A prospective comparison of four techniques for diagnosis of paroxysmal nocturnal hemoglobinuria.

Paroxysmal nocturnal hemoglobinuria (PNH) is an acquired clonal stem cell disorder with altered expression of glycosylphosphatidylinositol (GPI)-anchored proteins, resulting in the increased susceptibility of erythrocytes to complement-mediated lysis. This study compared the available laboratory methods for detection of PNH cells and evaluated their utility in routine clinical practice. Fifty patients were evaluated by flow cytometric immunophenotyping (FCMI) using CD55 and CD59 monoclonal antibodies, PNH gel card test (GCT), Ham test and sucrose lysis test (SLT). A PNH clone was detectable in erythrocytes in 14 (28%) patients by FCMI, 13 (26%) by GCT and 10 (20%) by Ham test and SLT. The GCT and lytic tests showed 100% specificity and sensitivity was 92.8% and 71.1%, respectively. The GCT results correlated with type III cells (positive for > or =3.21% type III cells) and lytic test results correlated with CD59(-) type III cells (positive for > or =5% CD59(-) type III cells). The GCT and lytic tests were comparable in their sensitivity to detect type II cells (positive for > or =18.5% type II cells). Among the available methods, FCMI is most sensitive, can quantify and delineate PNH cells with differential expression of GPI-anchored proteins. The GCT is a useful screening tool as it is fairly sensitive, easy to perform and interpret. Well-standardized lytic tests are fairly reliable as screening tests.