The CB Bank Düsseldorf has provided to date (July 2004) 224 CB units (216 unrelated and 8 related, 29% adults, 71% children) to transplant centers worldwide. Until now no correlation could be detected between the number of HLA-mismatches based on low resolution (LR) typing for HLA-A and-B and high resolution typing (HR) for DRB1 and the incidence of aGvHD as published previously by us and other groups. The lack of correlation between aGvHD occurrence and donor/recipient HLA diversity in patients given an unrelated CBT could be explained by the fact that some mismatches for HLA class I antigens (A, B and C) are not detected by LR typing. In order to determine the impact of HLA high resolution typing with outcomes, mainly aGvHD after UCBT we analysed DNA samples of 115 CB recipients (86 children; 29 adults; 66 male, 49 female; diagnosis ALL=43, AML=19, SecAL =1, MDS=5, CML=10, NHL=5, Hodgkin=1, AA=7, genetic and metabolic diseases= 24) and their unrelated CB grafts were HLA-typed for HLA-class I (A, B, C) and HLA-class II (DRB1 and DQB1) by sequencing. The transplant centers used their own protocols for GvHD prophylaxis, the most commonly used was the combination of CsA and steroids alone (60%), CsA alone (15%), or the combination with MTX (6%). 55 of 115 patients did not develop aGvHD (grade 0= 48%), 26 patients developed grade I (23%), 12 patients developed grade II (10%), 10 patients grade III (9%) and 12 patients grade IV (10%). When mismatches (MM) were analysed for HLA-A, B based on LR-typing and -DRB1 based on HR-typing in concordance with all published data so far, the following mismatch situation resulted: No MM (16 pairs, 13.9%), one MM (47 pairs, 40.9%), two MM (41 pairs, 35.7%), three MM (5 pairs, 4.3%), four MM (3 pairs, 2.6%). If the MM for A and B alleles detected by HR-typing were included, the situation was as follows: 0 MM (6 pairs, 5.2%), 1 MM (35 pairs, 30.4%), 2 MM (54 pairs, 47%), 3MM (14 pairs, 12.2%), 4 MM (5 pairs, 4.3%), 5 MM (1 pair, 0.9%). If analysing A, B, C, DR and DQ based on HR typing a high additional frequency of MM occurred: No MM (4 pairs, 3.5%), 1 MM (13 pairs, 11.3%), 2 MM (19 pairs, 16.5%), 3 MM (24 pairs, 20.9%), 4 MM (30 pairs, 26.1%), 5 MM (14 pairs, 12.2%), 6 MM (6 pairs, 5.2 %), 6 MM (6 pairs, 5.2%), 7 MM (3 pairs, 2.6%), 8 MM (2 pairs, 1.7%). There was no significant correlation between the number of MM (also analysed in GvHD or rejection direction) using high-resolution level for HLA-A, B and DRB1 as well as for HLA-A, B, C, DRB1 and DQB1 and the development of aGvHD grade III-IV. More interestingly, we have not found any significant correlation between numbers of MM with 2-year survival probability. Although the heterogeneity and number of patients analysed, it shows that the degree of mismatching is even higher than expected, also in comparison to unrelated BMT. It also shows that additional subtyping for HLA-A, B, C and DQ, not performed on a routine basis at present, does not improve the 2-year survival rate.

Author notes

Corresponding author