Predictors for Treatment Success and Population Vary for Rheumatoid Arthritis

A novel study provides a direct comparison of change, response, and state measures as predictors of rheumatoid arthritis (RA) outcomes.

Baseline disease activity markers in rheumatoid arthritis (RA) are significantly associated with state and change measures, but no prediction was confirmed for response margins, according to study findings published in the Annals of the Rheumatic Diseases.

Investigators conducted a systematic analysis of core set variables and derived indices to determine their relationship with achieving disease activity outcomes according to change, response, and state measures. They included data from 3 clinical trials in RA.

A total of 821 patients from the ASPIRE trial (patients with early RA, mean ± SD of disease duration 0.83 ± 0.7 years; 71.2% women; mean age 49 ± 13 years), 330 patients from the ATTRACT trial (established RA; mean disease duration 10.3 ± 8.1 years; 77.6% women; mean age 53 ± 12 years), and 573 patients from the GO-BEFORE trial (mean disease duration 3.5 ± 5.2 years; 83.4% women; mean age 50 ± 12 years). The main analysis only included 532 patients from the methotrexate (MTX) + infliximab (IFX) arm in the ASPIRE trial.

The researchers obtained the demographic characteristics and measures of disease activity from the original data sets, including tender joint counts, swollen joint counts, patient global and pain assessment, evaluator global assessment, erythrocyte sedimentation rate, C reactive protein (CRP), and Health Assessment Questionnaire (HAQ) scores as functional measures.

The study authors focused on Simplified Disease Activity Index (SDAI70) for response outcome in the main analysis, and low disease activity (LDA) by SDAI for state outcome in the MTX + IFX arm of the ASPIRE trial at week 30. SDAI change from baseline was the change outcome.

A higher baseline CRP level predicted a SDAI70 response, as patients in higher baseline CRP quartiles achieved SDAI70 responses more frequently than those in lower quartiles. However, the percentage of patients who achieved LDA (including remission) according to SDAI had an increasing tendency with lower baseline values of all composite indices and core set measures, with the exception of acute phase reactants and HAQ. Regarding SDAI change from baseline to 30 weeks, all core set measures had a significantly higher mean change in higher quartiles.

Multivariate analysis of the predictors of state outcomes determined that the best prediction was a combination of disease duration of less than 2 years, male sex, and lower disease activity. No consistently significant predictors were found regarding prediction of response, except for CRP, as higher baseline CRP levels were associated with better responses.

“As a take home message from our report, the achievement of response target remains without accurate predictors,” stated the researchers. “Nevertheless, state targets can be predicted by lower baseline disease activity, along with gender and disease duration, and these predictors may be used to inform the definition of study entry criteria to improve feasibility of recruitment.”


Capelusnik D, Aletaha D. Baseline predictors of different types of treatment success in rheumatoid arthritis. Ann Rheum Dis. Published online October 4, 2021. doi: 10.1136/annrheumdis-2021-220853