Single-cell transcriptome sequencing reveals the immune microenvironment in bronchoalveolar lavage fluid of checkpoint inhibitor-related pneumonitis.

BACKGROUND AND OBJECTIVES

Immune checkpoint inhibitors (ICIs) bring cancer patients tumor control and survival benefits, yet they also trigger immune-related adverse effects (irAEs), notably checkpoint inhibitor-related pneumonitis (CIP), affecting about 5% of patients among whom 1-2% experiencing severe grade 3 or higher pneumonitis. Current research points to potential links with T cell subset dysfunction and autoantibody increase, but the specific mechanisms underlying different grades of CIP are understudied.

METHODS

Herein, we employed single-cell RNA sequencing (scRNA-seq) on bronchoalveolar lavage fluid (BALF) from CIP patients across varying severity levels, aiming to elucidate underlying immune environment and mechanisms of CIP progression at cellular and molecular levels.

FINDINGS

Totally, 121,409 high qualified cells from BALF of 11 patients were annotated and categorized into five major cell types. Severe CIP (CIP-S) cases have a significant increase in the percentage of unreported epithelial cells in their bronchoalveolar lavage fluid compared with mild CIP (CIP-M) cases. These cells were defined as aberrant basaloid cells. They upregulated SOX9, increased the expression of CXCL3/5, recruited neutrophils, and activated the immune system. Additionally, macrophages in the CIP-S group had stronger antigen-presenting abilities and resulted in more CD8 + effective T cells infiltrated.

CONCLUSIONS

Utilizing single-cell sequencing of BALF, we discovered an enriched population of aberrant basaloid cells in CIP-S patients, which had not been previously reported. Aberrant basaloid cells may upregulate SOX9 via CXCL3/5-CXCR2 to recruit and activate neutrophils, and further activate the immune system, resulting in CIP-S. This finding could identify new targets for stratified treatment of CIP patients, holding promise of a novel approach for clinical guidance.