Luminex Antibody Threshold to Match Donor-Recipient Pairs Using Virtual Crossmatch in Kidney Paired Donation

Abstract

Introduction: We developed a new computer program to match maximal sets of incompatible live donor/recipient pairs from a national kidney paired donation (KPD) registry. The strength of Luminex mean fluorescence intensity (MFI) of antibody used to exclude from matching has not been validated. Methods: The Australian KPD matching program uses a stringent matching approach to minimise the number of positive crossmatches. The program ignores any HLA antigen matching rules and uses matching based on acceptable mismatches excluding donors from matching to recipients with DSA >2000 MFI for any class I or II antibodies. Following computer matching, all the identified donor-recipient pairs underwent CDC crossmatch (XM), matched donors had to be recalled to provide fresh cells for crossmatching. Within one year of inception of the Australian program four quarterly match runs were performed. In the first two rounds, recalled donors were crossmatched also against all unmatched recipients on the trays. Thus, an additional 101 T-cell XM (TCXM) and B-cell XM (BCXM) results were available for analysis, in which recipients were tested against donors despite the presence of single or multiple DSA. Results: During the first year the Australian KPD program included 53 pairs and 2 altruistic donors, HLA incompatibility accounted for 90% of the listed pairs. Following computer matching 27 recipients were matched using the 2000MFI cut-off for unacceptable mismatches and all these recipients had a negative CDC TCXM and BCXM. With the unmatched recipients there were 20 positive TCXM and 45 positive BCXM results. A positive TCXM was found in 90% of cases with DSA at >8000MFI, with one single case with weak DSA demonstrating a positive result despite a strength of ˜1500MFI. A positive BCXM was found in 60% of cases with DSA at >8000MFI and in 27% with DSA 2000-8000MFI. DSAs against C-locus were responsible for 15% of positive TCXM. DSA against DQB1 accounted for 50% of positive BCXM. Conclusion: The threshold of 2000MFI selected in our KPD program to exclude from matching recipients with HLA antibodies against donors is an excellent predictor of negative XM. Although a higher threshold may identify additional suitable pairs with negative XM, the risk of positive XM, in particular BCXM is significant and this could have significant implications for pairs that are links in an aborted chain.