摘要
目的探讨肺结核并发真菌感染的影响因素,构建基于Janus蛋白酪氨酸激酶(JAK)/信号转导因子和转录激活因子(STAT)信号通路的列线图预测模型并验证。方法回顾性选取我院诊治的肺结核患者308例,根据是否并发真菌感染将其分为并发组(35例)和非并发组(273例)。观察并发组患者病原性真菌分布情况,收集所有患者临床资料及实验室检查结果,采用单因素及多因素logistic回归分析评估肺结核并发真菌感染的危险因素,并以此构建列线图预测模型。采用受试者工作特征(ROC)曲线评估相关危险因素及列线图预测模型对肺结核并发真菌感染的预测价值。采用Bootstrap法对列线图预测模型进行内部验证,采用校准曲线评估列线图模型的校准度,采用决策曲线分析验证列线图模型的临床净收益率。结果35例并发真菌感染患者中,23例(65.71%)为曲霉菌,8例(22.86%)为白假丝酵母菌,3例(8.57%)为热带假丝酵母菌,1例(2.86%)为光滑假丝酵母菌。多因素logistic回归分析结果显示,合并糖尿病及慢性阻塞性肺疾病(COPD)、长期应用广谱抗生素、病变累及肺野数>3个、JAK2及STAT3蛋白相对表达水平较高均为肺结核并发真菌感染的独立危险因素(P<0.05)。ROC曲线分析结果显示,列线图预测模型的AUC均大于合并糖尿病及COPD、长期应用广谱抗生素、病变累及肺野数、JAK2及STAT3。通过Bootstrap法行内部验证,一致性指数(C-index)为0.909,表明肺结核并发真菌感染预测模型有良好的一致性。决策曲线分析结果显示,该列线图模型阈值概率在2%~100%时可获得临床净收益。结论合并糖尿病及COPD、长期应用广谱抗生素、病变累及肺野数>3个、JAK2及STAT3蛋白相对表达水平较高是肺结核并发真菌感染的独立危险因素,基于JAK/STAT信号通路及危险因素构建的列线图预测模型具有较高的预测价值。
Objective To explore the influencing factors of pulmonary tuberculosis complicated with fungal infection,construct and validate a nomogram prediction model based on the Janus kinase(JAK)/signal transducer and activator of transcription(STAT)signaling pathway.Methods A total of 308 patients with pulmonary tuberculosis diagnosed and treated in our hospital were retrospectively selected.They were divided into concurrent group(35 cases)and non⁃concurrent group(273 cases)according to whether there was concurrent fungal infection.Observe the distribution of pathogenic fungi in concurrent group of patients,collect the clinical data and laboratory test results of all patients.Univariate and multivariate logistic regression analyses were used to evaluate the risk factors of pulmonary tuberculosis complicated with fungal infection,and a nomogram prediction model was constructed based on this.The receiver operating characteristic curve(ROC)was used to evaluate the related risk factors and the predictive value of this used to conduct internal validation of the nomogram prediction model,the calibration degree of the nomogram model was evaluated by using the calibration curve,and the clinical net rate of return of the nomogram model was verified by using the decision curve analysis.Results Among the 35 cases of concurrent fungal infections,23 cases(65.71%)were Aspergillus,8 cases(22.86%)were Candida albicans,3 cases(8.57%)were Candida tropicalis,and 1 case(2.86%)was Candida glabra.Multivariate logistic regression analysis showed that combined diabetes and chronic obstructive pulmonary disease(COPD),long⁃term use of broad⁃spectrum antibiotics,the number of lung fields involved by the lesion>3,and the relatively high expression levels of JAK2 and STAT3 proteins were all independent risk factors for pulmonary tuberculosis complicated with fungal infection(P<0.05).The results of ROC curve analysis showed that the AUC of the nomogram prediction model was greater than that of combined diabetes and COPD,long⁃term use of broad⁃spectrum antibiotics,the number of lung fields involved by lesions,JAK2 and STAT3.Internal verification was conducted through the Bootstrap method,and the consistency index(C⁃index)was 0.909,indicating that the prediction model for pulmonary tuberculosis complicated with fungal infection has good consistency.Decision curve analysis results showed that the clinical net benefits could be obtained when the threshold probability of this nomogram model was between 2%and 100%.Conclusion Diabetes mellitus,chronic pulmonary obstruction,long⁃term use of broad⁃spectrum antibiotics,the number of lung fields involved in the disease>3,high JAK2 and high STAT3 protein relative expression are independent risk factors for pulmonary tuberculosis complicated with fungal infection.The prediction model based on the JAK/STAT signaling pathway and risk factors has high predictive value.
作者
曾德斌
钟永强
刘萍萍
吴杨宗
Zeng Debin;Zhong Yongqiang;Liu Pingping;Wu Yangzong(Departmentof Infectious Diseases,Huichang County People’s Hospital,Ganzhou 342600,China)
出处
《临床内科杂志》
2025年第6期474-478,共5页
Journal of Clinical Internal Medicine
基金
赣州市卫生局课题(GZWJW202402407)。
