Oxygenation deterioration and computed tomography findings in acute exacerbation of interstitial lung diseases: a prospective observational study

BACKGROUND: Acute exacerbation (AE) is a critical prognostic event in interstitial lung diseases (ILDs), initially recognized in idiopathic pulmonary fibrosis (IPF). Current AE diagnostic criteria do not account for oxygenation deterioration, and evidence of new computed tomography (CT) findings alone is insufficient to diagnose AE-ILD. We investigated a CT-based criterion for distinguishing AE from acute deterioration (AD)-ILD and assessed the prognostic significance of oxygenation deterioration in AE-ILD. METHODS: Patients with AD-ILD, defined by worsening shortness of breath and new lung abnormalities within 1 month, were prospectively enrolled. AD-ILD included both AE-ILD and other acute conditions, non-AE-AD-ILD. Oxygenation deterioration was assessed using modified Blancal criteria. CT-based AE criteria were developed through receiver operating characteristic (ROC) curve and multivariate logistic regression analyses based on CT findings suggestive of AE. RESULTS: Overall, 394 patients were analyzed, including 303 diagnosed with AE and 91 with non-AE-AD-ILD. The underlying ILDs included IPF (n = 177) and non-IPF ILDs (n = 217). The presence of at least two of three CT-based parameters, including bilateral shadows, ground-glass opacity predominance, and diffuse pattern, was identified as predictive of AE diagnosis in AD-ILDs using logistic and ROC analyses. Ninety-day survival was significantly lower in patients with AE-ILD with oxygenation deterioration than in those without deterioration and in patients with non-AE-AD-ILDs. Oxygenation deterioration independently predicted poor survival in multivariate Cox proportional hazard regression analysis. CONCLUSIONS: The developed CT-based AE criteria effectively predicted AE and oxygenation deterioration in AD-ILD. Additionally, oxygenation deterioration predicted worse survival in AE-ILD. This study was registered with the University Hospital Medical Information Network (UMIN000027844; registered 1 July 2017).