Article Text

Extended report
Effectiveness and retention rates of methotrexate in psoriatic arthritis in comparison with methotrexate-treated patients with rheumatoid arthritis
  1. Elisabeth Lie1,
  2. Désirée van der Heijde1,2,
  3. Till Uhlig1,
  4. Marte S Heiberg1,
  5. Wenche Koldingsnes3,
  6. Erik Rødevand4,
  7. Cecilie Kaufmann5,
  8. Knut Mikkelsen6,
  9. Tore K Kvien1
  1. 1Department of Rheumatology, Diakonhjemmet Hospital, Oslo, Norway
  2. 2Department of Rheumatology, Leiden University Medical Center, Leiden, The Netherlands
  3. 3Department of Rheumatology, University Hospital Northern Norway, Tromsø, Norway
  4. 4Department of Rheumatology, St Olav Hospital, Trondheim, Norway
  5. 5Department of Rheumatology, Buskerud Central Hospital, Drammen, Norway
  6. 6Lillehammer Hospital for Rheumatic Diseases, Lillehammer, Norway
  1. Correspondence to Dr Elisabeth Lie, Department of Rheumatology, Diakonhjemmet Hospital, Box 23 Vinderen, N-0319 Oslo, Norway; elisabeth_lie{at}


Objective To examine the effectiveness and 2-year retention rates of methotrexate (MTX) in MTX naïve patients with psoriatic arthritis (PsA).

Methods Data on 430 patients with PsA participating in an ongoing longitudinal observational multicentre study in Norway were analysed. 1218 MTX naïve patients with rheumatoid arthritis (RA) from the same study served as a reference population. Assessments included measures of disease activity (28 joint counts, acute phase reactants), health status and utility scores. Six-month effectiveness data were compared both by crude analyses and with adjustments for age, sex and the respective baseline values. Two-year drug survival was compared by Kaplan–Meier and Cox regression analyses.

Results After 6 months of MTX treatment, both patients with PsA and those with RA improved in most disease activity measures and patient reported outcomes. In the adjusted analysis, patients with PsA tended to have less improvement, but changes were in the same range as in patients with RA. Two-year retention rates of MTX therapy in patients with PsA and RA were 65% and 66%, respectively, with only minor differences in reported reasons for discontinuation. Lower age, longer disease duration and higher Modified Health Assessment Questionnaire (MHAQ) score and patient global assessment were independent predictors of MTX termination within the first 2 years of treatment.

Conclusion In this real-life study, MTX treatment was associated with improvement in disease activity and health-related quality of life in patients with PsA after 6 months of treatment. Retention rates of MTX were similar in PsA and RA.

View Full Text

Statistics from

Psoriatic arthritis (PsA) is a chronic inflammatory joint disease prevalent in up to one-third of patients with psoriasis.1 2 It is increasingly being recognised as a severe form of arthritis, and studies have shown that erosive progression and disability are quite frequent.3,,5

The efficacy of tumour necrosis factor α (TNFα) inhibitors in PsA is well documented in randomised controlled trials (RCTs).6,,8 The introduction of these agents has represented major progress in the treatment of PsA and has also led to increased interest in the mechanisms and management of the disease. However, biological agents are costly and normally restricted to patients who have failed conventional therapy with one or more non-biological disease-modifying anti-rheumatic drugs (DMARDs). Thus, optimal use of DMARDs is important.

Methotrexate (MTX) is the most frequently used DMARD; it is a cornerstone in the treatment of rheumatoid arthritis (RA) and its efficacy in RA is well established.9 10 Recommendations for use of MTX have recently been published.11 In PsA, however, the documentation of efficacy from RCTs is limited even though MTX is also one of the most widely used DMARDs for this disease.12 13

The objective of this longitudinal observational study was to examine the effectiveness of MTX in MTX naïve patients with PsA. A group of MTX naïve patients with RA treated in the same setting served as a reference population. Our hypothesis was that effectiveness was similar between the groups. A second objective was to compare retention rates of MTX between patients with PsA and those with RA.

Patients and methods


Since December 2000, adult (>18 years of age) patients with inflammatory arthropathies starting a new DMARD regimen in five Norwegian rheumatology departments have been included in the NOR-DMARD register. The study design is a prospective multicentre longitudinal observational study with assessments at baseline, after 3, 6 and 12 months and yearly thereafter. The completeness of the register has been approximately 85%, while the remaining 15% were missed for inclusion, refused enrolment or were excluded due to language barriers or inclusion in ongoing RCTs. By August 2007, 7045 treatment courses in 4792 individual patients had been included. An overview of the register is given in figure 1.

Figure 1

Selection of patients from the Norwegian disease-modifying antirheumatic drug (NOR-DMARD) register for the present analyses. MTX, methotrexate; PsA, psoriatic arthritis; RA, rheumatoid arthritis.


MTX naïve patients with a diagnosis of PsA or RA enrolled in the NOR-DMARD register before January 2007 and starting treatment with MTX monotherapy were selected for the current analyses. Rheumatologists in the respective departments made the diagnoses of PsA and RA by clinical judgement (ie, ICD-10 codes L40.5 and M07.0, M07.1, M07.2 or M07.3, and M05.8, M05.9 or M06.0).


Scores from assessments at baseline and after 6 months of treatment were included in the analyses. Assessments included 28 swollen and tender joint counts (28-SJC and 28-TJC) performed by trained research nurses or rheumatologists, erythrocyte sedimentation rate (ESR), C-reactive protein, 100 mm visual analogue scales (VAS) for physician's assessment of global disease activity and patient's assessment of joint pain, fatigue and global disease activity (100 mm=worst disease activity, fatigue and pain, respectively), Modified Health Assessment Questionnaire (MHAQ),14 and the Medical Outcomes Study 36-item Short-Form Health Survey (SF-36).15 16 The SF-6D is a utility measure derived from the SF-36 with scores ranging from 0 (dead) to 1 (perfect health).17 The disease activity score-28 (DAS28) was calculated from 28-SJC and 28-TJC, ESR and patient's assessment of disease activity.18 For each group we calculated proportions achieving European League Against Rheumatism (EULAR) good and moderate response, as well as DAS28 low disease activity and remission.19

Statistical analyses

The baseline demographic and disease variables were compared between patients with PsA and those with RA using the χ2 test for categorical variables. Continuous baseline variables and 6-month changes from baseline were compared by the two-sample t test or Mann–Whitney U test, as appropriate. Changes at 6 months were also compared by analysis of covariance (ANCOVA) with adjustments for age, sex and the respective baseline values. Proportions in remission and low disease activity state were compared by the χ2 test. Kaplan–Meier and Cox regression analyses were used for comparing continuation of treatment (‘drug survival’).

In the main analyses we performed no imputation for missing data. However, we also analysed the data using the last observation carried forward (LOCF) approach to replace missing 6-month data for continuous variables if data from the 3-month assessment were available. All statistical tests were two-sided with the level of significance set at 0.05 without correction for multiple comparisons. Statistical analyses were performed using SPSS for Windows Version 15.0 (SPSS, Chicago, Illinois, USA).


Demographic data and medication

We identified 430 patients with PsA and 1218 with RA fulfilling the inclusion criteria (figure 1). As expected, the patients with PsA were somewhat younger and more often male than the patients with RA, and they had a lower number of previously used DMARDs (table 1). Weekly MTX doses were not significantly different between the groups (mean starting dose 10.5 mg, 12.6 vs 13.1 mg at 3 months and 13.7 vs 13.9 mg at 6 months for patients with PsA and RA, respectively). At all time points a significantly higher proportion of patients with RA than with PsA used systemic corticosteroids (42% vs 16% at baseline, 35% vs 12% at 3 months and 31% vs 12% at 6 months; p<0.001 for all comparisons), but mean prednisolone doses were similar (9.3 vs 9.0 mg/day at baseline, 6.1 vs 6.6 mg/day at 3 months and 6.9 vs 5.9 mg/day at 6 months; differences not statistically significant).

Table 1

Baseline demographic data*

Patient-reported outcomes

Group differences in baseline patient-reported measures were generally minor (table 2). Seventy-seven per cent of patients with PsA and 76% of patients with RA attended the 6-month follow-up visit. An additional 14% of patients with PsA and 15% of those with RA became eligible for the 6-month data analyses when the LOCF approach was applied. The remaining 9% of patients in both groups had terminated treatment or were lost to follow-up before the 3-month assessment.

Table 2

Baseline and 6-month changes in patient-reported outcomes*

Only the data from the completer analysis are presented, since the completer and LOCF analyses gave similar results (see table S1 in online supplement). Patient-reported outcomes generally improved in patients with PsA, but a significantly larger improvement was observed in MHAQ score, SF-36 bodily pain, SF-36 vitality and the SF-6D in patients with RA. The adjusted analysis gave similar results (table 2).

Disease activity measures

Baseline 28-SJC, 28-TJC, inflammatory markers, physician's global assessment and DAS28 were all significantly higher in patients with RA than in those with PsA (table 3). All measures improved significantly in the patients with PsA (table 3), but the crude 6-month improvements were generally larger in patients with RA. Only the differences in DAS28 and 28-TJC improvement remained statistically significant after adjustment for age, sex and the respective baseline values (table 3). Again the LOCF analyses gave similar results (see table S2 in online supplement).

Table 3

Baseline and 6-month changes in measures of disease activity*

Disease activity states and response rates

At baseline 7% of patients with PsA and 3% of those with RA had a DAS28 <2.6 (p=0.001), the most commonly used threshold for remission in RA, while 13% vs 8% (p=0.001) had a DAS28 ≤3.2 (low disease activity in RA). At 6 months, 24% of patients with PsA vs 27% with RA (p=0.40) had a DAS28 <2.6 while 42% vs 43% (p=0.91) had a DAS28 ≤3.2. EULAR good/moderate response was achieved by 24%/57% of patients with PsA compared to 33%/70% of patients with RA.


Drug survival over the first 2 years is shown in a Kaplan–Meier plot (figure 2). Addition of other drugs (eg, TNF inhibitors) to MTX due to insufficient efficacy of MTX monotherapy was recorded as termination of MTX. 17% of patients in both groups terminated MTX during the first 6 months of therapy. During this period, adverse events (AEs) were the main reason for drug termination in both groups (47% of cases) followed by lack of efficacy (28% for PsA vs 30% for RA). Between 6 and 12 months, lack of efficacy was the main reason for discontinuation (51% vs 53%), followed by AEs (30% vs 27%). Among patients discontinuing MTX because of AEs, nausea was the most commonly reported AE followed by elevated liver enzymes. A slightly higher proportion of patients with PsA than with RA discontinued MTX because of elevated liver enzymes (4.7% vs 3.3%; χ2 p=0.19), and the overall reported incidence of liver enzyme elevations was higher in patients with PsA (registered in 8.1% vs 5.5% of patients; χ2 p=0.05).

Figure 2

Crude 2-year survival of methotrexate therapy in patients with psoriatic arthritis (PsA) and rheumatoid arthritis (RA) (Kaplan–Meier analysis; log rank test p=0.54).

When including age, sex and diagnosis in a Cox regression model, lower age was a significant predictor of MTX termination with a hazard ratio (HR) of 0.90 (95% CI 0.84 to 0.95) per 10-year increase in age (p<0.001), while male sex (HR 1.06 (95% CI 0.89 to 1.27)) and RA diagnosis (HR 1.04 (95% CI 0.85 to 1.27)) were not statistically significant. We also tested DMARD naivety, disease duration and baseline patient global assessment, disability (MHAQ score) and prednisolone use in the model. The final model is shown in table 4. Lower age, longer disease duration and higher MHAQ score and patient global assessment were all independent predictors of MTX termination within the first 2 years of treatment.

Table 4

Adjusted hazard ratios (HRs) for discontinuation of MTX monotherapy during the first 2 years of treatment (Cox regression analysis)

Secondary analyses

Twelve-month effectiveness data were available for 57% of patients with PsA and 60% of those with RA (24%/21% of PsA/RA patients had discontinued MTX, 9%/6% were lost to follow-up and 10%/12% had not yet had their 1-year follow-up visit). A tendency towards additional improvement was observed in both groups (data not shown). We performed subanalyses on patients with disease duration <3 years, and the 6-month effectiveness did not differ substantially from the results presented for the overall groups (see tables S3 and S4 in online supplement).


MTX is widely used both in RA and PsA, and the documentation of efficacy is strong for RA but only minimal for PsA. In a small crossover RCT published in 1964, Black et al found that three repeated doses of intravenous or intramuscular high-dose (1–3 mg/kg) MTX was more effective than placebo in improving joint and skin symptoms as well as lowering ESR in patients with PsA.12 Twenty years later, Willkens et al conducted another small RCT and concluded that orally administered MTX in doses of 7.5–15.0 mg/week in PsA was not superior to placebo, except for the physician's global assessment.13 Scarpa et al recently demonstrated that MTX (intramuscular 10 mg/week) plus a daily full-dose non-steroidal anti-inflammatory drug (NSAID) was superior to a step-up strategy with the addition of MTX to daily NSAID at 3 months.20 The gold standard for assessing efficacy of a drug is a randomised trial with appropriate power calculations and with placebo or a presumably less effective drug as a control. However, because of the lack of such a trial, we chose to present the MTX data for PsA together with data on a reference RA population, as the efficacy of MTX in RA is well established. The limited number of RCTs on MTX in PsA supports the view that longitudinal observational studies may provide important information. In this study we have shown that exposure to MTX was associated with improvement in disease activity and patient-reported outcomes in PsA. The improvements in several of the generic patient-reported outcomes were slightly more pronounced in RA (table 2).

However, our study has several potential limitations. The study was designed for follow-up of inflammatory arthropathies, but outcome measures were particularly selected for RA. Several recent initiatives have addressed the development of diagnostic and classification criteria as well as elaboration and validation of assessment tools for PsA.21,,23 Ideally, a full set of domains for assessment would include skin involvement, dactylitis, enthesitis, spinal involvement and radiographic outcomes.24

We have included 28-joint counts, which are considered less appropriate for PsA than more extensive joint counts which better capture the joint distribution of the disease.22 The DAS28 is a disease activity score developed for RA, but it has been found to be useful also in clinical trials of PsA, and so have the EULAR response criteria.25 However, DAS28 is not validated for PsA, and the statistical properties of this index have been shown to be different for PsA than for RA.26 As seen in our study (table 3), DAS28 scores are expected to be lower in PsA than in RA as oligoarticular joint affection is not uncommon and PsA is generally associated with less elevation of inflammatory markers.25 Consequently, patients with low DAS28 values could also be considered candidates to start MTX owing to clinical manifestations which are not captured by DAS28. The difference in DAS28 levels will also influence the potential to achieve the EULAR good and moderate responses which incorporate both magnitude of improvement and disease activity state in the response criteria. Hence, we do not present any statistical comparison of these responses. Interestingly, the baseline levels for the patient-reported outcomes were generally very similar in the two diseases despite differences in age and sex distributions (table 3). Patients with PsA actually scored significantly worse on the patient global VAS at baseline. This could be due to the added burden of skin disease not captured by the selected disease measures. However, patients with PsA did not score worse for the mental components of SF-36.

For patients with PsA as well as RA, inclusion in the study was based on the clinical diagnosis made by the treating rheumatologist. It is likely that the majority of patients with RA fulfilled the 1987 American College of Rheumatology classification criteria for RA at the time of inclusion. It is less clear how many of the patients with PsA fulfilled the more recently developed CASPAR (ClASsification criteria for Psoriatic ARthritis) criteria.21 For the purpose of this study, we find inclusion by clinical diagnosis appropriate even though it would result in less homogenous patient populations, perhaps especially for PsA.

The proportion of patients on concomitant prednisolone therapy was higher among patients with RA at all time points. In most prednisolone users, prednisolone had already been initiated at the time when MTX was started, and the dose was tapered during the first 6 months in many patients. It is therefore unlikely that this treatment strongly interfered with the recorded effectiveness. Furthermore, the results did not change notably when we adjusted for the prednisolone dose in the ANCOVA analyses (data not shown), neither did prednisolone use affect MTX retention in the Cox regression analyses.

Missing data are a common problem in observational studies. We found that 17% of patients with PsA and RA terminated MTX within 6 months and that the reasons for termination were similar across the diagnoses. As might be expected, the completer analysis gave somewhat larger 6-month improvements for patients with PsA or RA than the LOCF analysis. Still, the relative relationship between the diagnoses with regard to magnitude of improvements was very similar to the LOCF analysis and p values were also of the same magnitude (see tables S1 and S2 in online supplement). The LOCF approach is one of the most widely used methods for replacing missing data in rheumatology trials, but this approach has several limitations. The method assumes that data are missing completely at random (which is rarely the case in clinical research), it reduces variability (which can increase precision and lead to smaller p values) and it of course assumes that values remain constant for the rest of the trial.27 28 By using LOCF, effectiveness may actually be overrated.

The mean 6-month changes for several of the outcome measures were modest, sometimes smaller than the identified thresholds for minimally clinically important differences. This low magnitude of improvement could at least partially be explained by the selection of outcome measures not tailored for PsA. For both PsA and RA, mean baseline DAS28 was considerably lower in this study than is usually the case in RCTs, which may reflect the real-life situation of this study in contrast to the selection of patients with predetermined levels of severe and active disease for clinical trials. A previous study from NOR-DMARD examined the comparative effectiveness of MTX and TNF inhibitors in PsA, and the magnitude of the current 6-month changes is in good concordance with these previously published results.29 Generally, the access to TNF inhibitors in Norway is good for both patients with RA and those with PsA, and there is no demand for a specific high disease activity level. Before initiation of a TNF inhibitor, patients have to have failed at least one DMARD and, in RA, this has to be MTX (unless contraindicated).

We found that AEs were the most common reason for discontinuation within the first 6 months, while lack of efficacy was the most frequently reported reason for the first year overall. One hundred and twenty-eight patients (34 PsA and 94 RA) discontinued MTX due to AEs in the first 6 months. Over the next 6 months the corresponding number of patients was 39 (8 PsA and 31 RA). Even though about 6% of patients were not eligible for 1-year follow-up, these data indicate that AEs leading to treatment termination seem to occur more frequently during the first months of treatment with MTX. It is a common belief that liver enzyme elevations during MTX use are more frequent in PsA than in RA, but few studies have formally addressed this issue.30 31 We found a slightly increased incidence of liver enzyme elevations in patients with PsA compared to patients with RA. However, this well known AE was reported in a small percentage of patients in both groups, which could be partly due to the short follow-up time for many of the patients and relatively moderate doses of MTX. General underreporting of AEs is another possible reason. There was no fixed scheme for measuring liver enzymes in this study, but common monitoring practice in Norway has been every other week for the first 3 months and then monthly controls.32

In summary, we found that MTX therapy was associated with improvement in disease activity and health-related quality of life in patients with PsA after 6 months of therapy. Importantly, the rates of drug termination and the reasons for it during the first 2 years were similar in PsA and RA. With the lack of good RCTs, these data support the notion that MTX is an effective treatment in PsA, but more evidence is needed to establish the role of MTX in PsA in the new era of biological therapies.


View Abstract

Supplementary materials

  • Web Only Data ard.2009.113308

    Files in this Data Supplement:


  • Funding Supported by Eastern Norway Regional Health Authority. The Norwegian Disease-Modifying Antirheumatic Drug study has received unrestricted grant support from Abbott, Amgen, Wyeth, Aventis, MSD, Schering-Plough/Centocor, BristolMyers Squibb, Roche and the Norwegian Directorate for Health and Social Affairs.

  • Competing interests Hans Bijlsma was the Handling Editor for this article.

  • Patient consent Obtained.

  • Ethics approval Patients gave written informed consent before participation. The study was approved by the regional ethical committee and by the Data Inspectorate.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.