@article{Mao:2026:10.1016/j.pccm.2026.02.004,
author = {Mao, R and Zhang, H and Chang, Q and Teng, Y and Ni, Y and Chen, J and Li, M and Xu, N and Zhang, H and Chen, Y and Sun, J and Chung, KF and Renzoni, EA and Lu, Y and Dai, H and Li, F},
doi = {10.1016/j.pccm.2026.02.004},
journal = {Chin Med J Pulm Crit Care Med},
pages = {81--91},
title = {Progression of interstitial lung abnormalities and its impact on mortality in patients with lung cancer resection.},
url = {http://dx.doi.org/10.1016/j.pccm.2026.02.004},
volume = {4},
year = {2026}
}

Download

TY  - JOUR
AB  - BACKGROUND: Interstitial lung abnormalities (ILAs) are incidental abnormal lung findings on computed tomography that can co-exist with lung cancer, but risk factors for their progression and impact on survival after lung cancer resection remain unclear. This study aimed to identify risk factors for ILA progression, develop a radiomics-based model to predict it, and evaluate the association between ILA progression and mortality. METHODS: Patients with ILAs who underwent lung cancer resection in Shanghai Chest Hospital between January 2016 and May 2019 were retrospectively selected and divided into three subcategories at baseline: non-subpleural ILAs, subpleural non-fibrotic ILAs, and subpleural fibrotic ILAs; and three subcategories during follow-up: improved ILAs, unchanged ILAs, and progressive ILAs. Multivariate logistic regression was used to identify clinical and laboratory risk factors associated with ILA progression, and radiomics-based machine learning models were independently constructed to predict ILA progression using baseline CT imaging features. Survival data were analyzed using Kaplan-Meier curves and Cox proportional hazards regression. RESULTS: There were 1363 ILA cases among 10,295 patients who underwent primary lung cancer resection, with a proportion of 13.24%. After 2- and 4-year follow-up, the progression rates of ILAs were 9.86% (65/659) and 10.19% (43/422), respectively. Subpleural fibrotic ILAs (2-year follow-up: OR=6.078, 95% confidence interval [CI]: 2.633-14.028, P<0.001; 4-year follow-up: OR=3.339, 95% CI: 1.085-10.272, P=0.035), and radiotherapy (OR=13.595, 95% CI: 4.540-40.710, P<0.001; OR=11.496, 95% CI: 2.864-46.141, P=0.001) were risk factors for ILA progression. Using radiomics features extracted from baseline CT, the AutoGluon machine learning model demonstrated high performance in predicting ILA progression among patients with confirmed ILAs (for all ILAs, mean area under curve value: 0.790, accuracy: 0.801 &
AU  - Mao,R
AU  - Zhang,H
AU  - Chang,Q
AU  - Teng,Y
AU  - Ni,Y
AU  - Chen,J
AU  - Li,M
AU  - Xu,N
AU  - Zhang,H
AU  - Chen,Y
AU  - Sun,J
AU  - Chung,KF
AU  - Renzoni,EA
AU  - Lu,Y
AU  - Dai,H
AU  - Li,F
DO  - 10.1016/j.pccm.2026.02.004
EP  - 91
PY  - 2026///
SP  - 81
TI  - Progression of interstitial lung abnormalities and its impact on mortality in patients with lung cancer resection.
T2  - Chin Med J Pulm Crit Care Med
UR  - http://dx.doi.org/10.1016/j.pccm.2026.02.004
UR  - https://www.ncbi.nlm.nih.gov/pubmed/41970198
VL  - 4
ER  - 

Download