1007-LBP: A Novel Measure of Recipient-Donor HLA Mismatch Predicts Allograft Rejection After Heart Transplantation

Aim It has previously been shown that human leukocyte antigen mismatch (HLAMM) between a transplant recipient and donor at the level of allele groups (previously defined serologically) predicts mortality after heart transplantation. Because HLAMM is not a continuous measure it has limited prognostic value. We sought to develop a novel continuous measure of HLAMM for prediction of allograft rejection using the number of mismatched amino acids in the peptide binding region. Methods Records of patients receiving a heart transplant after 12/31/1999 with at least two recorded values for rejection were identified from the UNOS database. Patients receiving repeat and multi-organ transplants were excluded. A computational algorithm was used to convert allele groups to HLA haplotypes, dervied from the patient’s ethnicity and National Marrow Donor Program haplotype frequency data. The algorithm was validated using a data set of stem cell transplant recipients with high-resolution typing. The algorithm computes the number of mismatched amino acids over exons 2 and 3 for HLA-A and -B, and exon 2 for HLA-DRB1 and -DQB1. Results We identified 5230 patients in the database. Complete HLA-A, -B, -DRB1, and -DQB1 haplotypes were identified in 3197 patients (61%). HLAMM between patients with and without rejection during follow-up was significantly different for HLA-DRB1 ( p p  = 0.003). Conclusions A novel measure of HLAMM is able to predict rejection after heart transplantation. Thus, high-resolution HLA typing of donor and recipient may help predict the risk of allograft rejection after heart transplantation.