Institutional patterns in diagnosis and antifibrotic-era treatment uptake for IPF and PPF across U.S. centers

For ten years, nintedanib and pirfenidone have been the only approved therapies for idiopathic pulmonary fibrosis (IPF), with nintedanib subsequently also approved for progressive pulmonary fibrosis (PPF). Previous studies demonstrate low antifibrotic (AF) uptake, but differences by site-of-care and AF uptake after seminal trials (IPF: INPULSIS/ASCEND; PPF: INBUILD) are not well characterized. We evaluated temporal trends in IPF and PPF diagnosis and AF use at academic (AC) versus non-academic (NonAC) centers.

Using the U.S. TriNetX Collaborative Network, we compared IPF and PPF incidence, prevalence, and AF use by center type using linear regression across post-trial and post-FDA-approval windows (IPF: 2015-2017, 2015-2024; PPF: 2020-2022, 2020-2024). Slope differences (Δslope) were calculated as AC minus NonAC annual slopes; therefore, positive values indicate faster increases at ACs, whereas negative values indicate faster increases at NonACs.

Across both periods, IPF prevalence increased more rapidly at NonACs, while AF use was higher at ACs (2024: 33.5% vs 20.1%, p < 0.001). During the post-approval period (2015-2024), AF initiation and overall use rose faster at ACs: (AC-NonAC Δ slope +0.26 percentage points/year, p = 0.002; + 1.25 percentage points/year, p < 0.001). During the post-INPULSIS/ASCEND period (2015-2017), overall AF use rose faster at ACs (Δ slope +1.23 percentage points/year, p = 0.009), while AF initiation slopes did not differ significantly (Δ slope -0.22 percentage points/year, p = 0.155). PPF prevalence was similar across centers, with higher AF use at ACs in 2024 (17.1% vs 9.8%, p < 0.001). During the post-approval period (2020-2024), AF initiation and overall AF use increased faster at ACs (Δslope +1.16 percentage points/year, p < 0.001; + 1.94 percentage points/year, p = 0.009). During the post-INBUILD period (2020-2022), AF initiation rose faster at ACs (Δ slope +1.42 percentage points/year, p = 0.002), while overall AF slope differences were not significant. Outcome patterns for IPF revealed higher mortality and lung transplant in ACs, but lower respiratory failure and oxygen use. PPF showed higher respiratory failure and mortality in ACs, and no significant differences in lung transplant or oxygen use.

AF use in IPF and PPF was higher and rose faster at ACs, suggesting faster AC adoption and potential barriers at NonACs. Mixed outcome findings suggest referral and severity bias may partially explain differences, however, are unlikely to fully account for them. More research is needed to account for referral and severity bias. Given the small NonAC PPF cohort (n = 216), PPF findings should be interpreted with caution.