Predictive Value of PET-Derived Myocardial Flow Reserve for New-Onset Atrial Fibrillation

BACKGROUND: Atrial fibrillation (AF) commonly coexists with coronary artery disease, heart failure, and coronary microvascular dysfunction, sharing common risk factors and pathophysiologic mechanisms. Whether myocardial flow reserve (MFR), a positron emission tomography-derived measure of epicardial and microvascular function, predicts incident AF remains uncertain. METHODS: We included 10 500 patients without prior AF who underwent positron emission tomography MPI at 2 centers. The primary outcome was new-onset AF during follow-up. Multivariable Cox models adjusted for demographics, comorbidities, resting left ventricular ejection fraction (LVEF), and LVEF reserve were used to evaluate the association between MFR and incident AF. Incremental predictive value was assessed using χ 2 , Akaike’s information criterion, Bayesian information criterion, C-index, and a likelihood ratio test comparing models with and without MFR. RESULTS: Over a median follow-up of 640 days, 288 patients (2.74%) developed AF. Impaired MFR (<2) was associated with a higher incidence of AF (20.5 versus 8.3 per 1000 person years). In multivariable analysis, MFR was a significant predictor of AF (hazard ratio per 1 interquartile range increase, 0.58 [95% CI, 0.47–0.72]; P <0.0001), whereas stress and rest myocardial blood flow were not. MFR improved risk prediction beyond clinical variables and LVEF (C-index increased from 0.675–0.726 when added to clinical variables and from 0.758–0.770 when added to clinical variables, LVEF, and summed stress score; P <0.001). Similar incremental improvement was observed when MFR was added to the Cohorts for Heart and Aging Research in Genomic Epidemiology–Atrial Fibrillation risk score (C-index increased from 0.638 to 0.711). The association between MFR and incident AF was numerically greater in patients with preserved EF, although the interaction term was not statistically significant. CONCLUSIONS: Impaired positron emission tomography-derived MFR is independently associated with increased risk of new-onset AF and enhances risk prediction beyond clinical factors and LVEF, suggesting a role for coronary microvascular dysfunction in early AF pathogenesis.