erythropoietin
Pierre BOUZAT, M.D. 1,2,3 Anne MILLET, M.D. 1,2,4 Yvonnick BOUE, M.D. 1,2,3 Karin PERNET-GALLAY, Ph.D. 1,2 Thibaut TROUVE-BUISSON, M.D. 1,2,3 Lucie GAIDE-CHEVRONNAY, M.D. 1,2,3 Emmanuel L. BARBIER, Ph.D. 1,2 Jean-Francois PAYEN, M.D., Ph.D. 1,2,31 INSERM, U836, Grenoble, F-38042, France 2
Université Joseph Fourier, Grenoble Institut des Neurosciences, Grenoble, F-38042, France
3
Pôle d’Anesthésie-Réanimation, Hôpital Michallon, Grenoble, F-38043 France 4 Département de Réanimation Pédiatrique, Hôpital Couple-Enfant, Grenoble, F-38043, France
Correspondence: Dr J-F PAYEN, Grenoble Institut des Neurosciences, INSERM U836/Equipe 5. Université Joseph Fourier, Site Santé La Tronche. BP 170. Grenoble Cedex 9. 38042 France. Tel 33 4 56 52 05 89. Fax 33 4 56 52 05 98. Email: jfpayen@ujf-grenoble.fr
Received from INSERM U836 and Université Joseph Fourier, Grenoble, France
Running title: EPO and brain oxygenation after TBI
Keywords: erythropoietin; traumatic brain injury; brain edema; brain oxygenation; MRI; MR diffusion-weighted imaging; BOLD
Abstract
Objective. To investigate the effects of recombinant human erythropoietin
(rhEPO) on brain oxygenation in a model of diffuse traumatic brain injury (TBI).
Design. Adult male Wistar rats
Setting. Neurosciences and physiology laboratories
Interventions. Thirty minutes after diffuse TBI (impact-acceleration model), rats
were intravenously administered with either a saline solution (TBI-saline) or rhEPO (5,000 IU/kg; TBI-EPO). A third group received no TBI insult (sham-operated).
Measurements and main results. Three series of experiments were conducted
2h (H2) after TBI to investigate: (i) the effect of rhEPO on brain edema using diffusion-weighted MRI and measurements of apparent diffusion coefficient (n = 11 rats per group); local brain oxygen saturation (lSO2), mean transit time (MTT) and blood volume fraction (BVf) were subsequently measured using a multiparametric MR-based approach to estimate brain oxygenation and brain perfusion in the neocortex and caudoputamen; (ii) the effect of rhEPO on brain tissue PO2 (PbtO2) in similar experiments (n = 5 rats per group); and (iii) the cortical ultrastructural changes after treatment (n = 1 rat per group). Compared to the sham-operated group, TBI-saline rats showed a significant decrease in lSO2 and in PbtO2 alongside brain edema formation and microvascular lumen collapse at H2. Treatment with rhEPO reversed all of these TBI-induced changes. Brain perfusion (MTT and BVf) was comparable between the three groups of animals.
Conclusion. Our findings indicate that brain hypoxia can be related to
microcirculatory derangements and cell edema without evidence of brain ischemia. These changes were reversed with post-traumatic administration of rhEPO, thus offering new perspectives in the use of this drug in brain injury.
Introduction
Patients with severe traumatic brain injury (TBI) are treated by strict avoidance of conditions known to aggravate primary brain injury, such as arterial hypotension, hypertension, systemic hypoxia, hypercapnia or severe hypocapnia. The prevention of elevated intracranial pressure (ICP) and the maintenance of an appropriate cerebral perfusion pressure (CPP) are widely applied measures to minimize ischemic insult (1). However, recent studies have demonstrated that episodes of brain hypoxia as assessed by brain tissue PO2 (PbtO2) probes are common despite optimization of CPP, and are independently associated with poorer patient outcome (2,3). Aggressive treatment of low PbtO2 was associated with improved outcome compared to standard ICP/CPP-directed therapy in cohort studies of severely head-injured patients (4,5). In addition, PbtO2 could be primarily influenced by factors governing oxygen diffusion rather than total oxygen delivery (6). Collectively these findings suggest that brain tissue hypoxia, as reflected by low PbtO2, cannot be viewed as a simple consequence of macrovascular ischemic insult, but may reflect a more complex interaction between the injured tissue and its microvascular environment.
The transportation of oxygen into the brain tissue is driven by a diffusion gradient in oxygen concentration between blood and cell. Neurovascular coupling, the range of mechanisms that adjust arteriolar tone according to local metabolic needs, represents the basis of our current understanding on how this is achieved. It is also clear that the spatial organization of the vascular system is a key factor in maintaining effective oxygen transportation (7,8). In injured brain, the balance between oxygen demand and supply can be disrupted as a result of increased brain metabolism, reduction in cerebral blood flow, and/or disturbances in the microcirculation. Ultrastructural studies of brain microvessels after traumatic contusions have reported the occurrence of endothelial edema, microvascular collapse and perivascular swelling (9,10). The compression of microvessels by swollen glial-foot processes and/or their obstruction by microthromboses have been found responsible for peri-contusional ischemia (11-13). In TBI patients, such microvascular damage
accounted for the inability of pericontusional tissue to increase the oxygen extraction fraction (OEF) in response to moderate hypocapnia-induced hypoperfusion (14). Therefore, despite no evidence of frank cerebral ischemia, microvascular ischemia might develop after traumatic contusion, and is characterized by increased barriers to the diffusion of oxygen that reduce cellular oxygen delivery (14). Whether such relations among microcirculation disturbances, tissue edema and brain hypoxia exist without traumatic contusion is unknown.
Erythropoietin is a multifunctional agent with tissue-protecting antiapoptotic, anti-inflammatory, antioxidative and neurotrophic properties (15,16). Posttraumatic administration of recombinant human erythropoietin (rhEPO) was found to induce a significant reduction of inflammation and cell apoptosis, together with improved functional recovery after focal TBI in mice (17). Improved cognitive function was found in rats treated with erythropoietin-mimetic peptide even after mild TBI (18). Using the impact-acceleration model to induce diffuse TBI in rats, we found that rhEPO and carbamylated erythropoietin (CEPO), a modified erythropoietin lacking erythropoietic activity, blocked the development of posttraumatic cellular edema as early as 1 hr after the insult (19,20). In addition to its direct effects on neural cells, rhEPO may improve brain perfusion by acting on endothelial growth as well as on cerebral vasculature through an alteration in nitric oxide production (21,22). These findings prompted us to study the acute effects of rhEPO on brain oxygenation in a rat model of posttraumatic diffuse brain edema using a multiparametric MR-based approach to map both brain edema and oxygenation. We hypothesized in this study that the rhEPO-induced changes in brain edema and/or microvascular environment would in turn improve brain oxygenation after trauma.