Heterogeneity in Unclassifiable Interstitial Lung Disease. A Systematic Review and Meta-Analysis

Abstract

Background: Accurate diagnosis of interstitial lung disease (ILD) is necessary to identify the most appropriate management strategy and to inform prognosis. Many patients cannot be provided a confident diagnosis despite an exhaustive search for potential etiologies and review in a multidisciplinary conference; and are consequently labelled with unclassifiable ILD. Objective: to systematically review and meta-analyse previous studies reporting on the diagnostic criteria, prevalence, clinical features, and outcome of unclassifiable ILD. Data sources: Medline, Embase, and the Cochrane Central Register of Controlled Trials databases were systematically searched for all studies related to unclassifiable ILD published before September 1, 2017. Data extraction: Two authors independently screened each citation for eligibility criteria, serially reviewing the title, abstract, and full-text manuscript, and then abstracted data pertaining to the study objectives from eligible studies. Articles were stratified by risk of selection bias, whether the publication stated that patients were reviewed in a multidisciplinary discussion, and by the frequency of surgical lung biopsy. Meta-analyses and meta-regression were performed to calculate the pooled prevalence of unclassifiable ILD within an ILD population and within specific subgroups to identify reasons for across-study heterogeneity. Synthesis: The search identified 10130 unique citations, 313 articles underwent full-text review, and eligibility criteria were met in 88 articles. Twenty-two studies were deemed low risk of selection bias, including 1060 patients with unclassifiable ILD from a total of 10174 patients with ILD. The terminology and definition of unclassifiable ILD varied substantially across publications, with inconsistent diagnostic criteria and evaluation processes. The prevalence of unclassifiable ILD was 11.9% (95% confidence interval 8.5% to 15.6%), with lower prevalence in centers that reported use of a formal multidisciplinary discussion of cases (9.5% vs. 14.5%). Four articles reported survival of unclassifiable ILD, with 1-year, 2-year, and 5-year survival of 84-89%, 70-76%, and 46-70%, respectively. Conclusions: This systematic review and meta-analysis shows that unclassifiable ILD is common but has substantial heterogeneity and inconsistent definitions across ILD cohorts. These findings highlight important limitations in multicentre studies of fibrotic ILD and the need for a standardized approach to ILD diagnostic classification.