老年晚期肺癌患者化疗不良反应的列线图预测模型构建

Construction of a nomogram prediction model for chemotherapy adverse reactions in elderly patients with advanced lung cancer

目的分析老年晚期肺癌患者化疗不良反应的影响因素,构建化疗不良反应发生风险列线图预测模型。方法将156例老年晚期肺癌患者按化疗过程中是否发生不良反应分为不良反应(AR)组59例和无不良反应(NAR)组97例。分析老年晚期肺癌患者化疗不良反应的影响因素,根据危险因素或保护因素构建列线图预测模型,并分析列线图预测模型预测化疗不良反应的准确度与临床应用价值。结果AR组与NAR组患者年龄、化疗方案、化疗周期、肿瘤直径、微血管密度(MVD)、化疗后白蛋白水平比较,差异均有统计学意义(P﹤0.01)。年龄较大、联合化疗、化疗周期﹥2个、肿瘤直径较大、MVD较多均为老年晚期肺癌患者发生化疗不良反应的危险因素(OR﹥1,P﹤0.01),化疗后白蛋白水平较高为老年晚期肺癌患者发生化疗不良反应的保护因素(OR﹤1,P﹤0.01)。列线图预测模型校正曲线与工作曲线贴合度理想,内部校正曲线C指数=0.997;经受试者工作特征(ROC)曲线分析,列线图预测模型预测化疗不良反应的曲线下面积(AUC)为0.997。结论基于老年晚期肺癌患者化疗期间不良反应的危险因素与保护因素,构建列线图预测模型,能够量化性分析患者化疗不良反应的发生风险,具有较高的预测准确度,以指导临床尽早开展相应应对措施。

Objective To analyze the influencing factors of chemotherapy adverse reactions in elderly patients with advanced lung cancer,and to construct a prediction model of chemotherapy adverse reaction risk by nomogram.Method A total of 156 elderly patients with advanced lung cancer were divided into adverse reaction(AR)group(59 cases)and non-adverse reaction(NAR)group(97 cases)according to whether adverse reaction occurred during chemotherapy.To analyze the influencing factors of chemotherapy adverse reactions in elderly patients with advanced lung cancer,a nomogram prediction model was constructed based on risk factors or protective factors,and the accuracy and clinical application value of the nomogram prediction model in predicting chemotherapy adverse reactions was analyzed.Result There were significant differences in age,chemotherapy regimen,chemotherapy cycle,tumor diameter,microvascular density(MVD)and albumin level after chemotherapy between AR group and NAR group(P<0.01).Older age,combination chemotherapy,chemotherapy cycle>2,larger tumor diameter and more MVD were all risk factors for chemotherapy adverse reactions in elderly patients with advanced lung cancer(OR>1,P<0.01).Higher albumin level after chemotherapy was protective factor for chemotherapy adverse reactions in elderly patients with advanced lung cancer(OR<1,P<0.01).The calibration curve of the prediction model and the working curve fit well,and the C-index of the internal calibration curve was 0.997.Receiver operating characteristic(ROC)curve showed that the area under the curve(AUC)of the prediction model for chemotherapy adverse reactions was 0.997.Conclusion Based on the risk factors and protective factors of adverse reactions during chemotherapy in elderly patients with advanced lung cancer,the construction of a nomogram prediction model can quantitatively analyze the risk of adverse reactions during chemotherapy,and has high prediction accuracy,so as to guide the clinical implementation of appropriate countermeasures as soon as possible.