Replacement of HLA class II serology by the HLA-DR microtitre plate oligotyping assay: a one-year experience in unrelated bone marrow donor selection

Bone marrow transplantation with unrelated donors: what is the probability of identifying an HLA-A/B/Cw/DRB1/B3/B5/DQB1-matched donor?

Patients transplanted with marrow from an HLA-ABDR serologically matched unrelated donor suffer from more post-transplant complications than those who are transplanted with marrow from an HLA-identical sibling. This is most likely due to either HLA-ABDR incompatibilities not resolved by standard techniques and/or HLA polymorphisms not tested for by routine tissue typing (HLA-Cw,-DQ). By resolving these incompatibilities by molecular techniques combined with the in vitro cytotoxic T lymphocyte precursor frequency (CTLpf) test, we have shown that a high degree of HLA compatibility is associated with increased patient survival. However, higher requirements for HLA matching decrease the number of available donors. We have estimated the probability of finding an HLA-A/B/Cw/DRB1/DRB3/DRB5/DQB1 compatible donor based on 104 consecutive unrelated bone marrow donor searches initiated between January 1995 and December 1997, with December 1998 as the endpoint. For 96 patients (92.3%), one or more ABDR-identical donors were listed in the Bone Marrow Donor Worldwide Registry (BMDW). After contacting the registries, we obtained at least one (mean, 5.36; range, 1-20; total, 461) blood sample for 86 patients. A highly compatible donor was identified for 33/86 patients (38.4%), after testing an average number of 4.5 donors/patients (range, 1-13). However, by accepting an HLA-DRB3 or -DQB1 or -Cw incompatibility, this number would be as high as 68.6%. Approximately half of the patients (n = 40) for whom a search had been initiated have been transplanted: 22 patients with a perfectly matched donor, 15 patients with an HLA-DRB3 or -DQB1 or -Cw mismatch and three with other mismatches. The average time needed to identify the most compatible donor was 4 months. Extremely long searches seemed to be less useful, because after testing the first seven, a more compatible donor was seldom found. These results show that even when requirements for compatibility are high, the chances of finding a donor remain considerably low.

Search for an unrelated HLA-compatible stem cell donor

Through recent advancements in tissue-typing techniques, the true extent of HLA polymorphism has become evident. This has allowed a better assessment of the degree of matching of donor and recipient, which has reiterated the pivotal importance of HLA compatibility for the outcome of a stem cell transplantation. Now that most tissue typing laboratories are technically able to provide a flawless, high-resolution HLA typing, the challenge of the future will be to develop tissue typing-based search strategies that give an optimal chance to find the best donor within the time limits set by the disease of the patient.

DNA Typing for HLA-A and HLA-B Identifies Disparities Between Patients and Unrelated Donors Matched by HLA-A and HLA-B Serology and HLA-DRB1

High incidences of graft failure and graft-versus-host disease in the recipients of bone marrow transplantations (BMT) from unrelated donors (URD) may reflect the existence of allelic disparities between the patient and the URD despite apparent HLA identity at HLA-A, HLA-B, and HLA-DRB1 loci. To identify the extent and pattern of allelic disparities at HLA-A and HLA-B loci, 128 patients and 484 potential URD were evaluated by DNA typing. DNA typing for HLA-A, HLA-B, and HLA-DRB1 was performed at Memorial Sloan Kettering Cancer Center. HLA-A and HLA-B serotyping on URD was provided by the registries. By original typing (serology for HLA-A and HLA-B; DNA typing for DRB1) 187, 164, and 133 URD were 6/6, 5/6, and 4/6 matches, respectively. Following DNA typing, however, only 52.9% of the originally 6/6 matched URD remained 6/6, while 38.5%, 7.5%, and 1.1% were found to be 5/6, 4/6, and 3/6 matches. The level of disparity was higher in the originally 5/6 (P< .01) and 4/6 (P < .01) matched URD. A higher level of disparity was seen for HLA-B as compared to HLA-A. In addition, a serotype related variation was also noticed. For example, 24.1% of HLA-A2 and 60.1% of HLA-B35 seromatched URD were genotypically disparate, but no disparities were seen for HLA-A1 and HLA-B8. A higher percentage of HLA-A (67.4%) compared with HLA-B (35.4%) serologic homozygous URD remained genotypically homozygous (P = .01). The level of allelic disparity was lower (P < .01 for 6/6; P = .02 for 5/6) if the patient had one of the 15 most common haplotypes (A1B8DR3, A2B7DR15, A3B7DR15, etc) in comparison to the rest of the group. Outcome studies will answer the question whether these disparities are associated with a higher rate of immunological complications seen with URD-BMT.

DNA Typing for HLA-A and HLA-B Identifies Disparities Between Patients and Unrelated Donors Matched by HLA-A and HLA-B Serology and HLA-DRB1

High incidences of graft failure and graft-versus-host disease in the recipients of bone marrow transplantations (BMT) from unrelated donors (URD) may reflect the existence of allelic disparities between the patient and the URD despite apparent HLA identity at HLA-A, HLA-B, and HLA-DRB1 loci. To identify the extent and pattern of allelic disparities at HLA-A and HLA-B loci, 128 patients and 484 potential URD were evaluated by DNA typing. DNA typing for HLA-A, HLA-B, and HLA-DRB1 was performed at Memorial Sloan Kettering Cancer Center. HLA-A and HLA-B serotyping on URD was provided by the registries. By original typing (serology for HLA-A and HLA-B; DNA typing for DRB1) 187, 164, and 133 URD were 6/6, 5/6, and 4/6 matches, respectively. Following DNA typing, however, only 52.9% of the originally 6/6 matched URD remained 6/6, while 38.5%, 7.5%, and 1.1% were found to be 5/6, 4/6, and 3/6 matches. The level of disparity was higher in the originally 5/6 (P< .01) and 4/6 (P < .01) matched URD. A higher level of disparity was seen for HLA-B as compared to HLA-A. In addition, a serotype related variation was also noticed. For example, 24.1% of HLA-A2 and 60.1% of HLA-B35 seromatched URD were genotypically disparate, but no disparities were seen for HLA-A1 and HLA-B8. A higher percentage of HLA-A (67.4%) compared with HLA-B (35.4%) serologic homozygous URD remained genotypically homozygous (P = .01). The level of allelic disparity was lower (P < .01 for 6/6; P = .02 for 5/6) if the patient had one of the 15 most common haplotypes (A1B8DR3, A2B7DR15, A3B7DR15, etc) in comparison to the rest of the group. Outcome studies will answer the question whether these disparities are associated with a higher rate of immunological complications seen with URD-BMT.

Molecular characterization of HLA-C incompatibilities in HLA-ABDR-matched unrelated bone marrow donor-recipient pairs. Sequence of two new Cw alleles (Cw*02023 and Cw*0707) and recognition by cytotoxic T lymphocytes

While the influence of HLA-AB and -DRB1 matching on the outcome of bone marrow transplantation (BMT) with unrelated donors is clear, the evaluation of HLA-C has been hampered by its poor serological definition. Because the low resolution of standard HLA-C typing could explain the significant number of positive cytotoxic T lymphocyte precursor frequency (CTLpf) tests found among HLA-AB-subtype, DRB1/B3/B5-subtype matched patient/donor pairs, we have identified by sequencing the incompatibilities recognized by CD8+ CTL clones obtained from such positive CTLpf tests. In most cases the target molecules were HLA-C antigens that had escaped detection by serology (e.g. Cw*1601, 1502 or 0702). Direct recognition of HLA-C by a CTL clone was demonstrated by lysis of the HLA class I-negative 721.221 cell line transfected with Cw*1601 cDNA. Because of the functional importance of Cw polymorphism, a PCR-SSO oligotyping procedure was set up allowing the resolution of 29 Cw alleles. Oligotyping of a panel of 382 individuals (including 101 patients and their 272 potential unrelated donors, 5 related donors and 4 platelet donors) allowed to determine HLA-C and HLA A-B-Cw-DRB1 allelic frequencies, as well as a number of A-Cw, B-Cw, and DRB1-Cw associations. Two new HLA-Cw alleles (Cw*02023 and Cw*0707) were identified by DNA sequencing of PCR-amplified exon 2-intron 2-exon 3 amplicons. Furthermore, we determined the degree of HLA-C compatibility in 287 matched pairs that could be formed from 73 patients and their 184 potential unrelated donors compatible for HLA-AB by serology and for HLA-DRB1/ B3/B5 by oligotyping. Cw mismatches were identified in 42.1% of these pairs, and AB-subtype oligotyping showed that 30% of these Cw-incompatible pairs were also mismatched for A or B-locus subtype. The degree of HLA-C incompatibility was strongly influenced by the linkage with B alleles and by the ABDR haplotypes. Cw alleles linked with B*4403, B*5101, B18, and B62 haplotypes were frequently mismatched. Apparently high resolution DNA typing for HLA-AB does not result in full matching at locus C. Since HLA-C polymorphism is recognized by alloreactive CTLs, such incompatibilities might be as relevant as AB-subtype mismatches in clinical transplantation.

Factors influencing the outcome of bone marrow transplants using unrelated donors

The use of unrelated donors for bone marrow transplantation (BMT) is associated with an increased morbidity and mortality when compared with HLA-identical donors, primarily due to an increased rate of graft-versus-host disease, but also to increased susceptibility to infections and graft failure. HLA matching for donors and recipients is the single most important factor influencing the outcome of BMT. However, unrelated donor selection generally relies on matching only for HLA-A, -B and -DR antigens without considering potential incompatibility for other HLA loci, such as HLA-C, -DQ and -DP. In addition, other factors that affect the outcome of BMT need to be taken into consideration in selecting the best unrelated donor. In this review, we will focus on the effects of HLA-associated factors in determining the result of a transplant procedure. We will also mention other relevant factors, drawing on our experience of laboratory studies performed at The Anthony Nolan Research Institute and clinical studies at the Hammersmith Hospital in London.

HLA-B35-subtype mismatches in ABDR serologically matched unrelated donor-recipient pairs

We have characterized HLA incompatibilities in a group of 17 B35-positive patients who were ABDR matched (AB serology and oligotyping for DR1-14) with their 28 (unrelated) potential bone marrow donors. High-resolution oligotyping for DR subtypes disclosed that nine combinations were in fact DR mismatched. Cytotoxic T-lymphocyte (CTL) activity was detected in nine combinations (32%). In the group matched for DR subtypes, three (16%) of 19 combinations were CTL positive. Patient-specific cytotoxic activity appeared to be directed against HLA C (two cases) or against a subtype of B35. In the group of DR-subtype-mismatched combinations, CTL activity was found in six (67%) of nine pairs. In all four cases that were studied in detail, however, CTL reactivity appeared to be directed against a variant subtype of B35. We have studied the B35 incompatibilities recognized in five different combinations by specificity analysis of the B35-specific CTLs and by partially sequencing of relevant segments of B35 exon 3. Preliminary data show that, within this relatively small Caucasoid group, at least five B35-variant subtypes could be distinguished. This would make B35 an antigen that will be frequently subtype mismatched, in particular when DR matching is done with low resolution (DR1-14) only.