Three dimensional evaluation of the aortic annulus using multislice computer tomography: are manufacturer's...

To evaluate the effects of applying current sizing guidelines to different multislice computer tomography (MSCT) aortic annulus measurements on Corevalve (CRS) size selection.

Multislice computer tomography annulus diameters [minimum: D(min); maximum: D(max); mean: D(mean) = (D(min) + D(max))/2; mean from circumference: D(circ); mean from surface area: D(CSA)] were measured in 75 patients referred for percutaneous valve replacement. Fifty patients subsequently received a CRS (26 mm: n = 22; 29 mm: n = 28). D(min) and D(max) differed substantially [mean difference (95% CI) = 6.5 mm (5.7-7.2), P < 0.001]. If D(min) were used for sizing 26% of 75 patients would be ineligible (annulus too small in 23%, too large in 3%), 48% would receive a 26 mm and 12% a 29 mm CRS. If D(max) were used, 39% would be ineligible (all annuli too large), 4% would receive a 26 mm, and 52% a 29 mm CRS. Using D(mean), D(circ), or D(CSA) most patients would receive a 29 mm CRS and 11, 16, and 9% would be ineligible. In 50 patients who received a CRS operator choice corresponded best with sizing based on D(CSA) and D(mean) (76%, 74%), but undersizing occurred in 20 and 22% of which half were ineligible (annulus too large).

Eligibility varied substantially depending on the sizing criterion. In clinical practice both under- and oversizing were common. Industry guidelines should recognize the oval shape of the aortic annulus.