Background/Purpose: Current treatments for osteoarthritis (OA) are severely limited. Synovitis is prevalent in OA and is associated with pain. The slow-acting anti-rheumatic drug methotrexate (MTX) is the gold-standard treatment for synovitis in inflammatory arthritides. The primary aim of the PROMOTE trial was to determine the effectiveness of MTX versus placebo as an analgesic treatment for knee OA. 

Methods: PROMOTE was a multi-centre, randomised, placebo-controlled trial with 12 months follow-up. Participants with symptomatic (visual analogue scale (VAS) pain ≥4/10) and radiographic tibiofemoral knee OA, fulfilling clinical ACR criteria, were recruited across UK primary and secondary care. Participants were randomized on a 1:1 basis to MTX or placebo, in addition to ongoing usual care, with dose escalation from 10mg to 25mg over 8 weeks and maintenance at 25mg (or the highest tolerated dose) for the remainder of the study. The primary endpoint was average knee pain during the previous week (numerical rating scale [0-10], NRS) at 6-months. Secondary endpoints included WOMAC, quality-of-life and adverse events. Linear mixed models compared outcomes between groups on an intention-to-treat (ITT) basis. In a sub-study, contrast-enhanced MRI of the index knee was performed at baseline and 6 months.

Results: Of 207 patients screened, 155 participants (64% women, mean age 60.9 years, 50% K-L Grade 3-4) were randomized. Primary endpoint data at 6 months were available for 134 patients (86%); only ITT data are presented. At 6 months, average knee pain (as measured by NRS) was 6.2 in the placebo group and 5.1 in the MTX group, with a baseline adjusted treatment difference of -0.83 points (95% CI -1.55 to -0.10; p=0.025), equivalent to a standard effect size of 0.36. Statistically significant differences at 6 months were seen for WOMAC stiffness and physical function, but not pain. Treatment benefits were reduced by 12 months. 94 patients had analysable MRI data at baseline and 80 at 6 months; no change in synovial volume was found. Four serious adverse events were reported (MTX: 2 [not defined as possibly related], placebo: 2).

Conclusion: MTX added to usual care demonstrated significant reduction in knee OA pain at 6 months, and suggests improvements in WOMAC stiffness and function. Despite a moderate standard effect size, the treatment effect was smaller than thresholds considered clinically meaningful. Further analyses will explore predictors of response to understand if subsets with enhanced response can be identified.

We acknowledge Arthritis Research UK for funding support.

Table 1: Selected PROMOTE outcomes at baseline and 6 months follow-up (primary endpoint)