Buccheri Sergio, Capodanno Davide, Barbanti Marco, Popolo Rubbio Antonio, Di Salvo Maria Elena, Scandura Salvatore, Mangiafico Sarah, Ronsivalle Giuseppe, Chiarandà Marta, Capranzano Piera, Grasso Carmelo, Tamburino Corrado
Division of Cardiology, Ferrarotto Hospital, University of Catania, Catania, Italy.
Division of Cardiology, Ferrarotto Hospital, University of Catania, Catania, Italy.
Am J Cardiol. 2017 May 1;119(9):1443-1449. doi: 10.1016/j.amjcard.2017.01.024. Epub 2017 Feb 10.
There is a lack of specific tools for risk stratification in patients who undergo MitraClip implantation. We aimed at combining preprocedural variables with prognostic impact into a specific risk model for the prediction of 1-year mortality in patients undergoing MitraClip implantation. A total of 311 consecutive patients who underwent MitraClip implantation were included. A lasso-penalized Cox-proportional hazard regression model was used to identify independent predictors of 1-year all-cause mortality. A nomogram (GRASP [Getting Reduction of mitrAl inSufficiency by Percutaneous clip implantation] nomogram) was obtained from the Cox model. Validation was performed using internal bootstrap resampling. Forty-two deaths occurred at 1-year follow-up. The Kaplan-Meier estimate of 1-year survival was 0.845 (95% confidence interval, 0.802 to 0.895). Four independent predictors of mortality (mean arterial blood pressure, hemoglobin natural log-transformed pro-brain natriuretic peptide levels, New York Heart Association class IV at presentation) were identified. At internal bootstrap resampling validation, the GRASP nomogram had good discrimination (area under receiver operating characteristic curve of 0.78, Somers' D statistic of 0.53) and calibration (le Cessie-van Houwelingen-Copas-Hosmer p value of 0.780). Conversely, the discriminative ability of the EuroSCORE II (the European System for Cardiac Operative Risk Evaluation II) and the STS-PROM (the Society of Thoracic Surgeons Predicted Risk of Mortality score) was fairly modest with area under the curve values of 0.61 and 0.55, respectively. A treatment-specific risk model in patients who undergo MitraClip implantation may be useful for the stratification of mortality at 1 year. Further studies are needed to provide external validation and support the generalizability of the GRASP nomogram.
对于接受二尖瓣夹合术植入的患者，缺乏用于风险分层的特定工具。我们旨在将具有预后影响的术前变量组合成一个特定的风险模型，以预测接受二尖瓣夹合术植入患者的1年死亡率。总共纳入了311例连续接受二尖瓣夹合术植入的患者。使用套索惩罚Cox比例风险回归模型来识别1年全因死亡率的独立预测因素。从Cox模型中获得了一个列线图（GRASP[经皮夹合植入减少二尖瓣反流]列线图）。使用内部自助重采样进行验证。在1年随访时有42例死亡。1年生存率的Kaplan-Meier估计值为0.845（95%置信区间，0.802至0.895）。确定了四个死亡率的独立预测因素（平均动脉血压、血红蛋白、自然对数转换的脑钠肽前体水平、就诊时纽约心脏协会IV级）。在内部自助重采样验证中，GRASP列线图具有良好的辨别能力（受试者操作特征曲线下面积为0.78，Somers' D统计量为0.53）和校准（le Cessie-van Houwelingen-Copas-Hosmer p值为0.780）。相反，欧洲心脏手术风险评估系统II（EuroSCORE II）和胸外科医师协会预测死亡率评分（STS-PROM）的辨别能力相当有限，曲线下面积值分别为0.61和0.55。二尖瓣夹合术植入患者的特定治疗风险模型可能有助于1年死亡率的分层。需要进一步研究以提供外部验证并支持GRASP列线图的普遍性。
