Respiratory Research (Dec 2023)

Bronchoalveolar cytokine profile differentiates Pulmonary Langerhans cell histiocytosis patients from other smoking-related interstitial lung diseases

  • Silvia Barril,
  • Paloma Acebo,
  • Paloma Millan-Billi,
  • Alfonso Luque,
  • Oriol Sibila,
  • Carlos Tarín,
  • Abdellatif Tazi,
  • Diego Castillo,
  • Sonsoles Hortelano

DOI
https://doi.org/10.1186/s12931-023-02622-z
Journal volume & issue
Vol. 24, no. 1
pp. 1 – 13

Abstract

Read online

Abstract Background Pulmonary Langerhans cell histiocytosis (PLCH) is a rare interstitial lung disease (ILD) associated with smoking, whose definitive diagnosis requires the exclusion of other forms of ILD and a compatible surgical lung biopsy. Bronchoalveolar lavage (BAL) is commonly proposed for the diagnosis of ILD, including PLCH, but the diagnostic value of this technique is limited. Here, we have analyzed the levels of a panel of cytokines and chemokines in BAL from PLCH patients, in order to identify a distinct immune profile to discriminate PLCH from other smoking related-ILD (SR-ILD), and comparing the results with idiopathic pulmonary fibrosis (IPF) as another disease in which smoking is considered a risk factor. Methods BAL samples were collected from thirty-six patients with different ILD, including seven patients with PLCH, sixteen with SR-ILD and thirteen with IPF. Inflammatory profiles were analyzed using the Human Cytokine Membrane Antibody Array. Principal component analysis (PCA) was performed to reduce dimensionality and protein–protein interaction (PPI) network analysis using STRING 11.5 database were conducted. Finally, Random forest (RF) method was used to build a prediction model. Results We have found significant differences (p < 0.05) on thirty-two cytokines/chemokines when comparing BAL from PLCH patients with at least one of the other ILD. Four main groups of similarly regulated cytokines were established, identifying distinct sets of markers for each cluster. Exploratory analysis using PCA (principal component analysis) showed clustering and separation of patients, with the two first components capturing 69.69% of the total variance. Levels of TARC/CCL17, leptin, oncostatin M (OSM) and IP-10/CXCL10 were associated with lung function parameters, showing positive correlation with FVC. Finally, random forest (RF) algorithm demonstrates that PLCH patients can be differentiated from the other ILDs based solely on inflammatory profile (accuracy 96.25%). Conclusions Our results show that patients with PLCH exhibit a distinct BAL immune profile to SR-ILD and IPF. PCA analysis and RF model identify a specific immune profile useful for discriminating PLCH.

Keywords