Granulocyte transfusions for neonatal sepsis

Volume 99 Number 2 Editorial correspondence Granulocyte transfusionsfor neonatal sepsis To the Editor: The article by Laurenti et aP on the use of polymorphonuclear leukocyte (PMN) transfusion for the treatment of sepsis in the newborn infant is prompted by in vitro studies that indicate poor performance of polymorphonuclear leukocytes in healthy and infected newborn infants. Their results on 20 newborn infants demonstrate what appears to be a profound effect on the outcome of neonatal bacterial infection. The editorial comment by Hill makes reference to this point. Laurenti and coworkers indicate encouraging preliminary studies that have been published, all by their group. We feel that a cautionary comment is applicable. The method of granulocyte harvesting used was continuous flow filtration leukopheresis in which granulocytes are separated by reversible adhesion to nylon fibers? This procedure gives the highest yields (95% of the original granulocytes processed) but results in more reactions to both patients and donors, and is not the procedure of choice for granulocyte harvesting in the United States. More importantly, in the context of treating neonates with defective PMN function, the procedure of filtration leukopheresis yields granulocytes with decreased or defective function? -~ The results are consequently surprising. We have recently reviewed blood component therapy in neonates? The indications for the use of granulocytes are remote. More studies are needed; however, it would seem prudent that in the rare cases in which this is justified, the granulocyte preparation used should be derived from a centrifugation and not from a filtration procedure. Ronald A. Sacher, M.D., F.R.C.P.(C) Director, Blood Bank Georgetown University Medical Center 3800 Reservoir Rd, N W Washington, DC 20007 Bruce A. Lenes, M.D. Senior Staff Fellow Blood Bank, N.LH. K. N. Siva Subramanian, M.D. Attending Neonatologist Neonatal L C. U., Georgetown University Medical Center REFERENCES 1. Laurenti F, Ferro R, Isacchi G, Panero A, Savignoni PG, Malagnino F, Palermo D, Mandelli F, and Bucci G: Polymorphonuclear leukocyte transfusion for the treatment of sepsis in the newborn infant, J PEDIATR 98:118, 1981. 2. Djerassi I, Kim KS, Suvansri U, et al: Continuous flow filtration leukopheresis, Transfusion 12:75, 1972. 3. McCullough J, Weiblen BJ, Deinard AR, et al: In vitro function and post-transfusion survival of granulocytes collected by continuous flow and by filtration leukopheresis, Blood 48:315, 1976. 4. Steigbigel RT, Baum J, MacPherson JL, et al: Granulocyte bacteriocidal capacity and chemotaxis as affected by contin- 5. 6. 337 uous flow centrifugation and filtration leukopheresis, steroid administration and storage, Blood 52:197, 1978. Wright DG, Kauffman JC, Chusid M J, et al: Functional abnormalities of human neu~rophils collected by continuous flow leukopheresis, Blood 46:901, 1975. Lenes BA, and Sacher RA: Blood component ~ e r a p y in neonatal medicine, in Clinics in Laboratory Medicine, vol 2, Philadelphia, WB Saunders Company (in press). Reply To the Editor: Recent studies show severe impairment of granulocyte function in stressed and infected neonates. 1-~The survival of newborn rats infected with group B streptococci was significantly increased by the subsequent intraperitonea~ administration of imman adult (and less so of neonatal) granuIocytes. In a series of l l neonates with purulent meningitis treated by granulocyte transfusion by our group (G. Marzetti et al, unpi~blished data) death ,ccurred in only one patient and hydrocephalus developed in ofil); one of the survivors. Although qualitative abnormalities of granulocytes collected by flow filtration leukapheresis have been reported, most of the observed changes were mild, affecting only some of the functions explored, sometimes reversible, closely related to the length of the extraction procedure, and sometimes avoidable by improving the collection technique. We examined ten of our concentrates stored up to 48 hours after collection, and found no significant impairment of phagocytosis, killing, and NBT test, and only a 20% decrease of chemotaxis at the end of the storage period. ~ In any case, the favorable results obtained in human and animal studies suggest that the in vitro abnormalities observed in granulocytes collected by flow filtration leukapheresis are not necessarily associated with a significant impairment of their in vivo performance?. The centrifugation technique has the advantages o f a better preserved granulocyte function and of avoiding several problems with the donor. On the other hand, lymphocytes may represent up to 30 to 50% of the ceils collected by centrifugation, and we believe that following the transfusion of this preparation the risk of graft versus host disease in the sick n e o n a t e - m o r e so in the very preterm infant--should be of concern. Contaminating lymphocytes may be destroyed by irradiation, but when we planned our study we felt that too little information was available on the functional properties of granulocytes after irradiation. For all these reasons we selected the flow filtration procedure, which may provide a preparation with only 2 to 3% lymphocytes, and we feel that this should be the method of choice for the treatment of neonates until the uncertainties have been clarified. The preliminary results of an immunologic follow-up of infants who received granulocyte transfusion for neonatal sepsis shows no evidence of alloantibodies against granulocytes or other blood cells, and no differences between these infants and a control group with respect to T&B lymphocyte or granulocyte function tests (E. De Luca Carapella et al, unpublished data). The present evidence suggests that granulocyte transfusion