Comparison of four different in vitro assays for prediction of potential RBC antibody clinical significance

The monocyte monolayer assay (MMA) uses monocytes, but it is known that antibody-mediated hemolysis occurs in macrophages. Whether the use of macrophages would improve the predictive capacity of the MMA is unclear. Also, it is unclear if RBC alloantibodies contribute to hemolysis via antibody-dependent cellular cytotoxicity (ADCC). We have compared the MMA to an assay that utilizes monocytes cultured to two different macrophages, called the monocyte-macrophage assay (MOMA), and to ADCC using a ⁵¹chromium-release assay with natural killer cells.

MMA was performed using peripheral blood mononuclear cells (PBMCs) to adhere monocytes to chamber slides. MOMA was done using purified peripheral blood monocytes cultured in the presence of GM-CSF (M1 macrophages) or M-CSF (M2 macrophages) and then polarized using IL-4 (M2) or LPS and IFNγ (M1). NK cells were purified and incubated with alloantibody-opsonized RBCs loaded with 51-chromium. We examined 38 antibodies of different specificities, including anti-D, -E, -K, -Jk^a, -M, -S, -k, -Fy^a, -P, -Ge2, -Di^b, -Yt^a.

Most weakly reactive antibodies were negative in MMA but positive in MOMA and ADCC. M2 was generally more positive than M1. All anti-D, -K, -E gave strong IATs and showed reactivity in all four assays. Anti-P (2+ IAT) was positive in the MMA and MOMA but negative in ADCC. Twelve anti-Yt^as had very weak IATs and all were negative in the MMA and 3/12 positive in MOMA and none were positive in ADCC.

We show that both MOMA and ADCC are more sensitive than MMA for indicating potential for antibody clinical significance. However, validation with antibodies having clinical data as to their actual significance is necessary.