BACKGROUND: Current indications for surgery in patients with significant aortic regurgitation (AR) focus on symptoms and left ventricular dilation/dysfunction. However, prognosis is already reduced by this stage, and earlier identification of patients for surgery could be beneficial. Quantifying the regurgitation may help, but there are limited data on its link with outcome. Cardiovascular magnetic resonance (CMR) can accurately quantify AR, and we examined whether this was associated with the future need for surgery. METHODS AND RESULTS: One hundred thirteen patients with echocardiographic moderate or severe AR were monitored for up to 9 years (mean 2.6 ± 2.1 years) following a CMR scan, and the progression to symptoms or other indications for surgery was monitored. AR quantification identified outcome with high accuracy: 85% of the 39 subjects with regurgitant fraction >33% progressed to surgery (mostly within 3 years) in comparison with 8% of 74 subjects with regurgitant fraction ≤ 33% (P<0.0001); the area under the curve on receiver operating characteristic analysis was 0.93 (P<0.0001). This ability remained strong on time-dependent Kaplan-Meier survival curves. CMR-derived left ventricular end-diastolic volume >246 mL had good, although lower, discriminatory ability (area under the curve 0.88), but the combination of this measure with regurgitant fraction provided the best discriminatory power. CONCLUSIONS: High degrees of CMR-quantified AR were associated with the development of symptoms or other indications for surgery. Quantifying AR showed slightly better discriminatory ability than "gold standard" CMR ventricular volume assessment. This could provide a new paradigm for the timing of surgical intervention but requires confirmation in a clinical trial.
1452 - 1460
Adult, Aortic Valve Insufficiency, Disease Progression, Echocardiography, Female, Follow-Up Studies, Heart Valve Prosthesis, Humans, Kaplan-Meier Estimate, Magnetic Resonance Imaging, Male, Myocardial Perfusion Imaging, Preoperative Care, Proportional Hazards Models, ROC Curve, Reference Standards, Treatment Outcome, Ventricular Dysfunction, Left