Validation of the Health Assessment Questionnaire–Disability Index in Immune Checkpoint Inhibitor-Induced Inflammatory Arthritis

Objectives The use of immune checkpoint inhibitors (ICI) continues to expand across multiple tumor groups. ICIs can induce an inflammatory arthritis (ICI-IA), with significant impacts on patient function and quality of life. The Health Assessment Questionnaire–Disability Index (HAQ-DI) is a widely used measure of functional limitation in rheumatoid arthritis, but has not been studied in ICI-IA. This study assessed the reliability, construct validity, and responsiveness of the HAQ-DI in patients with ICI-IA. Methods Patients with ICI-IA enrolled in the Canadian Research Group of Rheumatology in Immuno-Oncology (CanRIO) multicenter prospective cohort who completed a baseline HAQ-DI were included. Demographic, clinical and laboratory data were extracted from the study database. Internal consistency reliability was assessed using Cronbach’s alpha coefficient (≥ 0.7 acceptable). Construct validity was examined by correlating HAQ-DI scores with other measures of disease activity. Responsiveness was evaluated using the standardized response mean (SRM) and effect size (ES) among participants with ≥ 3-point improvement on physician global assessment of disease activity (defined a priori) at 6 month follow-up. Results Ninety-six patients from 9 centers were included, 48 female, median age 67 years (range 39-84) with most common malignancies being melanoma (25%) and non-small cell lung cancer (20%). Polyarticular involvement was most frequent (64%) with median tender and swollen joint counts of 6 (range 0–36) and 2 (range 0–30), respectively at baseline. 54% of participants were on systemic glucocorticoids at baseline. Median HAQ-DI was 0.5 (range 0, 3) at baseline and 0.125 (range 0, 2.375) at 6 months (n=66). HAQ-DI scores were higher in those not on immunosuppression (median 0.5 vs 0.375). Floor and ceiling effects were 24% and 1%, respectively. Cronbach’s alpha ranged from 0.72-0.91 across the 8 HAQ-DI domains and was 0.94 for the 20 individual items, suggesting high internal consistency. The HAQ-DI had a significant positive correlation with C-reactive protein, swollen joint count, patient global score and physician global score (Table 1). In those with ≥ 3-point improvement on the physician global score at follow-up (n=23), responsiveness was moderate (SRM =1.4; ES = 0.3). Table 1: HAQ-DI correlations with clinical/demographic variables Conclusion In this prospective cohort study, the HAQ-DI was reliable, valid and sensitive to change in patients with ICI-IA. Over 50% of patients were on baseline immunosuppression likely impacting HAQ-DI scores and responsiveness to change assessment. Future research will explore content validity, the minimal clinically important difference in this population, and correlation to the physical functioning components of cancer-specific patient-reported outcome measures.