非小细胞肺癌患者发生脑转移风险列线图预测模型的建立

目的:基于非小细胞肺癌(non-small cell lung cancer,NSCLC)患者发生脑转移风险的影响因素分析，建立列线图预测模型。方法:选取2011年06月至2015年06月在本院确诊为NSCLC的患者452例作为研究对象，收集患者的临床资料并进行随访，如果出现脑转移情况或5年随访时间已满，则随访终止，通过单因素和多因素的Cox回归分析得出发生脑转移风险的影响因素，在此基础上根据筛选出的变量建立列线图预测模型。结果:Cox回归最终筛选出的变量为TNM分期、分化程度、病理类型、CA125水平和淋巴结转移数目，建立的列线图预测模型C-index为0.801（95%CI:0.778~0.882）。结论:本研究建立的NSCLC患者发生脑转移风险的列线图预测模型，可以根据多因素分析结果中有意义的变量，较为准确的预测患者的预后情况，临床应用前景较好。

Establishment of a nomogram prediction model for the risk of brain metastasis in patients with non-small cell lung cancer

Objective:To establish a nomogram prediction model by analyzing the factors affecting the risk of brain metastasis in patients with non-small cell lung cancer.Methods:452 patients diagnosed with NSCLC in our hospital from June 2011 to June 2015 were selected as the research objects，and the clinical data of the patients were collected and followed up.If brain metastases occurred or the 5-year follow-up time expired，follow up after termination，the influencing factors of brain metastasis risk were obtained through single-factor and multi-factor Cox regression analysis，and on this basis，a nomogram prediction model was established based on the selected variables.Results:The variables finally screened by Cox regression were TNM stage，degree of differentiation，pathological type，CA125 level and number of lymph node metastasis.The C-index of the established nomogram prediction model was 0.801（95%CI:0.778~0.882).Conclusion:The nomogram model for predicting the risk of brain metastases in NSCLC patients established in this study can accurately predict the prognosis of patients based on meaningful variables in the results of multivariate analysis and has a good clinical application prospect.