结缔组织病相关间质性肺疾病患者呈进展性肺纤维化表型的影响因素

Factors influencing the phenotype of progressive pulmonary fibrosis in patients with connective tissue disease-associated interstitial lung disease

目的分析结缔组织病相关间质性肺疾病患者呈进展性肺纤维化的临床特点,并对相应的影响因素实施分析,构建相应的数学模型进行预测效能计算。方法采用横断面研究,选取西安国际医学中心医院2020年1月至2023年12月治疗的结缔组织病相关间质性肺疾病患者219例的临床资料进行回顾性分析。观察组为进展性纤维化间质性肺疾病患者32例,对照组为非进展性纤维化间质性肺疾病患者187例。采用单因素分析(t检验、χ^(2)检验)、多因素分析(logisitic回归分析),构建进展性纤维化性间质性肺疾病的预测模型和预测效能分析[列线图模型、受试者工作特征曲线(ROC曲线)]。结果观察组咳嗽、呼吸急促的比例分别为40.63%(13/32)、40.63%(13/32),均显著高于对照组的16.58%(31/187)、16.04%(30/187)(χ^(2)=9.84、10.46,均P<0.05)。观察组白细胞计数、中性粒细胞比值、红细胞沉降率、C反应蛋白均高于对照组,白蛋白低于对照组,差异均有统计学意义(t=-2.69、-2.15、-9.27、-6.78、2.18,均P<0.05)。观察组免疫球蛋白A(IgA)[(3.18±0.44)g/L]显著高于对照组[(2.32±0.32)g/L](t=-13.23,P<0.05)。观察组癌胚抗原[(2.73±1.03)μg/L]、糖类抗原19-9(CA199)[(14.35±4.03)kU/L]均显著高于对照组[(1.53±0.24)μg/L、(8.98±2.71)kU/L](t=13.99、9.56,均P<0.05)。观察组条索状影的比例显著低于对照组、蜂窝状改变的比例显著高于对照组(χ^(2)=6.45、14.55,均P<0.05)。logistic多因素分析结果显示,IgA、CA199、蜂窝状改变是患者诊断为进展性纤维化性间质性肺疾病的独立影响因素(OR=17.13、1.42、8.01,均P<0.05)。经过验证,logistic多因素回归模型具有良好的拟合优度(Hosmer-Lemeshowχ^(2)=0.11,P>0.05)。依据多因素分析所筛选出来的变量构建列线图风险模型,C-index为0.71。利用Bootstrap自抽样法进行内部验证,校准曲线平均绝对误差为0.02。利用logistic回归模型的独立影响因素以及P值预测概率对患者诊断为进展性纤维化性间质性肺疾病的情况进行ROC曲线的预测,曲线下面积分别为0.73、0.73、0.68、0.86。回归模型P值预测概率的ROC曲线下面积明显高于其他独立影响因素指标(Z=-3.81、-3.66、-4.99,均P<0.05)。结论IgA浓度增加、CA199水平上升以及高分辨率CT显示的蜂窝状变化可能预示着结缔组织病相关间质性肺疾病的肺纤维化进展。对于这类患者,应严密监测临床表现,尽快完成全面的实验室和影像学评估,并适时采用抗纤维化治疗。

Objective To analyze the clinical characteristics of progressive pulmonary fibrosis in patients with interstitial lung disease associated with connective tissue disease,identify the relevant influential factors,and construct a corresponding mathematical model to evaluate prediction efficiency.MethodsIn a cross-sectional study,the clinical data of 219 patients with interstitial lung disease associated with connective tissue disease,who were treated at Xi'an International Medical Center Hospital from January 2020 to December 2023,were retrospectively collected and analyzed.Among these patients,32 were diagnosed with progressive fibrotic interstitial lung disease(observation group),while the remaining 187 patients were diagnosed with non-progressive fibrotic interstitial lung disease(control group).Univariate analysis(t-test andχ^(2)test)and multivariate analysis(logistic regression)were conducted to build a predictive model for progressive fibrotic interstitial lung disease.The prediction efficacy was evaluated using a column chart model and receiver operating characteristic curve analysis.Results In the observation group,the proportions of patients experiencing cough and shortness of breath were 40.63%(13/32)and 40.63%(13/32),respectively,which were significantly higher than those in the control group[16.58%(31/187),16.04%(30/187),χ^(2)=9.84,10.46,both P<0.05].The observation group also exhibited significantly higher levels of white blood cell count,neutrophil-to-lymphocyte ratio,erythrocyte sedimentation rate,and C-reactive protein,while albumin levels were significantly lower compared with the control group(t=-2.69,-2.15,-9.27,-6.78,2.18,all P<0.05).The immunoglobulin A levels in the observation group[(3.18±0.44)g/L]were significantly higher than those in the control group[(2.32±0.32)g/L,t=-13.23,P<0.05].Additionally,the levels of carcinoembryonic antigen[(2.73±1.03)μg/L]and carbohydrate antigen 19-9[(14.35±4.03)kU/L]in the observation group were significantly higher compared with the control group[(1.53±0.24)μg/L,(8.98±2.71)kU/L,t=13.99,9.56,both P<0.05].The proportion of reticular opacities in the observation group was significantly lower than that in the control group,whereas the proportion of honeycomb changes was significantly higher in the observation group compared with the control group(χ^(2)=6.45,14.55,both P<0.05).Multivariate logistic analysis revealed that immunoglobulin A,carbohydrate antigen 19-9,and honeycomb changes were independent influential factors associated with the diagnosis of progressive fibrotic interstitial lung disease(OR=17.13,1.42,8.01,all P<0.05).The multivariate logistic regression model demonstrated good fit(Hosmer-Lemeshowχ^(2)=0.11,P>0.05).A nomogram risk model was constructed based on the variables identified in the multivariate analysis,with a C-index of 0.71.Internal validation using the Bootstrap sampling method indicated an average absolute error of the calibration curve of 0.02.Receiver operating characteristic curve analysis based on the independent influential factors and their corresponding P-values showed areas under the curve of 0.73,0.73,0.68,and 0.86,respectively.The area under the curve for the predicted probability of the regression model was significantly higher than those for the other independent influential factors(Z=-3.81,-3.66,-4.99,all P<0.05).Conclusion Increased immunoglobulin A concentration,elevated carbohydrate antigen 19-9 levels,and honeycomb changes observed on high-resolution CT scans may indicate the progression of pulmonary fibrosis in patients with interstitial lung disease associated with connective tissue disease.Close monitoring of clinical symptoms,comprehensive laboratory and imaging evaluations,and timely antifibrotic treatment are recommended for these patients.