Article Text
PDF
Abstract
Background High resolution computed tomography (HRCT) scanning is able to detect abnormalities consistent with interstitial lung disease (ILD). However, if only a small proportion of lung is affected, radiologists variously report this as ‘minimal’, ‘minor’ or ‘early’ ILD. There is no definition of what constitutes ‘minimal’ ILD and the natural history of these patients is not known.
Aims To define ‘minimal’ ILD, test observer agreement with this definition and describe the characteristics and survival of these patients.
Hypothesis Minimal ILD can be defined by subjective quantification and has a benign course.
Methods Between 01.01.2002 and 31.12.2014 the Edinburgh Lung Fibrosis Database was prospectively populated with data for 1450 consecutively presenting patients with ILD. Of these, 56 were identified as presenting with ‘minimal’ disease according to HRCT. Three radiologists participated in a modified Delphi exercise and agreed on a definition of ‘minimal’ ILD. A sample (n = 38) of HRCT scans was provided to test inter- and intra-observer agreement according to this definition using Fleiss’ Kappa statistics. Survival was assessed using Kaplan-Meier curves.
Results The Delphi exercise resulted in ‘minimal’ disease being defined as ILD involving <5% of the total lung volume and/or <10% of the lung peripheries. Using this definition, inter-observer and intra-observer agreement was moderate (kappa 0.42 and 0.58 respectively). Of the 56 subjects originally deemed as ‘minimal’ ILD, 48 were unanimously described as minimal disease by post-definition criteria. One subject was biopsied (consensus after biopsy, unclassifiable). Forty-seven subjects were not biopsied and none met ATS/ERS consensus criteria for diagnosing IPF. Most subjects had ‘unclassifiable’ disease, but the working diagnoses were; IPF or other fibrotic idiopathic interstitial pneumonia (IIP) (n = 34), IIP without fibrosis (n = 7) and connective-tissue disease associated ILD (n = 7). The median age was 69yrs, 56% were male and 23% had never smoked. The mean (SD) %pred lung function was; FEV1 91.8% (19), VC 101% (18) TCO 62% (19). The median survival was 11.6 years, and all deaths (n = 12) were attributable to respiratory disease.
Summary Defining ‘minimal’ ILD is feasible and there was moderate radiological agreement. Minimal ILD is relatively benign, but the associated mortality was of respiratory cause.
Survival of minimal ILD