Integration of hierarchical clustering in hematologic biomarkers prediction for endocrine toxicity during immunotherapy in lung cancer patients

Immune check-point inhibitors (ICIs) have outstandingly changed the treatment for lung cancer patients providing longer progression-free (PFS) and overall survival (OS). Immune-related adverse events (irAEs) counteract this success, and different biomarkers are broadly investigated to predict their appearance. Hematologic predictive biomarkers for irAEs are extensively studied to better prevent, monitor and treat irAEs. This is a retrospective exploratory-in-nature cohort study of lung cancer patients treated with ICIs in a tertiary level hospital in Romania, from 1 November 2017 until 1 August 2024. We evaluate 173 non-small cell lung cancer (NSCLC) patients treated with at least three cycles of ICIs and we investigate the association between Endocrine-irAEs and hematologic variables, such as neutrophil-to-lymphocyte ratio (NLR), absolute neutrophils, lymphocytes and eosinophils count. We also revise the literature and use an Artificial Intelligence (AI) technique to interpret our endpoints. We identify a statistically significant correspondence between NLR value before ICIs initiation and all-irAEs [odds ratio (OR) 0.877, 95% confidence interval (CI) 0.763-0.979, p = 0.036], but not for Endocrine-irAEs (OR 0.898, 95% CI 0.781-1.031, p = 0.081). While significant barriers for identifying the patients at risk of developing irAEs are still present, the implementation of truly relevant and useful biomarkers remains challenging. NLR might be a useful and easy to assess tool in real-life clinical settings, but more research is needed to validate its prediction value in Endocrine-irAEs onset.