摘要
目的探讨术前锥体束受侵犯程度对胶质瘤患者术后替莫唑胺化疗疗效的影响。方法回顾性分析2019年5月至2022年5月我院收治的138例胶质瘤患者临床资料,将其按照随机数字表法以2∶1的比例分为训练集(92例)和验证集(46例),根据替莫唑胺化疗疗效,将训练集患者分为化疗有效组(65例)和化疗无效组(27例)。对比两组临床资料,采用受试者工作特征(ROC)曲线分析术前锥体束受侵犯程度对患者术后化疗疗效的预测价值;采用多因素Logistic回归分析探讨患者术后化疗疗效的影响因素。通过R软件构建贝叶斯网络模型,Netica软件进行贝叶斯网络模型推理预测;利用ROC曲线和校准曲线对模型的效能进行验证。结果术前锥体束受侵犯程度对患者术后化疗疗效具有一定的预测价值(AUC=0.795)。多因素Logistic回归分析结果显示,年龄、世界卫生组织(WHO)分级、术前卡氏功能状态评分(KPS)、手术切除程度、瘤周水肿程度、神经功能障碍、术前锥体束受侵犯程度是术后化疗疗效的独立危险因素(P<0.05);训练集和验证集ROC曲线和校准曲线结果均显示贝叶斯网络预测模型具有较好的区分度和准确性。结论术前锥体束受侵犯程度对胶质瘤患者术后替莫唑胺化疗疗效具有一定的影响。
Objective To investigate the influence of the degree of pyramidal tract invasion before operation on the efficacy of temozolomide chemotherapy in patients with glioma.Methods The clinical data of 138patients with glioma admitted to our hospital from May 2019to May 2022 were analyzed retrospectively.It is divided into training set(92cases)and verification set(46cases)in a ratio of 2:1according to the random number table method.According to the chemotherapy effect of temozolomide,the patients in the training set were divided into the effective chemotherapy group(65cases)and the ineffective chemotherapy group(27cases).Comparison of clinical data between the two groups.Predictive value of preoperative pyramidal tract invasion on postoperative chemotherapy efficacy of patients analyzed by ROC curve.Multivariate Logistic Regression Analysis on the Factors Influencing the Chemotherapy Effect of Patients after Operation.Building Bayesian Network Model through R Software,and use Netica software to infer and predict Bayesian network model.Use ROC curve and calibration curve to verify the effectiveness of the model.Results The degree of pyramidal tract invasion before operation has a certain predic-tive value for the postoperative chemotherapy effect of patients(AUC=0.795).Multivariate logistic regression analysis showed that age,World Health Organization(WHO)Grading,preoperative KPS,degree of surgical resection,degree of peritumoral edema,neurological dysfunction,and degree of preoperative pyramidal tract invasion were independent risk factors for postoperative chemotherapy(P<0.05).The results of ROC curve and calibration curve of training set and verification set show that the Bayesian network prediction model has good discrimination and accuracy.Conclusion The degree of pyramidal tract invasion before operation has a certain influence on the postoperative chemotherapeutic efficacy of temozolomide in patients with glioma.
作者
张照华
杨丽萍
钟浩龙
周铭
Zhang Zhaohua;Yang Liping;Zhong Haolong;无(Department of Pharmacy,Zhujiang Hospital,Southern Medical University,Guangzhou510280,China;Department of Neurosurgery,Zhujiang Hospital,Southern Medical University,Guangzhou510280,China)
出处
《立体定向和功能性神经外科杂志》
2023年第3期141-146,共6页
Chinese Journal of Stereotactic and Functional Neurosurgery
关键词
锥体束受侵犯程度
化疗疗效
影响因素
贝叶斯网络模型
Degree of pyramidal tract invasion
Chemotherapy effect
Influencing factors
Bayesian network model