肺癌患者合并深静脉血栓形成的列线图模型构建与评估

Construction and assessment of a Nomogram model for lung cancer patients with concurrent deep vein thrombosis

目的 构建肺癌患者合并深静脉血栓形成(DVT)的列线图模型并进行评估。方法 回顾性分析2022年1月至2023年1月蚌埠医科大学附属合肥市第二人民医院收治的330例肺癌患者的临床资料,根据入院时DVT发生情况将其分为DVT组(33例)和非DVT组(297例)。采用LASSO回归和多因素logistic回归筛选肺癌患者合并DVT的风险因素来构建列线图模型,并对该模型进行内部验证,评估模型的准确性、一致性和临床效用。结果 与非DVT组相比,DVT组年龄更大,吸烟、使用抗血管生成药物的人数比例更大,白细胞、中性粒细胞、红细胞分布宽度(RDW)、中性粒细胞-淋巴细胞比值(NLR)、乳酸脱氢酶(LDH)、脑钠尿肽(BNP)、D-二聚体、凝血酶原时间(PT)、C反应蛋白(CRP)和神经元特异性烯醇化酶(NSE)水平更高,血红蛋白、白蛋白水平更低,差异有统计学意义(P<0.05)。以DVT发生情况为因变量,通过LASSO回归和多因素logistic回归筛选出高血压、吸烟史、抗血管生成药物、D-二聚体、白蛋白5个肺癌合并DVT的风险因素,并以此构建列线图模型。受试者工作特征(ROC)曲线分析结果提示该模型具有良好的判别能力[AUC(95%CI)=0.880(0.811~0.950)]。校准曲线显示,模型的预测结果与实际观测值之间保持良好的拟合;Hosmer-Lemeshow拟合优度检验结果(χ^(2)=6.469,P=0.595)进一步证明了模型具有较好的一致性。决策曲线分析(DCA)结果显示,阈值概率处于0.07~0.95时,该模型具有较大的临床应用潜力。临床影响曲线(CIC)分析结果显示,在阈值概率范围内,预测发生DVT的患者数量总是大于实际发生DVT的患者数量,提示该列线图模型可以有效识别出DVT的高危患者。结论 该研究构建的肺癌合并DVT列线图模型具有良好的预测能力和临床实用性,有助于临床医师更好地识别高风险患者,并及时采取适当的干预措施,改善患者预后。

Objective To construct and assess a Nomogram model for lung cancer patients with concurrent deep vein thrombosis(DVT).Methods The clinical data of 330 cases of lung cancer who were admitted to the Second People′s Hospital of Hefei Affiliated to Bengbu Medical University from January 2022 to January 2023 were retrospectively analyzed,and the patients were divided into DVT group(33 cases)and non-DVT group(297 cases)according to the occurrence of DVT at admission.LASSO regression analysis and multivariate logistic regression analysis were used to screen the risk factors of the lung cancer patients with concurrent DVT to construct a Nomogram model,and the model was internally validated to assess its accuracy,consistency and clinical utility.Results Compared with the non-DVT group,the DVT group was older and had a greater proportion of smokers and users of antiangiogenic drugs,and had higher levels of leukocytes,neutrophils,red blood cell distribution width(RDW),neutrophil-lymphocyte ratio(NLR),lactate dehydrogenase(LDH),brain natriuretic peptide(BNP),D-dimer,prothrombin time(PT),C-reactive protein(CRP)and neuron-specific enolase(NSE),and lower levels of hemoglobin and albumin,and the differences were statistically significant(P<0.05).With the occurrence of DVT as the dependent variable,five risk factors of lung cancer complicated with DVT,including hypertension,smoking history,antiangiogenic drugs,D-dimer and albumin were screened by using LASSO regression and multivariate logistic regression and a Nomogram model was constructed based on these screened risk factors.The results of receiver operating characteristic(ROC)curve analysis indicated that the model had good discriminative ability[AUC(95%CI)=0.880(0.811-0.950)].The calibration curves showed that the predicted results of the model had a good fit with the actual observed values.The results of Hosmer-Lemeshow goodness-of-fit test(χ2=6.469,P=0.595)further proved that the model had good consistency.The results of decision curve analysis(DCA)showed that the model had a relatively large potential for clinical application when the threshold probability was in the range of 0.07-0.95.The results of clinical impact curve(CIC)analysis showed that within the range of threshold probability,the number of the patients who were predicted to have DVT was always greater than the number of the patients who actually had DVT,suggesting that the Nomogram model could effectively identify the high-risk patients with DVT.Conclusion The Nomogram model of lung cancer complicated with DVT constructed in this study has good predictive ability and clinical utility,which can help clinicians better identify high-risk patients and take appropriate intervention measures in time to improve the patients′prognosis.