# Prognostic Value of Atrial Phasic Dysfunction by CMR Feature Tracking for New-Onset Atrial Fibrillation in Patients with Cardiac Sarcoidosis

## TL;DR

This study shows that reduced left atrial function seen on MRI scans can predict the development of atrial fibrillation in patients with cardiac sarcoidosis.

## Contribution

The study introduces left atrial reservoir strain from CMR feature tracking as a novel predictor of new-onset AF in cardiac sarcoidosis.

## Key findings

- LA reservoir strain was a strong predictor of AF in newly diagnosed cardiac sarcoidosis patients.
- LA reservoir strain predicted AF best in clinically silent sarcoidosis and less so in clinically manifest cases.
- Age and LA reservoir strain were independent predictors of AF in clinically silent sarcoidosis.

## Abstract

Background/Objectives: It is unclear whether assessment of phasic atrial function could improve risk stratification for new-onset atrial fibrillation (AF) in patients with newly diagnosed cardiac sarcoidosis (CS). We aimed to investigate the prognostic value of left atrial (LA) phasic dysfunction by cardiac magnetic resonance (CMR) for new-onset AF in newly diagnosed patients with CS. Methods: 78 patients with CS, without a prior history of AF, were studied using CMR feature tracking. Over a 4-year follow-up period, AF was documented by Holter monitoring and interrogation of intracardiac devices. Clinically silent CS was defined as CS in patients with biopsy-proven extracardiac sarcoidosis, with no cardiac symptoms, but with abnormalities on CMR or positron emission tomography consistent with CS. Results: Patients with clinically manifest CS were younger (mean age 56 vs. 51 years, p = 0.018), had poorer ventricular function, higher extent of atrial late gadolinium enhancement and significantly lower LA reservoir, conduit and booster function compared to patients with clinically silent CS. Over a 4-year follow-up period, 39% of patients with clinically manifest CS and 29.7% of patients with clinically silent CS developed AF. LA reservoir strain was a strong predictor of AF in the entire cohort. In subgroup analysis, age (HR 1.30, 95% CI 1.02–1.65, p = 0.030) and LA reservoir strain (HR 0.63, 95% CI 0.44–0.90, p = 0.011) were independent predictors of AF in patients with clinically silent CS, whereas baseline NT-proBNP (HR 1.003, 95% CI 1.001–1.006, p = 0.017) predicted AF in patients with clinically manifest CS. Conclusions: Reduced LA reservoir strain on CMR predicts new-onset AF in patients with newly diagnosed CS. The predictive value of LA reservoir strain is strongest in clinically silent CS and decreases with disease progression in clinically manifest CS.