Non-leukodepleted red blood cell transfusion in sepsis patients: beyond oxygenation, is there a risk of inflammation?

No abstract

Donati and colleagues [1] evaluated the benefits of fresh, leukodepleted (LD) versus non-leukodepleted (nLD) erythrocyte transfusions on the microcirculation in sepsis patients. Oxygenation appeared equal in both groups. Differences between the two kinds of blood cells are presented in Table 1. In the reported study, there is a difference between the age of erythrocytes in the two groups (1.5 to 3 days for nLD versus 3.5 to 5 days for LD erythrocytes; a U test indicates P < 0.05); this suffices to influence oxygen delivery mediators ( Figure 1). The content of leukocytes prior to (and after) leukodepletion were not tested, nor was the freeing -which is usually very fast -of secreted or docked, soluble biological response modifiers. Considering the dynamics of secreted biological response modifiers in platelet components [2], there must be differences within the two groups, which possibly influenced sepsis conditions. Residual plasma within erythrocytes given by female donors may include antihuman leukocyte antigen (HLA) antibodies which attack the recipient's lung alveolar epithelium neutrophils, and as sepsis is characterized by the pathology of neutrophils that release microparticles and neutrophil extracellular traps that target lung epithelium [3], it may be feared here that the nLD condition aggravates pulmonary lesions.
The re-evaluation of procedures is infrequent in transfusion despite the rapid evolution of techniques and materials; for example, differential stresses are inflicted on erythrocytes, depending on the collection process (aphaeresis versus conventional whole blood), with consequences for neutrophils and (vascular) endothelial cells upon transfusion [4]. Clinical investigations and registered trials such as Donati and colleagues are valuable.

Authors' response
We thank Professor Garraud and colleagues for their interest in our study [1]. As underlined, the difference in the age of transfused red blood cells (RBCs) between the nLD group (4 (3.5 to 5) days) and the LD group (3 (1.5 to 3) days) may have influenced the RBC oxygen-delivery capacity. Stored RBCs lose their ability to release vasodilators (nitric oxide, ATP) during hypoxia [5,6]. We showed similar changes in microvascular reactivity (tissue oxygen saturation (StO 2 )-upslope) and oxygenation (StO 2 ) after nLD or LD RBC transfusions [1]. This may indicate that oxygen-delivery mediators were not sufficiently affected to determine relevant variations in the response observed.
Alternatively, heterogeneity in the study population prevented detection of subtle differences. Variability in the response to treatments is common during sepsis. The patient heterogeneity was underlined as a limitation of our investigation [1].
LD RBCs showed a more favorable effect on microcirculatory convective flow [1]. This may depend on the lower adhesiveness of LD RBCs to the endothelium [7]. The transfusion of nLD RBCs decreased blood flow velocity and increased glycocalyx damage markers [1]. Similar effects may reasonably occur in the lungs. As highlighted, anti-HLA antibodies in nLD blood from female donors may contribute to aggravate pulmonary lesions. Nonetheless, the evaluation of respiratory function went beyond our goals.   The efficacy of blood transfusion depends on multiple RBC-and patient-related factors. Understanding the response to transfusion during sepsis is a challenging task. Targeting predetermined hemoglobin levels and/or macrohemodynamics is clearly not sufficient. Monitoring the microcirculation may get us closer to the answer [8].

Competing interests
The authors declare that they have no competing interests.
Author details