The Clinical Significance of Some Specific HLA Class I Alleles in Japanese Patients with Bone Marrow Failure (BMF).

Abstract

Autoreactive T lymphocytes are implicated in the immune mechanisms involved in the bone marrow failure (BMF) syndrome, including aplastic anemia (AA), paroxysmal nocturnal hemoglobinuria (PNH), and myelodysplastic syndrome (MDS). However, the significance of the HLA class I alleles remains unknown in the BMF syndrome. Nevertheless, from many clinical and basic studies, it is certain that CD8+ T lymphocytes are implicated in some of the immune mechanisms involved in the occurrence of AA. To clarify some clinical significance of the HLA class I alleles in the BMF syndrome, we investigated the alleles using a high-resolution method of genotyping in 78 Japanese patients with BMF, including 32 AA, 24 PNH, and 22 MDS patients. Subsequently, we compared various clinical findings, including age, sex, white blood cell counts, absolute neutrophil counts, hemoglobin concentrations, reticulocyte counts, platelet counts, values of lactate dehydrogenase, durations of illness, chromosomal findings, and proportions of CD55− and CD59− erythrocytes, between the groups with and without some alleles. The diagnosis and grading of the severity of AA were based on the criteria of the International Agranulocytosis and Aplastic Anemia Study Group (Blood 1987; 70: 1718–21) and that of Frickhofen et al (N Engl J Med 1991; 324: 1297–304), respectively. A patient with a CD55− and CD59− population of more than 1% was judged to have PNH erythrocytes (Blood 1996; 87: 5332–40). The diagnosis of MDS was determined according to the FAB criteria (Br J Haematol 1982; 51: 189–99). The frequencies of the HLA-B* 4002 allele in AA patients (21.9%) and of the HLA-A* 0206 allele in PNH patients (22.9%) were significantly different from those in controls (n=371; 8.6%, p<0.002 and 7.7%, p<0.001, respectively), while we found no specific HLA class I alelles in MDS patients. The frequency of the HLA-DRB1*1501 allele in PNH patients (31.3%) was significantly higher than that in controls (6.1%, p<0.0001), while we could not find the high frequencies of the HLA-DRB1*1501 (10.9%) and *1502 (10.9%) alleles in AA patients. The proportion of severe or very severe AA patients with the HLA-B* 4002 allele (10/17, 58.8%) was significantly higher than that of non-severe AA patients with the allele (3/15, 20%; p<0.05). In contrast, the proportion of severe or very severe AA patients (4/17; 23.5%) with the HLA-DRB1* 1501 allele was not different from that of non-severe AA patients (3/15; 20%) with the allele. Subsequently, the reticulocyte counts (138 ± 73 x 10 9/L) and values of lactate dehydrogenase (2399 ± 235 IU/L) at the time of examination in PNH patients (n=10) with the HLA-A* 0206 allele were significantly higher than those in PNH patients (n=14) without the allele (78 ± 34 x 109/L, p<0.02 and 972 ± 770 IU/L, p<0.05, respectively). In addition, the frequency of PNH patients with over 30% of complement-sensitive erythrocytes, consisting of intermediate and negative populations of CD55 and CD59 expressions on erythrocytes by flow cytometry, was significantly higher in PNH patients with the HLA-A* 0206 than in those without the allele (80% versus 28.6%, p<0.05). In conclusions, our results suggest that the HLA-B* 4002 allele in AA or the HLA-A* 0206 allele in PNH is related to grading of the severity of AA or grading of hemolysis of PNH, respectively.