Introduction

The aim of haemodynamic resuscitation in haemorrhagic shock is to maintain systemic haemodynamics so as to limit microcirculatory hypoperfusion and tissue hypoxia and thus protect organ function. In the acute phase of haemorrhage, the priority is to stop the bleeding as soon as possible. As long as this bleeding is uncontrolled, aggressive fluid resuscitation may increase the risk of bleeding. Indeed, fluid resuscitation may promote coagulopathy by the dilution of coagulation factors and induction of hypothermia. In addition, increasing the arterial pressure may impede clot formation. Thus, a combination of low volume fluid resuscitation and permissive hypotension is used to avoid the adverse effects of early aggressive fluid resuscitation while maintaining a level of tissue perfusion that, although lower than normal, is tolerable over short periods . During this phase, occult microvascular hypoperfusion and tissue hypoxia may develop over time. After bleeding is controlled, the main therapeutic goals are to restore the blood pressure and macrovascular oxygen delivery to limit tissue hypoxia and organ dysfunction. It is expected that the optimization of macrovascular haemodynamic parameters will result in improved microcirculation and restore tissue oxygenation. However, this is the case only if the compensatory haemodynamic response to haemorrhagic shock, including hormonal, neural, biochemical and vascular regulatory control systems, remains intact without alterations in microvascular regulation. Haemodynamic coherence must be maintained to expect restoration of microcirculation and tissue oxygenation through systemic haemodynamic-driven resuscitation. However, haemorrhagic shock, reperfusion, traumatic injury and inflammation can damage the macrocirculatory and microcirculatory compensatory responses to haemorrhage and then lead to a loss of haemodynamic coherence. In these cases, a systemic haemodynamic-driven resuscitation would not be effective in restoring microcirculation and tissue oxygenation. Thus, it is obvious that there is a crucial need for an appropriate technique to monitor microcirculation at the bedside and guide the resuscitation on the basis of macrovascular and microvascular parameters. By doing so, it would be possible to adapt the resuscitation to optimize microcirculation parameters and detect the potential loss of haemodynamic coherence between macrocirculation and microcirculation when the optimization of macrocirculation fails to improve microcirculation. A real-time technique enabling microcirculation monitoring can become a part of haemodynamic algorithms of haemorrhagic shock resuscitation, thus creating the opportunity for microcirculatory haemodynamic-driven resuscitation to become the gold standard in the future.

Microcirculatory response to haemorrhage

In the acute phase of haemorrhage, macrocirculatory and microcirculatory responses rapidly compensate for blood loss and limit tissue hypoxia. The macrocirculatory compensatory response engages the autonomic nervous system. Decrease in venous return and arterial pressure leads to the unloading of cardiopulmonary and arterial baroreceptors. This induces a decrease in the activation of the vasomotor inhibitory centre in the brainstem that in turn leads to the activation of the vasomotor centre (sympathetic centre) and inhibition of vagal activity (sinoatrial node). The increased activity of the sympathetic nerves increases the heart rate, cardiac contractility, and arterial and venous tone and activates the renin-angiotensin-aldosterone system. The magnitude of the compensatory vasoconstriction that follows is the net result of the combined effects of norepinephrine from the peripheral nerves on the peripheral vascular adrenoceptors, epinephrine from the adrenal medulla and non-adrenergic mechanisms (i.e. angiotensin and vasopressin). Arterial vasoconstriction rapidly decreases non-vital organ blood flow (musculocutaneous, splanchnic and renal blood flow) to maintain perfusion pressure and blood flow to vital organs (heart and brain). It is important to keep in mind that the sympathetic stimulation activates both arterial and venous α-adrenergic receptors, which recruit blood from the venous unstressed volume to maintain venous return and cardiac output. Microcirculation regulates the distribution of blood flow within organs to balance oxygen delivery and oxygen demands. To do so, microcirculation limits blood flow in microcirculatory units with low oxygen demands and increases blood flow in microcirculatory units with high oxygen demands. This microvascular heterogeneity of blood flow is an essential property of normal microcirculation and is necessary to match oxygen delivery and metabolic needs. Such metabolic-driven heterogeneity of blood flow guarantees optimal oxygen extraction. During haemorrhagic shock simultaneously to the macrovascular redistribution of arterial blood flow at the expense of non-vital organs, blood flow is redistributed within the capillary networks of each organ dictated by the arteriolar tone, rheologic factors and oxygen demand. The local regulation of the arteriolar tone is a crucial factor in the microvascular matching of oxygen supply to oxygen demand. Several mechanisms contribute to the local regulation of the arteriolar tone including response to intra-luminal pressure (myogenic response), shear stress on the glycocalyx and endothelial cells (shear-dependent response), and tissue metabolite concentrations (metabolic response). An increasingly important role is played by the red blood cells (RBCs) and haemoglobin molecule in the regulation of the microvascular tone and in matching the oxygen supply to oxygen demand. Ellsworth et al. suggested that RBCs behave as mobile oxygen sensors and control the vascular tone through the release of ATP . ATP is released from RBCs in response to the mechanical deformation of their membranes, the transition of haemoglobin from oxygenated form to the deoxygenated form or receptor-mediated activation of RBC membrane-bound β-adrenergic receptors or prostacyclin receptors . The RBC-derived ATP interacts with endothelial purinergic receptors, thus inducing the release of vasodilator mediators. Vasodilation is conducted in a retrograde fashion, resulting in increased blood flow (oxygen supply) to areas of increased oxygen demand. RBC intra-cellular ATP is decreased in haemorrhage and corrected by transfusion. Other mechanisms involving RBCs in the regulation of the vascular tone have been proposed . Although this hypothesis is yet to be confirmed, it remains appealing in explaining the microvascular response to oxygen demand.

However, despite the microvascular response, oxygen delivery could be insufficient to cover oxygen demand during haemorrhagic shock, and tissues have to down-regulate their energy needs to limit tissue hypoxia.

The compensatory response of microcirculation to acute decrease in oxygen delivery during haemorrhagic shock could be impaired by the damage induced by the severity of the shock, reperfusion and inflammation. Restoring macrocirculation abnormalities can be ineffective in restoring microcirculation and correcting tissue oxygenation.

Microcirculatory response to haemorrhage

In the acute phase of haemorrhage, macrocirculatory and microcirculatory responses rapidly compensate for blood loss and limit tissue hypoxia. The macrocirculatory compensatory response engages the autonomic nervous system. Decrease in venous return and arterial pressure leads to the unloading of cardiopulmonary and arterial baroreceptors. This induces a decrease in the activation of the vasomotor inhibitory centre in the brainstem that in turn leads to the activation of the vasomotor centre (sympathetic centre) and inhibition of vagal activity (sinoatrial node). The increased activity of the sympathetic nerves increases the heart rate, cardiac contractility, and arterial and venous tone and activates the renin-angiotensin-aldosterone system. The magnitude of the compensatory vasoconstriction that follows is the net result of the combined effects of norepinephrine from the peripheral nerves on the peripheral vascular adrenoceptors, epinephrine from the adrenal medulla and non-adrenergic mechanisms (i.e. angiotensin and vasopressin). Arterial vasoconstriction rapidly decreases non-vital organ blood flow (musculocutaneous, splanchnic and renal blood flow) to maintain perfusion pressure and blood flow to vital organs (heart and brain). It is important to keep in mind that the sympathetic stimulation activates both arterial and venous α-adrenergic receptors, which recruit blood from the venous unstressed volume to maintain venous return and cardiac output. Microcirculation regulates the distribution of blood flow within organs to balance oxygen delivery and oxygen demands. To do so, microcirculation limits blood flow in microcirculatory units with low oxygen demands and increases blood flow in microcirculatory units with high oxygen demands. This microvascular heterogeneity of blood flow is an essential property of normal microcirculation and is necessary to match oxygen delivery and metabolic needs. Such metabolic-driven heterogeneity of blood flow guarantees optimal oxygen extraction. During haemorrhagic shock simultaneously to the macrovascular redistribution of arterial blood flow at the expense of non-vital organs, blood flow is redistributed within the capillary networks of each organ dictated by the arteriolar tone, rheologic factors and oxygen demand. The local regulation of the arteriolar tone is a crucial factor in the microvascular matching of oxygen supply to oxygen demand. Several mechanisms contribute to the local regulation of the arteriolar tone including response to intra-luminal pressure (myogenic response), shear stress on the glycocalyx and endothelial cells (shear-dependent response), and tissue metabolite concentrations (metabolic response). An increasingly important role is played by the red blood cells (RBCs) and haemoglobin molecule in the regulation of the microvascular tone and in matching the oxygen supply to oxygen demand. Ellsworth et al. suggested that RBCs behave as mobile oxygen sensors and control the vascular tone through the release of ATP . ATP is released from RBCs in response to the mechanical deformation of their membranes, the transition of haemoglobin from oxygenated form to the deoxygenated form or receptor-mediated activation of RBC membrane-bound β-adrenergic receptors or prostacyclin receptors . The RBC-derived ATP interacts with endothelial purinergic receptors, thus inducing the release of vasodilator mediators. Vasodilation is conducted in a retrograde fashion, resulting in increased blood flow (oxygen supply) to areas of increased oxygen demand. RBC intra-cellular ATP is decreased in haemorrhage and corrected by transfusion. Other mechanisms involving RBCs in the regulation of the vascular tone have been proposed . Although this hypothesis is yet to be confirmed, it remains appealing in explaining the microvascular response to oxygen demand.

However, despite the microvascular response, oxygen delivery could be insufficient to cover oxygen demand during haemorrhagic shock, and tissues have to down-regulate their energy needs to limit tissue hypoxia.

The compensatory response of microcirculation to acute decrease in oxygen delivery during haemorrhagic shock could be impaired by the damage induced by the severity of the shock, reperfusion and inflammation. Restoring macrocirculation abnormalities can be ineffective in restoring microcirculation and correcting tissue oxygenation.

Impact of haemorrhagic shock on microcirculation: pre-clinical research

Tissue hypoperfusion and oxygen deficit (i.e. decrease of oxygen delivery below required levels to support aerobic metabolism) have extensively been reported in experimental haemorrhagic shock models as crucial pathophysiological events leading to tissue hypoxia, inflammation, coagulopathy and multiple organ failures . A drop in cardiac output and oxygen delivery decreases microvascular blood flow and functional capillary density with an increase in flow heterogeneity in non-vital organs . In a progressive model of haemorrhage (three stepwise bleedings of 5 ml/kg at 30-min intervals) in anesthetized pigs, Dubin et al. reported that microvascular flow index (MFI) and capillary density in sublingual and ileal mucosa were progressively reduced with a progressive increase in heterogeneity flow index. These variations in microcirculation are accompanied by progressive decrease in cardiac output, superior mesenteric artery blood flow, lactate concentration and systemic and intestinal oxygen delivery . The fact that microcirculatory variations are closely related to macrocirculatory variables during the initial bleeding phase was confirmed in other haemorrhagic models. Krejci et al. observed that gastric and colon mucosal flow and liver and kidney flows decreased to a similar extent as superior mesenteric artery and systemic flows. In addition, van Iterson et al. reported that changes in microvascular pO ₂ in the gut and heart follow the changes in macrocirculatory parameters (cardiac index, mean arterial pressure and oxygen delivery) in the acute bleeding phase. Thus, in experimental models of haemorrhagic shock, haemodynamic coherence between macrocirculation and microcirculation is often preserved. However, it is important to notice that these models primarily explore the acute phase of haemorrhagic shock that is induced by blood spoliation without a prolonged analysis of the resuscitation phase.

However, some experimental studies suggested that haemorrhagic shock is capable of altering the compensatory microvascular response. For example, Kozar et al. demonstrated in a rodent model that haemorrhagic shock degrades the endothelial glycocalyx. Machiedo et al. reported that transfusion of RBCs from trauma haemorrhagic shock rats into naïve rats leads to impaired microcirculatory flow in several important organs, including the lungs, spleen, ileum and cecum. Finally, persistent microvascular alterations have been reported in haemorrhagic shock models despite adequate macrovascular resuscitation . Legrand et al. reported that fluid resuscitation with either normal or hypertonic saline targeting a low or high mean arterial pressure did not result in a correction of shock-induced renal microcirculatory hypoxia. Nevertheless, transfusion of fresh blood efficiently improved renal oxygenation, although it led to persistent hypoxic defects. This result reaffirms that microcirculatory oxygen delivery could be altered in haemorrhagic shock despite the restoration of macrocirculatory oxygen delivery, with a systemic haemoglobin level that is deemed adequate.

It is important to keep in mind that other factors in addition to the macrovascular haemodynamic factors could induce microvascular alterations in haemorrhagic patients. For example, tissue injury and severe hypoxemia in trauma patients can lead to a pro-inflammatory state through the release of endogenous factors termed damage-associated molecular patterns by activated immune cells or released from necrotic cells, which can contribute to microvascular alterations. Thus, microvascular alterations in haemorrhagic shock may be amplified by associated trauma injury , systemic inflammation or hypoxemia . Tissue injury can itself induce intra-vascular leukocyte accumulation and impairment of skeletal muscle microcirculation accompanied by tissue hypoxia . Hypoxemia can also contribute to microvascular dysfunction. This has been well demonstrated by Harrois et al. who found that during haemorrhagic shock, the occurrence of hypoxemia considerably alters villous intestinal perfusion as it decreases the fraction of perfused villi in a synergistic manner with blood loss, thereby increasing the risk of villous ischemia. Therefore, other factors in addition to macrovascular haemodynamic factors could induce microvascular alterations in haemorrhagic patients, and microcirculation can remain hypoperfused despite the restoration of macrocirculation. This has been extensively demonstrated in septic patients in whom the resuscitation response of microcirculation is often dissociated from the macrovascular response .