Objective: Alterations in cerebral microvascular blood flow may develop during sepsis, but the consequences of these abnormali-ties on tissue oxygenation and metabolism are not well defined. We studied the evolution of microvascular blood flow, brain oxy-gen tension (Pbo2), and metabolism in a clinically relevant animal model of septic shock. Design: Prospective randomized animal study. Setting: University hospital research laboratory. Subjects: Fifteen invasively monitored and mechanically ventilated female sheep.
Interventions: The sheep were randomized to fecal peritonitis (n = 10) or a sham procedure (n = 5), and craniectomies were per-formed to enable evaluation of cerebral microvascular blood flow, Pbo2, and metabolism. The microvascular network of the left frontal cortex was evaluated (at baseline, 6, 12, and 18 hr) using side- stream dark-field videomicroscopy. Using an off-line semiquantita-tive method, functional capillary density and the proportion of small perfused vessels were calculated. Pbo2 was measured hourly by a parenchymal Clark electrode, and cerebral metabolism was assessed by the lactate/pyruvate ratio using brain microdialysis; both these systems were placed in the right frontal cortex.
Measurement and Main Results: In septic animals, cerebral functional capillary density (from 3.1 ± 0.5 to 1.9 ± 0.4 n/mm,
p < 0.001) and proportion of small perfused vessels (from 98% ±
2% to 84% ± 7%, p = 0.004) decreased over the 18-hour study period. Concomitantly, Pbo2 decreased (61 ± 5 to 41 ± 7 mm Hg, p < 0.001) and lactate/pyruvate ratio increased (23 ± 5 to 36 ± 19, p < 0.001). At 18 hours, when shock was present, ani-mals with a mean arterial pressure less than 65 mm Hg (n = 6) had similar functional capillary density, proportion of small perfused vessels, and Pbo2 values but significantly higher lactate/pyruvate ratio (46 ± 18 vs 20 ± 4, p = 0.009) compared with animals with an mean arterial pressure of 65–70 mm Hg (n = 4). Conclusions: Impaired cerebral microcirculation during sepsis is associated with progressive impairment in Pbo2 and brain metab-olism. Development of severe hypotension was responsible for a further increase in anaerobic metabolism. These alterations may play an important role in the pathogenesis of brain dysfunction during sepsis. (Crit Care Med 2014; 42:e114–e122)
Key Words: brain oxygenation; cerebral microcirculation; microdialysis; sepsis; sidestream dark-field videomicroscopy
S
epsis and septic shock still represent major health issues, with persisting high morbidity and mortality rates (1), despite major progress in diagnosis and therapy. Inad-equate oxygen supply for metabolic needs, together with cell alterations, including mitochondrial dysfunction, may con-tribute to cellular injury and the subsequent development of multiple organ failure (2). Nevertheless, even after apparently stable global hemodynamics have been restored and markers of abnormal cell metabolism (e.g., blood lactate levels) corrected, organ failure may still occur, probably because regional flow (i.e., the microcirculation) remains impaired (3). The recent development of new imaging techniques has helped to better characterize this microvascular impairment (3–5).Acute encephalopathy is a frequent complication of sepsis, with a complex and incompletely understood pathophysiology Crit Care Med
Copyright © 2013 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins
DOI: 10.1097/CCM.0b013e3182a641b8 *See also p. 485.
1Department of Intensive Care, Erasme Hospital, Université Libre de Bruxelles, Bruxelles, Belgium. 2Department of Anesthesia and Intensive Care, ZOL St-Jan, Genk, Belgium. 3 Department of Neurosurgery, Erasme Hospital, Université Libre de Brux-elles, Bruxelles, Belgium. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (http://journals.lww.com/ccmjournal). Supported, in part, by grants from Fondation Erasme (Bourse de Recher-che 2008–2011) and Integra Lifesciences Service SAS. Dr. Taccone received the Basic Science Research Award from the Euro-pean Critical Care Research Network (ECCRN; October 2009). The remaining authors have disclosed that they do not have any potential con-flicts of interest.
For information regarding this article, E-mail: [email protected]
Sepsis Is Associated With Altered Cerebral
Microcirculation and Tissue Hypoxia in
Experimental Peritonitis*
Fabio Silvio Taccone, MD
1; Fuhong Su, MD, PhD
1; Cathy De Deyne, MD, PhD
2; Ali Abdellhai, MD
1;
Charalampos Pierrakos, MD
1; Xinrong He, MD
1; Katia Donadello, MD
1;
(6). Alterations in cerebral blood flow (CBF) may represent a key component in its development (7). Experimental data have also highlighted that endothelial dysfunction occurring during sepsis may lead to impairment of the cerebral micro-circulation, which may further reduce brain perfusion, so that CBF becomes inadequate to meet the brain’s oxygen needs (8). In septic animals, these microvascular alterations have been implicated in the development of electrophysiological abnor-malities and contribute to neurological dysfunction (9, 10).
Although several studies have evaluated the mechanisms involved in the development of cerebral microvascular abnor-malities during experimental sepsis, such as endothelial activa-tion, leukocyte rolling and adhesion, brain inflammaactiva-tion, and perivascular edema (10–12), it remains unclear whether these changes have an impact on tissue oxygenation and metabo-lism. Previous studies have shown that microcirculatory altera-tions, occurring for example after endotoxin injection or bowel anastomotic leakage, were associated with reduced tissue oxy-genation (13, 14) in the myocardium and gastrointestinal tract, but these observations cannot be extrapolated to the cerebral microcirculation, which has different metabolism and perfu-sion autoregulation (7).
The aim of this study was, therefore, to evaluate the rela-tionship of disturbances in brain tissue oxygenation and metabolism with microvascular alterations in a clinically rel-evant model of septic shock.
MATERIALS AND METHODS
For detailed Materials and Methods section, see supplemental data (Supplemental Digital Content 1, http://links.lww.com/ CCM/A738).
Experimental Animals
The study protocol was approved by the Institutional Review Board for Animal Care of the Free University of Brussels (Bel-gium). Care and handling of the animals were in accord with National Institutes of Health guidelines (Institute of Labora-tory Animal Resources). Experiments were performed on 15 female sheep (weight, 28–37 kg). On the day of the experi-ment, the animals were given intramuscular injections of mid-azolam (0.25 mg/kg—Dormicum; Roche SA, Beerse, Belgium) and ketamine hydrochloride (20 mg/kg—Imalgine; Merial, Lyon, France) as premedication. Tracheal intubation (Tracheal Tube, 8.0; Hi-Contour, Mallinckrodt Medical, Athlone, Ire-land) was performed after the IV injection of fentanyl (30 μg/ kg—Fentanyl; Janssen Pharmaceutica, Beerse, Belgium) and rocuronium (0.1 mg/kg—Esmeron; Organon, Oss, the Neth-erlands). Volume-controlled mechanical ventilation (Servo ventilator 900 C; Siemens-Elema, Solna, Sweden) was initiated as follows: tidal volume of 10 mL/kg, respiratory rate of 12–16 breaths/min, positive end-expiratory pressure of 5 cm H2O, Fio2 of 1, and inspiratory time/expiratory time of 1:2. General
anesthesia was maintained with a continuous IV infusion of ketamine hydrocloride (10 mg/kg/hr), morphine (0.5 mg/kg/ hr), and midazolam (0.5 mg/kg/hr). Muscular blockade was achieved using 10 μg/kg/hr of rocuronium. A Foley catheter
(14F; Beiersdorf AG, Hamburg, Germany) was placed in the bladder for continuous urine output monitoring.
Abdominal and Cerebral Surgery
The right femoral artery and vein were surgically exposed to place the arterial catheter (6F Vygon, Cirencester, UK) and the venous introducer. A 7F pulmonary artery catheter (Edwards Life Sciences, Irvine, CA) was advanced into a pulmonary artery via the introducer under monitoring of pressure wave-forms. The catheters were connected to pressure transducers (Edwards Life Sciences) with the zero pressure reference at mid-thorax level. Animals were randomized 2:1 to a sepsis group, in which peritonitis was induced, and a sham group. In the sepsis animals, a midline laparotomy was performed and a 2-cm incision made in the cecum for feces collection (1.5 g/kg of body weight). The cecotomy was then closed with a double suture and the cecum returned to the abdominal cav-ity after local disinfection. A plastic 25-cm tube (Beldico SA, Marche-En-Famenne, Luxembourg) was inserted through the abdominal wall among the intestinal loops for later feces injection, and the abdomen wall was closed in two layers. The animal was placed prone and allowed to stabilize. In the sham group, abdominal surgery was not performed to avoid the risk of postoperative infection.
A cruciate incision was made to open the scalp (eFig. 1, Supplemental Digital Content 3, http://links.lww.com/CCM/ A740), and two craniotomies were performed. On the left, a large bone flap was removed, the dura mater opened, and blood or cerebrospinal fluid gently removed using saline solu-tions and gauze. This left craniotomy was used to assess the microcirculation. A smaller 2-cm craniotomy was performed on the right, in the same area as that on the left side, and the dura mater was punctured to insert a microdialysis (MD) cath-eter (CMA 20, cut-off membrane 20 kDa, membrane length 10 mm; CMA Microdialysis AB, Solna, Sweden) and a Clark electrode (Licox catheter; Integra Lifesciences, Zaventem, Belgium) for tissue oxygen pressure (Pbo2) measurement.
Both catheters were placed under sterile conditions at a depth of 0.8–1 cm into the brain parenchyma.
Monitoring and Measurements
Mechanical ventilation was adjusted to ensure a Pao2 of 100–
130 mm Hg and Paco2 of 35–45 mm Hg. Blood gas analyses
Cerebral Microcirculation, Oxygen, and Metabolism Assessment
The microvascular network of the cerebral cortex was visualized using sidestream dark-field videomicroscopy (MicroScan; MicroVisionMedical, Amsterdam, the Neth-erlands), with a ×5 imaging objective giving ×326 magni-fication (8). At specific time points (Experimental Protocol section), five videos of minimum duration 20 seconds each were recorded from different areas using a computer and a video card (MicroVideo; Pinnacle Systems, Mountain View, CA). The images were stored by random number designa-tion, and an investigator blinded to the data analyzed these sequences semiquantitatively off-line (3). The mean flow index, the proportion of small perfused vessels (PSPV), and the functional capillary density (FCD) were calculated. The intracranial Pbo2 catheter was connected to a specific
monitor (Brain Tissue Oxygen Monitoring, AC31; Integra Lifesciences) and data collected hourly. The CMA 20 cath-eter was perfused with CNS perfusion fluid (148 mM NaCl, 2.7 mM KCl, 1.2 mM CaCl2, and 0.85 mM MgCl2; osmolal-ity 305 mOsm/kg; pH 6) at a flow rate of 1.0 μL/min by a miniaturized infusion pump (CMA 107; CMA Microdialysis AB). After 1 hour of stabilization, the perfusate was collected every 60 minutes in specific microvials. Samples were ana-lyzed for lactate, pyruvate, glycerol, glutamate, and glucose by an automatic analyzer (CMA 600 Microdialysis Analyzer; CMA Microdialysis AB). The lactate/pyruvate ratio (LPR) was calculated automatically.
Experimental Protocol
After the surgical procedures, baseline measurements, includ-ing cerebral microcirculation, oxygenation, and metabolism, were obtained. Feces were then spilled into the abdominal cav-ity (sepsis group). Ringer’s lactate solution and 6% hydroxy-ethyl starch solution (Voluven; Fresenius Kabi, Schelle, Belgium) were titrated to prevent hypovolemia. Refractory hypotension was defined as MAP less than or equal to 70 mm Hg despite fluid challenge. Septic shock was defined by the association of refractory hypotension with organ dysfunction (urine output < 0.5 mL/kg/hr and Pao2/Fio2 < 300) and lactate levels more than
2.0 mEq/L. Cerebral microcirculation was assessed at baseline and 6, 12, and 18 hours thereafter. All animals were observed for 18 hours after baseline, and then they were euthanized using IV potassium chloride.
Statistical Analysis
Statistical analysis was performed using SPSS 13.0 for Win-dows (2004 SPSS, Chicago, IL). Data are presented as mean ± sd or median (range). Normal distribution was confirmed
using the Kolmogorov-Smirnov test. Variables were com-pared using parametric Student t tests or Mann-Whitney U tests for nonparametric data. The significance of differences in the measured variables between groups was analyzed using a two-way (time and groups) analysis of variance for repeated measures, followed by a Bonferroni post hoc analy-sis. To estimate the correlation (expressed as r coefficient)
between different variables in the presence of repeated mea-surements, we used a mixed model with SAS system (version 9.2, SAS Institute, Paris, France). A p value of less than 0.05 was considered statistically significant.
RESULTS
At baseline, there were no significant differences between groups in hemodynamic and respiratory variables, cerebral microcirculation, oxygenation, or metabolic data (Table 1; eTable 1, Supplemental Digital Content 2, http://links.lww. com/CCM/A739). In the sham group, all these variables remained similar to baseline values throughout the study period. In the septic animals, the heart rate and CI increased and the systemic vascular resistance decreased within the first 6 hours after feces injection (Table 1; eFigs. 2–5, Supplemen-tal DigiSupplemen-tal Content 3, http://links.lww.com/CCM/A740). The MAP decreased by 20% from baseline within 12 hours; refrac-tory hypotension with lactic acidosis eventually occurred in all septic animals (median time, 16 hr [13–18 hr]), despite admin-istration of large amounts of fluid (median fluid balance, 4,550 mL [4,250–5,500 mL]). Stroke volume and LVSWI also decreased, particularly at the end stage of septic shock. Arte-rial blood gas analyses showed a progressive decrease in pH, Pao2/Fio2 ratio, and TPC over time, indicating sepsis-related
pulmonary dysfunction. Pao2 was maintained above 100 mm
Hg throughout the study period.
Cerebral FCD (from 3.1 ± 0.5 to 1.9 ± 0.4 n/mm,
p < 0.001) and PSPV (from 98% ± 2% to 84% ± 7%, p =
0.004) decreased significantly from baseline to shock onset (Figs. 1 and 2). FCD decreased significantly between baseline and 12 hours. Cerebral oxygenation decreased progressively over time, from 61 ± 5 mm Hg at baseline to 49 ± 7 mm Hg at 12 hours, with values that were significantly different from baseline during the last 5 hours of observation (Fig. 3). In septic animals, cerebral glucose levels also decreased progres-sively from baseline, despite relatively constant blood levels. In contrast, cerebral glycerol and glutamate levels remained relatively constant until shock onset, after which these levels increased sharply (Fig. 4). The LPR also increased, particularly at the onset of septic shock. Among animals in septic shock at 18 hours (n = 10), those with an MAP less than 65 mm Hg (n = 6) had similar lactate (3.7 ± 0.7 vs 3.3 ± 0.7 mEq/L, respectively), microcirculatory (FCD, 1.8 ± 0.3 vs 1.9 ± 0.5 n/ mm, respectively), and Pbo2 (41 ± 8 vs 41 ± 5 mm Hg,
respec-tively) values than did those with MAP 65–70 mm Hg (n = 4), but they significantly had higher LPR (46 ± 18 vs 20 ± 4,
p = 0.009), glycerol (51 ± 25 vs 16 ± 7 μmol/L, p = 0.009), and glutamate (44.6 ± 30.2 vs 3.3 ± 1.5 μmol/L, p = 0.008) levels (Fig. 5). There was a weak although not significant corre-lation between FCD and Pbo2 changes over time (r = 0.74, p = 0.08), but there was no correlation between FCD,
cere-bral glucose, Pbo2, LPR, and systemic hemodynamic
TAbLE 1.
Evolution of Systemic Hemodynamics, Respiratory, and biological Variables Over
Time in the Septic (n = 10) and Sham (n = 5) Animals
Variables Group baseline 6 Hr 12 Hr 18 Hr ANOVA Within Group ANOVA Sepsis vs Sham
Temperature, °C Sepsis 39.6 ± 0.4 40.1 ± 0.3 40.6 ± 0.5 41.3 ± 0.7 < 0.001 < 0.001
Sham 38.6 ± 0.4 39.1 ± 0.3 38.8 ± 0.5 38.6 ± 0.6 0.14
Heart rate, beats/min Sepsis 116 ± 13 151 ± 24 150 ± 18 139 ± 26 0.03 0.04
Sham 116 ± 14 117 ± 14 116 ± 8 112 ± 12 0.65
Cardiac index,
L/min/m2 Sepsis 4.7 ± 0.6 7.5 ± 2.1 6.7 ± 0.9 4.4 ± 1.0 < 0.001 < 0.001
Sham 4.2 ± 0.2 4.7 ± 0.2 4.6 ± 0.5 4.2 ± 0.4 0.21
Stroke volume, mL Sepsis 34 ± 6 41 ± 7 37 ± 5 27 ± 6 0.002 0.005
Sham 29 ± 4 29 ± 3 30 ± 6 30 ± 2 0.86
Left ventricular stroke work index, g-m/ beat
Sepsis 47 ± 9 52 ± 7 42 ± 9 20 ± 7 < 0.001 < 0.001
Sham 41 ± 4 42 ± 6 41 ± 7 40 ± 5 0.94
Mean arterial pressure,
mm Hg Sepsis 106 ± 8 98 ± 8 87 ± 13 57 ± 10 < 0.001 < 0.001 Sham 102 ± 4 97 ± 13 94 ± 8 91 ± 8 0.19 Mean pulmonary arterial pressure, mm Hg Sepsis 14 ± 2 13 ± 3 16 ± 3 18 ± 4 0.02 0.05 Sham 11 ± 3 12 ± 2 11 ± 3 12 ± 2 0.94 Pulmonary arterial occlusion pressure, mm Hg Sepsis 4 ± 1 3 ± 2 4 ± 2 4 ± 2 0.16 0.68 Sham 3 ± 2 4 ± 1 3 ± 1 4 ± 1 0.23
Right atrial pressure,
mm Hg Sepsis 1 ± 1 1 ± 1 1 ± 1 2 ± 1 0.09 0.18 Sham 2 ± 1 2 ± 1 2 ± 1 2 ± 1 0.34 Systemic vascular resistance index, dynes/s/cm5 Sepsis 1,804 ± 265 1,095 ± 286 1,043 ± 199 1,013 ± 167 < 0.001 < 0.001 Sham 1,820 ± 196 1,804 ± 177 1,839 ± 179 1,757 ± 151 0.52 Pulmonary vascular resistance index, dynes/s/cm5 Sepsis 177 ± 34 112 ± 34 146 ± 30 253 ± 93 < 0.001 0.008 Sham 167 ± 12 170 ± 22 177 ± 24 171 ± 26 0.44
Oxygen delivery, mL Sepsis 526 ± 63 947 ± 181 884 ± 120 601 ± 104 < 0.001 < 0.001
Sham 532 ± 113 514 ± 89 462 ± 48 474 ± 93 0.23 Oxygen consumption, mL Sepsis 132 ± 23 183 ± 28 193 ± 60 182 ± 45 0.009 0.12 Sham 111 ± 20 109 ± 17 116 ± 22 119 ± 18 0.94 Oxygen extraction rate, % Sepsis 24 ± 4 18 ± 4 20 ± 5 29 ± 5 0.001 0.003 Sham 19 ± 3 18 ± 4 17 ± 3 17 ± 5 0.52 Thoracopulmonary compliance, mL/ mm Hg Sepsis 19 ± 4 17 ± 3 16 ± 4 12 ± 3 0.01 0.05 Sham 19 ± 3 20 ± 2 19 ± 2 18 ± 3 0.09
Pao2/Fio2 Sepsis 403 ± 43 347 ± 99 312 ± 103 205 ± 41 < 0.001 < 0.001
DISCUSSION
The principal findings of this study are that cortical micro-vascular flow was progressively impaired during sepsis, espe-cially in septic shock, and that these alterations were associated with decreased cerebral oxygenation. Furthermore, cerebral metabolic disturbances compatible with tissue hypoxia (i.e., increased brain LPR) occurred mostly during shock, suggest-ing that hypotension is a critical factor in the development of anaerobic metabolism in the septic brain.
In experimental studies, microcirculatory alterations have been shown to be an early event in the changes of brain
perfusion associated with sepsis (8, 10). Vachharajani et al (11) showed that cerebral microvascular flow abnormalities were secondary to increased platelet and leukocyte adhesion to the vascular endothelium. In an ovine fecal peritonitis model, cor-tical microcirculatory abnormalities developed progressively and became significant at the onset of septic shock, despite aggressive fluid resuscitation (8). Furthermore, brain micro-circulatory changes were not related to changes in MAP, CI, or lactate in this study, suggesting that these alterations may occur even when systemic hemodynamics are maintained within
Paco2, mm Hg Sepsis 38 ± 3 37 ± 2 37 ± 3 43 ± 8 0.22 0.32
Sham 39 ± 3 37 ± 4 38 ± 3 39 ± 3 0.56
Fluid amount, mL Sepsis 314 ± 110 1,491 ± 413 2,911 ± 607 4,530 ± 560 < 0.001 < 0.001 Sham 241 ± 146 669 ± 131 1,194 ± 165 2,295 ± 288 < 0.001
Urine output, mL Sepsis 0 344 ± 277 712 ± 425 894 ± 460 < 0.001 0.21
Sham 0 313 ± 83 609 ± 117 1,098 ± 110 < 0.001
Glucose, mg/dL Sepsis 70 ± 14 81 ± 28 72 ± 19 71 ± 16 0.37 0.46
Sham 71 ± 20 67 ± 16 73 ± 17 66 ± 8 0.86
Hemoglobin, g/dL Sepsis 9.9 ± 1.2 11.4 ± 1.2 12.1 ± 1.4 12.9 ± 2.4 0.002 0.63
Sham 9.9 ± 1.0 11.4 ± 1.1 11.6 ± 0.9 12.1 ± 1.2 0.009
Lactate, mEq/L Sepsis 0.8 ± 0.3 1.1 ± 0.5 1.5 ± 1.1 3.5 ± 0.7 < 0.001 < 0.001
Sham 0.6 ± 0.3 0.8 ± 0.4 0.8 ± 0.4 0.8 ± 0.5 0.21
ANOVA = analysis of variance. Data are presented as mean ± sd.
TAbLE 1.
(Continued). Evolution of Systemic Hemodynamics, Respiratory, and biological
Variables Over Time in the Septic (n = 10) and Sham (n = 5) Animals
Variables Group baseline 6 Hr 12 Hr 18 Hr ANOVA Within Group ANOVA Sepsis vs Sham
Figure 1. Changes in cerebral functional capillary density (FCD) in the
septic (n = 10, open circle) and sham (n = 5, closed square) animals. Data are presented as mean ± sd. Analysis of variance for FCD: p value of
0.049 (time, sepsis group) and p value of less than 0.001 (group). p value of less than 0.05 versus baseline (*) or versus sham (#) with post hoc Bonferroni correction.
Figure 2. Changes in proportion of small perfused vessels (PSPV) in the
septic (n = 10, open circle) and sham (n = 5, closed square) animals. Data are presented as mean ± sd. Analysis of variance for PSPV: p value
of 0.02 (time, sepsis group) and p value of less than 0.001 (group).
p value of less than 0.05 versus baseline (*) or versus sham (#) with post
normal ranges. Also, lipopolysaccharide produces early delete-rious effects on brain microvascular endothelial cells, which are then responsible for alteration of the blood-brain barrier and, potentially, for tissue hypoperfusion (15).
Data suggest that microcirculatory alterations may have important consequences on brain cells and function. In endo-toxic rats, early microcirculatory failure preceded changes in evoked potential responses, suggesting that altered perfusion of active neurons was probably responsible for the develop-ment of these electrophysiological abnormalities during sep-sis (9). Using intravital microscopy in septic rats, Comim et al (10) showed a progressive increase in leukocyte and platelet adhesion within the brain microcirculation, associated with local production of proinflammatory cytokines and with abnormalities in locomotor function. Although the micro-circulation is determinant for the fine-tuning of oxygen and nutrient supply to organs, no data are available on the effects of microcirculatory alterations on brain oxygenation and metabolism in sepsis.
The effects of sepsis on tissue oxygenation have been described in various organs. In endotoxemic rats, Legrand et al (16) showed reduced renal cortical oxygenation, which was not prevented by fluid resuscitation. In murine skeletal muscle, Goldman et al (17) showed that increased microvas-cular flow heterogeneity could lead to a mismatch between local supply and demand, resulting in tissue hypoxia. After endotoxin administration in pigs, Humer et al (18) showed that increased heterogeneity of capillary transit times into the gut was an important cause of impaired tissue oxygenation. In a different model of porcine sepsis, an increased oxygen extraction rate with decreased local tissue oxygen tension was
found in sublingual and cuta-neous areas (19, 20). In septic rats, impaired microvascular flow was associated with tis-sue hypoxia in the myocar-dium, resulting in decreased cardiac contractility (14). Unfortunately, these studies usually did not report the rela-tionship between microcircu-lation density/flow and tissue oxygenation. In addition, the dysregulation of microvascu-lar flow and oxygen availability may be different in different organs and these results may not apply to the cerebral circulation.
An important question is whether systemic hemo-dynamics can influence the brain microcirculation and metabolism during sepsis. Some clinical studies have provided interesting informa-tion regarding this issue. Maekawa et al (21) showed reduced CBF and oxygen consumption in patients with sepsis when compared with healthy controls, and these changes were asso-ciated with a significant slowing of the electroencephalogram (EEG). In contrast, in mechanically ventilated septic patients with delirium (22), CBF was reported to be within the normal range (64 ± 29 mL/100 g/min), as was cerebral oxygenation. However, rather than absolute values, CBF autoregulation is the most important determinant of the adequacy of oxygen supply to neurons in response to changes in local demand or systemic hemodynamics; autoregulation is impaired in patients with sepsis and more so in patients with more severe disease, such as shock or delirium (23, 24). Impaired autoreg-ulation may leave brain tissue unprotected against the possibly harmful effects of arterial hypotension during sepsis, leading to cerebral ischemia. Interestingly, cerebral endothelial cells play an important regulatory role in CBF regulation (25, 26). Taken together, our data suggest that early brain microcir-culatory disturbances could be responsible for alterations in regional blood flow during sepsis, which would contribute to reduced tissue perfusion, as suggested by the decreased Pbo2,
and ultimately to cerebral ischemia, as indicated by increased LPR, when hypotension develops because of impaired auto-regulation and low CBF.
Pbo2 can decrease in various conditions associated with
primary brain injury, such as traumatic brain injury (TBI) or subarachnoid hemorrhage (SAH) (27). In this setting, changes in Pbo2 have been associated with changes in brain perfusion,
but they are also dependent on other more complex mecha-nisms inducing secondary brain hypoxia, such as reduced arte-rial oxygen content (e.g., hypoxemia or anemia) or increased
Figure 3. Changes in brain oxygen tension (Pbo2) in the septic (n = 10, open circle) and the sham (n = 5,
closed circle) animals. Data are presented as mean ± sd. Analysis of variance: p value of less than 0.001 (time,
demand (e.g., seizures or fever) (27). Furthermore, the mech-anisms involved in brain oxygenation disturbances may be different after TBI/SAH than during sepsis, so it is difficult to extrapolate the results. The decrease in Pbo2 we observed
was likely due to impaired microvascular perfusion because, first, we avoided hypoxemia and anemia in our study; second, seizures, although not monitored by continuous EEG, were unlikely to develop because of high sedatives doses; third, even though the solubility of oxygen is temperature dependent, we corrected Pbo2 values for blood temperature to eliminate
any confounding. Indeed, it has already been shown that Pbo2
should be considered as the reflection of dissolved plasma oxy-gen diffusion across the blood-brain barrier rather than as a measure of total oxygen delivery and metabolism (28). The decrease in FCD and PSPV we observed suggests an increase in the diffusion distance for oxygen from the capillaries to the target tissues. Furthermore, the increase in the capillary het-erogeneity reflects the nonadaptive pathologic changes occur-ring within the microcirculation and responsible for cellular dysfunction.
We evaluated cerebral metabolism using MD, which is a validated technique to monitor brain metabolism after acute brain injury. In TBI, neurochemical disturbances typically include elevated glutamate (reflecting excitotoxicity), glyc-erol (reflecting degradation of cellular membrane), and LPR, as well as low glucose (reflecting systemic hypoglycemia or low regional flow) concentrations, and have been related to poor outcome (29). In particular, an LPR greater than 40 has been frequently reported as a feature of brain ischemia (30, 31). However, LPR may be elevated even if adequate cerebral perfusion is maintained (32) and could indicate an increased glycolytic response in cerebral cells, such as astrocytes, or mito-chondrial dysfunction (33, 34). In our study, we first observed a progressive decrease in brain glucose levels, independent of changes in blood glucose, suggesting decreased glucose supply, probably because of limited regional flow due to microcircula-tory alterations (35). As such, the decrease in pyruvate levels, associated with the increased LPR, glutamate, and glycerol lev-els, strongly suggests tissue ischemia. Mitochondrial dysfunc-tion, characterized by an impairment of cellular respiration in the presence of normal oxygen levels, was unlikely because of the concomitantly reduced Pbo2 (34). It is possible that these
metabolic alterations were secondary to other mechanisms inducing cellular injury; as such, lactic acidosis may increase infarct size in animal model of cerebral ischemia through the exacerbation of oxidative glutamate toxicity on both neuronal and astrocyte cells (36). Nevertheless, cerebral microvascular abnormalities may significantly contribute to reduced glucose and oxygen supply to brain cells and promote brain injury. An increased muscle-to-serum lactate gradient has been proposed as an important marker of sepsis progression in clinical stud-ies (37). Furthermore, in experimental endotoxemia, reduced lactate and glycerol levels characterized the metabolic tissue response (38). Our data are, therefore, the first to report cere-bral metabolism disturbances occurring during sepsis in an experimental model of peritonitis.
Figure 4. Changes in cerebral glucose (A), glycerol (b), glutamate (C), and
lactate/pyruvate ratio (LPR, D) in the septic (n = 10, open circle) and sham (n = 5, closed circle) animals. Data are presented as mean ± sd. Analysis
of variance (ANOVA) for cerebral glucose: p value of less than 0.001 (time, sepsis group) and p value of 0.01 (group). ANOVA for cerebral glycerol:
p value of 0.002 (time, sepsis group) and p value of less than 0.001 (group).
ANOVA for cerebral glutamate: p value of less than 0.001 (time, sepsis group) and p value of less than 0.001 (group). ANOVA for cerebral LPR:
This study has some limitations. First, we only analyzed cortical brain microcirculation, and it remains unknown whether these results can be extrapolated to other cerebral areas. Second, we studied previously healthy animals under tightly controlled conditions, in contrast to the clinical setting in which patients often have underlying illnesses and other risk factors for brain hypoperfusion. Third, most metabolic disturbances occurred during the hypotensive phase, and the effect of MAP correction, for example, by administering vaso-pressors, on alterations of microvascular flow, oxygenation, and metabolism need to be further evaluated. Furthermore, whether the induction of hypotension in healthy brain could reproduce the same metabolic changes remains to be deter-mined. However, the use of vasodilators could per se induce changes in microcirculation (39), independently from MAP, so that it would be difficult to test this hypothesis in our model. Furthermore, other important confounders, such as cardiac output, should also be controlled to make microcirculatory analysis comparable. Importantly, an alternative approach could be the induction of bleeding or heart failure to decrease blood pressure. Of note, Wang et al (39) already showed that brain microcirculation is preserved in severe hemorrhagic and in cardiogenic shock, suggesting that sepsis per se, and not only hypotension, may impair cerebral microvascular flow. Fourth, our model is lethal and we could not evaluate any association between microvascular disturbances and clinical neurological abnormalities. Also, we did not assess whether microcirculatory disturbances would have any impact on brain function, as evaluated by EEG, or cellular morphology, using neuropathologic examination of the frontal lobe. Fifth, we did not evaluate cerebral autoregulation or directly mea-sure CBF. Furthermore, the Pbo2 catheter was placed in the
gray cortical matter, and base-line values in our study were significantly different from those reported from the white matter in clinical settings, and any specific cut-off or defini-tion of brain hypoxia was not possible. The use of a MD flow rate of 1.0 μL/min (rather than 0.3 μL/min in the clinical setting) provided only partial recovery of metabolites, and no comparison of concentra-tions of different molecules with other studies was pos-sible. Sixth, all animals were deeply sedated and one could argue that LPR abnormali-ties may have occurred earlier with higher metabolic rates, as occurs in clinical practice with a minimal sedation strategy. Finally, these data are difficult to extrapolate to the human setting; the brain microcirculation is still impossible to moni-tor and visualize in clinical practice without direct exposure of the cerebral cortex after craniectomy so that few data on microvascular abnormalities are available and those that do exist are only for primary brain injury.
CONCLUSION
In this clinically relevant model of septic shock, alterations in brain microcirculation were associated with a progressive decrease in cerebral oxygenation and, in the shock state, with impaired brain metabolism.
ACKNOWLEDGMENT
We thank Isabelle De Neve, Department of Sterilization at Erasme Hospital, Brussels, Belgium, for providing the neuro-surgical material.
REFERENCES
1. Vincent JL, Sakr Y, Sprung CL, et al; Sepsis Occurrence in Acutely Ill Patients Investigators: Sepsis in European intensive care units: Results of the SOAP study. Crit Care Med 2006; 34: 344–353
2. Sair M, Etherington PJ, Peter Winlove C, et al: Tissue oxygenation and perfusion in patients with systemic sepsis. Crit Care Med 2001; 29:1343–1349
3. De Backer D, Creteur J, Preiser JC, et al: Microvascular blood flow is altered in patients with sepsis. Am J Respir Crit Care Med 2002; 166:98–104
4. Doerschug KC, Delsing AS, Schmidt GA, et al: Impairments in micro-vascular reactivity are related to organ failure in human sepsis. Am J
Physiol Heart Circ Physiol 2007; 293:H1065–H1071
5. Sakr Y, Dubois MJ, De Backer D, et al: Persistent microcirculatory alterations are associated with organ failure and death in patients with septic shock. Crit Care Med 2004; 32:1825–1831
Figure 5. Differences in cerebral lactate/pyruvate ratio, cerebral glutamate, and cerebral glycerol between
6. Iacobone E, Bailly-Salin J, Polito A, et al: Sepsis-associated encephalopathy and its differential diagnosis. Crit Care Med 2009; 37:S331–S336
7. Taccone FS, Scolletta S, Franchi F, et al: Brain perfusion in sepsis.
Curr Vasc Pharmacol 2013; 11:170–186
8. Taccone FS, Su F, Pierrakos C, et al: Cerebral microcirculation is impaired during sepsis: An experimental study. Crit Care 2010; 14:R140 9. Rosengarten B, Hecht M, Auch D, et al: Microcirculatory dysfunction in
the brain precedes changes in evoked potentials in endotoxin-induced sepsis syndrome in rats. Cerebrovasc Dis 2007; 23:140–147 10. Comim CM, Vilela MC, Constantino LS, et al: Traffic of leukocytes
and cytokine up-regulation in the central nervous system in sepsis.
Intensive Care Med 2011; 37:711–718
11. Vachharajani V, Russell JM, Scott KL, et al: Obesity exacerbates sepsis-induced inflammation and microvascular dysfunction in mouse brain. Microcirculation 2005; 12:183–194
12. Papadopoulos MC, Lamb FJ, Moss RF, et al: Faecal peritonitis causes oedema and neuronal injury in pig cerebral cortex. Clin Sci (Lond) 1999; 96:461–466
13. Diller R, Stratmann U, Helmschmied T, et al: Microcirculatory dysfunc- tion in endotoxemic bowel anastomosis: The pathogenetic contribu-tion of microcirculatory dysfunction to endotoxemia-induced healing impairment. J Surg Res 2008; 150:3–10
14. Bateman RM, Tokunaga C, Kareco T, et al: Myocardial hypoxia-induc-ible HIF-1alpha, VEGF, and GLUT1 gene expression is associated with microvascular and ICAM-1 heterogeneity during endotoxemia.
Am J Physiol Heart Circ Physiol 2007; 293:H448–H456
15. Cardoso FL, Kittel A, Veszelka S, et al: Exposure to lipopolysaccha-ride and/or unconjugated bilirubin impair the integrity and function of brain microvascular endothelial cells. PLoS One 2012; 7:e35919 16. Legrand M, Bezemer R, Kandil A, et al: The role of renal hypoperfusion
in development of renal microcirculatory dysfunction in endotoxemic rats. Intensive Care Med 2011; 37:1534–1542
17. Goldman D, Bateman RM, Ellis CG: Effect of sepsis on skeletal mus-cle oxygen consumption and tissue oxygenation: Interpreting capil-lary oxygen transport data using a mathematical model. Am J Physiol
Heart Circ Physiol 2004; 287:H2535–H2544
18. Humer MF, Phang PT, Friesen BP, et al: Heterogeneity of gut capillary transit times and impaired gut oxygen extraction in endotoxemic pigs.
J Appl Physiol 1996; 81:895–904
19. Awan ZA, Häggblad E, Wester T, et al: Diffuse reflectance spec-troscopy: Systemic and microvascular oxygen saturation is linearly correlated and hypoxia leads to increased spatial heterogeneity of microvascular saturation. Microvasc Res 2011; 81:245–251 20. Knotzer H, Maier S, Dünser MW, et al: Arginine vasopressin does not
alter mucosal tissue oxygen tension and oxygen supply in an acute endotoxemic pig model. Intensive Care Med 2006; 32:170–174 21. Maekawa T, Fujii Y, Sadamitsu D, et al: Cerebral circulation and
metabolism in patients with septic encephalopathy. Am J Emerg Med 1991; 9:139–143
22. Thees C, Kaiser M, Scholz M, et al: Cerebral haemodynamics and carbon dioxide reactivity during sepsis syndrome. Crit Care 2007; 11:R123 23. Pfister D, Siegemund M, Dell-Kuster S, et al: Cerebral perfusion in sepsis-associated delirium. Crit Care 2008; 12:R63 24. Taccone FS, Castanares-Zapatero D, Peres-Bota D, et al: Cerebral autoregulation is influenced by carbon dioxide levels in patients with septic shock. Neurocrit Care 2010; 12:35–42 25. Zonta M, Angulo MC, Gobbo S, et al: Neuron-to-astrocyte signaling is central to the dynamic control of brain microcirculation. Nat Neurosci 2003; 6:43–50 26. Panerai RB: Assessment of cerebral pressure autoregulation in humans— A review of measurement methods. Physiol Meas 1998; 19:305–338 27. Maloney-Wilensky E, Gracias V, Itkin A, et al: Brain tissue oxygen and outcome after severe traumatic brain injury: A systematic review. Crit Care Med 2009; 37:2057–2063 28. Rosenthal G, Hemphill JC 3rd, Sorani M, et al: Brain tissue oxygen tension is more indicative of oxygen diffusion than oxygen delivery and metabolism in patients with traumatic brain injury. Crit Care Med 2008; 36:1917–1924
29. Vespa P, Prins M, Ronne-Engstrom E, et al: Increase in extracellu-lar glutamate caused by reduced cerebral perfusion pressure and seizures after human traumatic brain injury: A microdialysis study.
J Neurosurg 1998; 89:971–982
30. Sarrafzadeh A, Haux D, Küchler I, et al: Poor-grade aneurysmal sub-arachnoid hemorrhage: Relationship of cerebral metabolism to out-come. J Neurosurg 2004; 100:400–406 31. Nordström CH, Reinstrup P, Xu W, et al: Assessment of the lower limit for cerebral perfusion pressure in severe head injuries by bedside monitor-ing of regional energy metabolism. Anesthesiology 2003; 98:809–814 32. Vespa PM, O’Phelan K, McArthur D, et al: Pericontusional brain tissue exhibits persistent elevation of lactate/pyruvate ratio independent of cerebral perfusion pressure. Crit Care Med 2007; 35:1153–1160
33. Nielsen TH, Bindslev TT, Pedersen SM, et al: Cerebral energy metab-olism during induced mitochondrial dysfunction. Acta Anaesthesiol
Scand 2013; 57:229–235
34. Oddo M, Levine JM, Frangos S, et al: Brain lactate metabolism in humans with subarachnoid hemorrhage. Stroke 2012; 43:1418–1421 35. Frykholm P, Hillered L, Långström B, et al: Relationship between cere-bral blood flow and oxygen metabolism, and extracellular glucose and lactate concentrations during middle cerebral artery occlusion and reperfusion: A microdialysis and positron emission tomography study in nonhuman primates. J Neurosurg 2005; 102:1076–1084
36. Lewerenz J, Dargusch R, Maher P: Lactacidosis modulates glutathi-one metabolism and oxidative glutamate toxicity. J Neurochem 2010; 113:502–514
37. Levy B, Perez P, Gibot S, et al: Increased muscle-to-serum lactate gradient predicts progression towards septic shock in septic patients.
Intensive Care Med 2010; 36:1703–1709
38. Klaus S, Staubach KH, Heringlake M, et al: Tissue metabolism dur-ing endotoxin shock after pretreatment with monophosphoryl lipid A.
Cardiovasc Res 2003; 59:105–112
