Blood components are essential to regulate microcirculatory blood flow.

We read with interest the article by Jacob and colleagues [1] on the regulation of microcirculatory blood flow. In our opinion, the authors missed an important parameter, i.e., blood components and, particularly, red blood cells (RBC). This is regrettable since blood composition is the main parameter that we can influence by our therapies (fluid resuscitation, blood or albumin transfusion) in shocked patients.

RBC may use different contradictory pathways to modulate microvascular flow by modifying nitric oxide (NO) bioavailability, which is responsible for precapillary dilatation and capillary perfusion.

On the one hand, erythrocytes may reduce NO bioavailability through hemoglobin NO scavenging; on the other hand, enhancing hematocrit may increase viscosity and wall shear stress (WSS), a crucial agonist enabling endothelium NO release.

Thus, in hypoxic conditions, RBC may sense tissue oxygen tension and release vasodilatory agents such as NO or ATP.

In moderate hemodilution, blood viscosity is reduced, cardiac output increases, wall shear rate (WSR; blood velocity) increases and WSS is unchanged (WSS = WSR × viscosity). NO bioavailability increases because of reduced NO scavenging by erythrocytes, leading to systemic vasoplegia. In extreme hemodilution [2] (hematocrit at 11%), viscosity, WSR, and WSS drop. Functional capillary density (FCD), which reflects microcirculatory flow, decreases. In this model, enhancing blood viscosity with high viscosity-plasma expander increased FCD. In the same way, in a murine model of hemorrhagic shock, transfusion of fresh RBC without oxygen-carrying capacity restored blood viscosity, and enhanced FCD, microvascular flow, and systemic hemodynamics. Recently, Tanaka and colleagues [3] observed that RBC transfusion improved sublingual FCD in humans in hemorrhagic shock with active blood loss.

We have known since 1999 from the results of a trial using modified human hemoglobin in traumatic hemorrhagic shock [4] that modifying blood properties could be harmful. Patients in this study had a better blood systolic pressure than control patients, but more of them died. Free hemoglobin scavenges NO, enhancing precapillary vasoconstriction and impairing capillary perfusion.

We do not share the authors’ opinion that it is “better to choose hemorrhagic rather than septic shock for yourself”. Guidelines for traumatic hemorrhagic shock management [5] do not consider RBC transfusion for their microcirculatory properties and viscosity improvement, but only for their oxygen-delivering capacity, which is altered for many hours after transfusion, when blood is stored. Recommending a low hemoglobin trigger delays RBC transfusion when microcirculation flow is already impaired.

Nicolas Morel and Marie Moisan have drawn attention to the special but confounding role played by RBC in modulating regional microcirculatory flow distribution. As they correctly point out, contributory effects include changes in blood viscosity and scavenging or provision of vasoactive metabolites. Such influences must be expected, especially in conjunction with transfusions of RBC in situations of treatment or prevention of shock [6]. Although the effect of blood transfusion on blood viscosity and flow resistance has fascinated physiologists and experimental clinicians for over 100 years, the action of RBC on blood flow distribution has not been resolved satisfactorily. This is, in part, due to the fact that microvascular flow resistance arises mainly from the leukocyte component of whole blood, and not from RBC-mediated changes in viscosity [7]. With respect to scavenging of endothelium-derived vasodilatory NO by the hemoglobin of RBC, such an effect may be countered by the recently described presence of an active type-3 NO synthase in RBC [8]. In addition, supplying oxygen to the tissue provokes the production of vasoconstrictors in vivo. This realization has disproved other hypotheses proposing oxygen-dependent destruction or removal of vasodilators by RBC [9].

These few examples of the problems still associated with understanding the mechanisms by which RBC may influence regional flow distribution illustrate the continued necessity of conducting well-planned clinical studies for the improvement of patient outcome in situations of shock. Although at first sight the approach to management of traumatic hemorrhagic shock seems more clear than that of septic states, Morel and Moisan have provided good reasons why this is not resolved.