Intervention with methotrexate in patients with arthralgia at risk of rheumatoid arthritis to reduce the development of persistent arthritis and its disease burden (TREAT EARLIER): a randomised, double-blind, placebo-controlled, proof-of-concept trial
Doortje I Krijbolder,Marloes Verstappen,Bastiaan T van Dijk,Yousra J Dakkak,Leonie E Burgers,Aleid C Boer,Yune Jung Park,Marianne E de Witt-Luth,Karen Visser,Marc R Kok,Esmeralda T H Molenaar,Pascal H P de Jong,Stefan Böhringer,Tom W J Huizinga,Cornelia F Allaart,Ellis Niemantsverdriet,Annette H M van der Helm-van Mil
DOI: https://doi.org/10.1016/S0140-6736(22)01193-X
2022-07-23
Abstract:Background: Rheumatoid arthritis is the most common autoimmune disease worldwide and requires long-term treatment to suppress inflammation. Currently, treatment is started when arthritis is clinically apparent. We aimed to evaluate whether earlier intervention, in the preceding phase of arthralgia and subclinical joint inflammation, could prevent the development of clinical arthritis or reduce the disease burden. Methods: We conducted a randomised, double-blind, placebo-controlled, proof-of-concept-trial at the Leiden University Medical Centre, Leiden, Netherlands. Adults aged 18 years or older with arthralgia clinically suspected of progressing to rheumatoid arthritis and MRI-detected subclinical joint inflammation were eligible for enrolment across 13 rheumatology outpatient clinics in the southwest region of the Netherlands and randomly assigned (1:1) to a single intramuscular glucocorticoid injection (120 mg) and a 1-year course of oral methotrexate (up to 25 mg/week), or placebo (single injection and tablets for 1 year). Participants and investigators were masked to group assignment. Follow-up continued for 1 year after the end of the 1-year treatment period. The primary endpoint was development of clinical arthritis (fulfilling the 2010 rheumatoid arthritis classification criteria or involving two or more joints) that persisted for at least 2 weeks. Patient-reported physical functioning, symptoms, and work productivity were secondary endpoints, which were measured every 4 months. Additionally, the course of MRI-detected inflammation was studied. All participants entered the intention-to-treat analysis. This trial is registered with EudraCT, 2014-004472-35, and the Netherlands Trial Register, NTR4853-trial-NL4599. Findings: Between April 16, 2015, and Sept 11, 2019, 901 patients were assessed for eligibility and 236 were enrolled and randomly assigned to active treatment (n=119) or placebo (n=117). At 2 years, the frequency of the primary endpoint was similar between the groups (23 [19%] of 119 participants in the treatment group vs 21 [18%] of 117 in the placebo group; hazard ratio 0·81, 95% CI 0·45 to 1·48). Physical functioning improved more in the treatment group during the first 4 months and remained better than in the placebo group (mean between-group difference in Health Assessment Questionnaire disability index over 2 years: -0·09, 95% CI -0·16 to -0·03; p=0·0042). Similarly, pain (on scale 0-100, mean between-group difference: -8, 95% CI -12 to -4; p<0·0001), morning stiffness of joints (-12, -16 to -8; p<0·0001), presenteeism (-8%, -13 to -3; p=0·0007), and MRI-detected joint inflammation (-1·4 points, -2·0 to -0·9; p<0·0001) showed sustained improvement in the treatment group compared with the placebo group. The number of serious adverse events was equal in both groups; adverse events were consistent with the known safety profile for methotrexate. Interpretation: Methotrexate, the cornerstone treatment of rheumatoid arthritis, initiated at the pre-arthritis stage of symptoms and subclinical inflammation, did not prevent the development of clinical arthritis, but modified the disease course as shown by sustained improvement in MRI-detected inflammation, related symptoms, and impairments compared with placebo. Funding: Dutch Research Council (NWO; Dutch Arthritis Society).