Eosinophil Dynamics as Prognostic Indicators in Immunotherapy for Pulmonary Lymphoepithelioma-Like Carcinoma

Pulmonary lymphoepithelioma-like carcinoma (pLELC) is a rare form of non-small cell lung cancer. As immunotherapy (IO) becomes an important treatment for advanced pLELC, identifying reliable prognostic markers is crucial for guiding clinical decisions and resource allocation.

This multicentre retrospective cohort study included 26 patients with advanced or metastatic pLELC who received palliative IO at two tertiary hospitals in Hong Kong from 2010 to 2023. Clinical and haematological data, particularly eosinophil peak timing, eosinophil-to-neutrophil ratio (ENR), and neutrophil-to-lymphocyte ratio (NLR), were obtained at each treatment cycle. Progression-free survival (PFS) and overall survival (OS) were evaluated using Kaplan-Meier analysis. Cox regression models, complemented by landmark and sensitivity analyses, were used to determine independent prognostic factors.

Early eosinophil peak (≤5 weeks after starting IO) was significantly linked to shorter PFS (HR 7.0, p = 0.007), while ENR <0.054 independently predicted poorer PFS (HR 2.9, p = 0.04). Neither ENR nor NLR showed significant associations with OS in multivariate analysis. Delayed eosinophil peaks were associated with more favourable PFS, supporting the prognostic value of eosinophil kinetics. Sensitivity analyses confirmed these findings were robust across patient subgroups.

Eosinophil peak timing is a practical, independent biomarker for identifying pLELC patients less likely to benefit from IO and may enhance patient prognostication and management. Further prospective studies in larger, multi-centre cohorts are needed to validate the clinical use of eosinophil dynamics in IO monitoring for pLELC.