OBJECTIVETo estimate whether the expression of glycosylphosphatidylinositol (GPI)-anchored protein in B lymphocytes could be an indicator for detecting paroxysmal nocturnal hemoglobinuria (PNH).

METHODSFlow cytometry and two-color McAbs were used to detect CD59 expression in B lymphocytes and granulocytes of 92 clinical samples. The feasibility of the indicators for diagnosis of PNH was assessed.

RESULTSB lymphocytes from healthy individuals mainly showed a single population of cells strongly positive for CD59, as the percentage of CD59 expression was (96.3 +/- 1.2)%. In PNH patients, B lymphocytes showed decreased expression of CD59, as well as negative cells and partly-positive cells were observed, as the percentage of CD59 expression was 10.5%-92.3%. In 26 of 28 non-PNH anemia patients, CD59 expression in B lymphocyte was normal, and in the other 2 patients, CD59 expression were decreased (< 95%), but no negative cell peak was observed. The precision CV of CD59 in B lymphocyte assay is less than 4.4%, stained samples could be kept stable for 24 h, the CD59 in B lymphocyte assay was 100% in accordance with routine assay (CD59 in granulocyte assay). The sensitivity of the indicators (CD59 in B lymphocytes and granulocytes) was 100%, the specificity was 97.4%, misdiagnosis rate was 2.6%, the rate of failing to diagnosis was 0%, the rate of positive forecast was 87.5%, the rate of negative forecast was 100%, the accuracy was 97.8%.

CONCLUSIONSThe test of CD59 expression in B lymphocytes by flow cytometry was simple, accurate and reproduceable. It could be a good marker for diagnosis of PNH because of high sensitivity and specificity.