(P-524) Conduction Disorders/dysrhythmias and Hydroxychloroquine in Rheumatoid Arthritis

Background: Concerns regarding hydroxychloroquine (HCQ)-induced conduction disorder/dysrhythmia have emerged during its evaluation as a potential COVID-19 treatment.

Objectives: To determine whether HCQ (versus methotrexate, MTX) is associated with increased conduction disorders/dysrhythmias in rheumatoid arthritis (RA) in the pre-COVID-19 era.

Methods: Using MarketScan databases (Jan 2011-Dec 2018), we identified adult RA new users of HCQ or MTX. Time-zero was set as the date of the first prescription; subjects were required to have medical/pharmacy plan coverage one year before. Conduction disorders or dysrhythmia episodes, ascertained from the hospitalization diagnoses, were measured: i) within 30 days of time-zero (short-term effect); or ii) anytime after time-zero. Patients were followed until the first event or censored (for loss of health plan coverage, discontinuation/switching of initial HQN/MTX, or end of study). Cox regression was used to compare outcomes for HCQ versus MTX exposure. Models were adjusted for age, sex, calendar year at time-zero, health plan type (commercial or Medicare), comorbidities, previous ER visits/hospitalizations, and relevant medications (corticosteroids, biologics, DMARDs other than HCQ/MTX, and antimicrobials).

Results: The RA patients included 70,995 new users HCQ and 95,230 new users of MTX. Subjects were mostly women (77%), averaging 53.3 (standard deviation 13.2) years old. Within the 30-days after time-zero, 151 patients experienced conduction disorder/dysrhythmias, with an event rate of 1.01/100 p-y (95% confidence interval, CI= 0.78-1.31) among HCQ users vs 1.24/100 p-y (95% CI = 1.01-1.52) for MTX users. We did not detect a clear difference in 30-day risk for HCQ versus MTX (adjusted hazard ratio, HR= 0.84, 95% CI= 0.60–1.17). We identified 1,020 events occurring any time after time zero, with an event rate of 1.04/100 p-y (95% CI = 0.94-1.15) among HCQ users vs 1.08/100 p-y (95% CI = 1.00-1.16) for MTX users. We were also unable to detect a clear difference in HR for HCQ versus MTX (adjusted hazard ratio, HR 0.99, 95% CI= 0.87–1.13). Older age, Medicare plan type, and past ER visits/hospitalizations were associated with a higher risk of conduction disorders/dysrhythmia at both 30 days and in the longer-term. For corticosteroids, DMARDs, biologics and comorbidities, there were variations in HR estimates for short-term versus longer-term effects, potentially due to multicollinearity.

Conclusions: In this RA population, HCQ was not clearly associated with higher conduction disorders/dysrhythmia risk versus MTX. Greatest risk was seen in patients with older age, health plan type, and previous ER visits/hospitalization.