Elsevier

Joint Bone Spine

Volume 81, Issue 2, March 2014, Pages 107-109
Joint Bone Spine

Editorial
Methotrexate in psoriatic arthritis

https://doi.org/10.1016/j.jbspin.2013.10.005Get rights and content

Introduction

Psoriatic arthritis (PsA) can manifest as a broad range of clinical presentations, of which the most common by far include peripheral joint involvement (symmetric or asymmetric polyarthritis, asymmetric oligoarthritis, distal interphalangeal arthritis, and/or mutilating arthritis) [1]. Nevertheless, axial skeleton involvement, enthesitis, and dactylitis are also among the possible manifestations, and PsA is classified among the spondyloarthropathies [2]. Methotrexate is the main disease-modifying antirheumatic drug used to treat rheumatoid arthritis (RA). The clinical and structural benefits of methotrexate in RA have been convincingly established and an additive effect with TNFα antagonists has been demonstrated [3]. Similar proof of efficacy is not available for PsA. In plaque psoriasis, methotrexate is known to be an effective maintenance drug, with a PASI75 response rate of 60% after 16 weeks [4]. However, methotrexate has not demonstrated beneficial effects on the axial manifestations of spondyloarthritis [5]. The issue is therefore whether methotrexate is effective on the peripheral joint manifestations of PsA.

The list of disease-modifying antirheumatic drugs (DMARDs) that help to control the joint manifestations of PsA includes methotrexate in some textbooks, [1] but not in others, which point out the limited data available on the efficacy of methotrexate in PsA [6]. A metaanalysis based on treatment withdrawal rates in randomized controlled trials (RCTs) published in 1966-2006 does not list methotrexate among effective DMARDs for PsA, as no adequate studies were identified [7]. The International Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) recommends methotrexate among DMARDs for PsA based on level B evidence, implying the existence of a well-conducted clinical trial but not of an RCT [8]. The 2007 update of the recommendations issued by the French Society for Rheumatology (SFR) indicate that methotrexate is among the effective DMARDs for PsA, based only on studies of methotrexate in combination with other drugs [9]. The 2009 French clinical practice guidelines based on expert opinion advocate the use of DMARDs, for instance with methotrexate, in PsA with peripheral arthritis refractory to nonsteroidal antiinflammatory drugs (NSAIDs) [10]. The guidelines specify that proof of efficacy for methotrexate was obtained in a single study in 37 patients. A literature review by EULAR experts underlined the paucity of studies of methotrexate in PsA and concluded that methotrexate produced clinical benefits, although potential structural benefits could not be determined [11]. However, the British multicenter RCT MIPA published in August 2012 [12] with an editorial by Maxime Dougados [13] found no evidence that methotrexate was effective, thus reinvigorating the controversy about the use of methotrexate in PsA.

Before publication of the MIPA trial, only four controlled studies evaluating the efficacy of methotrexate in PsA were available [14], [15], [16], [17]. Only 128 patients were included in all. Table 1 lists the main results. The earliest placebo-controlled trial is difficult to interpret, as methotrexate was used intravenously in high doses, resulting in marked bone marrow toxicity [14]. Nevertheless, methotrexate clearly improved the clinical synovitis and erythrocyte sedimentation rate (ESR). The first study of low-dose methotrexate (7.5-15 mg/week) versus placebo lasted 3 months and included only 37 patients with a mean PsA duration of 10 years [15]. Whereas the cutaneous psoriasis improved significantly, among the parameters used to evaluate the joint involvement only the physician global assessment showed benefits from methotrexate therapy. An RCT reported in 2008 [16] included 35 patients with recent-onset PsA. The randomized phase involved treatment for 3 months with an NSAID alone or combined with methotrexate and was followed by 3 months of combined treatment in all patients. The results after the 3-month randomized phase were mixed, with lower swollen and tender joint counts in the methotrexate + NSAID arm but no significant difference for the other clinical parameters or laboratory variables. Another Italian study compared low-dose methotrexate to cyclosporine for 12 months in 35 patients [17]. At the 6- and 12-months time points, both arms showed significant improvements in clinical joint parameters, serum C-reactive protein (CRP), and cutaneous psoriasis. However, the only difference between the study drugs was a significant ESR decrease with methotrexate but not cyclosporine. Although these four studies seem to suggest clinical benefits from methotrexate in PsA, their design precludes definitive conclusions, explaining the differences in interpretation depending on whether emphasis is put on the favorable clinical experience or on the lack of incontrovertible scientific evidence.

Given this uncertainty, the MIPA trial generated considerable interest as soon as its results were first reported at the 2010 ACR meeting [12]. This double-blind 6-month RCT compared methotrexate (n = 109) to a placebo (n = 112) in patients with recent-onset PsA (median disease duration, 1 year). The inclusion criteria were broad: all patients having cutaneous psoriasis and at least one peripheral joint with synovitis were eligible. Methotrexate was given orally in increasing dosages, from 7.5 mg/week initially to 10 mg/week after 4 weeks and 15 mg/week after 8 weeks. If needed, the rheumatologist could further increase the dosage to 20 mg/week at 4 months and 25 mg/week at 5 months. However, after the first 3 months most patients took the target dosage of 15 mg/week. Overall, the study drug was discontinued prematurely in 12% of patients, which was within the expected range. In contrast, the 20% (44/221 patients) rate of patients lost to follow-up seems unusually high. At 6 months, the 23% missing data rate required the use of multiple imputations for the intention-to-treat analysis.

The primary evaluation criterion in the MIPA trial was the response assessed using the ARC PsA response criteria (PsARC) at 6 months in the intention-to-treat population. The odds ratio for methotrexate vs. placebo was 1.77 but the 95% confidence interval (95%CI) was 0.97-3.23 and the P value was 0.06. Similarly, nonsignificantly higher values were seen with methotrexate at 6 months for the proportion of patients meeting ACR20 criteria (OR, 2.00; 95%CI, 0.65-6.22; P = 0.23) or the DAS28 response (OR, 1.70; 95%CI, 0.90-3.17; P = 0.10).

Table 2 reports the results of the intention-to-treat analysis of changes at 6 months versus baseline. The only parameters that were significantly better in the methotrexate arm were the global assessments by the patients and physicians. No significant differences were found for the other clinical parameters or for the laboratory markers of inflammation. Nevertheless, the changes in a favorable direction were consistently larger in the methotrexate arm. Subgroup analyses in patients with oligoarthritis or polyarthritis produced similar results. When the analysis was confined to patients who completed the study, a significant difference in favor of methotrexate was found for the PsARC but not for the ACR20 or DAS28 response. Methotrexate was significantly better than the placebo in improving the cutaneous psoriasis lesions as assessed using the PASI but not the PASI75 response, perhaps because the baseline PASI values were low. The authors concluded that their trial failed to provide proof of efficacy of methotrexate on active peripheral synovitis due to PsA.

The editorial for the MIPA trial report focuses on the efficacy of methotrexate in peripheral joint involvement due to spondyloarthropathies [13]. It specifies that, in contrast to sulfasalazine, methotrexate is not recommended by the ASAS for the treatment of peripheral arthritis in patients who have PsA with or without concomitant skin psoriasis. The trial is criticized for the mismatch between using active synovitis in a single peripheral joint as the inclusion criterion and the PsARC response as the primary endpoint. The swollen and tender joint counts are two of the four items used to assess the PsARC response. In addition, the long patient recruitment period of 7 years is puzzling. When a patient-reported outcome was used, namely, the patient's global assessment, a significant difference in favor of methotrexate was found, as well as for the physician's global assessment. Thus, the MIPA trial leaves considerable room for skepticism. Philip Mease stated that it does not challenge the efficacy of methotrexate in PsA given its methodological flaws and the lessons derived from everyday clinical practice [18].

The other available data on the efficacy of methotrexate in PsA deserve discussion. In the methotrexate arm of the RESPOND trial comparing infliximab plus methotrexate to methotrexate alone in methotrexate-naive patients with PsA, the response rates in the arm given methotrexate alone in a dosage of 15 mg/week orally were considerably higher than in the MIPA trial: 66.7% for the ACR20, 39.6% for the ACR50, 18% for the ACR70, and 54.3% for the PASI75 [19]. This discrepancy may be related to differences in baseline activity and severity of the PsA: mean values in the MIPA and RESPOND trials were 6 vs. 14 for the swollen joint count and 0.88 vs. 1.49 for the HAQ score, respectively. A study in the Norwegian cohort NOR-DMARD assessed outcome data from 430 patients with PsA and 1218 patients with rheumatoid arthritis [20]. All patients were methotrexate-naive at baseline and were given similar mean methotrexate dosages (13.5 to 14 mg/week). At 6 months, after adjustment on age, gender, and baseline disease activity parameters (higher in the RA group), only two of the six efficacy criteria were significantly better in the RA group than in the PsA group: the tender joint count (−3.9 vs. −3.2; P = 0.04) and the DAS28 (−1.38 vs. −1.13; P = 0.006). No significant differences were found for the swollen joint count, ESR, CRP, physician's global assessment, function, or quality of life. The 2-year methotrexate continuation rates were similar in the PsA and RA groups (65 and 66%, respectively). This comparison provides indirect evidence that methotrexate attenuates the inflammation and improves quality of life in patients with PsA.

Given the pathophysiological differences between PsA and RA, the clinical efficacy of methotrexate may differ between these two conditions. Data from RCTs in patients with PsA do not support combining methotrexate with a TNFα antagonist, either etanercept [21] or adalimumab [22]. A study of the Danish registry DANBIO indicates that concomitant methotrexate therapy in PsA only marginally improves the effect of TNFα antagonist therapy on the ACR20 response and has no benefits on the ACR50 or ACR70 response or on the treatment continuation rate [23]. In the Norwegian registry NOR-DMARD [24], concomitant methotrexate therapy did not improve the efficacy of TNFα antagonists in PsA. The combination increased the treatment continuation rate for infliximab but not for etanercept; for adalimumab, a non-significant increase in this parameter was apparent.

The evidence supporting a structural effect of methotrexate in PsA is extremely weak. In the prospective Toronto cohort of patients with peripheral erosive disease, the percentages of patients with radiographic disease progression after 2 and 4 years were significantly higher with methotrexate than with TNFα antagonist therapy [25]. A study [26] evaluated the progression of erosions and peripheral osteophytes after 1 year of methotrexate (n = 13) or TNFα antagonist therapy (n = 28). Although both treatments improved the clinical parameters and decreased the progression of erosions at 1 year, neither prevented the development of osteophytes.

Methotrexate remains one of the most widely used conventional DMARDs in PsA. That definite proof of efficacy is lacking cannot be construed as proof of inefficacy. Well-designed studies are needed to compare methotrexate not only to placebo therapy, but also to TNFα antagonists or other biotherapies (e.g., anti-IL12/23) used alone or with methotrexate. The issue of the optimal methotrexate dosage may be of paramount importance. The effects of methotrexate on quality of life deserve evaluation. Also of interest would be detailed assessments of the effects of methotrexate on specific manifestations such as enthesitis and dactylitis, as well as on structural disease progression.

Until new data become available, methotrexate should remain among the preferred DMARDs for chronic peripheral joint involvement in patients with PsA.

Section snippets

Disclosure of interest

The author declares that he has no conflicts of interest concerning this article.

References (26)

  • Z. Ash et al.

    A systematic literature review of drug therapies for the treatment of psoriatic arthritis: current evidence and meta-analysis informing the EULAR recommendations for the management of psoriatic arthritis

    Ann Rheum Dis

    (2012)
  • G.H. Kingsley et al.

    A randomized placebo-controlled trial of MTX in PsA

    Rheumatology

    (2012)
  • M. Dougados

    Methotrexate in peripheral spondyloarthritis including psoriatic arthritis: a need for further evaluation

    Rheumatology

    (2012)
  • Cited by (3)

    • Treatment response, drug survival and safety of anti-tumour necrosis factor α therapy in 193 patients with psoriatic arthritis: A twelve-year "real life" experience

      2015, Joint Bone Spine
      Citation Excerpt :

      However, cardiovascular comorbidities in PsA have been the subject of numerous publications illustrating the need to manage these comorbidities in routine clinical practice [31,32]. The value of concomitant methotrexate and anti-TNFα therapy remains controversial [33]. No advantage of co-prescription of cs-DMARDs was demonstrated in our study, but only a few patients received this co-prescription.

    View full text