Correlation between blood cell count and outcome of melanoma patients treated with anti-PD-1 antibodies.

Background: Anti-programmed cell death protein 1 monoclonal antibodies (αPD-1mAbs) have been shown to be effective for advanced malignant melanoma. Treatment with αPD-1mAbs can also cause immune-related adverse events (irAEs). However, clinical predictive factors for treatment responses or irAE risk remain unclear.

Objective: To identify useful blood biomarkers for response and occurrence of irAEs with αPD-1mAbs treatment.

Methods: We retrospectively collected data from patients with melanoma treated with αPD-1mAbs at the University of Tsukuba Hospital. Clinical data including age, sex, clinical type, metastatic site, treatment course, blood laboratory tests, irAEs and treatment outcome were collected.

Results: Multivariate logistic regression analysis showed that increased baseline neutrophil-lymphocyte ratio (NLR) was significantly associated with poor response (odds ratio [OR]: 2.638, P = 0.0227, cutoff value = 2.8). Similarly, multivariate Cox regression analysis revealed that NLR at baseline were significantly associated with shorter progression survival (hazard ratio: 1.343, P = 0.0095). As for irAEs, logistic regression analysis revealed that baseline absolute eosinophil count was positively associated with occurrence of endocrine irAEs (OR: 1.601, P = 0.045, cutoff value = 240/μL). Additionally, a higher relative eosinophil count at 1 month was significantly correlated with occurrence of endocrine irAEs (OR: 1.229, P = 0.0296, cutoff value = 3.2%).

Conclusion: Our results suggested that NLR > 2.8 could be a useful baseline biomarker for indicating poor response to αPD-1mAbs treatment and that absolute eosinophil count >240/μL at baseline and relative eosinophil count at 1 month >3.2% could be useful biomarkers to predict endocrine irAEs in patients receiving αPD-1mAbs.