Electrocardiographic criteria for the diagnosis of left ventricular hypertrophy in patients with severe aortic stenosis

Abstract

Abstract Introduction There is a lack of data regarding the performance of current electrocardiographic (EKG) criteria for left ventricular hypertrophy (LVH) in patients with severe aortic stenosis (AS). Purpose To test the performance of the current EKG scores to diagnosis LVH in AS patients, and also to compare the performance of a new proposed score. Methods The present study evaluated 80 patients with severe AS that underwent cardiac magnetic resonance to diagnose LVH, defined as left ventricular indexed mass of 95 g/m² for female and 115 g/m² for male patients. The EKG criteria used for left ventricular hypertrophy were Sokolow-Lyon criteria, Romhilt-Estes point score system, Cornell voltage criteria and Peguero-Lo Presti criteria. Results Among the 80 patients included in analysis, 58.7% were male and the mean age was 64±8 years. There was a high prevalence of comorbidities, highlighting diabetes (32.5%), hypertension (67.5%), atrial fibrillation (5.0%) and coronary artery disease (11.3%). Regarding echocardiogram, the median left ventricle ejection fraction was 64 (56-68)% and the median indexed aortic valve area was 0.43 (0.35-0.49) cm²/m². Cardiac magnetic resonance analysis demonstrated a mean left ventricle mass of 168±55g, and the LVH incidence was 26.3%. Regarding EKG, Strain pattern was found in 49.1% of the patients, Cornell voltage criteria was positive in 51.3%, Sokolow-Lyon criteria in 51.3%, Romhilt-Estes point score system in 42.5% and Peguero-Lo Presti criteria in 61.3%. Discriminative capacity of electrocardiographic criteria for LVH is shown in Figure 1. Peguero-Lo Presti score had the best diagnostic accuracy in the overall population and in male sex (AUC 0.776), while Cornell voltage criteria had the best diagnostic accuracy in female patients (AUC 0.776). Based on linear regression model, the proposed score was created to predict LVH at cardiac magnetic resonance: 48 + (1.3 x Deepest S wave) + (0.5 x S wave in V1 or V2 plus R wave in V5). The proposed score cutoff value of 90.5 mm had a sensitivity of 0.765 and specificity of 0.707 (for male sex: 0.769 and 0.655, respectively; for female sex: 0.750 and 0.759, respectively). Compared to the scores described in the literature, the proposed score had the best area AUC for the overall population (Figure 1), for female and male patients (AUC 0.796, 0.819 and 0.780, respectively). Conclusion In patients with severe AS, the proposed score had the best diagnostic accuracy compared to the scores described in the literature for LVH identification, followed by Peguero-Lo Presti score, even when analyzed based on gender. Besides, the proposed score had also acceptable sensitivity and specificity for LVH detection.AUC of electrocardiographic scores.