Optimal pre-TAVR annulus sizing in patients with bicuspid aortic valve. Area-derived perimeter by CT is the best-correlated measure with intraoperative sizing

To clarify the optimal measurements for patients with bicuspid aortic valve (BAV) preferred for transcatheter aortic valve replacement (TAVR), our study compared intraoperative sizing with five different approaches by transthoracic echocardiography (TTE), three-dimensional transesophageal echocardiography (3DTEE) and computed tomography (CT). We enrolled 104 BAV patients prescreened for TAVR but who underwent surgery with direct intraoperative annulus sizing. All five approaches [2DTTE, 3DTEE, area-derived perimeter (CTarea), perimeter-derived diameter (CTperi) and mean diameter (CTmean)] were compared with intraoperative sizing, respectively. Agreements on theoretical valve selections by five methods with those by intraoperative sizing were analyzed. CTarea showed the highest correlation (r = 0.932) and the best agreement with intraoperative sizing. Agreement for theoretical surgical and TAVR prosthesis selection was found in 84.6% and 74.0% BAVs by CTarea (κ = 0.791, κ = 0.585). CTperi-based prosthesis selection led to overestimation of 26.9% for surgical valves (κ = 0.589) and 36.5% for TAVR valves (κ = 0.425). Good correlations were observed between CT measurements and intraoperative sizing regardless of the predominant site of aortic valve calcification (r = 0.860-0.953). The CTarea, which demonstrated the optimal approach to annulus sizing and prosthesis choice of BAVs with high eccentricity, should be included into the BAV-specific annulus sizing recommendation. The insufficiency of CTperi lay in overestimation of surgical or TAVR valve selections. Good agreement of 3DTEE sizing proved its superiority in annulus sizing for BAVs unsuitable for CT, but it should be used with caution for patients with a calcified annulus, where partial acoustic shadowing could lead to image inaccuracy. (orig.)