Brief Report: Clinical Trials Aiming to Prevent Rheumatoid Arthritis Cannot Detect Prevention Without Adequate Risk Stratification: A Trial of Methotrexate Versus Placebo in Undifferentiated Arthritis as an Example.

Abstract

Prevention of rheumatoid arthritis (RA) was the aim of several trials in undifferentiated arthritis (UA), with overall negative results. As preparatory work has revealed that only ∼30% of UA patients progress to having RA, we hypothesized that inclusion of patients without imminent RA could lead to false-negative results. We undertook this study to evaluate this hypothesis by reinvestigating the Probable Rheumatoid Arthritis: Methotrexate versus Placebo Treatment (PROMPT) trial (a 1-year course of methotrexate [MTX] versus placebo in UA) after excluding patients without a high risk of developing RA. A validated prediction model was used to determine the risk of RA in all patients included in the PROMPT trial. Patients with a prediction score of ≥8 (positive predictive value of ≥84% for developing RA) were considered to have a high risk of developing RA. The effect of a 1-year course of MTX during 5 years of follow-up was reinvestigated in these patients. Twenty-two of the 110 patients in the PROMPT trial had a high risk of RA at baseline. In the MTX arm, 6 of 11 patients (55%) developed RA, compared to 11 of 11 patients (100%) in the placebo arm (P = 0.011). Time to RA development was longer in the MTX arm than in the placebo arm (median 22.5 months versus 3 months; P < 0.001). Drug-free remission was achieved by 4 of 11 patients (36%) in the MTX arm compared to 0 of 11 patients (0%) in the placebo arm (P = 0.031). These beneficial effects of MTX were observed both in anti-citrullinated protein antibody (ACPA)-positive and in ACPA-negative UA patients with a high risk of RA, but not in UA patients without a high risk of RA. In retrospect, 43 of 110 patients fulfilled the American College of Rheumatology/European League Against Rheumatism 2010 classification criteria for RA at baseline. In addition, beneficial effects were observed only in patients with a high prediction score. A 1-year course of MTX delayed and prevented RA development in high-risk UA patients. This emphasizes the importance of adequate risk prediction in trials that aim to prevent RA.