Construction and evaluation of a nomogram model of progressive pulmonary fibrosis phenotype in interstitial lung disease associated with primary Sj?gren's syndrome
10.3969/j.issn.1671-8348.2025.03.018
- VernacularTitle:原发性干燥综合征相关间质性肺疾病呈进展性肺纤维化表型的列线图模型构建及评价
- Author:
Chuanrong ZHANG
1
;
Yongping CHEN
Author Information
1. 锦州医科大学研究生学院,辽宁 锦州 121001;重庆大学附属三峡医院风湿免疫科,重庆 404000
- Keywords:
primary Sj?gren's syndrome;
interstitial lung disease;
progressive pulmonary fibrosis;
risk factors;
nomogram
- From:
Chongqing Medicine
2025;54(3):659-667
- CountryChina
- Language:Chinese
-
Abstract:
Objective To identify risk factors of progressive pulmonary fibrosis(PPF)development in patients with primary Sj?gren's syndrome-associated interstitial lung disease(pSS-ILD),and to develop and validate a clinically applicable nomogram prediction model.Methods Enrolled 168 pSS-ILD patients from Three Gorges Hospital affiliated to Chongqing University from January 2019 to January 2023,randomly allo-cated into training(n=117)and internal validation(n=51)cohorts at 7∶3 ratio.The training cohort was stratified into the PPF group(n=38)and the non-PPF group(n=79)based on one-year radiographic pro-gression.An external validation cohort(n=24)was prospectively enrolled from February 2023 to April 2024.Multivariable analyses incorporated demographic characteristics,clinical manifestations,laboratory parame-ters,pulmonary function tests,and high-resolution CT(HRCT)imaging features.A nomogram was developed using independent predictors identified through logistic regression.Receiver operating characteristic(ROC)curve,area under curve(AUC),calibration curve,Hosmer-Lemeshow test and decision curve were used to e-valuate the prediction efficiency of the model.Results PPF developed in 31.0%(52/168)of patients within one year.Univariate and multivariate logistic regression analysis showed that age,25 hydroxyvitamin D level and high resolution CT fibrosis score were independent risk factors for PPF phenotype in pSS-ILD patients(P＜0.05).Based on the above predictors,a column-line diagram was constructed.Internal verification showed that the column-line differentiation ability was good(AUC=0.88,95%CI:0.79-0.98),and exter-nal verification showed that the predicted probability of the calibration curve was close to the actual probabili-ty(HL test χ2=9.516,P=0.301),indicating that the model had good consistency.Conclusion Age,serum 25 hydroxyvitamin D level and high resolution CT fibrosis score are independent risk factors for PPF phenotype in pSS-ILD patients.The nomogram model constructed according to the above three predictors has good differentiation and consistency,and could provide a reference for the clinical prediction of PPF in pSS-ILD patients.
