• P Mohan and P Brocklehurst.
• Pediatrics, Baylor College of Medicine, Houston, Texas 77030, USA. suseela12@hotmail.com
• Cochrane Db Syst Rev. 2003 Jan 1(4):CD003956.

BackgroundNeonatal sepsis causes significant neonatal mortality and morbidity. Neonates, especially preterm infants, have an immaturity of granulopoeisis and have a limited capacity for progenitor cell proliferation. This results in the frequent occurrence of neutropaenia in septic neonates. Neutropaenic septic neonates have a higher mortality than neonates who are septic but not neutropaenic. Transfusion of granulocytes to septic neutropaenic neonates, therefore, may help reduce mortality and morbidity.ObjectivesThe primary objective of this review was to determine the efficacy and safety of granulocyte preparations (granulocyte and buffy coat transfusions) as adjuncts to antibiotics for the treatment of confirmed or suspected sepsis in neonates with neutropaenia in reducing all-cause mortality during hospital stay and adverse neurological outcome at a year of age or later. Secondary objectives were to determine the effects of granulocyte transfusions on length of hospital stay in survivors to discharge, adverse effects (fluid overload, transmission of blood borne infections, pulmonary complications and sensitisation to donor leukocyte antigens) and immunologic outcomes at a year of age or later.Search StrategySearches for eligible trials were made in September 2002 of the following electronic databases: The Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 3, 2002), MEDLINE, EMBASE, CINAHL. The proceedings of the Pediatric Academic Societies, which were published in the journal, Pediatric Research, were searched from 1987 onwards. The Cambridge Database of Scientific Abstracts (1982 onwards) and Dissertation Abstracts on disc (1960 onwards) were also searched. Communications with prominent authors in the field and additional searches in the reference lists of identified trials were performedSelection CriteriaRandomised and quasi-randomised studies were included, if the participants were neonates with suspected or confirmed sepsis and neutropaenia who received granulocyte preparations (granulocyte concentrates prepared by leukophoresis or buffy coat transfusions prepared by simple centrifugation of blood) at any dose or duration compared with placebo or no granulocyte transfusion, each as adjuncts to antibiotics or compared to other adjuncts to antibiotics to treat sepsis (e.g. intravenous immunoglobulin). Studies were included if any of the following outcomes were reported: all-cause mortality during hospital stay, neurological outcome at one year of age or later, length of hospital stay in survivors at discharge, adverse effects or immunological outcome at one year or laterData Collection And AnalysisTrials identified by the search strategy were assessed for eligibility for inclusion in the review. Full text versions of the eligible trials were obtained. Data of interest were extracted on paper forms. Relative risk (RR) and risk difference (RD) with 95% confidence intervals using the fixed effects model were reported for dichotomous outcomes. Pre-specified subgroup analyses were done and reported.Main ResultsFour eligible studies were identified. A total of 44 infants with sepsis and neutropaenia on antibiotics were randomised in three trials to granulocyte transfusions OR placebo/no transfusion. In another trial, 35 infants with sepsis and neutropaenia on antibiotics were randomised to granulocyte transfusion or intravenous immunoglobulin. When granulocyte transfusions were compared with placebo or no transfusion, there was no significant difference in all-cause mortality [typical RR 0.89 (95% CI 0.43, 1.86), typical RD -0.05 (95% CI -0.31, 0.21)]. There was no statistical heterogeneity in the results of the included trials. When granulocyte transfusions were compared with intravenous immunoglobulin, there was a reduction in all-cause mortality of borderline statistical significance [RR 0.06 (95% CI 0.00, 1.04), RD -0.34(95% CI -0.60, -0.09), NNT 2.9 (95% CI 1.7, 11.1)]. Pulmonary complication, seen in four infants (4%), was the only adverse effect reported. This adverse effect was noticed in the two trials which used buffy coat transfusions. None of the trials reported on neurological outcome at one year of age or later, length of hospital stay in survivors to discharge or immunological outcome at one year of age or later.Reviewer's ConclusionsCurrently, there is inconclusive evidence from RCTs to support or refute the routine use of granulocyte transfusions in neonates with sepsis and neutropaenia to reduce mortality and morbidity. Researchers can be encouraged to conduct adequately powered multicentre trials of granulocyte transfusions to clarify their role in neonates with sepsis and neutropaenia. Other adjuncts to antibiotics aimed at improving host defence mechanisms such as colony stimulating factors, IVIG and pentoxifylline should also be tested in RCTs.

12 keywords...

hide…