Circulating tumor cells (CTCs) can provide non-invasive insight into how a cancer patient responds to therapy. Their role in disease monitoring of advanced melanoma patients treated with immune checkpoint inhibitors (ICI) is unknown. CTC protein expression of human leukocyte antigen class-I (HLA I) and programmed death ligand-1 (PD-L1) may give insight into how a patient's disease evolves over the course of treatment. In our study, we utilize microfluidic Exclusion-based Sample Preparation (ESP) technology to isolate and characterize CTCs from patients with advanced-stage melanoma. CTC samples from melanoma patients are collected, captured, and stained. A range of 2 to 35 CTCs is observed in a cohort of 16 samples from 10 advanced-stage melanoma patients treated with ICI therapy. Single-cell protein expression data is generated from image cytometry analysis and used to calculate mean HLA I and PD-L1 expression. Using our ESP capture approach, we successfully detect phenotypic and numerical heterogeneity in CTCs from melanoma patients. Our assay shows sufficient capture sensitivity and promising prognostic and predictive information, as we illustrate in our case example. A greater clinical sample size will be necessary to confirm the diagnostic sensitivity and specificity of the assay in predicting clinical outcomes for patients with advanced-stage melanoma.