Hemodynamic assessment of severe aortic stenosis: MRI evaluation of dynamic changes of vena contracta.

ABSTRACT

PURPOSE: Direct magnetic resonance imaging (MRI) planimetry of the maximal systolic aortic valve area does not consider temporal variations of the opening area during the ejection period. We evaluated an MRI-based methodology for the assessment of valvular dynamics in patients with severe aortic stenosis by measuring the systolic variability of the valvular blood stream, that is, the "vena contracta." MATERIALS AND METHODS: With institutional review board approval, we examined 22 patients (13 male, 9 female; mean age, 68 ±10 years) with severe aortic stenosis using 1.5 T MRI and a standardized scanning protocol consisting of gradient-echo phase-contrast velocity imaging and steady-state free precession-cine MRI before and after valve replacement therapy. Temporal changes of the aortic valve area, represented by systolic variations of the area of poststenotic turbulent flow at its smallest convergence, that is, the proximal vena contracta, were determined by MRI and quantified by a calculated parameter of temporal valve dynamics (T). T was defined as the period which the aortic valve spent over its maximal opening area (>85%) during systole. MRI was also used to determine left ventricular hypertrophy before (LVMI) and its regression (LVMR) after valve replacement. Findings were compared with transthoracic echocardiography and cardiac catheterization. RESULTS: All patients had an echocardiographic effective orifice area, EOATTE, of <1.0 cm2. The comparison of T to LVMI and LVMR revealed significant correlations (LVMI r = -0.62; P = 0.002; LVMR r = 0.62; P = 0.002). Further significant correlations with aortic stenosis severity were observed in the comparison with manual planimetry, invasive measurements, and echocardiographic valve areas, as well as with pressure gradients. CONCLUSIONS: MRI can measure systolic variations of the aortic valve area. Quantitative parameters of the hemodynamic relevance of valve dynamics obtained by this method correlate with established parameters of aortic stenosis severity and LVMR.