD naive patients the balance of letters and review articles, and all languages were efficacy/toxicity favours MTX monotherapy. In DMARD included. inadequate responders the evidence is inconclusive. Trials are needed that compare currently used MTX doses and Study selection combination therapies. Two reviewers (WK, JT) independently screened the titles and abstracts of the citations and retrieved relevant articles. The following selection Methotrexate (MTX) is among the most effective criteria were used: (a)randomisedcontrolled disease-modifying antirheumatic drugs (DMARDs) trials of MTX monotherapy versus MTX com- in rheumatoid arthritis (RA) with less toxicity and bined with other DMARDs of at least 12 weeks of better tolerability. Unfortunately, MTX alone may trial duration (open-label extensions were not fully control disease activity. Increasingly, excluded as well as studies comparing DMARDs MTX is used in combination with other non- not currently used—for example, oral gold); (b) biological DMARDs.1–3 patients with RA >18 years old; (c)dataavailable Many MTX and traditional DMARDs combina- on one or more of following prespecified out- tion regimens have been studied, but several comes: ACR core set8;ACR20,50or70 This paper is freely available 4–6 9 10 online under the BMJ Journals important questions remain. What is the relative responses ; ACR remission ;DiseaseActivity 11 12 13 unlocked scheme, see http:// benefit and toxicity of MTX mono versus MTX Score (DAS) ; EULAR response ;withdrawal ard.bmj.com/info/unlocked.dtl combination with DMARDs? When should the duetolackofefficacyoradverseevents(AEs); Ann Rheum Dis 2009;68:1105–1112. doi:10.1136/ard.2008.099861 1105 Extended report Ann Rheum Dis: first published as 10.1136/ard.2008.099861 on 3 December 2008. Downloaded from Figure 1 Results of the literature search and disposition of the potentially relevant studies. *Number is not equal to the sum of the number from each database owing to duplication among databases. ACR, American College of rheumatology; CCRT, Cochrane Central Register of Controlled Trials; DMARD, disease-modifying antirheumatic drug; EULAR, European League Against Rheumatism; MTX, methotrexate; RA, rheumatoid arthritis; RCT, randomised controlled trial. number of total or individual AEs (only commonly reported questions, including randomisation, blinding procedure individual AEs are presented). (patients, provider and outcome assessor), concealed treatment allocation, similarity of the important baseline characteristics, Data abstraction and quality assessment co-intervention, timing of the outcome assessment, compliance Two reviewers (WK, VP) independently extracted the data and and withdrawals and intention-to-treat analysis. Each item is assessed the quality of relevant studies. Study quality was rated as ‘‘yes’’ = 1 and ‘‘no or do not know’’ = 0. The score http://ard.bmj.com/ assessed using van Tulder ‘s scale.14 This scale comprises 11 ranges from 0 to 11. Table 1 Excluded studies and reason for exclusion Data synthesis Study Reason for exclusion We used RevMan 4.2.10 for analysis. The efficacy analysis was stratified into three groups based on previous DMARD use. The Calguneri, 1999 No methotrexate monotherapy arm (data combined with sulfazalazine and hydroxychloroquine monotherapy) ‘‘DMARD naı¨ve, parallel strategy’’ refers to trials in which on September 30, 2021 by guest. Protected copyright. Clegg, 1997 No outcome of interest patients who never received DMARDs (including MTX) were Haagsma, 1995 Summary of Haagsma et al. (included in this review)27 randomised to start MTX alone or MTX plus another DMARD; Kremer, 2004 Open-label extension of randomised controlled trial The ‘‘MTX inadequate response, step-up strategy’’ refers to Maillefert, 2003 Open-label extension of randomised controlled trial trials in which patients with inadequate response to MTX were Matucci-Cerinic, 2003 Summary of Kremer et al. (included in this review)20 randomised to continue the use of MTX alone or to add a Mottaghi, 2005 Non-randomised controlled trial second DMARD. The ‘‘non-MTX DMARDs inadequate Mroczkowski, 1999 Open-label extension of randomised controlled trial response, step-up strategy’’ refers to trials where patients with Nagashima, 2006 Non-randomised controlled trial inadequate response to DMARDs (other than MTX) were Nisar, 1994 Non-randomised controlled trial randomly switched to MTX alone or MTX plus another O’Dell, 1996 Open-label extension of randomised controlled trial DMARD. The toxicity analysis was stratified by DMARD Rau, 1998 Non-randomised controlled trial combination and pooled across trials for each combination. Stein, 1997 Open-label extension of randomised controlled trial For continuous measures of efficacy, we used either end of Trnavsky, 1993 No methotrexate monotherapy arm trial data or change from baseline and pooled them as weighted Willkens, 1996 Published in the journal supplements and the key data have 15 been reported in Willkens et al (included in this review)25 mean differences using a random effect model. For the Krause D et al (German), Non-randomised controlled trial categorical measures of efficacy, the end of trial results were 1998 pooled and estimated using the relative risk (RR) with random Geokoop-Ruiterman No MTX monotherapy arm effect model. An RR .1 favours MTX combination therapy. For