Abstract
Objective. To develop recommendations for the use of methotrexate (MTX) in patients with rheumatoid arthritis.
Methods. Canadian rheumatologists who participated in the international 3e Initiative in Rheumatology (evidence, expertise, exchange) in 2007–2008 formulated 5 unique Canadian questions. A bibliographic team systematically reviewed the relevant literature on these 5 topics. An expert committee consisting of 26 rheumatologists from across Canada was convened, and a set of recommendations was proposed based on the results of systematic reviews combined with expert opinions using a nominal group consensus process.
Results. The 5 questions addressed drug interactions, predictors of response, strategies to reduce non-serious side effects, variables to assess clinical response, and incorporating patient preference into decision-making. The systematic review retrieved 93 pertinent articles; this evidence was presented to the expert committee during the interactive workshop. After extensive discussion and voting, a total of 9 recommendations were formulated: 2 on drug interactions, 1 on predictors of response, 2 on strategies to reduce non-serious side effects, 3 on variables to assess clinical response, and 1 on incorporating patient preferences into decision-making. The level of evidence and the strength of recommendations are reported. Agreement among panelists ranged from 85% to 100%.
Conclusion. Nine recommendations pertaining to the use of MTX in daily practice were developed using an evidence-based approach followed by expert/physician consensus with high level of agreement.
- RHEUMATOID ARTHRITIS
- METHOTREXATE
- GUIDELINE
- RECOMMENDATION
- EVERYDAY PRACTICE
- SYSTEMATIC REVIEW
The need for recommendations
Methotrexate (MTX) is among the most effective and the most commonly prescribed disease modifying antirheumatic drugs (DMARD) in the treatment of rheumatoid arthritis (RA)1–3. Despite the advent of new effective biologic agents, MTX is still used as an anchor drug to enhance or maintain the efficacy of biologic agents4–10. Although MTX has been commonly used in patients with RA over the last 2 decades, clinical practice varies considerably among rheumatologists. It is unclear whether this variation reflects conflicting evidence in the literature or variable application of the evidence in clinical practice.
3e Initiative in Rheumatology
The 3e Initiative in Rheumatology (evidence, expertise, exchange) is a multinational effort aimed at promoting evidence-based medicine by formulating detailed recommendations addressing clinical problems11. The objective of the 3e Initiative 2007–2008 was to develop practical recommendations for the use of MTX in rheumatic disorders, by integrating systematically generated evidence and expert opinion of a broad panel of international rheumatologists. Ten clinical questions on MTX were selected by rheumatologists from 17 countries in Europe and North and South America. The Canadian participants selected 5 additional questions pertaining to drug interaction, monitoring, predictor of response, patient preference, and management of nuisance side effects. The recommendations for the 10 international questions have been published12. This article presents the summary of the evidence and the recommendations for the additional Canadian questions.
MATERIALS AND METHODS
Stakeholders
The Canadian 3e Initiative group consisted of a steering group, a bibliographic team, and an expert committee. The steering group included the principal investigator (CB) and 8 members (VB, BH, SL, DM, JP, KS, JT, and CT). The bibliographic team included 5 rheumatology fellows (WK, JB, JD, TD, and GR) who undertook a systematic review of literature assisted by 3 mentors (CB, BH, and JP). Twenty-six Canadian rheumatologists from across Canada representing academic and community practices formed the expert committee. They reviewed the evidence from the systematic reviews prepared by the bibliographic team and formulated practice recommendations.
Evidence based approach
The methodology for the systematic review and for the practice recommendations is presented in Figure 1. The 5 Canadian questions (Table 1) were selected by the Canadian Steering Committee at the international meeting held April 27–28, 2007. A systematic search of Medline, Embase, and Cochrane Central Register of Controlled Trials to September 2007 was carried out by the bibliographic team assisted by experienced librarians. The sensitive search strategy included MeSH terms, keywords, and text words related to RA and MTX, and other terms specific to each of the 5 questions (Table 1) included: drug toxicity, adverse effects, drug interaction, patient preference, monitor, treatment outcome, and predictor; there were no restrictions on language. To supplement these electronic bibliographic databases, abstracts from annual scientific meetings were also searched (American College of Rheumatology and European League Against Rheumatism 2005–2007). The reference lists of retrieved articles and reviews were also reviewed. To identify eligible articles, prespecified inclusion and exclusion criteria were applied to the citations obtained from the search strategies. These included population (RA), drug (MTX), and for each question specific interventions and outcome measures. Retained studies were systematically reviewed for quality assessment, data extraction, and synthesis. The evidence was summarized. The level of evidence and grade of recommendation were scored using “The Oxford Centre for Evidence-based Medicine Level of Evidence (May 2001)” (URL: www.cebm.net/index.aspx?o=1047) (⇔Table 2). A series of full systematic reviews13–16 underpins the recommendations for the Canadian questions. Four are published in this issue of The Journal14–16.
Expert opinion approach
Summaries of the systematic reviews on the 5 topics were presented to the Canadian expert committee at a national meeting in January 2008. Draft recommendations were formulated by the expert committee based on the results of the systematic review. These recommendations were discussed and reworded using the nominal group approach17. The final statements were established using a touch-pad voting process with prespecified cutoff agreement. Additionally, participants expressed their level of agreement with the final recommendation using a numeric scale from 0 to 100.
RESULTS
For the 5 questions, the literature search identified 9603 citations. After applying the inclusion and exclusion criteria, 93 full-length articles were retained for systematic review. Table 3 presents the final set of 5 recommendations, their level of evidence, the strength of the recommendations, and the agreement among experts based on touch-pad voting.
Recommendation 1: Drug Interactions
-
The majority of drugs including nonsteroidal antiinflammatory drugs (NSAID) may be used safely in combination with MTX in rheumatic diseases (Grade of recommendation C).
-
Trimethoprim and sulfamethoxazole (TMP-SMX) should be avoided in patients treated with MTX (Grade of recommendation C).
These recommendations are based on the systematic review of 21 pharmacokinetics studies, 5 observational studies, and 78 case reports (Level of evidence 4)16. Cytopenia and elevation of liver enzymes were the main reported toxicities. Most reports of cytopenia were attributed to the use of concomitant NSAID or high-dose aspirin (ASA)18–31. Other medications, e.g., antibiotics, gastroprotective agents, and antihypertensive drugs, have been noted in case reports.
Most NSAID and selective cyclooxygenase-2 inhibitors did not significantly affect the pharmacokinetic profile of MTX32–44. For ibuprofen and naproxen, studies showed conflicting results33–35. Four studies evaluating high-dose ASA (1.3–4.5 g/day) reported an increase of serum concentration of MTX45–48 (Level of evidence 4).
The use of TMP-SMX was mentioned as a risk factor for developing bone marrow suppression in one retrospective case-control study49 and in 17 case reports18,50–62 (Level of evidence 4).
Cytopenia and elevated liver enzymes were reported with several medications other than NSAID and TMP-SMX, but in only one to a few cases each. Experts agreed that the evidence was not strong enough to make a recommendation. Some experts also proposed that drugs that affect renal function should be used cautiously in patients receiving MTX; however, there was no evidence directly supporting this statement, and the expert committees’ agreement for this statement was only 41%. Consequently, it was not included in the final recommendation.
Recommendation 2: Prognostic Factors for Response to MTX
-
In determining treatment strategy of patients treated with MTX, characteristics of poor prognosis should be considered, such as female sex and persistent disease activity (Grade of recommendation B).
This recommendation is based on the systematic review15 consisting of 2 metaanalyses63,64, 3 cohorts of MTX-treated RA65–67, and 4 cohorts using data from randomized controlled trials68–71 (Level of evidence 2b). Both early RA64,65,68–71 and long-standing RA66,67 were included. The dose of MTX used in these studies ranged from 15 to 25 mg/wk. Poor clinical response was defined as a lack of evidence of achieving a low disease activity state, measured by Disease Activity Score (DAS) < 2.465,66, DAS28 < 3.267, or Simplified Disease Activity Index score (SDAI) ≤ 1164 at the end of followup, while poor radiographic outcome was defined as having evidence of significant radiographic progression, measured by Sharp score70,71, Modified Sharp/van der Heijde score69, or Modified Larsen score68 at the end of followup.
Predictors of poor response to MTX include female sex65–67, prior use of DMARD67, high disease activity at baseline measured by DAS65,66 or SDAI64, and high tender joint count67. Other predictors considered in the published literature were not found to be independent predictors of clinical response in both early and established RA.
Predictors of poor radiographic outcome include high baseline erythrocyte sedimentation rate (ESR)68,69, particularly in patients with persistent evidence of inflammation, e.g., high DAS2869, ESR69,71, and C-reactive protein (CRP)69,71.
These studies consistently identify disease activity as a predictor of poor response to MTX. Since different measures of disease activity were used across the studies, the expert panel decided to use the general term “persistent disease activity” instead of specifying the individual parameters of disease activity (e.g., DAS28, tender joint count, etc.).
Recommendation 3: Management of Non-serious Side Effects
-
To minimize non-serious gastrointestinal side effects of MTX one could try to switch from oral to parenteral (subcutaneous or intramuscular) MTX (Grade of recommendation D).
-
Other strategies to minimize non-serious side effects could include splitting of the dose of MTX (Grade of recommendation D).
The systematic review “Strategies to Reduce “Nuisance” Side-Effects of Methotrexate”13 failed to find direct evidence to support the benefit of modalities to reduce nuisance side effects. Experimental evidence from appropriately designed clinical trials was not available; these recommendations are, therefore, based on extrapolation from studies demonstrating that intramuscular (im) form is more tolerable than oral form in 2 cohorts (Level of evidence 4). In a survey of patients forced to switch to oral MTX when the supply of im MTX ran out72, 69 (48%) patients who tolerated im MTX could not tolerate taking it orally due to nausea (p < 0.001). In a RA cohort of 212 patients73, switching from oral to im MTX was found to decrease gastrointestinal side effects: after 6 months only 9% terminated im MTX due to adverse events.
The recommendation for splitting the dose of MTX was based entirely on expert opinion due to a lack of evidence addressing this issue. The only evidence related to the dose of MTX was from an open-labeled RCT74 showing that starting MTX treatment at a dose of 25 mg/week was associated with a higher rate of minor toxicity (gastrointestinal except liver toxicity) as compared to 15 mg/week (28% vs 17%, p < 0.05, for 25 vs 15 mg/wk, respectively) (Level of evidence 4).
Other strategies or modalities have been studied; however, the expert panel chose not to make recommendations on these modalities due to insufficient evidence75–77.
Recommendation 4: Parameter Used in the Assessment of Clinical Response
-
Use of validated outcome measures to reach a target of low disease activity or remission is recommended (Grade of recommendation A).
-
Joint counts should be included in the assessment of disease activity in RA (Grade of recommendation B).
-
In addition to joint counts, other parameters in the assessment of disease activity in RA could include validated measures of global assessments and acute-phase reactants (Grade of recommendation B).
The systematic review on this topic showed that there was no evidence for which parameters should be used in management of patients with RA to assess a clinically meaningful response in daily practice14. These recommendations were extrapolated from 3 randomized controlled trials of tight control strategy in RA identified by experts (Level of evidence 1a). These studies78–80 demonstrated that targeted care aiming at remission or low disease activity and frequent followup (every month) may lead to more aggressive treatment, resulting in better disease states/clinical outcomes than usual care in RA. These studies used different outcome measures to assess the clinical response. DAS and DAS28 were used in the TICORA study78 and in Fransen, et al79, respectively; whereas in the CAMERA study80, a computer program was used to calculate the 50% improvement response in swollen joint count and the improvement of at least 2 out of 3 of the following variables: number of tender joints, ESR, and patient global assessment of general well-being.
Based on the results of these 3 RCT, all experts but one agreed that in clinical practice validated outcome measures aiming at remission or at least low disease activity should be used to assess clinically meaningful response. Although there is a lack of evidence on which parameter should be used in daily practice, the group reached the following consensus: the use of patient or physician global assessments alone is not sufficient; joint assessment is the most important parameter reflecting disease activity; and other variables including patient global assessment, physician global assessment, and inflammatory marker, e.g., ESR or CRP, should be considered in addition to joint count in the assessment of RA disease activity.
Recommendation 5: Patient preference
-
Patients need to be educated on their disease and treatment options and involved in the decision-making process (Grade of recommendation D).
This recommendation is entirely expert-based. There is no evidence in the literature that incorporating RA patients’ preference in the therapeutic decision improves treatment outcomes, adherence to medications, or patient satisfaction in those taking MTX. Nonetheless, the expert panels agreed that shared decision-making should incorporate patient’s preference, research evidence, and knowledge of the patient’s clinical state. In addition, the expert panel was also aware of the role and impact of patient education on the treatment outcomes based on a Cochrane review, “Patient education for adults with RA”81 (Level of evidence 2b). The results of this review supported a beneficial effect of patient education programs in terms of pain (small benefit, 4% or 0.2 cm in the visual analog scale), functional impairment (moderate benefit, 10% or 0.16 points on the Health Assessment Questionnaire score), tender joint count (moderate benefit, 9% or 1.3 points on the Ritchie index), patient’s overall assessment [moderate benefit, 12% or 0.28 points on the Arthritis Impact Measurement Scales 2 (AIMS2) arthritis subscale], and psychological status (moderate benefit, 5% or 0.15 points on the AIMS2 affect subscale and 12% or 0.14 points on the Hospital Anxiety and Depression Scale). However, no lasting benefits were found at one year after the end of the educational program. These effects were related chiefly to educational programs, as opposed to simple patient information.
DISCUSSION
These recommendations were developed using an evidence-based approach. A methodology team conducted systematic reviews using a comprehensive search in 2 bibliographic databases, Medline and Embase, plus screening of abstracts of scientific meetings. A group of clinical experts considered the quality of the evidence from these systematic reviews as well as the clinical relevance, applicability, and values and preferences of patients and practitioners to ensure that recommendations meet their needs.
We followed an established group decision method, the nominal group process. This included a representative expert panel of academic and community rheumatologists from across Canada, who openly discussed the evidence from the literature followed by a silent voting process. We used the touch-pad methodology with prespecified cutoff levels of agreement to generate the final recommendations. Several rounds of rewording and revoting were sometimes required to reach the agreed cutoff. This process ensured that the final recommendations were evidence-driven as well as clinically relevant.
Of the 15 questions initially proposed by the Canadian steering committee at the international meeting of the 3e Initiative, 10 were also rated highly by the 17 participating countries. Recommendations for these 10 top-rated international questions have been published12. This article addressed the 5 remaining Canadian questions. Although these 5 questions are clinically important, the 10 international questions addressing MTX initiation, monitoring, and safety were considered of higher priority by the international experts. In their selection of the top 10 questions, experts may also have taken into account whether there would be sufficient evidence in the literature to generate robust recommendations. Indeed, we found that many of the 5 questions lacked high-quality studies, or studies were not specifically related to MTX treatment; for instance, no study directly addressed which of the objective parameters should be used to assess the clinical response to MTX, or should patient preference be taken into account in MTX treatment decisions. Our recommendations were, therefore, based on expert opinion, resulting in the lowest “grade of recommendation” on the Oxford scale. Nevertheless, our recommendations emphasize the need for future research in these clinically important areas.
Of several recent guidelines available to assist in the management of patients with RA82–86, none addressed our question on drug interactions. Most guidelines addressed none or just a few of our 5 questions, but where our questions were addressed, the result was generally congruent with our recommendations.
In conclusion, using a nominal group process and scientific evidence, we provide recommendations for the use of MTX in patients with RA to assist specialists in everyday practice. These 9 Canadian recommendations complement the 10 recommendations from the international 3e Initiative expert panel. These recommendations are intended to benefit all patients with RA who receive MTX therapy.
Acknowledgments
Canadian 3e Initiative Consensus Group: Kenneth Blocka, Cathy Elizabeth Flanagan, Paul Davis, Wojciech Olszynski, Vandana Ahluwalia, Michel Zummer, Gilles Boire, Jamie Henderson, Barry Koehler, Philip Baer, Zareen Ahmad, Anne St-Pierre, Martin Cohen, Maysan Abu-Hakima, Majed Khraishi, Elaine Soucy, and Jacob Karsh.
Footnotes
-
Full Release Article. For details see Reprints/Permissions at jrheum.org
-
Supported by an unrestricted educational grant from Abbott. Dr. Bombardier holds a Canada Research Chair in Knowledge Transfer for Musculoskeletal Care.
- Accepted for publication December 13, 2009.
Free online via JRheum Full Release option
REFERENCES
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
- 28.
- 29.
- 30.
- 31.
- 32.
- 33.
- 34.
- 35.
- 36.
- 37.
- 38.
- 39.
- 40.
- 41.
- 42.
- 43.
- 44.
- 45.
- 46.
- 47.
- 48.
- 49.
- 50.
- 51.
- 52.
- 53.
- 54.
- 55.
- 56.
- 57.
- 58.
- 59.
- 60.
- 61.
- 62.
- 63.
- 64.
- 65.
- 66.
- 67.
- 68.
- 69.
- 70.
- 71.
- 72.
- 73.
- 74.
- 75.
- 76.
- 77.
- 78.
- 79.
- 80.
- 81.
- 82.
- 83.
- 84.
- 85.
- 86.