Using the derived 28-joint disease activity score patient-reported components (DAS28-P) index as a discriminatory measure of response to disease-modifying anti-rheumatic drug therapy in early rheumatoid arthritis

Abstract

Background: The 28-joint disease activity score (DAS28) is a widely used measure to assess disease activity in rheumatoid arthritis (RA). The DAS28-P index, a derived proportion of the patient-reported components (joint tenderness and patient global assessment) within the DAS28, has been utilized as a discriminatory measure of non-inflammatory pain mechanisms in RA. This study aimed to evaluate the use of the DAS28-P index as a predictor of treatment response in early RA.

Methods: Patients with early RA enrolled in a supplemental fish oil clinical trial received a combination of disease-modifying anti-rheumatic drugs (DMARDs) according to a 'treat-to-target' protocol. First, consecutive measures of the DAS28-P index, derived from the DAS28-erythrocyte sedimentation rate (DAS28-ESR), at each visit over a 1-year period were estimated for each patient. Then, distinct subgroups of treatment responders based on the trajectories of the DAS28-P indices were identified using bivariate k-means cluster analysis. Data on baseline predictors as well as longitudinal outcomes of disease impact and DMARD use over a 1-year period and radiographic progression over a 3-year period were collected and analyzed using a random intercept, population-averaged generalized estimating equation model.

Results: 121 patients were included (74% female; mean age of 57; median of 16 weeks of active disease) and a 3-cluster model was identified-the 'Responders' group (n = 58; 48%), the 'Partial Responders' group (n = 32; 26%), and the 'Non-Responders' group (n = 31; 26%). The 'Partial Responders' group had consistently higher proportions of the DAS28-P index throughout the study period and had minimal radiographic progression over time, with the lowest joint erosion score of 0.9 [95% confidence interval (CI) 0.2, 1.6], observed at the 3-year follow-up. At 52 weeks, the methotrexate dose was higher for both 'Partial Responders' and 'Non-Responders' groups (18.5 mg [95% CI 15.5, 21.5] and 18.6 mg [95% CI 15.3, 21.8] respectively), when compared with the 'Responders' group (12.8 mg [95% CI 14.7, 20.9]).

Conclusions: Persistently high DAS28-P index scores are useful to distinguish poor patient global assessment and excessive treatment escalation in early RA, suggestive of underlying non-inflammatory pain contributing to higher disease activity score. Early identification of patients with discordant subjective and objective components of composite disease activity measures may allow better tailoring of treatment in RA.

Keywords: DAS28-P index; Pain; Patient global assessment; Rheumatoid arthritis.

Fig. 1

DAS28 and DAS28 component scores (predicted marginal means) during the first year of treatment for the three responder groups. A DAS28-ESR B DAS28—Objective Component C DAS28—Subjective Component D DAS28-P (an index defined by DAS28—Subjective Component/DAS28-ESR)

Fig. 2

Self-reported scores (predicted marginal means) during the first year of treatment for the three responder groups: A modified Health Assessment Questionnaire (mHAQ) disability B Fatigue (measured in 0–100 mm visual analogue scale) C Helplessness in coping with arthritis (measured in a 5–30 subscale scoring system in a Validated 5-item Rheumatology Attitudes Index (VALI-RAI))

Fig. 3

DMARD treatment (predicted marginal means) during the first year for the three responder groups: A Methotrexate (MTX) Dose (in milligram, mg of weekly dosing) B Leflunomide Use (proportion of patients)

Fig. 4

Total Erosion Scores (predicted marginal means) over 3 years of follow-up for the three responder groups