Abstract
The current study aimed at the optimization of circulating tumor cell (CTC) enrichment for downstream protein expression analyses in non-small cell lung cancer (NSCLC) to serve as a tool for the investigation of immune checkpoints in real time. Different enrichment approaches-ficoll density, erythrolysis, their combination with magnetic separation, ISET, and Parsortix-were compared in spiking experiments using the A549, H1975, and SKMES-1 NSCLC cell lines. The most efficient methods were tested in patients (n = 15) receiving immunotherapy targeting programmed cell death-1 (PD-1). Samples were immunofluorescently stained for a) cytokeratins (CK)/epithelial cell adhesion molecule (EpCAM)/leukocyte common antigen (CD45), and b) CK/programmed cell death ligand-1 (PD-L1)/ indoleamine-2,3-dioxygenase (IDO). Ficoll, ISET, and Parsortix presented the highest yields and compatibility with phenotypic analysis; however, at the patient level, they provided discordant CTC positivity (13%, 33%, and 60% of patients, respectively) and enriched for distinct CTC populations. IDO and PD-L1 were expressed in 44% and 33% and co-expressed in 19% of CTCs. CTC detection was associated with progressive disease (PD) (p = 0.006), reduced progression-free survival PFS (p = 0.007), and increased risk of relapse (hazard ratio; HR: 10.733; p = 0.026). IDO-positive CTCs were associated with shorter PFS (p = 0.039) and overall survival OS (p = 0.021) and increased risk of death (HR: 5.462; p = 0.039). The current study indicates that CTC analysis according to distinct immune checkpoints is feasible and may provide valuable biomarkers to monitor NSCLC patients treated with anti-PD-1 agents.