# Clinical profiles associated with rapidly progressive interstitial lung disease in antisynthetase syndrome: A multicentric cohort study (TYPASS study)

## TL;DR

This study identifies clinical factors and distinct patient groups in antisynthetase syndrome linked to rapidly progressive lung disease.

## Contribution

A novel unsupervised clustering approach reveals distinct phenotypes in antisynthetase syndrome patients with interstitial lung disease.

## Key findings

- 39% of antisynthetase syndrome patients had rapidly progressive interstitial lung disease at diagnosis.
- Male sex, fever, organizing pneumonia pattern, and pleural effusion were independently associated with rapid progression.
- Clustering identified four distinct patient groups, including a severe inflammatory phenotype with higher mortality.

## Abstract

To assess factors associated with rapidly progressive interstitial lung disease (ILD) (RP‐ILD) at time of ILD diagnosis in a multicentric retrospective cohort study of antisynthetase syndrome (ASyS). We used a complementary unsupervised approach, hierarchical clustering, to delineate distinct phenotypes among ASyS patients with ILD.

A total of 132 patients with ASyS, defined according to the 2024 ACR/European Alliance of Associations for Rheumatology (EULAR) ASyS classification criteria, and ILD, diagnosed by CT scan, were included. RP‐ILD was defined by the presence of respiratory failure at ILD diagnosis or rapid ILD progression during the first 3 months.

In our study, 39% of patients had RP‐ILD at ILD diagnosis. Multivariate logistic regression analysis with cluster‐robust SE identified the factors associated with RP‐ILD at ILD diagnosis as male sex (aOR = 9.7 [1.6–59.5], p = 0.006), fever (aOR = 128 [12.6–1300], p < 0.001), organizing pneumonia (OP) pattern (aOR = 66.8 [3.4–1316], p = 0.006), and pleural effusion (aOR = 20.2 [1.1–373], p = 0.04), whereas RP‐ILD was associated with lower likelihood of severe muscle disease (aOR = 0.004 [0.0001–0.13], p = 0.002). Clustering analysis identified four distinct groups: Cluster 1 (n = 62) included patients with systemic presentation, non‐RP‐ILD at ILD diagnosis, and anti‐Jo‐1 antibodies with good prognosis; Cluster 2 (n = 40) included older age patients with more RP‐ILD at ILD diagnosis, pleuropericarditis, and a higher mortality rate.

Fever, pleural effusion, and OP pattern were independently associated with RP‐ILD in ASyS patients. Unsupervised cluster analysis identified a severe inflammatory phenotype in ASyS patients with ILD.