Natural history of paroxysmal nocturnal hemoglobinuria clones in patients presenting as aplastic anemia

Abstract

To investigate the natural history of paroxysmal nocturnal hemoglobinuria (PNH) clones in patients with acquired aplastic anemia (AA). Twenty-seven patients with AA and a detectable PNH clone were monitored for a median of 5.7 years (range 1.5-11.5 years). Twenty-two patients received high-dose cyclophosphamide (HiCy) therapy. The erythrocyte and granulocyte PNH clone sizes were measured using flow cytometry and analyzed via CellQuest software. PE-conjugated anti-glycophorin A, anti-CD15, FITC-conjugated anti-CD59, and FLAER staining were used to define glycosylphosphatidylinositol-AP-deficient cells. We found a linear relationship between PNH clone size and the development of intravascular hemolysis, assessed by lactate dehydrogenase (LDH) values (Pearson correlation coefficient = 0.80, P < 0.001 for erythrocyte PNH clones; and Pearson correlation coefficient = 0.73, P < 0.0001 for granulocyte PNH clones). An erythrocyte PNH size of 3-5% and granulocyte PNH size of 23% were the thresholds to predict hemolysis as measured by an elevated LDH (receiver operating characteristic analyses with AUC = 0.96 for erythrocyte PNH clone sizes and AUC = 0.88 for granulocyte PNH clone sizes). Patients with small (≤15%) initial PNH clone sizes were less likely to develop an elevated LDH (mean ± SD: 236.9 ± 109.9 vs. 423.1 ± 248.8; P = 0.02). Over time, the PNH clone sizes remained stable in 25.9% of patients; 48.1% experienced a rise in the PNH clone size; and 25.9% experienced a decrease. The risk of developing clinically significant PNH after HiCy therapy appears to be low in AA patients with PNH clones, especially for those with small initial PNH clones and for those who respond to HiCy therapy.