Associations between Patient Global Assessment scores and pain, physical function, and fatigue in rheumatoid arthritis: a post hoc analysis of data from phase 3 trials of tofacitinib

Background Tofacitinib is an oral Janus kinase inhibitor for the treatment of rheumatoid arthritis (RA). We examined the degree to which Patient Global Assessment of Disease Activity (PtGA) was driven by patient-reported assessments of pain (Pain), physical function, and fatigue in patients receiving tofacitinib 5 mg twice daily or placebo, each with conventional synthetic disease-modifying antirheumatic drugs (csDMARDs). Methods This post hoc analysis used data pooled from three randomized controlled trials in csDMARD-inadequate responder (csDMARD-IR) patients (ORAL Scan: NCT00847613; ORAL Standard: NCT00853385; ORAL Sync: NCT00856544). Using subgroup analysis from 2 x 2 tables, associations between PtGA and Pain, Health Assessment Questionnaire-Disability Index (HAQ-DI), and Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F) at month 3 were evaluated using Pearson's Phi correlation coefficients. To support the main analysis, associations between select patient-reported outcomes (PROs) were also evaluated in csDMARD-naive (ORAL Start; NCT01039688) and biologic (b)DMARD-IR (ORAL Step; NCT00960440) patients. Results Across csDMARD-IR treatment groups, low disease activity (defined as PtGA <= 20 mm), and moderate (>= 30%) and substantial (>= 50%) improvements from baseline in PtGA were associated with mild Pain (Visual Analog Scale score <= 20 mm), and moderate (>= 30%) and substantial (>= 50%) improvements from baseline in Pain; lack of Pain improvement was associated with little/no improvement in PtGA. In contrast, large proportions of csDMARD-IR patients who reported PtGA improvements did not report HAQ-DI or FACIT-F scores >= normative values (<= 0.25 and >= 43.5, respectively) or changes in HAQ-DI or FACIT-F scores >= minimum clinically important difference (>= 0.22 and >= 4.0, respectively). Generally, PtGA and Pain outcomes were moderately-to-strongly correlated at month 3 in csDMARD-IR patients, with weaker correlations evident between PtGA and HAQ-DI/FACIT-F outcomes. Similar findings were generally evident in csDMARD-naive and bDMARD-IR patients. Conclusions This analysis supports the role of Pain as a key driver of PtGA in RA; physical function and fatigue play lesser roles in patients' perceptions of disease activity. These findings corroborate the importance of improved PROs and attainment of low symptom states for optimizing patient care.