Inconsistent echocardiographic grading of aortic stenosis: is the left ventricular outflow tract important?

Objective Discrepancy in the echocardiographic severity grading of aortic stenosis (AS) based on current guidelines has been reported. We sought to investigate the left ventricular outflow tract diameter (LVOTd) as a source of inconsistencies, and to explore hypothetical alternatives for discrepancy improvement. Design Retrospective echocardiographic cross-sectional analysis. Setting From 2000 to 2010, we identified all AS patients with left ventricular EF >= 50%, mean gradient (MG) >= 20 mm Hg, aortic valve area (AVA) <= 2.5 cm(2), = 2.3 cm. In each group, inconsistency of data for classification of severity of AS was assessed and alternative thresholds explored. Results Of 9488 total patients, 58% were men, LVOTd 2.18 +/- 0.19 cm, peak velocity (Vmax) 3.9 +/- 0.8 m/s, MG 37 +/- 16 mm Hg, and AVA 1.09 +/- 0.34 cm(2). Small LVOTd patients were older women (91%) with worse systemic haemodynamics and more prevalent paradoxical low-flow, compared with average and large LVOTd patients (all parameters p <0.001). Despite clinically similar MG and Vmax across all groups, mean AVA ranged from 0.88 to 1.25 cm(2) (p <0.001), classifying small LVOTd patients as severe, average LVOTd as moderate-severe and large LVOTd as moderate. For patients with large, average and small LVOTd, an AVA of 1 cm(2) corresponded to MG of 42, 35 and 29 mm Hg, Vmax of 4.1, 3.8 and 3.5 m/s and dimensionless index (DI) of 0.22, 0.29 and 0.36, respectively. An AVA cut-off of 0.8 cm(2) reduced severe AS inconsistency from 48% to 26% for small LVOTd patients. An AVA cut-off of 0.9 cm(2) reduced severe AS inconsistency from 37% to 26% for average LVOTd patients. The current AVA cut-off of 1 cm(2) was consistent for large LVOTd patients. Conclusions The LVOTd is associated with significant inconsistencies in AS assessment by current guidelines. For patients with normal EF and normal flow, current guideline definition of severe AS is most consistent for patients with large LVOTd, but not so for patients with average or small LVOTd in whom lower AVA cut-offs should be further studied. The DI cut-off for severe AS is highly variable depending on the LVOTd and guideline revision of this threshold should be considered.