Abstract
BACKGROUND: Immune checkpoint inhibitors (ICIs) enhance survival rate in non-small-cell lung cancer (NSCLC), but lead to immune-related adverse events, among which checkpoint-inhibitor pneumonitis (CIP) is a leading cause of death. The utility of bronchoalveolar-lavage (BAL) lymphocytosis for diagnosing CIP and predicting its severity remains uncertain. METHODS: A systematic review and meta-analysis were conducted in accordance with PRISMA guidelines. RESULTS: A total of ten studies were included in the meta-analysis, comprising a total of 168 CIP samples. The pooled estimate for BAL lymphocyte percentage was 56.43% (95% CI, 43.89–68.98, I(2) = 99.1%) in CIP, 14.85% (95% CI, 12.04–17.66, I(2) = 9.0%) in idiopathic pulmonary fibrosis (IPF), and 43.76% (95% CI, 24.43–63.09, I(2) = 99.3%) in NSCLC patients with no ICP or IPF. Significant differences were observed between CIP and IPF (p < 0.0001), NSCLC patients with no ICP or IPF (p = 0.0024). Additionally, the BAL lymphocyte percentage varied between low-grade and high-grade CIP (SMD = -1.15, 95% CI, -2.26–-0.04, I(2) = 82.4%). Similar results were also calculated on the BAL lymphocyte CD4:CD8 ratio. The included studies demonstrated significant heterogeneity. CONCLUSION: BAL lymphocyte percentage is elevated and CD4:CD8 ratio is reduced in NSCLC patients with CIP, with higher lymphocyte percentages associated with severe disease. A deeper understanding of the relationships between immunophenotyping, clinical factors and lymphocytosis will guide the use of BAL in the diagnose and evaluation of CIP. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12890-026-04224-z.