Dobutamine stress echocardiography remains indispensable for the diagnosis of low-flow low-gradient severe aortic stenosis

Abstract Background In patients with low-gradient aortic stenosis (AS) and low transvalvular flow, current guidelines recommend dobutamine stress echocardiography (DSE) to determine valve stenosis severity. Additionally, aortic valve calcification (AVC) scores are also used with this intent. Previous studies have determined cut-off values of ≥2000AU for men and ≥1200AU for women to distinguish between true severe AS and pseudosevere AS. Purpose To assess the relationship between AVC and DSE in low-flow low-gradient aortic stenosis. Methods We performed a multicenter international registry study including patients that underwent DSE to determine AS severity and multi-slice computed tomography (MSCT) of sufficient quality to determine AVC scores. Severe AS was diagnosed when resting peak aortic valve velocity was ≥4.0 m/s, or stress aortic valve area (AVA) remained <1.0cm2 at peak dobutamine stress levels; moderate AS was diagnosed when resting AVA was ≥1.0cm2 and peak velocity was < 4.0m/s or stress AVA exceeded 1.0cm² at peak stress. Using the previous cut-off values, MSCT-AVC scores were compared to DSE-derived AS diagnoses to determine sensitivity and specificity of AVC, and the positive and negative predictive values of high and low scores. Subsequently, we performed a receiver-operating characteristics (ROC)-curve analysis for men and women separately. Results We included 271 patients from 8 centers, of which 231 patients had sufficiently interpretable DSE and MSCT. Median age was 79 years (Interquartile range (IQR): 71-84), median BMI was 26.4 kg/m2 (IQR: 24.1-30.0) and median BSA was 1.9 m2(IQR: 1.8-2.1). Left ventricular ejection fraction was reduced (<50%) in 209 (90.5%) patients. Bicuspid valve morphology was present in 20 patients (8.7%). DSE confirmed severe AS in 103 patients, 4 of whom already had resting peak aortic valve velocity ≥4.0m/s. Moderate AS was diagnosed in 128 patients, of which 76 had a resting AVA ≥ 1.0cm2. Characteristics of these patients are summarized the table. Of the patients with severe AS, 49 (47.6%) had a high AVC score (≥2000 for men or ≥1200 for women) and of patients with moderate AS, 75 (58.6%) had an AVC score below those cut-off values. This corresponds to a sensitivity of 47.6%, a specificity of 58.4% and positive and negative predictive values of 53.5% and 51.6%, respectively. The ROC-curves are shown in the figure; the area under the curve was 0.55 for men and 0.52 for women, signifying poor discriminative performance. Conclusion In this multicenter international registry we found that aortic valve calcification scores correlated poorly with AS severity and therefore DSE remains the preferred technique to assess borderline AS severity.Table.Patient characteristicsFigure.ROC curve