Validation of Flow Cytometric Measurements of NK Cytotoxicity Against Chromium Release Assay for Use in Clinical Monitoring of Immune Function.

Chromium (⁵¹Cr) release assays have long been a standard technique for measuring the cytolytic activity of various immune effector cells. Although relatively easy to perform, these assays suffer from problems with sensitivity and require special radiation precautions limiting their routine use in clinical laboratories. Several flow cytometric techniques measuring cytotoxicity have been introduced and used as a measure of immune response. These assays are more rapid than the ⁵¹Cr based assay and have the added advantage of freeing the laboratory from the precautions required when working with radioactivity. Such testing has however largely been restricted to the research setting. In an attempt to extend these benefits to a clinical laboratory, we attempted to validate two commercially available flow cytometry kits (CyToxiLux Plus® & GranToxiLux®, OncoImmunin, Gathersburg, MD) measuring either caspase or granzyme B activity inside K562 target cells induced by interaction with NK cells. These assays were then compared with a standard NK cytotoxicity assay using ⁵¹Cr release technique. With K562 as a target, we parallel tested three CD56+CD3-KIR- NK clones; both the ⁵¹Cr and caspase based cytotoxicity values were highly positive. With WH autologous EBV cell line as target, we parallel tested five NK clones and five normal control donors as effectors. As expected, all cytotoxicity values for both assays were <10%. The caspase and granzyme assays were then compared and the degree of separation in the flow cytometric plots of the granzyme based assay were found to be marginally superior facilitating analysis of the data. The granzyme B assay therefore was chosen for investigation in the clinical samples. The granzyme and ⁵¹Cr assays were then compared both in 49 normal control donors and in 10 patients. There was a high level of concordance between the two assays. Comparing all results using linear correlation yielded a squared correlation (R²) value of 0.6878. However, given the wide range of NK activity in normal control donors the degree of positivity may not be clinically significant. Therefore, the data was examined in the context of a positive ⁵¹Cr response. Using a 2×2 table to exam the level of agreement, the kappa statistic was 0.63, indicating good agreement between the assays. When progressively fewer effector cells were added to a fixed number of target cells, the % cytotoxicity decreased from 39% to 0% with the flow cytometric assay and from 26% to 1% with the ⁵¹Cr assay. In the normal control donors the positive values ranged from 17% to 65% cytotoxicity using the flow assay compared with the previous range of 15% to 72% established with ⁵¹Cr. Patients having cytotoxicity values below 17% were deemed as having an unreactive or abnormal low result. One of the patients with few CD56+CD3- NK cells had consistently low cytotoxicity each time tested. For the patients who have either undergone allogeneic stem cell transplantation or have a known immunodeficiency, further clinical follow-up will serve to establish the clinical relevance of these measurements. These preliminary data suggest that the flow cytometric technique can be used in the clinical setting for immune monitoring.