A95-04 Associations Between Changes in HRCT Scores and Transplant-Free Survival in Patients With Progressive Pulmonary Fibrosis

Abstract Rationale Predicting the clinical trajectory of patients with progressive pulmonary fibrosis (PPF) remains challenging. High-resolution computed tomography (HRCT) offers detailed visualization of lung architecture in patients with interstitial lung disease (ILD). However, the prognostic significance of changes in radiographic extent of lung fibrosis in patients with PPF, particularly after accounting for change in lung function, has not been well established. Methods The cohort included 138 patients with PPF enrolled in the ILD-PRO Registry. All patients had an ILD other than idiopathic pulmonary fibrosis, met criteria for ILD progression, and had undergone at least two HRCT scans. A previously developed machine learning algorithm was employed to determine the total burden of ILD (quantitative QILD QILD score) as a sum of the extent of lung fibrosis (quantitative lung fibrosis QLF score), ground-glass opacity, and honeycombing. The QILD and QLF scores were expressed as percentages of total lung involvement. To evaluate associations between change in QILD or QLF score and time to death or lung transplant, Cox proportional hazards models were fit. Models were landmarked at the time of the second scan and adjusted for QILD or QLF score on the first scan and time between the scans. In a subset of patients (n = 91) who had pulmonary function tests within 3 months of both scans, models were also adjusted for change in FVC % predicted. Results The cohort had a median (Q1, Q3) age of 66.0 (55.0, 72.0) years; 63% were female. Median FVC and DLCO % predicted were 63.8 (51.4, 72.6) and 39.8 (30.9, 51.1), respectively. The median time between the HRCT scans was 12.1 (7.1, 17.6) months, and median changes in QLF and QILD scores were 0.9 (-1.9, 5.1) and 1.1 (-8.2, 8.0), respectively. Over a median follow-up of 23.4 (10.5, 33.2) months, 49 (35.5%) patients experienced the outcome of death or lung transplant. Each 5% absolute increase in QLF score was significantly associated with a 47% higher risk of death or lung transplant (HR 1.47 1.34, 1.62; Figure). Similar associations were observed for increase in QILD score (HR 1.45 1.29, 1.63; Figure). These associations remained significant in models adjusting for change in FVC % predicted (Figure). Conclusions In patients with PPF, increases in the radiographic extent of lung fibrosis were associated with worse transplant-free survival, independent of changes in lung function. Our findings support the prognostic value of serial quantitative HRCT assessments in clinical and research settings. This abstract is funded by: The IPF-PRO/ILD-PRO Registry is supported by Boehringer Ingelheim Pharmaceuticals, Inc and run in collaboration with the Duke Clinical Research Institute and enrolling centers.