Abstract
Background: Both a probable usual interstitial pneumonia (UIP) pattern (p-UIP) and a definite UIP pattern (d-UIP) on high-resolution computed tomography (HRCT) are sufficient to establish a diagnosis of idiopathic pulmonary fibrosis (IPF) without the need for a surgical lung biopsy, according to the 2022 IPF guidelines. However, it remains unknown whether patients with p-UIP and d-UIP have similar disease behaviors and clinical courses. Material and Methods: We retrospectively collected clinical data of patients with IPF and divided the patients into two groups according to their HRCT patterns: a p-UIP group and a d-UIP group. The baseline characteristics, survival rates, and pulmonary function tests were compared between the two groups. The risk factors for mortality were determined by Cox regression in p-UIP and d-UIP separately. Results: There were 304 patients in the p-UIP group and 480 patients in the d-UIP group. Patients in the d-UIP group were more likely to have smoking histories (p < 0.001) and had lower baseline FVC% (74% vs. 77%, p = 0.021) and DLCO% (50% vs. 58%, p < 0.001). Survival rates were higher in p-UIP compared with d-UIP (p = 0.004). There were no differences in changes in FVC% or DLCO% between the two groups. Baseline DLCO% was the only independent risk factor for mortality in p-UIP. Baseline FVC% was independently associated with mortality in d-UIP. Symptom of cough was a risk factor for disease progression (OR = 1.2, p = 0.002) in p-UIP, while symptom of dyspnea might be associated with disease progression in d-UIP (OR = 2.7, p = 0.065). Male patients (OR = 1.88, p = 0.002) with a smoking history (OR = 1.16, p = 0.002) were at higher risk of developing d-UIP. Conclusions: We observed the different disease trajectories between p-UIP and d-UIP. P-UIP on HRCT might identify a subgroup of IPF patients who are in the early stage with a better prognosis.