辽宁中医药大学学报2026，Vol.28Issue(1)：7-12,6.DOI:10.13194/j.issn.1673-842X.2026.01.002
非小细胞肺癌脑转移风险的临床预测模型构建
Construction of a Clinical Prediction Model for Brain Metastasis Risk of Non-Small Cell Lung Cancer
摘要
Abstract
Objective To observe the effects of clinical and imaging data on brain metastasis of lung cancer,to discover the risk factors of brain metastasis,to analyze the correlation between risk factors and brain metastasis,and to make nomograms to predict the probability of brain metastasis of non-small cell lung cancer(NSCLC),in order to achieve the purpose of early prevention.Methods Differential analysis was performed to find the risk factors of brain metastasis of lung cancer.Multivariate regression analysis was used to find the correlation factors of brain metastasis,and a nomogram was established to test the predictive performance.Results(1)There were 45 patients in the brain metastasis group and 136 patients in the non-brain metastasis group,and there were statistically significant differences in gender(P=0.010),age(P=0.013),pathological type(P＜0.001),lobulation sign(P=0.004),burr sign(P=0.048),pleural depression sign(P=0.006),clinical stage(P=0.017),and smoking index(P=0.047)between the two groups(P＜0.05).There was no significant difference in pneumatic bronchial signs between the two groups(P＞0.05).(2)Among the 181 patients,43 had brain metastases,with an incidence of 23.8％.Binary logistic regression analysis showed that the TCM syndrome type P(0.04),maximum P(0.027),kurtosis P(0.022),pathological P(0.002),age P(0.038),and stage classification P(0.024)were significant(P＜0.05).(3)The ROC curve analysis results showed that the area under the ROC curve of the model was 0.799,indicating that the model had good prediction performance.The sensitivity and specificity were 0.719 and 0.879,respectively.(4)The bootstrap sampling verification(1000 times of internal bootstrap sampling verification)was carried out on the constructed nomogram model,and the calibration curve was drawn,and the results showed that the prediction probability of the model was basically consistent with the actual incidence,and the average absolute difference was 0.029,indicating that the accuracy of the model was good.At the same time,the goodness-of-fit test(HL)of the model was P＞0.05,indicating that the model fit well.(5)Within the risk threshold probability range of 0.08-0.77,the model had a net benefit,and the net benefit of intervention based on the constructed model was higher than that of intervention for all and non-intervention for all.Conclusion(1)There are significant differences in gender,age,pathological type,lobulation sign,burr sign,pleural depression sign,clinical stage,smoking index,TCM syndrome type,lung mass mass,volume and skewness,which may be important risk factors for brain metastasis.(2)TCM syndrome type,maximum,kurtosis,pathology,age,and stage classification are the relevant risk factors for brain metastasis in NSCLC,and based on multivariate regression analysis,the established nomogram model has good predictive performance and can provide reference value for the prevention and treatment of brain metastases in NSCLC.关键词
非小细胞肺癌/脑转移/危险因素/列线图Key words
non-small cell lung cancer/brain metastases/risk factors/nomogram分类
医药卫生引用本文复制引用
张亚密,贾永军,张唯,何伟,杨昭,安书芬,张佩云,常智勇,梁兰..非小细胞肺癌脑转移风险的临床预测模型构建[J].辽宁中医药大学学报,2026,28(1):7-12,6.基金项目
陕西省重点研发计划资助项目(2022SF-316) （2022SF-316）
