Analysis of Platelets By Flow Cytometry in Patients with Paroxysmal Nocturnal Hemoglobinuria (PNH)

Paroxysmal Nocturnal Hemoglobinuria (PNH) is characterized by a clonal population of hematopoietic stem cells with an acquired somatic mutation in the PIG-A gene, giving rise to populations of circulating mature cells that are unable to synthesize glycosylphosphatidylinositol (GPI). The disease is most readily diagnosed by flow cytometry analysis of red blood cells, using antibodies specific for the GPI-linked protein CD59, or analysis of granulocytes, using antibodies specific for the GPI-linked protein CD24, along with the FLAER reagent, a fluorescent protein that binds to the GPI structure and which is detected only on the surface of GPI (+) cells. However, other mature blood lineages can be derived from the PNH clone. Notably, thrombosis is a major life threatening complication of PNH and may be triggered by complement activation on platelets that belong to the GPI-negative stem cell clone. The PNH clone size generally predicts thrombosis, but sometimes the proportion of PNH red cells and granulocytes are highly discordant, in which case there might be a role for the determination of the proportion of PNH platelets. Historically, flow cytometry analysis of platelets in patients with PNH has been technically difficult. Here is described a method to do this that avoids technical challenges by using aspirin and sepharose gel filtration of platelets to prevent their activation as well as simultaneous determination of CD59 expression and uptake of the FLAER reagent. Red cells were analyzed based on CD59 expression and granulocytes based on CD24 and FLAER. We analyzed blood samples from 48 patients with PNH and or AA/PNH who provided informed consent, 16 of whom had a prior history of thrombosis. To separate platelet rich plasma (PRP), whole blood collected in EDTA tubes was centrifuged at 200g for 7 minutes at room temperature with the brake turned off. After this step, there was no further centrifugation or vortexing of the platelets. A solution of aspirin was made up immediately prior to use and was added to the PRP at a final concentration of 0.5mMolar. Aspirinated PRP was then loaded on top of a sepharose-2B column using Tyrode's buffer. The platelet-rich turbid drops were collected, to isolate platelets from red cells and coagulation proteins. 50 ul of platelet rich buffer was then incubated with FLAER-Alexa-488 (Pinewood, 1:20 dilution) and CD59-PE (Serotec, 1:10 dilution) in the dark for 30' at room temperature. To prevent doublet events from confounding the analysis, the platelet suspension was diluted 1:200 in Hanks with 0.1% BSA. The sample was passed through a 35 uM Falcon cell strainer, and platelets were identified by forward/side scatter acquired on a log-log scale on a BD Facscan. The median proportion of PNH red cells, granulocytes and platelets was 24%, 86%, 76% respectively in the group without a history of thrombosis and 23% ,82%, and 65% in the group with a history of thrombosis. The proportion of PNH platelets was highly correlated with the proportion of PNH granulocytes (r=0.84). In two patients with almost undetectable PNH red cells and over 90% PNH granulocytes, the proportion of PNH platelets was over 90%; both were on prophylaxis and neither had thrombosis. It is predicted that this technique may be useful for determining thrombosis risk, particularly when the results from the analysis of rbc's and granulocytes are discordant.