回顾性分析主动脉瓣置换术后近端主动脉扩张的危险因素

Retrospective analysis of risk factors for proximal aortic dilation after aortic valve replacement

目的:探索主动脉瓣置换术(aortic valve replacement,AVR)后近端主动脉≥45mm患者的危险因素并构建模型,以预测此类患者术后近端主动脉扩张的风险。方法:回顾性分析2018年1月至2022年10月,北京安贞医院接受主动脉瓣置换外科手术治疗患者的临床资料。以术后近端主动脉≥45mm为终点,随机将这些患者的70%划分为建模组,剩余30%的患者划分验证组。在建模组中应用二元多因素Logistic回归探寻危险因素,并构建模型,绘制列线图,并在验证组中验证模型的区分度和校准度。结果:本研究共纳入的979例患者,120例(12.7%)患者发生终点事件(术后近端主动脉≥45mm)。多因素Logistic回归分析表明男性、高血压、主动脉窦指数(aortic sinus index,DAS/BSA)、升主动脉指数(ascending aortic index,DAA/BSA)、LVEDD是主动脉瓣置换术后近端主动脉≥45mm的危险因素。根据以上5个预测因子构建出的模型在建模组中区分度良好,一致性指数(concordance index,C-index)为0.718(95%CI:0.665~0.771),且模型的准确度较高。模型在验证组中的C-index为0.727(95%CI:0.640~0.816)。在预测的主要终点事件发生风险低于50%的患者中,校准曲线表明预测风险和观测风险基本一致。结论:建立AVR术后近端主动脉≥45mm的风险预测模型,可有效预测此类患者的发生率,有助于在此类人群中识别出高危主动脉瓣置换手术患者进行手术策略的优化。

Objective:To explore the risk factors in patients with proximal aortic dilatation after undergoing aortic valve replacement(AVR),and to construct a model that predicts the risk of postoperative proximal aortic dilatation in these patients.Methods:A retrospective analysis was conducted on the clinical data of patients undergoing aortic valve replacement surgery at Beijing Anzhen Hospital from January 2018 to October 2022.Using postoperative proximal aortic dilatation as the endpoint,70%of these patients were randomly divided into a training set,while the remaining 30%were allocated to a validation set.Apply binary multivariate Logistic regression to explore risk factors in the training set,construct a model,draw a column chart,and validate the model's discrimination and calibration in the validation set.Results:Of the 979 patients included in this study,120(12.7%)had endpoint events(postoperative recurrent aortic lesions).Multivariate logistic regression analysis showed that male,hypertension,aortic sinus index(DAS/BSA),ascending aortic index(DAA/BSA),and LVEED were risk factors for recurrent aortic disease after aortic valve replacement surgery.The model constructed based on the above five predictive factors has good discrimination in the training set,with a consistency index(C-index)of 0.718(95%CI:0.665-0.771),and high accuracy.The C-index of the model in the validation set is 0.727(95%CI:0.640-0.816).In patients with a predicted risk of less than 50%of the primary endpoint events,the calibration curve indicates that the predicted risk is basically consistent with the observed risk.Conclusions:Establishing a risk prediction model for recurrent aortic disease after AVR surgery can effectively predict the incidence of such patients and help identify high-risk aortic valve replacement patients in this population for optimizing surgical strategies.