Filtration versus gravity leukapheresis in febrile granulocytopenic patients: a randomized prospective trial

Forty-eight patients with fever greater than 38.3 degrees C for at least 24 hr despite broad spectrum antibiotics and an absolute granulocyte count less than 1000/microliter were randomly allocated to 4 days of granulocyte transfusions obtained by leukapheresis using filtration (n = 27) or gravity (n = 21) techniques, the latter permitting simultaneous nonmechanical collection of granulocytes and platelets utilizing hydroxyethyl starch as a sedimenting agent. Patient characteristics and dose of granulocytes obtained from both techniques were similar. Complete response to granulocyte transfusions was established by a reduction in temperature to less than 37.2 degrees C sustained for at least 48 hr after the fourth transfusion with sterilization of cultures where previously positive and diminution of measurable infection when present. This occurred in 6/21 (29%) for gravity leukapheresis and 9/27 (33%) for filtration leukapheresis. An additional group had diminution in temperature and clinical improvement during transfusions (6/21 gravity leukapheresis versus 10/27 filtration leukapheresis). Eighty-six percent of patients transfused with gravity leukapheresis cells were alive at day 20 compared with 81% for filtration leukapheresis cells. Transfusion reactions were comparable. Thus, gravity leukapheresis appears to be as efficacious as filtration leukapheresis for treating granulocytopenic febrile patients, with the added advantages of availability to any blood bank without new equipment, of having platelets as by-products, and of not requiring donor heparinization.