Towards personalized treatment: predictors of short-term HAQ response in recent-onset active rheumatoid arthritis are different from predictors of rapid radiological progression

Abstract

Objective: Personalized treatment depends on the treatment goals. Current prediction models to guide initial treatment choices focus on radiological damage progression. However, for some patients this outcome is less relevant, whereas short-term functional ability is relevant to all. Do these various treatment goals share the same predictors?Methods: Data for 497 patients from the Dutch Behandel Strategieen (BeSt) study of treatment strategies for early rheumatoid arthritis (RA), randomized to initial monotherapy or combination therapy, were used. Predictors of short-term functional disability [Health Assessment Questionnaire (HAQ) score ≥ 1 after 3 months of treatment] were identified with logistic regression analyses. Predicted risks of a HAQ score ≥ 1 were determined for each treatment group and for each subpopulation.Results: At baseline, 76% of patients had a HAQ score ≥ 1 (mean 1.7 ± 0.5). After 3 months of treatment this score was achieved by 40% (mean HAQ score 1.5 ± 0.5). Baseline HAQ score, pain, the Ritchie Articular Index (RAI), and treatment group were significant independent predictors for a HAQ score ≥ 1; the presence of rheumatoid factor (RF), anti-cyclic citrullinated peptide (anti-CCP) antibodies, and baseline radiological damage were not. With cut-offs of 35% and 60%, the risk of a HAQ score ≥ 1 was high for 47% and low for 20% of the patients treated with initial monotherapy. Risks were markedly reduced in the combination therapy groups, also in unfavourable risk profiles.Conclusion: In recent-onset active RA, baseline HAQ score, pain, and initial treatment are predictors for a HAQ score ≥ 1 after 3 months. Known predictors of radiological damage were not predictive of short-term functional disability. The choice of the best initial treatment thus depends on the relevance of various outcome measures for an individual patient.