Evaluation of discrepancies between Luminex single antigen bead assays

Abstract

Currently there are two manufacturers offering HLA antibody testing using the Luminex platform. Our laboratory incorporates both methods in our testing protocol. We have documented discrepancies and use them to help us identify samples that need follow-up verification of unacceptable antigens. More discrepancies are found with Class I than with Class II and some appear to be major differences. The purpose of this study was to examine the discrepancies with Class I single antigen testing between the two manufacturers with respect to correlation with Flow crossmatch results and MFI values. For this study discrepant samples from a single run of Class I single antigen beads were chosen. An antibody of interest was chosen for each sample and surrogate cells were found which did not have HLA detected by other Class I antibodies reported in the same sample. Flow crossmatch was performed using a BD FACsCaliber instrument on a 1024 scale. All discrepancies in this study resulted from an antibody being called by Method A that was not detected using Method B. Surrogate crossmatches were done to determine if there was a sensitivity problem with Method B or a specificity problem with Method A. Twenty-seven crossmatches were examined which included eleven different specificities. Fourteen samples were T cell positive and thirteen were T cell negative. Of the fourteen that were T cell positive, six were B cell negative, which could indicate non-HLA reactivity. Therefore, it is possible that 19 of the 27 specificities detected by Method A may not be clinically relevant. The discrepancies between methods is of concern and emphasizes the necessity of including manufacturers and MFI cutoffs when publishing Luminex single antigen bead results. Most HLA laboratories use single antigen results for determining unacceptable antigens and virtual crossmatches. More specificities were called with Method A using our current cutoff, but at least half of these did not result in a positive Flow crossmatch. Re-evaluation of the cutoff suggests that the cutoff may be different for certain loci or certain specificities. The use of both methods may have value in fine-tuning the unacceptable antigens and illuminating those specificities that should be confirmed.