摘要
目的 分析肺腺癌骨转移的危险因素,构建预测肺腺癌患者发生骨转移风险的列线图模型。方法 回顾性收集244例肺腺癌患者的临床资料及血液学指标,按8∶2比例随机分为训练集(195例)与验证集(49例)。基于训练集,通过Logistic回归分析筛选肺腺癌骨转移的危险因素,并据此构建列线图模型。随后,采用受试者工作特征(ROC)曲线、校准曲线及决策曲线分析(DCA)评估模型准确性和临床实用性。结果 训练集肺腺癌患者中骨转移81例,无骨转移114例。单因素分析显示:无骨转移组、骨转移组患者T分期、N分期、癌胚抗原(CEA)、细胞角蛋白19片段(CYFRA21-1)、神经元特异性烯醇化酶(NSE)、胃泌素释放肽前体(ProGRP)、纤维蛋白原(FIB)、碱性磷酸酶(ALP)和乳酸脱氢酶(LDH)比较差异均有统计学意义(P<0.05)。多因素Logistic回归分析结果显示:T分期(T_(3~4)期)、N分期(N_(1~3)期)及CEA、NSE、ALP、LDH水平高是肺腺癌患者骨转移的独立危险因素(P<0.05)。通过多因素Logistic回归模型计算回归系数,建立肺腺癌患者骨转移风险的列线图模型:P=-7.493+1.085×T分期+1.298×N分期+0.013×CEA+0.090×NSE+0.017×ALP+0.011×LDH。训练集:曲线下面积(AUC)为0.937,敏感度为77.8%,特异度为87.7%;验证集:AUC为0.917,敏感度为68.4%,特异度为86.2%。校准曲线与DCA均表明其具有良好的预测效能与临床应用价值。结论 T分期、N分期、CEA、NSE、ALP和LDH构建的肺腺癌患者骨转移的列线图模型具有较高的预测价值,可为肺腺癌患者的个体化风险评估提供参考。
Objective To analyze the risk factors for bone metastasis in lung adenocarcinoma,and construct a nomogram model for predicting the risk of bone metastasis in patients with lung adenocarcinoma.Methods Clinical data and hematological parameters of 244 lung adenocarcinoma patients were retrospectively collected.The patients were proportionally randomized into a training set(195 cases)and a validation set(49 cases)in a 8∶2 ratio.Based on the training set,risk factors for bone metastasis were identified through Logistic regression analysis,and a nomogram model was subsequently developed.The model's accuracy and clinical utility were evaluated using receiver operating characteristic(ROC)curves,calibration curves,and decision curve analysis(DCA).Results Among the training set of lung adenocarcinoma patients,81 cases had bone metastases and 114 cases did not.Univariate analysis revealed that there were differences in T stage,N stage,carcinoembryonic antigen(CEA),cytokeratin 19 fragment(CYFRA21-1),neuron-specific enolase(NSE),pro-gastrin-releasing peptide(ProGRP),fibrinogen(FIB),alkaline phosphatase(ALP),and lactate dehydrogenase(LDH)between the non-bone metastasis group and the bone metastasis group(P<0.05).Multivariate Logistic regression analysis revealed that T stage(T_(3-4)),N stage(N_(1-3)),CEA,NSE,ALP,and LDH were independent risk factors for bone metastasis in patients with lung adenocarcinoma(P<0.05).The regression coefficients were calculated through the multivariate Logistic regression model to establish a nomogram model for the risk of bone metastasis in patients with lung adenocarcinoma:P=-7.493+1.085×T stage+1.298×N stage+0.013×CEA+0.090×NSE+0.017×ALP+0.011×LDH.Training set:The area under the curve(AUC)was 0.937,the sensitivity was 77.8%,and the specificity was 87.7%.Validation set:AUC was 0.917,the sensitivity was 68.4%,and the specificity was 86.2%.Both calibration curve and DCA demonstrated good predictive performance and clinical application value.Conclusion The nomogram model for predicting bone metastasis in lung adenocarcinoma patients,constructed based on T stage,N stage,CEA,NSE,ALP,and LDH,demonstrates high predictive value and can serve as a reference for individualized risk assessment in these patients.
作者
吴静
肖燕萍
罗晓莉
WU Jing;XIAO Yan-ping;LUO Xiao-li(Clinical Oncology School of Fujian Medical University,Fujian Cancer Hospital,Fuzhou 350014,China)
出处
《中国现代药物应用》
2026年第5期10-14,共5页
Chinese Journal of Modern Drug Application
关键词
肺腺癌
骨转移
预测模型
列线图
Lung adenocarcinoma
Bone metastasis
Predictive model
Nomogram
