Diagnostic Significance of Haptoglobin Concentration in Patients with Bone Marrow Failure Syndromes.

Paroxysmal nocturnal hemoglobinuria (PNH) is one of the bone marrow failure syndromes, including aplastic anemia (AA) and myelodysplastic syndromes (MDS). Recently, the International PNH Interest Group proposed that evidence of a population of erythrocytes and granulocytes deficient in glycosylphosphatidylinositol (GPI) proteins and assessment of hemolytic parameters, including haptoglobin concentration, are important as minimal essential diagnostic criteria of PNH and that less than 1.0% GPI-deficient erythrocytes and granulocytes identifies subclinical PNH from classic PNH (Parker C et al, Blood, 2005). To know whether haptoglobin can be a hallmark which expects the occurrence of classic PNH during the clinical course in AA and MDS patients, we examined the expressions of CD59 on erythrocytes and granulocytes by flow cytometry and relationship between proportions of negative populations of them and various clinical parameters, including haptoglobin concentrations, in Japanese patients with AA (n=23; M:F=11:12; 50.5 ± 19.1 years), PNH (n=28; M:F=14:14; 42.7 ± 16.1 years), and MDS (n=29; M:F=20:9; 66.1 ± 13.9 years). Less than 20 mg/dl of haptoglobin were judged as significant decrease. Flow cytometry showed that the proportions of CD59⁻ erythrocytes (38.11 ± 35.49%) and granulocytes (52.57 ± 42.39%) from PNH patients were significantly higher than those from AA and MDS patients and healthy individuals (n=21; M:F=12:9; 41.3 ± 12.2 years). The values of serum asparatate aminotransferase (AST) and lactate dehydrogenase (LDH) were significantly higher in PNH patients (54.9 ± 53.1U/l and 1035 ± 1052 U/l, respectively) than AA (20.8 ± 9.1 U/l and 205.8 ± 45.0 U/l, respectively) and MDS (22.4 ± 16.9 U/l and 217.7 ± 64.0 U/l, respectively) patients. In contrast, the concentrations of serum haptoglobin were significantly lower in PNH patients (12.9 ± 27.6 mg/dl) than AA (84.6 ± 81.9 mg/dl) and MDS (77.7 ± 47.3 mg/dl) patients. When comparing PNH patients (n=9; minimal PNH) with less than 5% of CD59⁻ erythrocytes with those (n=19; bulky PNH) with over 5% of CD59⁻ erythrocytes, the values of AST (70.7 ± 58.2U/l) and LDH (1021 ± 1083U/l) of the latter were significantly higher than those of the former (21.4 ± 6.2U/l and 220.3 ± 45.1U/l, respectively), but the concentrations of haptoglobin were similar between the latter (11.8 ± 28.9mg/dl) and the former (15.8 ± 26.0mg/dl). In addition, all the PNH patients, but not AA and MDS patients, with over 1% of CD59⁻ erythrocytes had significant decrease of haptoglobin concentration, suggesting that low concentrations of serum haptoglobin may predict occurrence of classic PNH during the clinical course in AA and MDS patients. In conclusion, over 1% of CD59⁻ erythrocytes in PNH patients certainly cause clinical hemolysis and serum haptoglobin is a useful marker which can predict the occurrence of classic PNH.