Epidemiology of interstitial lung diseases and their progressive-fibrosing behaviour in six European countries

The PERSEIDS study aimed to estimate incidence/prevalence of interstitial lung diseases (ILDs), fibrosing interstitial lung diseases (F-ILDs), idiopathic pulmonary fibrosis (IPF), systemic sclerosis-associated ILD (SSc-ILD), other non-IPF F-ILDs and their progressive-fibrosing (PF) forms in six European countries, as current data are scarce. This retrospective, two-phase study used aggregate data (2014-2018). In Phase 1, incident/prevalent cases of ILDs above were identified from clinical databases through an algorithm based on codes/keywords, and incidence/prevalence was estimated. For non-IPF F-ILDs, the relative percentage of subtypes was also determined. In Phase 2, a subset of non-IPF F-ILD cases was manually reviewed to determine the percentage of PF behaviour and usual interstitial pneumonia-like (UIP-like) pattern. A weighted mean percentage of progression was calculated for each country and used to extrapolate incidence/prevalence of progressive-fibrosing ILDs (PF-ILDs). In 2018, incidence/10⁵ person-years ranged between 9.4 and 83.6 (ILDs), 7.7 and 76.2 (F-ILDs), 0.4 and 10.3 (IPF), 6.6 and 71.7 (non-IPF F-ILDs), and 0.3 and 1.5 (SSc-ILD); and prevalence/10⁵ persons ranged between 33.6 and 247.4 (ILDs), 26.7 and 236.8 (F-ILDs), 2.8 and 31.0 (IPF), 22.3 and 205.8 (non-IPF F-ILDs), and 1.4 and 10.1 (SSc-ILD). Among non-IPF F-ILDs, sarcoidosis was the most frequent subtype. PF behaviour and UIP-like pattern were present in a third of non-IPF F-ILD cases each and hypersensitivity pneumonitis showed the highest percentage of progressive behaviour. Incidence of PF-ILDs ranged between 2.1 and 14.5/10⁵ person-years, and prevalence between 6.9 and 78.0/10⁵ persons. To our knowledge, PERSEIDS is the first study assessing incidence, prevalence and rate of progression of ILDs across several European countries. Still below the threshold for orphan diseases, the estimates obtained were higher and more variable than reported in previous studies, but differences in study design/population must be considered.