基线HRCT影像特征对进行性肺纤维化的有限预测能力及其增量价值评估

Limited Predictive Performance and Incremental Value of Baseline HRCT Features for Progressive Pulmonary Fibrosis

摘要: 目的：评估间质性肺病(interstitial lung disease, ILD)患者基线高分辨率CT (high-resolution computed tomography, HRCT)影像特征在进行性肺纤维化(progressive pulmonary fibrosis, PPF)风险识别中的作用及其增量价值。方法：回顾性纳入2018至2025年我院确诊并完成基线HRCT及随访的ILD患者77例，根据2022年ATS/ERS/JRS/ALAT指南分为PPF组(n = 26)和非PPF组(n = 51)。收集基线临床及HRCT特征，采用Logistic回归分析相关因素，并以ROC曲线评价判别效能。结果：PPF组年龄高于非PPF组(P < 0.001)。除年龄外，基线肺功能、KL-6及主要HRCT征象差异均无统计学意义(均P > 0.05)。年龄与PPF相关(AUC = 0.759)，单个HRCT变量判别能力有限(AUC = 0.503~0.606)。年龄联合纤维化相关累及肺叶数后，模型AUC提高至0.819。结论：在本研究特定的单中心、小样本ILD队列中，单个基线HRCT变量对PPF的判别能力有限。年龄与PPF相关，加入纤维化相关累及肺叶数后AUC仅数值上升；鉴于样本量有限、队列异质性及OR方向异常，该结果仅具探索性，尚不能证明稳定的独立或增量预测价值。PPF评估仍需结合临床、肺功能及纵向影像，且本结论不宜直接外推至其他ILD人群。

Abstract: Objective: To evaluate the role and incremental value of baseline high-resolution computed tomography (HRCT) features in risk assessment for progressive pulmonary fibrosis (PPF) in patients with interstitial lung disease (ILD). Methods: A total of 77 patients with ILD who were diagnosed at our hospital between 2018 and 2025 and completed baseline HRCT and follow-up examinations were retrospectively enrolled. According to the 2022 ATS/ERS/JRS/ALAT guideline criteria, patients were classified into the PPF group (n = 26) and non-PPF group (n = 51). Baseline clinical and HRCT features were collected. Logistic regression analysis was performed to identify associated factors, and receiver operating characteristic (ROC) curves were used to evaluate discriminative performance. Results: Patients in the PPF group were older than those in the non-PPF group (P < 0.001). No significant differences were observed in baseline pulmonary function, KL-6 levels, or major HRCT features between the two groups (all P > 0.05). Age was associated with PPF (AUC = 0.759), whereas individual HRCT variables showed limited discriminative performance (AUC = 0.503~0.606). After adding the number of lobes involved by fibrosis-related abnormalities to age, the model AUC increased to 0.819. Conclusion: In this specific single-center, small ILD cohort, individual baseline HRCT features showed limited discrimination for PPF. Age was associated with PPF, and adding the number of lobes involved by fibrosis-related abnormalities only numerically increased the AUC. Given the limited sample size, cohort heterogeneity, and unexpected OR direction, this finding remains exploratory and does not establish stable independent or incremental predictive value. PPF assessment should integrate clinical, pulmonary function, and longitudinal imaging data, and these findings should not be generalized beyond this cohort.