D23-18 Change in Oxygen Saturation During the Six-minute Walk Test Predicts Lung Function Decline and Outcomes in Idiopathic Pulmonary Fibrosis - Analyses From the JIPS Registry and K-Junko Study

Abstract Rationale Idiopathic pulmonary fibrosis (IPF) exhibits heterogeneous progression despite an overall poor prognosis. Although forced vital capacity (FVC) decline is an established surrogate for disease progression, individual FVC trajectories vary considerably, limiting its predictive precision. We evaluated whether the change in oxygen saturation (ΔSpO2) between pre- and post-six-minute walk test (6MWT) could predict an absolute ≥5% FVC decline and transplant-free survival (TFS), serving as an early physiological marker complementary to baseline FVC and six-minute walk distance (6MWD). Methods The IPF cohort in the JIPS Registry (ClinicalTrials.gov identifier NCT03041623) was used as the training cohort, and the K-Junko Study served as the validation cohort. The JIPS Registry included patients with idiopathic interstitial pneumonias newly enrolled from 85 institutions across Japan, and the K-Junko Study included those with IPF from the Kanagawa Cardiovascular and Respiratory Center. Parameters from the 6MWT (minimum SpO2, 6MWD, and ΔSpO2) and their optimal thresholds for predicting a ≥ 5% FVC decline were identified using a univariate Cox proportional hazards model. Predictive ability was confirmed in the validation cohort using a multivariate Cox model. A nomogram was then constructed to predict a ≥ 5% FVC decline and its performance for predicting both FVC decline and TFS was evaluated using the concordance index (C-index). Results Among 866 patients with idiopathic interstitial pneumonias in the JIPS Registry, 457 had IPF; the K-Junko Study included 318 IPF patients. A ΔSpO2 ≥6% showed the strongest association with ≥5% FVC decline (HR 1.301), whereas minimum SpO2 and 6MWD were not significant. In multivariate analysis, ΔSpO2 ≥6% remained an independent predictor for a ≥ 5% FVC decline. Using this threshold, time to ≥ 5% FVC decline was significantly shorter in both the JIPS (HR 1.292, 95% CI 1.034-1.615, P = 0.024) and K-Junko cohorts (HR 1.480, 95% CI 1.095-2.000, P = 0.010). Moreover, more patients with ΔSpO2 ≥6% met progressive pulmonary fibrosis criteria within 1 year (46.3% vs 36.0%, P = 0.039). The nomogram achieved C-indices of 0.619 and 0.632 in the training and validation cohorts, respectively. The favorable group (FVC ≥80%, ΔSpO2 6%, 6MWD ≥420 m) had better TFS (HR 5.028, 95% CI 2.726-9.273, P 0.001; HR 2.751, 95% CI 1.580-4.793, P 0.001). Conclusions ΔSpO2 during the 6MWT is an independent predictor of ≥ 5% FVC decline in IPF. Integrating ΔSpO2 with baseline FVC and 6MWD enables early identification of subgroups with favorable outcomes, supporting its use in risk stratification and clinical trial enrichment. This abstract is funded by: Boehringer Ingelheim