Amiodarone and pulmonary toxicity in atrial fibrillation: a nationwide Israeli study

Abstract Background and Aims Amiodarone-related interstitial lung disease (ILD) is the most severe adverse effect of amiodarone treatment. Most data on amiodarone-related ILD are derived from periods when amiodarone was given at higher doses than currently used. Methods A nationwide population-based study was conducted among patients with incident atrial fibrillation (AF) between 1 December 1999 and 31 December 31 2021. Amiodarone-exposed patients were matched 1:1 with controls unexposed to amiodarone based on age, sex, ethnicity, and AF diagnosis duration. The final patient cohort included only matched pairs where amiodarone therapy was consistent throughout follow-up. Directed acyclic graphs and inverse probability treatment weighting (IPTW) modelling were used. Patients with either prior ILD or primary lung cancer (PLC) were excluded. The primary outcome was the incidence of any ILD. Secondary endpoints were death and PLC. Results The final cohort included 6039 amiodarone-exposed patients who were matched with unexposed controls. The median age was 73.3 years, and 51.6% were women. After a mean follow-up of 4.2 years, ILD occurred in 242 (2.0%) patients. After IPTW, amiodarone exposure was not significantly associated with ILD [hazard ratio (HR): 1.45, 95% confidence interval (CI): 0.97, 2.44, P = 0.09]. There was a trivial higher relative risk of ILD among amiodarone-exposed patients between Years 2 and 8 of follow-up [maximal risk ratio (RR): 1.019]. Primary lung cancer occurred in 97 (0.8%) patients. After IPTW, amiodarone was not associated with PLC (HR: 1.18, 95% CI: 0.76, 2.08, P = 0.53). All-cause death occurred in 2185 (18.1%) patients. After IPTW, amiodarone was associated with reduced mortality risk (HR: 0.65, 95% CI: 0.60, 0.72, P < 0.001). The results were consistent across a variety of sensitivity analyses. Conclusion In a contemporary AF population, low-dose amiodarone was associated with a trend towards increased risk of ILD (15%-45%) but a clinically negligible change in absolute risk (maximum of 1.8%), no increased risk of PLC, and a lower risk of all-cause mortality.