The impact of leukoreduced red blood cell transfusion on mortality of neonates undergoing extracorporeal membrane oxygenation

Abstract

Background: Blood products containing leukocytes have been associated with negative immunomodulatory and infectious effects. Transfusion-related acute lung injury is partially explained by leucocyte agglutination. The Food and Drug Administration has therefore recommended leukoreduction strategies for blood product transfusion. Our institution has been using leukocyte-reduced blood via filtration for neonates on Extracorporeal Membrane Oxygenation (ECMO). We hypothesized that the use of leukocyte-reduced blood would decrease mortality and morbidity of neonatal ECMO patients.

Methods: Retrospective review of noncardiac ECMO in neonates from 1984-2011, stratified into year groups I and II (≤1996 and ≥1997). Demographics, duration and type of ECMO, complications, and outcome data were collected. Blood product use data was collected. Univariate, bivariate, and multivariate analyses determined predictors of risk-adjusted mortality by year group.

Results: Patients (827) underwent ECMO with 65.3% (540) in group I. Overall median blood product use in mL/kg/d was 36.2 packed red blood cells (pRBC), 8.1 platelets, and 0 cyroprecipitate and/or fresh-frozen plasma. Overall mortality was 16.4%. Median pRBC used or transfused was 42.1 mL/kg/d in group I versus 19.1 mL/kg/d group II (P <0.001). On bivariate analysis, there was no difference in crude mortality between the 2 year groups (17.2% versus 16.0%, P = 0.66). However, on multivariate analysis adjusting for demographics, diagnosis, complications, and blood product use other than pRBCs, each additional transfusion of 10 mL/kg/d of pRBC was associated with a 33% increase in mortality in group I (P <0.05). Group II also showed an increase in mortality with each additional transfusion (21%) but this was not statistically significant (P = 0.07). Days on ECMO were not associated with pRBC transfusion in group I but increased in group II (additional 3 d for each 10 mL/kg/d transfused). There was no difference in infectious complications between groups I and II.

Conclusions: Blood transfusion requirement has diminished in newborns undergoing ECMO at our institution. Transfusion of non leukocyte-reduced blood is associated with an increase in mortality whereas transfusion of leukocyte-reduced blood provided no benefit with a trend toward increased mortality. Further research is recommended to understand these trends.