{"title":"Usual interstitial pneumonia on CT: same pattern, different prognosis","authors":"Fasihul Khan","doi":"10.1136/thorax-2025-223598","DOIUrl":null,"url":null,"abstract":"A radiological pattern of usual interstitial pneumonia (UIP) has long been recognised as a marker of poor prognosis in interstitial lung disease (ILD).1 While UIP is the defining pattern in idiopathic pulmonary fibrosis (IPF), it can also be seen in other ILDs, including connective tissue disease ILD (CTD-ILD) and fibrotic hypersensitivity pneumonitis (fHP).2–4 It is well understood that CTD-ILD typically follows a more indolent course than IPF, but the prognostic implications of a UIP pattern within these non-IPF diagnoses remain incompletely understood.5 6 Does the presence of UIP override disease-specific biology, or does diagnostic context remain paramount? Kim and colleagues7 explore this question using two large, well characterised multicentre cohorts—the Pulmonary Fibrosis Foundation (PFF) Patient Registry and an international meta-cohort drawn from four academic centres. Across more than 2500 participants, the authors evaluated lung function decline and transplant-free survival in patients with IPF, CTD-ILD or fHP, stratified by the presence or absence of a radiological UIP pattern. They also incorporated data-driven texture analysis (DTA), a machine-learning approach to quantify fibrosis on CT, highlighting the evolving role of artificial intelligence in ILD phenotyping.8–10 Their findings demonstrated that a UIP pattern does not carry the same clinical course across ILD subtypes. Despite sharing a …","PeriodicalId":23284,"journal":{"name":"Thorax","volume":"15 1","pages":""},"PeriodicalIF":7.7000,"publicationDate":"2025-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Thorax","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1136/thorax-2025-223598","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"RESPIRATORY SYSTEM","Score":null,"Total":0}
引用次数: 0
Abstract
A radiological pattern of usual interstitial pneumonia (UIP) has long been recognised as a marker of poor prognosis in interstitial lung disease (ILD).1 While UIP is the defining pattern in idiopathic pulmonary fibrosis (IPF), it can also be seen in other ILDs, including connective tissue disease ILD (CTD-ILD) and fibrotic hypersensitivity pneumonitis (fHP).2–4 It is well understood that CTD-ILD typically follows a more indolent course than IPF, but the prognostic implications of a UIP pattern within these non-IPF diagnoses remain incompletely understood.5 6 Does the presence of UIP override disease-specific biology, or does diagnostic context remain paramount? Kim and colleagues7 explore this question using two large, well characterised multicentre cohorts—the Pulmonary Fibrosis Foundation (PFF) Patient Registry and an international meta-cohort drawn from four academic centres. Across more than 2500 participants, the authors evaluated lung function decline and transplant-free survival in patients with IPF, CTD-ILD or fHP, stratified by the presence or absence of a radiological UIP pattern. They also incorporated data-driven texture analysis (DTA), a machine-learning approach to quantify fibrosis on CT, highlighting the evolving role of artificial intelligence in ILD phenotyping.8–10 Their findings demonstrated that a UIP pattern does not carry the same clinical course across ILD subtypes. Despite sharing a …
期刊介绍:
Thorax stands as one of the premier respiratory medicine journals globally, featuring clinical and experimental research articles spanning respiratory medicine, pediatrics, immunology, pharmacology, pathology, and surgery. The journal's mission is to publish noteworthy advancements in scientific understanding that are poised to influence clinical practice significantly. This encompasses articles delving into basic and translational mechanisms applicable to clinical material, covering areas such as cell and molecular biology, genetics, epidemiology, and immunology.