Real-World Management and Outcomes of Probable and Indeterminate UIP Patterns on HRCT: A Retrospective Cohort Study and Implications for the 2022 IPF Guideline

Background: The 2018 ATS/ERS/JRS/ALAT guideline conditionally recommended surgical lung biopsy (SLB) for patients with probable or indeterminate usual interstitial pneumonia (UIP) patterns on high-resolution computed tomography (HRCT). This study assessed the real-world use of this recommendation and outcomes of patients with probable or indeterminate UIP. Methods: HRCT scans of 1255 patients with interstitial lung disease (ILD) followed between 2012–2021 were retrospectively reviewed. A total of 185 patients met the inclusion criteria. We re-evaluated the HRCT scans and classified them through multidisciplinary discussion (MDD) involving one thoracic radiologist, one pulmonary pathologist, one rheumatologist, and at least three pulmonologists specializing in ILD. We analyzed clinical features, biopsy rates, final diagnoses, and survival outcomes. Results: SLB was performed in 10.8% of patients and met the predefined criteria for clinical biopsy success in 50%. Among patients with probable UIP who were ultimately diagnosed with IPF, 32 of 38 (84.2%) received the diagnosis without SLB. The remaining six (15.8%) underwent SLB, of whom five met predefined criteria for clinical biopsy success. Final diagnoses included IPF (23.8%), hypersensitivity pneumonitis (16.8%), and connective tissue disease-associated ILD (16.2%). Crude mortality proportions did not differ significantly between the probable and indeterminate UIP groups (p = 0.173). However, time-to-event analyses demonstrated shorter survival in the probable UIP group (log-rank p = 0.019), and probable UIP was associated with a higher mortality hazard after adjustment for age, sex, and smoking history (adjusted HR 1.91, 95% CI 1.06–3.42; p = 0.030). This association was attenuated and no longer statistically significant after additional adjustment for cardiovascular disease (adjusted HR 1.71,95% CI 0.93–3.14; p = 0.082). Conclusions: These findings support an individualized, multidisciplinary diagnostic approach and are consistent with the diagnostic flexibility introduced in the 2022 IPF guideline.