A simple, effective method for removing granulocytes from stored blood is described. Microaggregate filtration removes approximately 95% of the granulocytes from blood which has been stored for 2 weeks, centrifuged and filtered. The mean number of remaining leukocytes is 8 +/- 3.7 x 10(8)/unit. The residual white cell population, which is composed almost entirely of lymphocytes, is substantially less than the average number of cells previously associated with febrile reactions. 45 patients were selected for the study. All had significant febrile transfusion reaction histories, and averaged one reaction for every 3.6 U of conventional red cell product transfused. Administration of 212 units of microaggregate filtered granulocyte poor red cells caused a 95% reduction in the incidence of fibrile reactions. The technique is inexpensive, easily incorporated into the routine of the clinical blood bank, and does not require "open-system' processing. These considerations make microaggregate filtration a logical first choice method for the preparation of granulocyte-poor red blood cells.

A patient with sickle cell disease who concomitantly developed red cell autoimmunity and alloimmunization is reported. The implied but 'wrong' specificity of the autoantibody mimicked one of the alloantibodies in the patient's serum. Although the patient's red blood cells phenotyped at Ro4, anti-rh" was eluted from them on several occasions. Absorption and secondary elution from selected cells proved the cell bound antibody had a unique and independent specificity from the anti-rh" in his serum. Standard antibody identification procedures did not distinguish these differences.

Serologic procedures were performed in parallel using hydrophilized polystyrene cuvettes and conventional glass test tubes. Consistent and comparable results were obtained with both containers. No significant differences in antibody titer, specificity, neutralization, or elution were observed. The effect of a proteolytic enzyme on red blood cell antigen was not altered by the hydrophilized plastic tubes. There was no difference in the stability of immune sera stored in glass or plastic at 4 degrees C and--30 degrees C. However, hemolysis of stored red blood cells was significantly slower in the hydrophilized plastic cuvettes.

A simple, effective method for removing granulocytes from stored blood is described. Microaggregate filtration removes approximately 95% of the granulocytes from blood which has been stored for 2 weeks, centrifuged and filtered. The mean number of remaining leukocytes is 8 ± 3.7 x 10^8 / unit. The residual white cell population, which is composed almost entirely of lymphocytes, is substantially less than the average number of cells previously associated with febrile reactions. 45 patients were selected for the study. All had significant febrile transfusion reaction histories, and averaged one reaction for every 3.6 U of conventional red cell product transfused. Administration of 212 units of microaggregate filtered granulocyte poor red cells caused a 95% reduction in the incidence of fibrile reactions. The technique is inexpensive, easily incorporated into the routine of the clinical blood bank, and does not require ‘open-system’ processing. These considerations make microaggregate filtration a logical first choice method for the preparation of granulocyte- poor red blood cells.