INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

AFTER THE MAXIMUM hemorrhage volume is reached during severe experimentally controlled arterial hemorrhage, if the shed blood or other volume is not infused, there is a rapid decrease in the mean arterial pressure (MAP), resulting in life-threatening hypotension and cardiovascular collapse. This decompensatory phase is characterized by a decrease in the peripheral vascular resistance, insensitivity to -adrenergic agonists, and irreversible loss of vascular contractility (32, 36). Although not fully elucidated, the mechanisms for these vascular changes are primarily related to K⁺-channel activation (27, 33). This can be caused by increases in nitric oxide (32) or decreases in cytosolic ATP, especially in combination with acidosis (25, 27). Whereas experimental strategies to prevent or delay vascular decompensation have focused on the inhibition of nitric oxide synthase or K⁺-channel activation (27, 36), there has been little focus on minimizing cellular energetic deterioration during hemorrhagic shock. In vitro studies have shown that pyruvate, a metabolic substrate, provides remarkable protection of organ and cellular energetics and function during hypoxia, ischemia, or reperfusion (7, 11, 18, 21). Because hypotension, systemic acidosis, and cellular de-energization cause cardiovascular failure during hemorrhagic shock, we investigated the hypothesis that pyruvate administered during hemorrhagic shock could enhance organ tolerance to ischemia and delay cardiovascular collapse.

Pyruvate in millimolar concentrations can improve cellular function during metabolic stress by decreasing the cytoplasmic redox potential ([NADH_c]/[NAD⁺_c]) and maintaining the cellular phosphorylation potential ([ATP]/[ADP][P_i]). During hemorrhagic shock, critical decreases in oxygen delivery cause an increase in cytosolic redox ratio (lactate/pyruvate [NADH_c]/ [NAD⁺_c]), depletion of the ATP pool, a fall in the cytosolic phosphorylation potential, and cellular de-energization (36). Substrate-induced changes in the redox state have beneficial effects on both the contractile state and energetics of myocardial and vascular smooth muscle (VSM) (3, 22, 28). A pyruvate-induced decrease in [NADH_c]/[NAD⁺_c] favors an increase in the phosphorylation potential due to coupling of the cytoplasmic redox state to the phosphorylation potential ([ATP]/[ADP][P_i]) through the glyceraldehyde-3-phosphate dehydrogenase/phosphoglycerate kinase system (5, 28). A second positive effect of pyruvate on oxidative metabolism is improvement in mitochondrial energetics by increasing the driving force for ATP production, mitochondrial NADH (NADH_m). More important than the overall ATP content, however, is the phosphorylation potential. It is the thermodynamically relevant metabolite ratio of [ATP]/[ADP][P_i] that readily responds to physiological, metabolic, and pathological conditions (7, 13, 16, 28). For example, vital enzyme functions and ionic pump capabilities are stoichiometrically linked to the cytosolic phosphorylation potential (13, 22). Thus maintaining [ATP]/[ADP][P_i] within physiological limits is considered essential for cellular adaptation, integrity, and function, as this ratio has been shown to directly correlate with recovery from ischemic stress (4, 13). Indeed, the provision of exogenous pyruvate to both the heart and VSM has been shown to cause anaplerosis of the citric acid cycle, increase NADH_m, oxygen consumption, and the phosphorylation potential (2, 3, 7, 28). These beneficial effects are present even during hypoxia (7, 28). However, it is unknown whether the beneficial effects of pyruvate in in vitro systems are present in a whole animal situation.

In the present study, we examined whether intravenous pyruvate could improve vascular, cardiac, and neuronal function during severe hemorrhage and improve survival in swine. To assess metabolic stress during controlled hemorrhage, we measured global indexes of the cytosolic redox state [lactate/pyruvate ratio (L/P)] and indirectly monitored the cytosolic phosphorylation potential of the brain through extracellular (neocortical microdialysis) levels of adenosine and its degradation products, inosine and hypoxanthine. The extracellular accumulation of adenosine purine nucleosides is an indirect indicator of ATP catabolism and decrease in the phosphorylation potential (12, 15). Microdialysis was chosen because it is minimally invasive and is suitable for repetitive measurements. In the brain, the microdialysis adenosine indexes were also compared with cerebral oxidative metabolism (cerebral oxygen consumption), the net adenosine purine balance across the cortex (arterial-sagittal sinus), and the maintenance of electrophysiological function (electrocorticogram, ECOG).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

After Institutional Animal Care and Use Committee approval, 20 adolescent female Yorkshire swine (Sus scrofa) were randomized to receive either an intravenous infusion of 30% Na⁺ pyruvate (n = 6), 0.9% NaCl (n = 8, volume control), or 10% NaCl (n = 6, osmotic control) before and during controlled arterial hemorrhage. All animals were cared for according to the United States Department of Agriculture Animal Welfare Act and the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Animal Preparation and Monitoring

Anesthesia and hemodynamic monitoring procedures. The swine were fasted overnight, sedated with an intramuscular injection of ketamine (500 mg), and anesthetized with halothane by nose cone to facilitate tracheal intubation. For surgical preparation, anesthesia was maintained with halothane (1.0% end-tidal concentration) and ventilation controlled (Narkomed 2B, North American Drager, Telford PA) to maintain the end-tidal carbon dioxide tension at 35-40 mmHg. The inspired gas was a mixture of 25% oxygen and 75% nitrogen. Neuromuscular blocking drugs were not administered. Inspired and expired oxygen, carbon dioxide, and halothane concentrations were continuously monitored (Hewlett-Packard M1026A Gas Analyzer; Hewlett-Packard, Andover, MD). Leads II and V5 of the electrocardiogram were monitored. The right external jugular vein was isolated and an 8.5-Fr introducer sheath was inserted. A continuous thermodilution cardiac output pulmonary artery catheter was advanced through the sheath to measure pulmonary artery pressure and continuous cardiac output measurements (Abbott Critical Care, North Chicago, IL). Both femoral arteries and a femoral vein were isolated and cannulated with 8.5-Fr introducer sheaths. A micromanometer (MPC-500; Millar Instruments, Houston, TX) was inserted into the right femoral artery and advanced to the mid-thoracic aorta for the measurement of MAP. A second micromanometer (MPC-650) was inserted through the left femoral artery sheath and advanced into the left ventricle for measurement of left ventricular end-diastolic pressure (LVEDP) and calculation of the first derivative of LV pressure (dP/dt_max). The femoral vein was used for infusions. Physiological data were displayed on an eight-channel Hewlett Packard model 68 clinical monitor.

Cerebral blood flow, ECOG, and microdialysis instrumentation. After hemodynamic instrumentation, the swine were placed in ventral recumbent position and the cranium was stabilized. The scalp and galea were reflected, and a wide craniectomy was performed. A 6.0-mm ultrasound flow probe (6SB; Transonic Instruments, Ithaca, NY) was placed around the distal sagittal sinus for the measurement of cerebral blood flow (CBF) (14, 24). A 1.5-in. 22-gauge Teflon catheter was inserted into the sagittal sinus to obtain blood samples for cerebrovenous blood gases, purine nucleosides, and L/P measurements. Bilateral gold cup electrodes were placed on the surface of the dura (Fp_1,2 and C_3,4) for recording of a two-channel bipolar ECOG. The ECOGs were recorded (Aspect 1000, Natick, MA), filtered (1-30 Hz), digitized, and stored on hard disk for review and analysis. Two microdialysis probes (CMA-10; CMA/Microdialysis, Acton, MA) of concentric design (polycarbonate fiber length 4 mm, diameter 0.5 mm, 20,000 Dalton molecular cutoff) were inserted 6 mm into the right frontal neocortex. The location of insertion was 1 cm lateral to the sagittal sinus and 2 and 2.5 cm anterior to the coronal suture. After insertion, the probes were perfused with artificial cerebrospinal fluid (in mM, 155 Na⁺, 1.1 Ca²⁺, 0.83 Mg²⁺, 2.9 K⁺, pH 7.4) at 2 µl/min using a precision, multisyringe pump (BAS Bee; Bioanalytical Systems, West Lafayette, IN). A micromanometer (MPC-500) was inserted into the subdural space for measurement of intradural pressure. Brain temperature was monitored by an epidural thermistor and maintained at 38.5 ± 0.5°C with heating lamps and a forced-air warming device.

Experimental Protocols

Controlled hemorrhage protocol. One hour before the initiation of hemorrhage, the expired halothane concentration was reduced to 0.5% and the animals were allowed to ventilate spontaneously. Hemodynamic data were recorded every 30 min for the duration of the protocol. Controlled arterial hemorrhage was accomplished by removal of 45% of the estimated blood volume over 15 min (2.0 ml · kg¹ · min¹). After the first 15 min, blood was withdrawn as necessary to maintain the MAP at 40 mmHg. With the exception of the protocol infusions of Na⁺ pyruvate or NaCl, no fluids or shed blood was administered to support the MAP at 40 mmHg. After spontaneous vascular decompensation, death was defined as a MAP <10 mmHg and cessation of spontaneous respiratory effort.

Treatment groups and infusion protocols. The animals were assigned to one of three treatment groups. Group 1 had 30% Na⁺ pyruvate (pH titrated to 7.4 with 1 N NaOH) infused into the femoral vein 1 h before (infusion A, 1 g · kg¹ · h¹) and for 2 h (infusion B, 0.5 g · kg¹ · h¹) after the start of hemorrhage. To control for volume and osmolarity effects of the hypertonic Na⁺ pyruvate, an equivalent volume of 0.9% NaCl (group 2) or Na⁺ load (10% NaCl, group 3) was infused before (infusion A, 1 h) and after (infusion B, 2 h) the initiation of hemorrhage.

Analytic sampling procedures. Microdialysis samples were continuously collected over 30-min periods (60-µl aliquots) and stored at 80°C until HPLC analysis. One hour after insertion of the microdialysis probes, arterial, mixed venous, and sagittal sinus blood were sampled every 30 min for measurement of pH, base excess, blood gases, hemoglobin, and oxygen content (IL 1610 and IL 682 CoOx, Instrumentation Laboratories, Lexington, KY). Global oxygen consumption was calculated as [(arterial oxygen content mixed  venous oxygen content)(cardiac output)]/100. The cerebral metabolic rate for oxygen consumption was calculated as [(arterial oxygen content  sagittal sinus oxygen content)(CBF)]/100. Blood samples for serum electrolytes (Na⁺, K⁺, Cl, HCO₃), osmolality, lactate, and pyruvate were obtained every 30 min and immediately centrifuged. Blood (1.5 ml arterial, mixed venous, sagittal sinus) for purine analysis was collected with 1.5-ml ice-cold dipyridamole (400 µM) to minimize purine uptake and metabolism by cellular blood components. The dipyridamole-treated samples were immediately centrifuged and 0.5 ml of the supernatant was mixed with 0.5 ml 6 N HClO₄ on ice to denature and precipitate plasma proteins. Deproteination was followed by immediate neutralization with KHCO₃ to a slightly acidic pH of 6.0-7.0. After standing on ice for 30 min, the KClO₄ precipitate was removed by centrifugation. Samples were stored at 80°C until HPLC analysis.

Analytic Methods

HPLC sample preparation and analysis. Analysis of hypoxanthine (HX), inosine (Ino), and adenosine (ADO) in the dialysate and blood was performed using a multiwavelength reverse-phase HPLC technique. Neocortical microdialysis samples (30 µl) were injected directly into a C-18 column. Sample aliquots (10 µl) were injected into the HPLC workstation using a refrigerated automated injector system (Waters Wisp 712; Waters Associates, Milford, MA). Peak identification and quantification were accomplished by comparing known retention times with peak absorbance ratios at four different wavelengths (254, 263, 273, 293 nm) using a programmable multi-wavelength detector (model 490, Waters Associates). Chromatograms and absorbance data were recorded and analyzed using the Water's Maxima software (V 6.11, Waters Associates).

Microdialysis calibration procedures. In vitro relative recovery for purines for each microdialysis probe was determined in triplicate before and after each experiment. The probes were immersed in a purine calibration standard solution at 38.5°C and perfused with artificial cerebrospinal fluid. The calibration standards contained HX, Ino, and ADO dissolved in double-distilled deionized water at a final concentration of 10 µM for each compound. The HPLC peak heights of the nondialyzed standards were compared with the peak heights of the in vitro microdialysis samples to determine the relative recovery. The relative recovery for each compound was used to estimate the extracellular concentration of the nucleosides in the immediate vicinity of the probes.

Lactate and pyruvate analysis. After collection, the in vivo blood samples were centrifuged and deproteinated as described for the nucleoside analyses. Analysis of the in vitro perfusion medium used in the L/P titration protocols was performed directly on the perfusate samples. Lactate and pyruvate were determined spectrophotometrically with lactate dehydrogenase. Lactate and pyruvate standards were run in parallel with the blood extracts.

ECOG analysis. The ECOG was analyzed using a five-point visual scale previously reported (9). Each ECOG was reviewed by an investigator blinded to animal grouping status. The ECOG was assigned a numeric score: 5, unchanged from baseline; 4, mildly depressed (25%) amplitude and frequency compared with baseline; 3, moderate depression (50%) of amplitude and frequency from baseline; 2, markedly depressed (75%) amplitude and frequency from baseline; 1, burst suppression or an isoelectric ECOG. In addition, the median frequency and amplitude from the digitized ECOG were averaged for the last 5 min of each measurement period.

Data Presentation and Statistics

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The animal weights in the pyruvate, 0.9% NaCl (volume control), and 10% NaCl (osmotic control) groups were similar (44.9 ± 1.1, 45.2 ± 0.8, and 45.1 ± 1.1 kg, respectively). The intravenous pyruvate infusion raised the arterial pyruvate concentration from 0.12 ± 0.01 to 5.68 ± 0.91 mM 30 min after starting the infusion (Fig. 1) and this concentration was maintained during the infusion. After cessation of the pyruvate infusion, the concentration decreased from 6.25 ± 0.73 to 1.02 ± 0.14 mM within 30 min. In both the volume and osmotic control groups, the prehemorrhage pyruvate levels ranged from 0.08 to 0.18 mM and increased to 0.32 ± 0.03 and 0.33 ± 0.04 mM, respectively, during hemorrhage (Fig. 1).

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Arterial pyruvate levels (means ± SE) before and during controlled hemorrhage. Pyruvate was infused at 1 g · kg1 · h1 for 1 h before hemorrhage (infusion A) and 0.5 g · kg1 · h1 for 2 h (infusion B) after initiation of hemorrhage. To control for volume and osmolarity effects of hypertonic Na+ pyruvate solution, an equivalent volume of 0.9% NaCl (group 2) or an equivalent Na+ load (10% NaCl, group 3) was infused for 1 h before (infusion A) and 2 h after (infusion B) initiation of controlled hemorrhage. Pyruvate levels beyond 90 min of hemorrhage were not available for control groups due to death of animals.

Intravenous Pyruvate Effects on Hemodynamics, Oxygen Delivery, and Oxygen Consumption

During the hemorrhage phase, the CI was higher in the pyruvate group at H₃₀ (Table 1) with no differences between the groups at H₆₀ (Table 1). During the first hour of hemorrhage (H₃₀ and H₆₀), there were no differences between groups in the MAP. At H₉₀, the MAP in the volume and osmotic control groups was 14 and 20% lower, respectively, than in the pyruvate group, indicating the onset of vascular decompensation. At H₁₂₀ the control groups had no measurable MAP, whereas the pyruvate-treated animals still maintained a MAP >30 mmHg. After the pyruvate infusion was stopped at H₁₂₀, the pyruvate animals showed a rapid fall in MAP, CI, and dP/dt_max. The average time until cardiovascular decompensation and death in the pyruvate-treated group was 151.2 ± 10.0 min, whereas the time to death was 82.7 ± 5.5 and 74.8 ± 8.2 min in the volume and osmotic control groups, respectively (P < 0.001). The prolonged survival in the pyruvate group was not secondary to differences in volume status or protocol management because the peak hemorrhage volumes (44.9 ± 1.6, 41.8 ± 1.1, and 43.2 ± 1.5 ml/kg) and the times to reach the peak hemorrhage volume (59 ± 4, 55 ± 4, and 58 ± 5 min) were similar in the pyruvate, volume control, and osmotic control groups, respectively.

Table 2 shows the systemic and cerebral oxygen delivery and consumption data. Before hemorrhage (H₀), there was an increase in systemic oxygen delivery in the pyruvate and osmotic control groups, which was secondary to the augmented CI (Table 1, H₀). There were no differences between groups for the other parameters. At H₃₀ and H₆₀, systemic oxygen consumption was 40 and 30% higher in the pyruvate group compared with the control groups. In addition, despite the comparable decreases in the cerebral perfusion pressure (Table 1), the pyruvate group maintained the cerebral oxygen consumption 75% higher than the controls at H₃₀, H₆₀, and H₉₀.

Intravenous Pyruvate Effects On Blood Electrolytes, Acid-Base Status, Lactate, and L/P

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Means ± SE for pH (A), base excess (B), lactate (C), and lactate/pyruvate ratio (L/P; D) in control groups compared with pyruvate group before and during hemorrhage. In control groups, there was a decrease in pH and base excess with marked increase in L/P. In contrast, pyruvate infusion caused initial increase in pH and base excess that peaked at 30 min into hemorrhage period (A and B). Although pyruvate group had highest lactate levels over the course of the experiment (C), L/P remained <6 while pyruvate was infused during hemorrhage phase (D). After termination of pyruvate infusion, L/P rapidly climbed to a similar level as the highest levels in control animals. * P < 0.05 pyruvate group compared with control groups.

The initial lactate levels (1.2 ± 0.1 mM) and L/P ratios (7-15) were within the physiological range for all animals (Fig. 2, C and D). In the control groups, the lactate levels and the L/P ratios were unchanged before hemorrhage and increased progressively to near 12 mM and 40 during hemorrhage (H₉₀). In contrast, the pyruvate infusion before hemorrhage caused an increase in lactate to 8.6 ± 0.8 mM and a decrease in L/P to 1.0 ± 0.2. During hemorrhage the lactate concentration increased to 30-40 mM, whereas the L/P rose to only 5.6 (Fig. 2, C and D). After termination of the pyruvate infusion at H₁₂₀ the L/P rose to the peak levels (~40) observed in the volume and osmotic control groups. This suggests that while pyruvate substantially increased the lactate levels, it prevented the marked rise in L/P observed during hemorrhage in the control groups.

Data on hemoglobin concentration, osmolality, and serum electrolytes are shown in Table 3. The pyruvate and osmotic control group had small but significant decreases in the hemoglobin concentration before hemorrhage. During hemorrhage, all groups had significant decreases in the hemoglobin concentration with no differences between the groups. During the hemorrhage phase the volume control group (0.9% NaCl) had a progressive severe hyperkalemia (Table 3, H₉₀) and did not exhibit major changes in osmolality, Na⁺, or Cl concentrations. The osmotic control group (10% NaCl) also had a significant hyperkalemia in addition to increases in osmolality (20%), Na⁺ (+18%), and Cl (+40%). In the pyruvate group, hyperkalemia did not occur. However there were significant changes in the serum osmolality (+20%), Na⁺ (+25%), and Cl (10%).

Intravenous Pyruvate Effects On Neocortical Microdialysis and Blood Purines

View larger version (32K): [in this window] [in a new window]   Fig. 3.   A: means ± SE for changes in extracellular concentrations of adenosine (ADO) and its degradative products inosine (Ino) and hypoxanthine (HX) are shown from first collection period after insertion until last full collection period for all animals. Pyruvate- treated group had no increase in purine concentrations at 30 min and decrease rise at 60 min after start of hemorrhage phase. B: arterial-sagittal sinus blood differences in purines during hemorrhage phase (means ± SE). * P < 0.05 pyruvate group compared with control groups.

Intravenous Pyruvate Effects on ECOG

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Representative electrocorticogram (ECOG) epochs recorded from pyruvate and a 0.9% NaCl control animal at 30-min intervals during hemorrhage. Control animal had moderate loss of high-frequency signals at 30 min, severe slowing of frequency and loss of amplitude at 60 min, and essentially isoelectric ECOG at 90 min. In contrast, ECOG of pyruvate-treated animal shows only mild slowing of ECOG frequency at 90 min with progression to isoelectric tracing at 240 min.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Acute hemorrhage causes a shock state characterized by hypotension caused by a decreased intravascular volume and cardiac output that is compensated for by the release of endogenous vasoconstrictors and the recruitment of extravascular volume. Severe prolonged hemorrhage, however, initiates a sequence of events that ultimately results in refractory vascular failure, cardiovascular decompensation, and death. The specific aim of this investigation was to determine if the infusion of a metabolic substrate, pyruvate, could improve functional and metabolic status and delay the time until death during severe hemorrhagic shock in swine. We used an isobaric arterial hemorrhage model (40 mmHg) to induce the acute loss of approximately 66% of the estimated blood volume. With the exception of the experimental (30% pyruvate) and control (0.9% or 10% NaCl) infusions, cardiovascular decompensation was not treated and resulted in death. The shock state induced in the control animals was characterized by a substantial metabolic acidosis (pH 7.2, HCO₃ 17 meq/l), accumulation of lactate (12 mM), an increase in the L/P (40), hyperkalemia (7-8 meq/l), and death within 90 min. However, in animals treated with pyruvate, the time to cardiovascular collapse and rise in the specific indicators of cellular dysfunction and metabolic decompensation were delayed by 75 min. Whereas this highly instrumented swine model does not represent an uncontrolled hemorrhage scenario, it did allow for the delineation of the effects of pyruvate on metabolic and functional status during global and neocortical delivery dependent oxygen consumption. Furthermore, it was feasible to evaluate the neocortical energetic status qualitatively using the phosphorylation potential-related purine nucleoside parameters.

Cardiovascular Effects on Pyruvate During Controlled Arterial Hemorrhage

Another important effect of pyruvate in extending short-term survival was the obvious delay in the onset of vascular failure. Pyruvate produced a 30-min delay from peak hemorrhage volume until spontaneous decline in the MAP. Possible explanations for this pyruvate-induced maintenance of vascular performance are the absence of a metabolic acidosis, stabilization of the cytoplasmic redox potential (L/P), and the known beneficial effects of pyruvate on oxidative metabolism and the phosphorylation potential during metabolic stress. In VSM, decreases in pH and the cellular phosphorylation potential are known to cause K_ATP-channel activation with a decrease in intracellular Ca²⁺ concentration and contractility (1, 25). K⁺-channel activation has been shown to be a major component in the vascular insensitivity to -adrenergic agonists, during both hemorrhagic and endotoxic shock (27, 33).

Although direct evidence supporting differences in K⁺-channel activation or intracellular Ca²⁺ concentration caused by pyruvate is not provided by this investigation, the absence of hyperkalemia in the pyruvate group is consistent with reduced K⁺-channel activity. Mechanistically, pyruvate may have attenuated K⁺-channel activation by minimizing the development of acidosis and depletion of ATP in VSM. Whereas pyruvate, as much as lactate, causes intracellular acidosis when applied to isolated incubation systems (10), we observed systemic alkalinization in the whole animal and no acidosis during shock. The prevention of acidosis could be caused by the cytoplasmic reduction of pyruvate to lactate with a net decrease of two hydrogen ions. Additionally, pyruvate import into the mitochondria requires H⁺ cotransport and the hydrogen ion is removed when pyruvate is oxidized to carbon dioxide and water during respiration. In addition to alkalinization, pyruvate enhances the thermodynamic state of ATP, stimulates anaplerosis of the citric acid cycle, and increases mitochondrial NADH and oxygen consumption (2, 3). In VSM, glycolytic ATP production is the primary source of energy for maintenance of transsarcolemmal ion gradients (35). Nevertheless, pyruvate, a direct oxidative metabolic substrate, can also maintain normal ionic gradients, Ca²⁺ flux, and vascular contraction even in the absence of glycolysis (2, 3, 20). Thus the prevention of acidosis in conjunction with the stimulatory effects of pyruvate on oxidative metabolism and cellular energetics may have prevented or attenuated K⁺-channel activation and delayed the onset of a hypocontractile state.

A third possible beneficial effect of exogenous pyruvate on VSM function may be related to changes in the cytoplasmic redox potential as indicated by the arterial L/P. In contracting porcine carotid arteries, 10 mM lactate induces changes in the cytoplasmic redox state (L/P 68) that are associated with a 30% decrease in sustained KCl contraction compared with pyruvate (10 mM) (2). If this applies in vivo, then cytoplasmic redox changes induced by pyruvate (L/P < 6) during hemorrhagic shock, compared with controls (L/P >30), may have helped to maintain the contractile state of the vascular system.

Neocortical Effects of Pyruvate During Hemorrhagic Shock

Before hemorrhage in the current investigation, pyruvate administration had no measurable effects on CBF, the ECOG, or the microdialysis adenosine purine levels (H₀, Tables 2 and 4, Fig. 3A). However, during hemorrhagic shock the pyruvate-treated animals showed definite evidence of improved cerebral metabolic and electrophysiological function (Tables 2 and 4, Fig. 3). The enhanced neocortical metabolic function and stability were demonstrated by 1) delayed and reduced early accumulation of microdialysis purines (Fig 3A), 2) lack of net purine release into the sagittal sinus throughout the hemorrhage phase (Fig. 3B), and 3) maintenance of a normal ECOG for more than 60 min of hemorrhage (Table 4). In this study, pyruvate delayed (30 min) and markedly attenuated (60 min) the rise in the microdialysis purine accumulation. Although there were no differences in the purine levels between the pyruvate and control groups at 90 min, the ECOG was only moderately depressed compared with severe depression or complete absence of ECOG activity in the controls. Under the same conditions, pyruvate prevented the net release in purines into the sagittal sinus from the neocortex. These observations indicate that the local neuronal ATP in the immediate vicinity of the microdialysis probes was degraded and the cells de-energized at 90 min in all groups. However, the lack of net efflux into the sagittal sinus indicates that pyruvate preserved or stabilized the global neocortical ATP pool. In addition, the global improvement in the energy state and ischemic tolerance in the pyruvate group are consistent with the prolonged maintenance of the ECOG (Table 2 and Fig. 4) (30). In the pyruvate group, the improved energetic and functional parameters occurred in conjunction with a significantly higher CBF and cerebral oxygen consumption at similar perfusion pressures. Because CBF and cerebral oxygen consumption are coupled to cerebral metabolic demand, the pyruvate enhancement of the metabolic status appears to be responsible for these differences. This may represent improved CBF autoregulation at these low perfusion pressures or perhaps a redistribution of CBF and substrate within the brain. This interpretation is speculative because regional CBF was not measured.

In this study the measured indicators of cerebral ATP, energetics, and ECOG function provide evidence that pyruvate enhances cerebral ischemic tolerance during severe hemorrhagic shock. Previous studies have shown that millimolar pyruvate protects cerebral striatal cells from hydrogen peroxide toxicity and glutamate-mediated neurotoxicity in respiring hippocampal slices (11, 18). In vivo studies of partial or complete cerebral ischemia show that metabolic support through pyruvate dehydrogenase activation with dichloroacetate or acetylcarnitine normalizes postischemic levels of ATP and phosphocreatine, improves the clearance of intracellular acidosis, and substantially improves neurological outcomes (8, 26). However, pyruvate dehydrogenase activation alone does not prolong survival during hemorrhagic shock (31). Thus this is the first whole animal study to show that metabolic substrate administration enhances neuronal ischemic tolerance and prolongs survival during the ischemic conditions of hemorrhagic shock.

Limitations of Study

The administration of the Na⁺ salt of pyruvate in this study also caused biochemical and acid-base changes. These were the increase in serum lactate, the metabolic alkalosis, and the increases in osmolality and serum Na⁺. Normally, substantial lactate accumulations would be a major concern because, in conjunction with a high L/P and metabolic acidosis, it is a marker of cellular hypoxia and anaerobic glycolysis. However, in this investigation the high lactate level was driven by exogenous pyruvate and was associated with an L/P below basal levels and the prevention of a metabolic acidosis despite severe hemorrhagic shock (Fig. 2). The other biochemical changes were the increases in osmolality and Na⁺. These changes were secondary to the hypertonic 30% Na⁺ pyruvate formulation and substituting part of the Na⁺ content with other cations (Ca²⁺, choline, or ammonium) or a metabolizable pyruvate ester could minimize these effects.

Another limitation is that evidence in support of the metabolic mechanisms underlying the beneficial effects of pyruvate is indirect, because the actual phosphorylation potential, the cytosolic redox state, and the intracellular pH were not directly measured. However, other studies have established that an increase in extracellular purine nucleosides results from decreases in the cytosolic phosphorylation potential or energy status in the heart (6, 16). In addition, there are firmly established effects of millimolar pyruvate on the cytosolic redox status, cellular anaplerosis, improved oxygen utilization, and increases in the phosphorylation potential in intact myocardium (4, 7, 21, 28) and VSM (2).

In conclusion, the administration of intravenous pyruvate before and during hemorrhagic shock may have a major therapeutic potential for the delay, if not prevention, of global and cerebral metabolic failure, cardiovascular decompensation, and death. Our data show that these benefits are not related to volume or osmotic effects. Currently, widespread application of the principle of metabolic support during shock states is hampered by the incomplete understanding of the mechanism(s) by which pyruvate exerts its beneficial effects in the whole animal. Also, the minimum effective dose needs to be defined. Future work will determine if pyruvate will be useful only as a prophylactic agent or also as part of a resuscitative regimen.