Division of Cardiology, New-York-Presbyterian Hospital and Columbia University Medical Center, New York, NY, United States; Division of Cardiology, Hospital Charles Nicolle, University Hospital of Rouen, Rouen, France.
Division of Cardiology, New-York-Presbyterian Hospital and Columbia University Medical Center, New York, NY, United States.
J Cardiovasc Comput Tomogr. 2017 Jul-Aug;11(4):302-308. doi: 10.1016/j.jcct.2017.04.009. Epub 2017 Apr 22.
The threshold for the optimal computed tomography (CT) number in Hounsfield Units (HU) to quantify aortic valvular calcium on contrast-enhanced scans has not been standardized. Our aim was to find the most accurate threshold to predict paravalvular regurgitation (PVR) after transcatheter aortic valve replacement (TAVR).
104 patients who underwent TAVR with the CoreValve prosthesis were studied retrospectively. Luminal attenuation (LA) in HU was measured at the level of the aortic annulus. Calcium volume score for the aortic valvular complex was measured using 6 threshold cutoffs (650 HU, 850 HU, LA × 1.25, LA × 1.5, LA+50, LA+100). Receiver-operating characteristic (ROC) analysis was performed to assess the predictive value for > mild PVR (n = 16). Multivariable analysis was performed to determine the accuracy to predict > mild PVR after adjustment for depth and perimeter oversizing.
ROC analysis showed lower area under the curve (AUC) values for fixed threshold cutoffs (650 or 850 HU) compared to thresholds relative to LA. The LA+100 threshold had the highest AUC (0.81), and AUC was higher than all studied protocols, other than the LA x 1.25 and LA + 50 protocols, where the difference approached statistical significance (p = 0.05, and 0.068, respectively). Multivariable analysis showed calcium volume determined by the LAx1.25, LAx1.5, LA+50, and LA+ 100 HU protocols to independently predict PVR.
Calcium volume scoring thresholds which are relative to LA are more predictive of PVR post-TAVR than those which use fixed cutoffs. A threshold of LA+100 HU had the highest predictive value.
在对比增强扫描中，量化主动脉瓣钙质的最佳计算机断层扫描（CT）数 Hounsfield 单位（HU）阈值尚未标准化。我们的目的是找到最准确的阈值来预测经导管主动脉瓣置换术（TAVR）后瓣周漏（PVR）。
回顾性分析 104 例行 TAVR 并植入 CoreValve 假体的患者。在主动脉瓣环水平测量 HU 的管腔衰减（LA）。使用 6 个阈值（650 HU、850 HU、LA×1.25、LA×1.5、LA+50、LA+100）测量主动脉瓣复合体的钙体积分数。进行受试者工作特征（ROC）分析以评估对>轻度 PVR（n=16）的预测价值。进行多变量分析，以确定在调整深度和周长过度扩张后预测>轻度 PVR 的准确性。
ROC 分析显示，与 LA 相关的固定阈值（650 或 850 HU）的曲线下面积（AUC）值较低。LA+100 阈值的 AUC 值最高（0.81），高于所有研究方案，除了 LA x 1.25 和 LA+50 方案，差异接近统计学意义（分别为 p=0.05 和 0.068）。多变量分析显示，LAx1.25、LAx1.5、LA+50 和 LA+100 HU 方案确定的钙体积可独立预测 PVR。
与 LA 相关的钙体积评分阈值比使用固定阈值更能预测 TAVR 后 PVR。LA+100 HU 阈值具有最高的预测价值。
