An alternative flow cytometry strategy for peripheral blood dendritic cell enumeration in the setting of repetitive GM-CSF dosing

BACKGROUND: Enumeration of circulating peripheral blood dendritic cells (DCs) is complicated by the absence of a unique cell surface marker expressed on all DC subsets and by the use of various biological adjuvants to modulate the DC compartment, including granulocyte macrophage colony stimulating factor (GM-CSF). Common methods employ a cocktail of antibodies, typically including anti-CD14, to define a lineage negative, MHC class II positive, putative DC population. Reported flow cytometry protocols include highly variable gating strategies and DC identification criteria. Increasing appreciation of DC pleiomorphism, GM-CSF biology, and recognition of CD14 expression in some DC subsets led us to consider an alternative lineage cocktail to improve identification of the circulating DC pool. METHODS: Standard whole blood staining with appropriate fluorochrome conjugated antibodies to MHC class II and either standard CD14 containing, or an alternate CD66acde containing, lineage cocktail was performed on samples obtained from normal donors and breast cancer patients before and after administration of dose-dense, cytotoxic chemotherapy with daily GM-CSF hematopoetic growth factor support. Putative DCs were enumerated by standard flow cytometry. Data set differences were evaluated using two tailed Mann-Whitney or Wilcoxon signed rank tests. Cellular morphology was examined in cell-sorted populations from post GM-CSF samples. RESULTS: Use of either antibody cocktail defined comparably sized lineage negative, MHC class II positive populations in normal donors and at baseline in cancer patients. However, selection of lineage negative subsets with increasing MHC class II expression levels yielded larger putative DC populations identified with the alternate cocktail. Both cocktails yielded highly reproducible data. Use of the alternate cocktail: 1) yielded a putative DC population, post GM-CSF that was more homogenous and consistent with DCs, 2) resulted in less data variation across gating strategies, and 3) resulted in more uniform and concordant longitudinal data, consistent with established GM-CSF biological activity. CONCLUSION: An alternative lineage negative cocktail substituting anti-CD66 antibody for anti-CD14 is a viable option for enumerating the circulating DC population, potentially more accurately defining the circulating DC pool by including CD14 positive immature DCs, and thus, may give more reliable data, particularly in the setting of sustained GM-CSF administration.