Usefulness of Electrocardiographic Strain to Predict Survival After Surgical Aortic Valve Replacement for Aortic Stenosis
Résumé
Electrocardiographic (ECG) strain has been reported as a specific marker of midwall left ventricular (LV) myocardial fibrosis, predictive of adverse clinical outcomes in aortic stenosis (AS), but its prognostic impact after aortic valve replacement (AVR) is unknown. We aimed to assess the impact of ECG strain on long-term mortality after surgical AVR for AS. From January 2005 to January 2014, patients with interpretable preoperative ECG who underwent isolated AVR for AS were included. ECG strain was defined as ≥1-mm concave downslopping ST-segment depression with asymmetrical T-wave inversion in lateral leads. Mortality was assessed over a follow-up period of 4.8 ± 2.7 years. Among the 390 patients included, 110 had ECG strain (28%). They had significantly lower body mass index, higher mean transaortic pressure gradient and Cornell-product ECG LV hypertrophy than in those without ECG strain. There was also a trend for lower LV ejection fraction in patients with ECG strain as compared with those without. Patients with ECG strain had significantly lower 8-year survival than those without. ECG strain remained associated with reduced survival both in patients with and without LV hypertrophy (p <0.0001 for both). After adjustment, ECG strain remained a strong and independent determinant of long-term survival (hazard ratio 4.4, p <0.0001). Similar results were found in patients with LV hypertrophy or without LV hypertrophy. In the multivariate model, the addition of ECG strain provided incremental prognostic value (p <0.0001). In conclusion, in patients with AS, ECG strain is associated with 4-fold increased risk of long-term mortality after isolated AVR, regardless of preoperative LV hypertrophy.