Methotrexate Maintenance After Initiation of Biological or Targeted Synthetic DMARDs in Rheumatoid Arthritis: Results from the 2-Year Longitudinal Prospective Non-interventional STRATEGE2 Study

This longitudinal, prospective, non-interventional, multicentre study conducted in France examined MTX therapeutic strategies at the initiation of b/tsDMARDs for the treatment of RA in real-world clinical settings. At baseline, the 173 participants had been diagnosed with RA for approximately 5.6 (7.3) years, with a DAS28 score of 4.3 (1.2). Approximately two-thirds of participants had been receiving MTX therapy for over 1 year, with a mean weekly dose of 18.8 (4.2) mg, and 72.3% were treated via subcutaneous administration. At the time of b/tsDMARD initiation, 97.7% of participants continued MTX therapy. After 1 year, 83.2% of participants remained on MTX, with 39.9% (95% CI 32.5–47.6) maintaining the same treatment modalities. MTX discontinuation was primarily attributed to participant preferences or adverse events. By the 2-year follow-up, the mean change from baseline in the DAS28 score was − 2.0 (1.3), with 66.0% of participants achieving remission. According to EULAR criteria, 65.2% of participants demonstrated a good response to treatment, 19.1% had a moderate response, and 15.6% had an inadequate response.

At baseline, prior to b/tsDMARD initiation, most participants were receiving MTX via subcutaneous injection, while nearly one-third were using oral administration. These participants may have benefited from optimising their treatment modalities before considering b/tsDMARD therapy. This is supported by real-world data from the ESPOIR cohort study, which reported 3–4 times higher remission rates and improved functional outcomes after MTX optimisation [16]. Additionally, a recent study by Jung et al. suggested that incomplete utilisation of MTX and other csDMARDs may impede the achievement of remission or low disease activity [17]. All national and international treatment guidelines—including the recommendations of the French Society for Rheumatology (SFR) from 2014 [7], 2019 [4], and 2024 [11], the EULAR recommendations (2022 update) [10], and the 2021 ACR guideline [18]—consistently advise that if a patient demonstrates an inadequate response following MTX optimisation, a combination therapy with MTX and a b/tsDMARD should be initiated, as maintaining MTX is the preferred strategy. In the STRATEGE2 study, three-quarters of participants continued MTX therapy 2 years after initiating b/tsDMARD, regardless of any changes in treatment modalities. This outcome aligns with the SFR guidelines available at the time, which recommended prescribing b/tsDMARD in combination with MTX [4].

Firstly, within 2 years after initiating b/tsDMARD, rheumatologists maintained the initial MTX treatment modalities for one-third of participants. Factors associated with this decision-making process were investigated but only a few stood out. Participant age was a factor associated with the maintenance of MTX treatment modalities. Younger patients with a recent RA diagnosis may be more inclined to request MTX discontinuation, whereas older patients typically exhibit better treatment adherence, as they are more accustomed to it. Furthermore, a previous study has shown that older patients have a reduced risk of suspending b/tsDMARD treatment [19]. Finally, perceived disease activity reported by participants was also identified as a contributing factor. It can be argued that patients who sense a flare are more likely to advocate for the continuation of MTX to manage their symptoms effectively.

Secondly, rheumatologists reduced the dosage, adapted the route of administration of MTX, or both, for 40% of participants within 2 years of initiating targeted therapy. When remission is achieved with combination therapy, the question arises as to whether MTX or b/tsDMARD therapy should be tapered or discontinued. Evidence from the TARA trial indicates no significant difference in clinical outcomes when tapering either the csDMARD or the bDMARD first. The proportion of patients experiencing a flare-up within 2 years was similar between the two groups: 61% (95% CI 50–71) following csDMARD discontinuation versus 62% (95% CI 52–72) following anti-TNF discontinuation (p = 0.840) [20]. Although tapering bDMARDs was previously recommended for cost-related considerations, this assumption is still being investigated [21]. According to the latest SFR guidelines (2024), progressive tapering of cs/b/tsDMARDs should only be considered once sustained remission is achieved without the use of corticosteroids [11]. However, the definition of remission varies depending on the assessment tool employed. Achieving remission based on the DAS28 score has been associated with sustained remission, improved quality of life, and reduced cardiovascular issues. As a result, DAS28 remission is recognised as the treatment target across all guidelines. In this study, a significant proportion of participants achieved DAS28 remission, with 53.3% at 12 months and 66.0% at 24 months of follow-up. In parallel, a clinically meaningful improvement in participants’ health was reported in half of the study population, consistent with the association between DAS28 remission and functional, clinical, and structural improvements [22]. Remission as defined by the DAS28 score is considered the most permissive criterion compared to stricter frameworks, such as the EULAR criteria, Clinical Disease Activity Index (CDAI), or Simplified Disease Activity Index (SDAI). This may explain why treatments for some participants remained unchanged, with no tapering, despite achieving DAS28 remission. EULAR recommendations suggest tapering either the background csDMARD or the targeted therapy (bDMARD), without a preference, for patients in DAS28 remission, as flare rates are comparable after 1 year of follow-up [23]. However, caution is warranted when tapering MTX, particularly when administered subcutaneously. Tapering subcutaneous MTX is associated with a higher risk of flare-ups and requires closer monitoring compared to tapering oral MTX [24]. Findings from the RETRO study further confirmed the ability to maintain DAS28 remission after tapering or discontinuing either biologic DMARDs or conventional synthetic DMARDs [25]. However, relapses were most likely to occur within the first 6 months following treatment adjustments and were associated with anticitrullinated peptide antibody-positive (ACPA) status [25].

Thirdly, in this study, approximately one-quarter of participants discontinued MTX at the 2-year follow-up. This was frequently a therapeutic adaptation requested by participants, with adverse events cited as the primary reason for discontinuing MTX therapy. A systematic review published in 2021 identified predictors of adverse events associated with MTX discontinuation, including rheumatoid factor status, body mass index, and HAQ scores [26]. MTX was also discontinued by participants attempting to conceive. However, the latest data suggest that MTX does not cause testicular toxicity, indicating that it is safe for men to continue treatment while attempting conception [27]. Remission was cited only twice as a reason for discontinuing MTX, likely due to the short-term follow-up period of 2 years, as sustained remission of at least 6–12 months is typically required before discontinuation can be considered as a therapeutic option. Additionally, there was a notable decline in the proportion of participants using anti-TNF as part of b/tsDMARD treatment strategies. Despite this decrease, anti-TNF therapies still accounted for half of the prescribed b/tsDMARDs at the 24-month follow-up visit. Between the initiation of the STRATEGE2 study and the 2-year follow-up visit, a safety alert was issued concerning the use of JAKi in patients with cardiovascular risks or those who smoke. During this time, tofacitinib became a recommended third-line treatment [10, 28, 29]. The decision to discontinue MTX is influenced by the level of disease activity observed during combination therapy, as demonstrated by a retrospective Canadian study examining MTX de-escalation patterns in patients with RA treated concomitantly with b/tsDMARDs [30]. In patients with an inadequate response to treatment, MTX withdrawal has been reported to be either less effective or comparable to maintaining combination therapy with anti-TNF and MTX [31]. For patients with a moderate response to bDMARD and MTX combination therapy, continuing MTX is preferable as it helps prevent radiographic progression. In contrast, for patients achieving low disease activity, MTX withdrawal does not significantly worsen disease activity or progression in the short term [30‐32]. However, when patients experienced relapse following MTX discontinuation, regaining remission became a prolonged process, with some patients only able to return to low disease activity. Therefore, tapering or discontinuing treatments should be approached gradually and with great caution.

Finally, one of the key benefits of using MTX in combination with targeted therapies is its ability to prevent the formation of antibodies against bDMARDs, thereby improving the survival of anti-TNF drugs. The CONCERTO study demonstrated that an MTX dose of 10 mg weekly is beneficial when combined with the anti-TNF drug adalimumab [33]. Indeed, anti-TNF monotherapy has been linked to a high risk of treatment failure. However, the immunogenicity-limiting effect of MTX may be questioned in older individuals, potentially due to a decline in immunogenicity associated with immuno-senescence in aging [34].

Concerning treatments adjunct to MTX, half of the participants were receiving CS at baseline, with a relatively high mean daily dose of 9.6 mg. While this dosage remained consistent throughout the study, the proportion of participants treated with CS decreased by 24 months. The long-term impacts of low-dose CS have been investigated over a 10-year period, revealing cumulative negative outcomes after more than 5 years of continuous use [35]. Furthermore, side effects were found to increase with each additional year of use [35]. CS therapy should ideally be prescribed for short-term use during the initiation or modification of csDMARD therapy, with the aim of tapering off CS as rapidly as clinically feasible. This approach minimises safety risks such as osteoporotic fractures, serious infections, and diabetes, which are exacerbated by higher doses and prolonged use [36]. In patients with early inflammatory arthritis, the initial prescribed dose of CS tends to be higher compared to its use as background treatment. Yet, when used in combination with DMARDs, CS therapy does not appear to increase mortality or hospitalisations rates [37]. Current recommendations indicate that the persistent need for CS therapy suggests inadequate efficacy of DMARD treatment. Furthermore, they advise against the concurrent use of CS alongside b/tsDMARDs, despite evidence supporting its benefits in RA management [36]. Potential reasons for maintaining CS therapy as a background treatment include its stimulant effect, its ability to address concomitant diseases, and the risk of withdrawal syndrome, which can lead to a resurgence of pain. Nevertheless, tapering and eventually discontinuing CS therapy has been demonstrated to be both feasible and safe [38].

A key limitation of this study lies within its observational nature, which relies on physician-guided treatment decision without standardised protocol. This approach may introduce potential selection bias, limit external validity, and imposes cautiousness with the generalization of the results. Furthermore, the study design restricts our ability to assess the direct impact of MTX modifications on clinical outcomes. These highlight the need for future analytical or longitudinal studies to validate and expand upon these findings. Another limitation of this study was the reliance on the DAS28 score, which may overestimate remission compared to more stringent criteria, such as the ACR/EULAR criteria (Boolean definition), SDAI, or CDAI [39, 40]. Additionally, detailed data on structural progression of RA were not monitored (e.g., radiologic surveillance frequency, date of the latest radiographic examination, and radiographic progression), and tolerance-related data were also not collected during the study period. Lastly, the final part of the enrolment period coincided with the coronavirus disease (COVID-19) pandemic (March 2020 to December 2020). A recent retrospective study evaluated the impact of COVID-19 on the initiation of biologics for chronic inflammatory diseases in France. The authors reported a marked increase in the prescription of IL-6 antagonists [41], medications thought at the time to offer potential benefits on COVID-19 outcomes [42]. Consequently, one cannot rule out that the COVID-19 pandemic may have impacted the selection of targeted therapy during the period covered by STRATEGE2.