Platelet transfusions and mortality in necrotizing enterocolitis

Abstract

Background: Prior studies have suggested an association between platelet transfusions (PTXs) and worse outcomes among infants with necrotizing enterocolitis (NEC), potentially mediated by proinflammatory factors released by platelets. However, the effects of storage on platelet proinflammatory factor release and the confounding role of illness severity on NEC outcomes have not been determined.

Study design and methods: First, neuropeptide Y (a potent splanchnic vasoconstrictor released by platelets) was measured by enzyme-linked immunosorbent assay in fresh frozen plasma and in the supernatant of leukoreduced apheresis-derived platelets at different times during storage. Next, we evaluated the relationship between PTX rates and death in a multicenter cohort of very-low-birth-weight infants with NEC, adjusting for illness severity.

Results: Neuropeptide Y levels increased over time in the supernatant of leukoreduced apheresis-derived platelets and were 4.4-fold and 8.9-fold higher than in fresh frozen plasma on Days 2 and 3 of storage, respectively (p < 0.001). Among 598 very-low-birth-weight infants, 44 developed NEC. In unadjusted analysis, PTX rate was 30.3 (95% confidence interval [CI], 11.5-80.1) per 100 infant-days among infants who died, compared to 6.0 (95% CI, 3.2-11.2) among survivors (incidence rate ratio, 5.1; 95% CI, 1.6-16.2; p = 0.006). In multivariable analysis, there was no association between PTX rate and mortality (incidence rate ratio, 3.0; 95% CI, 0.6-15.0; p = 0.18), although estimation was imprecise.

Conclusion: Proinflammatory mediators accumulate in platelet suspensions during storage. Although PTX rates were not associated with increased mortality among infants with NEC in our study, our estimates suggest the potential for such an association that needs evaluation in larger studies.

Figure 1.. NPY levels increase in LR-A-PLTs during storage.

A. NPY levels were measured in individual LR-A-PLTs right before release for transfusion. Columns and error bars represent the geometric mean and 95% confidence interval of 5 (Day 2), 11 (Day 3) and 4 (Day 4/5) samples per time point; test for linear trend, p < 0.001. B. NPY levels were serially measured in two LR-A-PLT units on days 3 and 4/5 of storage; p = 0.02 for increase over time).

Figure 2.. PF4 levels increase in LR-A-PLTs during storage.

A. Columns and error bars represent the mean and 95% confidence interval of 5 (Day 2), 11 (Day 3) and 4 (Day 4/5) samples per time point; test for linear trend, p < 0.001. B. Association between NPY (y-axis) and PF4 (x-axis) protein levels from 8 samples of fresh frozen plasma (FFP) and 17 samples of platelet (PLT) with both NPY and PF4 levels (p < 0.001 by linear regression, R² = 0.63).

Figure 3.. Transfusion rates after NEC.

Figure depicts the rates of transfusion of blood components after onset of medical and surgical NEC. Platelet transfusion rates were higher among 14 infants with surgical NEC, compared to 25 infants with medical NEC (IRR 12; 95% CI 4.1–34). Infants who died on the day of NEC onset not included. Abbreviations: NEC, necrotizing enterocolitis; IRR, incidence rate ratio; CI, confidence interval; RBC, Red blood cell; PLT, platelet; FFP, fresh frozen plasma; CRYO, cryoprecipitate; ALL, any blood component.

Figure 4.. Mean platelet count at first transfusion.

Figure shows platelet trajectory and count at timing of first platelet transfusion to reflect general platelet transfusion practices at centers involved in this study (108 infants and 1078 platelet count measurements). Model-based means and 95% confidence intervals were estimated from repeated measures model.