Incidence and Predictors of Atrial Fibrillation in Cardiac Sarcoidosis: A Multimodality Imaging Study

BackgroundIn cardiac sarcoidosis (CS), the risk and predictors of new-onset atrial fibrillation (AF) are poorly known.ObjectivesThe authors evaluated the incidence and characteristics of AF in newly diagnosed CS.MethodsThe authors studied 118 patients (78 women, mean age 50 years) with AF-naive CS having undergone cardiac ¹⁸F-fluorodexoyglucose positron emission tomography (¹⁸F-FDG PET) at the time of diagnosis. Details of patient characteristics and medical or device therapy were collected from hospital charts. The PET scans were re-analyzed for presence of atrial and ventricular inflammation, and coincident cardiac magnetic resonance (CMR) studies and single-photon emission computed tomography (SPECT) perfusions were analyzed for cardiac structure and function, including the presence and extent of myocardial scarring. Detection of AF was based on interrogation of intracardiac devices and on ambulatory or 12-lead electrocardiograms.ResultsAltogether 34 patients (29%) suffered paroxysms of AF during follow-up (median, 3 years) with persistent AF developing in 7 patients and permanent AF in 4. The estimated 5-year incidence of AF was 55% (95% CI: 34%-72%) in the 39 patients with atrial ¹⁸F-FDG uptake at the time of diagnosis vs 18% (95% CI: 10%-28%) in the 79 patients without atrial uptake (P < 0.001). In cause-specific Cox regression analysis, atrial uptake was an independent predictor of AF (P < 0.001) with HR of 6.01 (95% CI: 2.64-13.66). Other independent predictors were an increased left atrial maximum volume (P < 0.01) and history of sleep apnea (P < 0.01). Ventricular involvement by PET, SPECT, or CMR was nonpredictive. Symptoms of AF prompted electrical cardioversion in 12 patients (35%). Three of the 34 patients (9%) experiencing AF suffered a stroke versus none of those remaining free of AF.ConclusionsIn newly diagnosed CS, future AF is relatively common and associated with atrial inflammation and enlargement on multimodality cardiac imaging.