Rheumatoid Arthritis Prediction Helps Prevention

July 11, 2017

With adequate risk stratification in a reanalysis of PROMPT data, the preventative benefits of methotrexate became apparent.

New treatments have dramatically improved rheumatoid arthritis disease outcomes, but most patients still require lifelong therapy. With the ultimate goal of avoiding the disease altogether, several innovative trials have attempted to prevent disease progression from undifferentiated arthritis to classifiable rheumatoid arthritis. Thus far, these trials have been unsuccessful.

In a recent study, however, a new approach to risk stratification has had significant results.

For high-risk patients with undifferentiated arthritis, a one-year course of methotrexate was associated with an absolute risk reduction of 45% in rheumatoid arthritis development.

In addition, rheumatoid arthritis development was significantly delayed in the methotrexate arm compared with the placebo arm and drug-free remission was achieved significantly more often with methotrexate.

Led by Leonie E. Burgers, MD, of Leiden University Medical Center in Leiden, The Netherlands, researchers writing in Arthritis & Rheumatology reached these conclusions after a postrandomization exclusion of uninformative patients from the Probable Rheumatoid Arthritis: Methotrexate versus Placebo Treatment (PROMPT) trial.

The original study, a prospective double-blind, randomized, placebo-controlled multicenter study, evaluated the effect of a 1-year course of methotrexate versus placebo in 110 patients with undifferentiated arthritis. After a follow-up period of 5 years, the original study found that methotrexate had no preventative effect.

However, because only about 30% of patients with undifferentiated arthritis ever progress to rheumatoid arthritis, researchers hypothesized that the inclusion of patients who were not at high risk for rheumatoid arthritis could lead to false-negative results.

The study

In the present study, researchers reanalyzed the PROMPT data using the well-validated Leiden prediction rule, a 9-item instrument that includes clinical and biomarker factors that result in a prediction score ranging from 0 to 14. The current study included those patients from PROMPT who had a prediction score of 8 or higher, which resulted in 22 high-risk undifferentiated arthritis patients. By chance, these high-risk patients had been equally randomized between the 2 treatment arms.

In the reanalysis, the primary outcome measurement was fulfillment of the American College of Rheumatology (ACR) 1987 revised criteria. The secondary outcome measure was the proportion of patients who achieved drug-free remission.

With adequate risk stratification in place, the benefits of methotrexate as a form of prevention became apparent. In the methotrexate arm, rheumatoid arthritis developed in only 6 of 11 patients (55%), compared with 11 of 11 patients (100%) in the placebo arm. Rheumatoid arthritis was also significantly delayed in the methotrexate arm, with a median of 22.5 months versus 3 months in the placebo arm.

Drug-free remission was also achieved significantly more often in the methotrexate arm (36% versus 0%). The beneficial effects of methotrexate were observed in both anti-citrullinated protein antibody (ACPA)–positive and in ACPA-negative patients at high risk for rheumatoid arthritis.

“The current findings suggest that the PROMPT trial yielded false-negative results due to the inclusion of [undifferentiated arthritis] patients without a high risk of developing [rheumatoid arthritis],” the authors concluded. “These data therefore illustrate how inclusion of uninformative patients can blur highly relevant study outcomes, such as prevention of [rheumatoid arthritis].”

Implications for future trials

Up to this point, none of the clinical trials performed in undifferentiated arthritis have used extensive risk stratification, leading researchers to speculate on whether the results of those trials might be different if only patients at high risk for rheumatoid arthritis were evaluated. Although selecting only high-risk patients could make trial recruitment even more challenging, the current data suggest it may positively affect the power of a study.

However, despite the successful reevaluation of the PROMPT data, the authors of an editorial that accompanied the article in Arthritis & Rheumatology cautioned that this was a very small study and used retrospective analyses that may introduce bias and therefore much larger studies will be needed to determine the best approaches for future prevention studies.

The editorial authors, from the University of Colorado Denver School of Medicine, recommended that “great efforts should be put into identifying the right set of inclusion criteria for clinical prevention trials, with these criteria being based on a sound understanding of the natural history of disease and its underlying biology.”

In addition, it isn’t yet clear how the findings from the PROMPT study will directly apply to prevention trials that target earlier stages of the disease, such as the prearthritis phase of arthralgia. Currently there are no validated models to accurately predict the risk of rheumatoid arthritis in patients who present with arthralgia, and the methods used in the PROMPT study, such as the Leiden score, may not be appropriate. Different strategies may also be necessary to identify patients with rheumatoid arthritis even earlier in the natural history of the disease.

Likewise, methotrexate may not be appropriate to use in the earlier stages of rheumatoid arthritis prevention because the drug has been optimized for the treatment of clinically apparent synovitis, a disease process that may be distinctly different from preclinical autoimmunity.

Studies will also be needed to address how changes in biomarkers or other measures over time can be used to assess the efficacy of an intervention, and targets may need to change to reflect underlying immunopathology rather than counts of tender and swollen joints.

Still, the reevaluation of the PROMPT data demonstrates the importance of adequate risk prediction models for preventing false-negative study results in the future. “It’s exciting that rheumatology is at the point where prevention in [rheumatoid arthritis] can be addressed more comprehensively,” the editorial authors concluded.

References:

Leonie E. Burgers, Cornelia F. Allaart, Tom W. J. Huizinga, et al. “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.” Arthritis Rheumatol. 2017;69:926-931. doi: 10.1002/art.40062. Epub 2017 Mar 31.

Kevin D. Deane, Christopher C. Striebich, V. Michael Holers. “Editorial: Prevention of Rheumatoid Arthritis: Now Is the Time, but How to Proceed?Arthritis Rheumatol. 2017;69:873-877. doi: 10.1002/art.40061. Epub 2017 Apr 11.