Objectives To compare the effectiveness and safety of rituximab alone or in combination with either methotrexate or leflunomide.
Methods 10 European registries submitted anonymised datasets with baseline, 3, 6, 9 and 12-month clinical data from patients who started rituximab.
Results 1195 patients were treated with rituximab plus methotrexate, 177 with rituximab plus leflunomide and 505 with rituximab alone. Significantly more patients achieved a European League Against Rheumatism good response at 6 months when treated with rituximab plus leflunomide (29.1%) compared with rituximab plus methotrexate (21.1%) and rituximab alone (19.3%; p=0.02 and p=0.01, respectively). Similar results were observed at 12 months. Adverse events occurred in 10.2%, 13.2% and 13.9% of patients on rituximab plus leflunomide, rituximab plus methotrexate and rituximab alone, respectively.
Conclusions Leflunomide is an effective and safe alternative to methotrexate as concomitant treatment with rituximab. Slightly better results were obtained by the combination of rituximab and leflunomide than rituximab and methotrexate, raising the possibility of a synergistic effect of leflunomide and rituximab.
Statistics from Altmetric.com
Rituximab (Mabthera, Rituxan) has been approved for the treatment of rheumatoid arthritis (RA). Randomised controlled clinical trials with rituximab have shown significant clinical improvements for patients who failed tumour necrosis factor (TNF) inhibition and biological agent naive patients.1,–,3 In all trials rituximab was administered in combination with methotrexate.
Methotrexate and leflunomide are synthetic disease-modifying antirheumatic drugs (DMARD) that inhibit purine and pyrimidine synthesis, respectively. Both drugs have been proved effective for the treatment of RA,4,–,7 but the exact pharmacological mechanisms of action are still unclear. Leflunomide is often used in clinical practice as an alternative DMARD, either alone or in combination with biological agents for patients intolerant to methotrexate, but limited data on the efficacy and safety of combination treatment with leflunomide and anti-TNF or other biological agents from clinical or epidemiological studies are available.8
Controlled clinical trials are the major source of information on the efficacy and safety of medications. However, since the inclusion of patients in these studies is governed by strict inclusion and exclusion criteria, the study populations can differ from real-life patients in several respects. We know from experience that a significant number of patients do not tolerate methotrexate, but the vast majority of clinical trials conducted to assess the efficacy of rituximab used methotrexate as anchor therapy. Furthermore, rituximab is sometimes used alone in the case of patients intolerant to several synthetic DMARD. There is therefore a need for evidence of the effectiveness and safety of rituximab either alone or in combination with synthetic DMARD other than methotrexate.
Patients and methods
The European Collaborative Registries for the Evaluation of Rituximab in Rheumatoid Arthritis (CERERRA) is an investigator-led, industry-supported initiative with the aim of evaluating the clinical aspects of rituximab use in patients with RA (K Chatzidionysiou, E Lie, E Nasonov, et al, unpublished observation). All 10 participating European registries submitted fully anonymised datasets with baseline characteristics of all RA patients who had been treated with rituximab. Levels of rheumatoid factor and anti-cyclic citrullinated peptide (CCP) were determined by local laboratories and local cut-off values for positivity were applied. Disease activity markers and disability score at baseline and after 3, 6, 9 and 12 months were also provided (see supplementary data, available online only).
Patients were separated into three groups: rituximab plus methotrexate, rituximab plus leflunomide and rituximab alone. A small number of patients was excluded (see supplementary data, available online only).
Overall differences between the groups for baseline values were tested with analysis of variance (ANOVA). For those variables that ANOVA showed to have significant heterogeneity pairwise comparison was performed with the two-tailed Student's t test (for normally distributed continuous data) and the χ2 test (for categorical variables). The univariate and multivariate logistic regression analyses are described elsewhere (see supplementary data, available online only).
Among the 2265 RA patients who started treatment with rituximab, 1195 were treated with rituximab plus methotrexate, 177 with rituximab plus leflunomide and 505 with rituximab alone. The baseline characteristics of the patients in the three treatment groups are shown in table 1. The majority of patients were treated with 2×1000 mg rituximab according to recommendations: 91.5% of patients in the rituximab plus methotrexate, 91.4% in the rituximab plus leflunomide and 86.4% in the rituximab monotherapy group (the difference being significant only between rituximab plus methotrexate and rituximab monotherapy, p=0.003). The rest of the patients received 2×500 mg rituximab. The mean dose (±SD) of methotrexate was 14.4±5.4 mg a week. The majority of patients on rituximab plus leflunomide (95.1%) were treated with 20 mg leflunomide a day (mean±SD daily dosage 17.6±3.6 mg).
Patients on rituximab plus leflunomide achieved a greater reduction in disease activity score in 28 joints (DAS28) values from baseline to the 6 and 12-month assessments compared with rituximab plus methotrexate and rituximab alone (figure 1A). At 6 months the mean±SD ΔDAS28 from baseline in the rituximab plus leflunomide, rituximab plus methotrexate and rituximab-alone groups were 2.1±1.3, 1.9±1.5 and 1.7±1.5, respectively (p=0.02 for rituximab plus leflunomide vs rituximab alone). At 12 months ΔDAS28 from baseline in the rituximab plus leflunomide, rituximab plus methotrexate and rituximab-alone groups were 2.2±1.6, 1.8±1.5 and 1.7±1.3, respectively (p=0.06 for rituximab plus leflunomide vs rituximab plus methotrexate and p=0.05 for rituximab plus leflunomide vs rituximab alone). The mean±SD health assessment questionnaire (ΔHAQ) from baseline in the rituximab plus leflunomide, rituximab plus methotrexate and rituximab-alone groups were 0.5±0.7, 0.5±0.6 and 0.4±0.6, respectively, at 6 months, and 0.5±0.6, 0.4±0.6 and 0.3±0.6 at 12 months, respectively. No significant differences were observed in the HAQ reductions at 6 and 12 months between the treatment groups.
Significantly more patients achieved a European League Against Rheumatism (EULAR) good response at 6 months when treated with rituximab plus leflunomide (29.1%) compared with rituximab plus methotrexate (21.1%) and rituximab alone (19.3%), p=0.02 and p=0.01, respectively (figure 1B). At 12 months an even higher percentage of good responders was observed in the rituximab plus leflunomide group (42.6%), while the percentage of good responders remained stable in the rituximab plus methotrexate and rituximab-alone groups (p=0.001 and p=0.08 respectively; figure 1B). The number of good responders was significantly higher in patients treated with rituximab alone than in the rituximab plus methotrexate group (p=0.04).
The combination rituximab plus leflunomide was significantly associated with a good EULAR response compared with rituximab plus methotrexate and rituximab monotherapy in a univariate regression analysis at 6 and 12 months adjusted for age and sex. In the multivariate analysis model adjusted for age, sex and country, a lower number of previous biological agents and concomitant leflunomide remained predictive at 6 months (table 2). At 12 months concomitant leflunomide and positive anti-CCP were significant predictors of EULAR good response to rituximab treatment (table 2).
Fewer patients with rituximab plus leflunomide required retreatment during the first 12 months (21.5%) compared with rituximab plus methotrexate (31.9%) or rituximab monotherapy (27.9%), but these differences were not statistically significant.
Adverse events occurred in 10.2%, 13.2% and 13.9% of patients in rituximab plus leflunomide, rituximab plus methotrexate and rituximab monotherapy, respectively. No difference in the infection rate was observed (6.2% for rituximab plus leflunomide, 6.6% for rituximab plus methotrexate and 7.9% for rituximab alone). One death was reported due to aspiration pneumonia in the rituximab plus methotrexate group. In the same group, one patient was diagnosed with prostate cancer 3 months after the initiation of rituximab and one patient had an acute myocardial infarction after 6 months. In the rituximab-alone group, one acute myocardial infarction and one haemorrhagic stroke were reported. A stroke was also reported in the rituximab plus leflunomide group, 6 months after the start of rituximab treatment.
The results of this large observational cohort of RA patients support the findings of previous smaller studies reporting good results with the combination of rituximab and leflunomide.9 10 At both 6 and 12 months a significantly greater number of patients treated with rituximab plus leflunomide achieved a EULAR good response than patients treated with rituximab plus methotrexate and rituximab alone, and fewer patients in the rituximab plus leflunomide group required retreatment during the first year. The latter observation may suggest a longer duration of response to therapy, but this difference was not statistically significant and the criteria for retreatment were not defined according to any specific protocol. However, a consensus had been published guiding physicians about how rituximab could be used according to available evidence and expert opinion.11 Co-medication with leflunomide did not lead to more adverse events. These findings thus indicate that leflunomide is a good alternative to methotrexate when used in combination with rituximab in RA.
Studies have shown that leflunomide has an impact on B-cell proliferation and cell cycle progression, thus suppressing B-cell antibody responses.12 13 These findings may provide a rationale regarding the mechanism of action of leflunomide in RA with a possible synergistic effect when used in combination with rituximab. Of note, methotrexate and/or TNF antagonists were not able to decrease the frequency of autoreactive B-cell clones, indicating that these anti-inflammatory therapies are not able to correct the presence of defective B-cell tolerance.14
We did not observe any difference in efficacy between patients treated with rituximab plus methotrexate or rituximab alone. Owczarczyk et al15 came to a similar conclusion in a small study with 40 patients.
The strengths of our study are the inclusion of a large number of patients from 10 participating countries and the ‘real-life’ character of the study, which provided the possibility to examine the combination of rituximab with other DMARD apart from methotrexate. This analysis has potential limitations inherent to the analysis of observational data. Almost all of the patients treated with leflunomide had either potential contraindications to methotrexate treatment or had previously not tolerated methotrexate, which could result in selection bias or confounding by indication. However, confounding by indication would most likely bias the results towards the null, as leflunomide is usually prescribed to more difficult patients with either problems of tolerance or inadequate response to methotrexate. While we could adjust our analysis for many important disease characteristics, we cannot exclude the possibility of residual confounding or confounding by unmeasured factors. Missing data are another concern with observational studies. Nevertheless, the proportion of patients at these time points with available effectiveness data was similar for the rituximab plus methotrexate group (57.5% and 31.9%) and the rituximab plus leflunomide group (51% and 29.7%) at 6 and 12 months, respectively. The absence of radiological data was also a limitation of this study. The mean dose of methotrexate in the rituximab plus methotrexate group was relatively low (14.4 mg/week). Higher doses could lead to better results, but they are not always well tolerated.
Concomitant treatment with leflunomide remained significant in the multivariate analysis, which was adjusted for country, as well as for age, sex and the number of previous biological agents, thus allowing us to draw more robust conclusions. The weaknesses of observational studies have to be put in the perspective that no optimal randomised controlled trial has been performed to address the effectiveness and safety of co-therapy with synthetic DMARD apart from methotrexate in rituximab-treated patients. In addition, this study provides important information on patients intolerant to various DMARD who can be treated with rituximab monotherapy.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Funding The study was funded by industry (Roche).
Competing interests None.
Ethics approval The study was approved by different ethics committees in the different countries involved.
Provenance and peer review Not commissioned; externally peer reviewed.
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.