Methods for detecting local intestinal ischemic anaerobic metabolic acidosis by PCO2

Monitoring the tissue perfusion during hemorrhagic shock and resuscitation: tissue-to-arterial carbon dioxide partial pressure gradient in a pig model

Background

Despite much evidence supporting the monitoring of the divergence of transcutaneous partial pressure of carbon dioxide (tcPCO₂) from arterial partial pressure carbon dioxide (artPCO₂) as an indicator of the shock status, data are limited on the relationships of the gradient between tcPCO₂ and artPCO₂ (tc-artPCO₂) with the systemic oxygen metabolism and hemodynamic parameters. Our study aimed to test the hypothesis that tc-artPCO₂ can detect inadequate tissue perfusion during hemorrhagic shock and resuscitation.

Methods

This prospective animal study was performed using female pigs at a university-based experimental laboratory. Progressive massive hemorrhagic shock was induced in mechanically ventilated pigs by stepwise blood withdrawal. All animals were then resuscitated by transfusing the stored blood in stages. A transcutaneous monitor was attached to their ears to measure tcPCO₂. A pulmonary artery catheter (PAC) and pulse index continuous cardiac output (PiCCO) were used to monitor cardiac output (CO) and several hemodynamic parameters. The relationships of tc-artPCO₂ with the study parameters and systemic oxygen delivery (DO₂) were analyzed.

Results

Hemorrhage and blood transfusion precisely impacted hemodynamic and laboratory data as expected. The tc-artPCO₂ level markedly increased as CO decreased. There were significant correlations of tc-artPCO₂ with DO₂ and COs (DO₂: r = − 0.83, CO by PAC: r = − 0.79; CO by PiCCO: r = − 0.74; all P < 0.0001). The critical level of oxygen delivery (DO_2crit) was 11.72 mL/kg/min according to transcutaneous partial pressure of oxygen (threshold of 30 mmHg). Receiver operating characteristic curve analyses revealed that the value of tc-artPCO₂ for discrimination of DO_2crit was highest with an area under the curve (AUC) of 0.94, followed by shock index (AUC = 0.78; P < 0.04 vs tc-artPCO₂), and lactate (AUC = 0.65; P < 0.001 vs tc-artPCO₂).

Conclusions

Our observations suggest the less-invasive tc-artPCO₂ monitoring can sensitively detect inadequate systemic oxygen supply during hemorrhagic shock. Further evaluations are required in different forms of shock in other large animal models and in humans to assess its usefulness, safety, and ability to predict outcomes in critical illnesses.

A Comparison of Gastric and Rectal CO₂ in Cardiac Surgery Patients

Critical care nurses assess and treat clinical conditions associated with inadequate oxygenation. Changes in regional organ (gut) blood flow are believed to occur in response to a decrease in oxygenation. Although the stomach is a widely accepted monitoring site, there are multiple methodological and measurement issues associated with the gastric environment that limit the accuracy of P CO2 detection. The rectum may provide nurses with an alternative site for monitoring changes in PCO2 without the limitations associated with gastric monitoring. This pilot study used a repeated measures design to examine changes in gastric and rectal PCO2 during elective coronary artery bypass grafting with cardiopulmonary bypass (CPB) and in the immediate 4-hr postoperative period in 26 subjects. The systemic indicators explained little variation in the regional indicators during protocol. A comparison of rectal and gastric PCO2 revealed no statistically significant differences in the direction or magnitude of change over any phase of cardiac surgery (baseline, CPB, post-CPB). A reduction in both rectal and gastric PCO2 occurred during CPB, and both values trended upward during the post-CPB phase. However, poor correlation and agreement was found between the measures of PCO2 at the two sites. Although clinically important, the cause is unclear. Possible explanations include variation in CO2 production between the gastric and rectal site, differences in sensitivity of the two monitoring instruments, or the absence of hemodynamic complications, which limited the extent of change in PCO2. Further investigation using patients with more profound changes in oxygenation are needed to identify response patterns and possible mechanisms.

Regional susceptibility to stress-induced intestinal injury in the mouse

Injury to the intestinal mucosa is a life-threatening problem in a variety of clinical disorders, including hemorrhagic shock, trauma, burn, pancreatitis, and heat stroke. The susceptibility to injury of different regions of intestine in these disorders is not well understood. We compared histological injury across the small intestine in two in vivo mouse models of injury, hemorrhagic shock (30% loss of blood volume) and heat stroke (peak core temperature 42.4°C). In both injury models, areas near the duodenum showed significantly greater mucosal injury and reductions in villus height. To determine if these effects were dependent on circulating factors, experiments were performed on isolated intestinal segments to test for permeability to 4-kDa FITC-dextran. The segments were exposed to hyperthermia (42°C for 90 min), moderate simulated ischemia (Po2 ∼30 Torr, Pco2 ∼60 Torr, pH 7.1), severe ischemia (Po2 ∼20 Torr, Pco2 ∼80 Torr, pH 6.9), or severe hypoxia (Po2 ∼0 Torr, Pco2 ∼35 Torr) for 90 min, and each group was compared with sham controls. All treatments resulted in marked elevations in permeability within segments near the duodenum. In severe hypoxia or hyperthermia, permeability was also moderately elevated in the jejunum and ileum; in moderate or severe ischemia, permeability was unaffected in these regions. The results demonstrate increased susceptibility of proximal regions of the small intestine to acute stress-induced damage, irrespective of circulating factors. The predominant injury in the duodenum may impact the pattern of acute inflammatory responses arising from breach of the intestinal barrier, and such knowledge may be useful for designing therapeutic strategies.

Tissue gas tensions and tissue metabolites for detection of organ hypoperfusion and ischemia

BACKGROUND

The aim of this study was to evaluate how tissue gas tensions and tissue metabolites measured in situ can detect hypoperfusion and differentiate between aerobic and anaerobic conditions during hemorrhagic shock. We hypothesized that tissue PCO(2) (PtCO(2)) would detect hypoperfusion also under aerobic conditions and detect anaerobic metabolism concomitantly with or earlier than other markers.

METHODS

Prospective experimental animal study with eight anesthetized pigs subjected to a continuous blood loss ∼8% of total blood volume per hour until death. We measured cardiac index, organ blood flows, and tissue levels of PO(2), PCO(2), glucose, pyruvate, lactate, and glycerol in intestine, liver, kidney, and skeletal muscle.

RESULTS

With reduction in blood flow to the organs under aerobic conditions, PtCO(2) increased ∼1-4 kPa from baseline. With the onset of tissue hypoxia there was a pronounced increase of PtCO(2), lactate, lactate-pyruvate (LP) ratio, and glycerol. Tissue pH and bicarbonate decreased significantly, indicating that metabolic acid was buffered by bicarbonate to generate CO(2).

CONCLUSION

Moderate tissue hypoperfusion under aerobic conditions is associated with increased PtCO(2), in contrast to metabolic parameters of ischemia (lactate, LP ratio, and glycerol) which remain low. From the onset of ischemia there is a much more rapid and pronounced increase in PtCO(2), lactate, and LP ratio. PtCO(2) can be used as a marker of hypoperfusion under both aerobic and anaerobic conditions; it gives an earlier warning of hypoperfusion than metabolic markers and increases concomitantly with or earlier than other markers at the onset of tissue anaerobiosis.

Small volume resuscitation with tempol is detrimental during uncontrolled hemorrhagic shock in rats

BACKGROUND

In a previous study, titration of a hypertonic saline (HTS) solution during severe uncontrolled hemorrhagic shock (UHS) failed to reduce mortality. In a separate study, a novel antioxidant, polynitroxylated albumin (PNA) plus tempol (4-hydroxy-2,2,6,6-tetramethylpiperidine-N-oxyl), infused during shock increased long-term survival. We hypothesized that combining potent antioxidants with a hypertonic solution during UHS would preserve the logistical advantage of small volume resuscitation and improve survival.

METHODS

An UHS outcome model in rats was used. UHS phase I (90 min) included blood withdrawal of 30 ml/kg over 15 min, followed by tail amputation for uncontrolled bleeding. At 20 min, rats were randomized to four groups (n=10 each) for hypotensive resuscitation from 20 to 90 min (mean arterial pressure [MAP] > or = 40 mmHg): HTS/starch group received 7.2% NaCl/10% hydroxyethyl starch; HTS/albumin group received 7.5% NaCl/20% albumin; HTS/PNA group received 7.5% NaCl/20% PNA; HTS/albumin+tempol group received 7.5% NaCl/20% albumin plus tempol. Resuscitation phase II (180 min) included hemostasis, return of shed blood and administration of fluids to restore MAP > or = 80 mmHg. Observation phase III was to 72 h.

RESULTS

The total amount of fluid required to maintain hypotensive MAP during HS was low and did not differ between groups (range: 3.4+/-1.9 to 5.3+/-2.5 ml/kg). The rate of fluid administration required was higher in the HTS/albumin+tempol group compared to all other groups (p=0.006). Additional uncontrolled blood loss was highest in the HTS/PNA group (16.2+/-5.7 ml/kg [p=0.01] versus 10.4+/-7.9 ml/kg in the HTS/starch group, 7.7+/-5.2 ml/kg in the HTS/albumin group and 8.2+/-7.1 ml/kg in the HTS/albumin+tempol group). MAP after start of resuscitation in phase I was lower in the HTS/albumin+tempol group than the HTS/albumin or HTS/PNA groups (p<0.01). This group was also less tachycardic. Long-term survival was low in all groups (2 of 10 after HTS/starch and 1 of 10 after HTS/albumin, 3 of 10 after HTS/PNA, 1 of 10 after HTS/albumin+tempol). Median survival time was shortest in the HTS/albumin+tempol group (72 min [CI 34-190]) compared to all other groups (p=0.01).

CONCLUSIONS

Despite its benefits in other model systems, free tempol is potentially hazardous when combined with hypertonic fluids. PNA abrogates these deleterious effects on acute mortality but may lead to increased blood loss in the setting of UHS.

Characterization of ischemic colitis associated with myocardial infarction: an analysis of 23 patients

PURPOSE

The study characterizes the clinical presentation of ischemic colitis (IC) associated with myocardial infarction (MI) and helps determine whether the primary mechanism for this association is thrombus, embolus, or localized nonocclusive mesenteric ischemia (NOMI) associated with systemic hypotension.

METHODS

We compared 23 study patients presenting with IC occurring simultaneously with or within 3 days after MI who were admitted to 5 medical centers versus (1) 32 patients with IC without MI (IC-controls) or (2) 32 patients with MI without IC (MI-controls).

RESULTS

Of 17,500 patients admitted to the study sites with MI, 23 (0.13%) had IC. Study patients had a high in-hospital mortality of 39%. An Acute Physiology and Chronic Health Evaluation (APACHE) II score greater than 15 was a significant predictor of mortality in these patients (P<.04). Compared with the IC-controls, study patients had a significantly lower mean arterial pressure (MAP) (76.0 +/- 17.1 mm Hg vs 98.3 +/- 18.6 mm Hg, P<.0001) and a significantly higher rate of hypotension (57% vs 9%, odds ratio [OR] = 12.6, confidence interval [CI]: 3.10-49.7, P<.001). The 2 groups, however, had a similar mean number of risk factors for thromboembolism per patient. Study patients had more severe illness than IC-controls, as demonstrated by mean APACHE II scores (19.0 +/- 5.5 vs 10.4 +/- 4.8, P<.0001). Study patients had a significantly higher incidence of complications, including respiratory failure (57% vs 13%, P=.001), altered mental status (48% vs 13%, P<.01), and renal insufficiency or failure (61% vs 28%, P<.04). Study patients had a significantly lower minimum hematocrit. Study patients had a significantly higher rate of prolonged hospitalization (>30 days) or in-hospital death (74% vs 19%, OR = 12.3, CI: 3.47-43.5, P<.0001). Compared with MI-control patients, study patients had a significantly lower MAP, significantly higher rate of hypotension, much higher mean APACHE II score, much higher incidence of complications, and significantly worse hospital outcome.

CONCLUSIONS

Patients with both IC and MI present as a clinically distinct group from patients with either IC alone or MI alone. They have significantly more complications and worse in-hospital prognoses. They present with a dramatically lower MAP and a higher frequency of hypotension. This last finding suggests that the most common and most important mechanism for IC with MI may be hypotension from cardiogenic shock. Hypotension is the cardinal risk factor for generalized NOMI with acute mesenteric ischemia and may be an important risk factor for localized NOMI with IC. An APACHE II score greater than 15 may be a predictor of mortality from IC after MI.

Titrated hypertonic/hyperoncotic solution for hypotensive fluid resuscitation during uncontrolled hemorrhagic shock in rats

BACKGROUND

In volume- or pressure-controlled hemorrhagic shock (HS) a bolus intravenous infusion of hypertonic/hyperoncotic solution (HHS) proved beneficial compared to isotonic crystalloid solutions. During uncontrolled HS in animals, however, HHS by bolus increased blood pressure unpredictably, and increased blood loss and mortality. We hypothesized that a titrated i.v. infusion of HHS, compared to titrated lactated Ringer's solution (LR), for hypotensive fluid resuscitation during uncontrolled HS reduces fluid requirement, does not increase blood loss, and improves survival.

METHODS

We used our three-phased uncontrolled HS outcome model in rats. HS phase I began with blood withdrawal of 3 ml/100g over 15 min, followed by tail amputation. Then, hydroxyethyl starch 10% in NaCl 7.2% was given i.v. to the HHS group (n=10) and LR to the control group (n=10), both titrated to prevent mean arterial pressure (MAP) from falling below 40 mmHg during HS time 20-90 min. At HS 90 min, resuscitation phase II of 180 min began with hemostasis, return of all the blood initially shed, plus fluids i.v. as needed to maintain normotension (MAP>or=70 mmHg). Liver dysoxia was monitored as increase in liver surface pCO2 during phases I and II. Observation phase III was to 72 h.

RESULTS

During HS, preventing a decrease in MAP below 40 mmHg required HHS 4.9+/-0.6 ml/kg (all data mean+/-S.E.M.), compared to LR 62.2+/-16.6 ml/kg (P<0.001), with no group difference in MAP. Uncontrolled blood loss during HS from the tail stump was 13.3+/-1.9 ml/kg with HHS infusion, versus 12.6+/-2.5 ml/kg with LR infusion (P=0.73). Serum sodium concentrations were moderately elevated at the end of HS in the HHS group (149+/-3 mmol/l) versus the LR group (139+/-1 mmol/l) (P=0.001), and remained elevated throughout. Liver pCO2 increased during HS in both groups equally (P<0.001 versus baseline), and tended to return to baseline levels at the end of HS. Blood gas and lactate values throughout did not differ between groups. During HS, 2 of 10 rats in the HHS group versus 0 of 10 in the LR group died (P=0.47). There was no difference between HHS and LR groups in survival rates to 72 h (3 of 10 in the HHS group versus 2 of 10 in the LR group) (P=1.0). Survival times, by life table analysis, were not different (P=0.75).

CONCLUSION

In prolonged uncontrolled HS, a titrated i.v. infusion of HHS can maintain controlled hypotension with only one-tenth of the volume of LR required, without increasing blood loss. This titrated HHS strategy may not increase the chance of long-term survival.

Lingual, splanchnic, and systemic hemodynamic and carbon dioxide tension changes during endotoxic shock and resuscitation

Sublingual and intestinal mucosal blood flow and Pco2 were studied in a canine model of endotoxin-induced circulatory shock and resuscitation. Sublingual Pco2 (PsCO2) was measured by using a novel fluorescent optrode-based technique and compared with lingual measurements obtained by using a Stowe-Severinghaus electrode [lingual Pco2 (PlCO2)]. Endotoxin caused parallel changes in cardiac output, and in portal, intestinal mucosal, and sublingual blood flow (Q̇s). Different blood flow patterns were observed during resuscitation: intestinal mucosal blood flow returned to near baseline levels postfluid resuscitation and decreased by 21% after vasopressor resuscitation, whereas Q̇s rose to twice that of the preshock level and was maintained throughout the resuscitation period. Electrochemical and fluorescent Pco2 measurements showed similar changes throughout the experiments. The shock-induced increases in PsCO2 and PlCO2 were nearly reversed after fluid resuscitation, despite persistent systemic arterial hypotension. Vasopressor administration induced a rebound of PsCO2 and PlCO2 to shock levels, despite higher cardiac output and Q̇s, possibly due to blood flow redistribution and shunting. Changes in PlCO2 and PsCO2 paralleled gastric and intestinal Pco2 changes during shock but not during resuscitation. We found that the lingual, splanchnic, and systemic circulations follow a similar pattern of blood flow variations in response to endotoxin shock, although discrepancies were observed during resuscitation. Restoration of systemic, splanchnic, and lingual perfusion can be accompanied by persistent tissue hypercarbia, mainly lingual and intestinal, more so when a vasopressor agent is used to normalize systemic hemodynamic variables.

Gut dysoxia: comparison of sites to detect regional gut dysoxia

Dysoxia, a state in which O2 supply is inadequate to meet tissue metabolic needs, is often first detected in regional organs such as the gut. An increase in PCO2 is believed to reflect the development of gut dysoxia. The stomach is a well-documented clinical site for detecting gut PCO2; however, measurement issues make this a less than ideal monitoring site. Other sites along the GI tract may be equally sensitive to detect changes in PCO2. Rectal CO2 measurement may have the advantage of being less invasive, low risk, and continuous without the limitations associated with gastric monitoring. In this study, we compared PCO2 at two sites (gastric, rectum) at baseline and during a dysoxic challenge, cardiac arrest. We obtained similar values of PCO2 at both sites. Ten male Wistar rats were anesthetized with 1%-2% Isoflurane/50% nitrous oxide/balanced O2 and the tail artery and right atrium were cannulated. Severinghaus-type active tip PCO2 electrodes (Microelectrode Inc, Bedford, NH) were calibrated and one electrode was surgically inserted into the stomach (G-PCO2) and a second electrode was placed in the rectum (R-PCO2). Animals were stabilized following surgery. Cardiac arrest was induced by administering a rapid injection of norcuron (0.1-0.2 mg/kg) and potassium chloride solution (0.5 M/L; 0.12 mL/100 gm of body weight). Animals were monitored for 15 minutes post-arrest. Data were collected at one minute intervals using the software Data Collect. All data are reported as mean +/- SD. Baseline G-PCO2 was 64 +/- 17 torr, not significantly different from R-PCO2, 58 +/- 7 torr. After 15 minutes of cardiac arrest, G-PCO2 rose to 114 +/- 42 torr, again not significantly different from R-PCO2, which reached 112 +/- 35 torr. Monitoring PCO2 in the rectum is less invasive and might provide similar information when compared with gastric monitoring at baseline and during a dysoxic challenge.

Tissue capnometry: does the answer lie under the tongue?

Increases in tissue partial pressure of carbon dioxide (PCO(2)) can reflect an abnormal oxygen supply to the cells, so that monitoring tissue PCO(2) may help identify circulatory abnormalities and guide their correction. Gastric tonometry aims at monitoring regional PCO(2) in the stomach, an easily accessible organ that becomes ischemic quite early when the circulatory status is jeopardized. Despite substantial initial enthusiasm, this technique has never been widely implemented due to various technical problems and artifacts during measurement. Experimental studies have suggested that sublingual PCO(2 )(P(sl)CO(2)) is a reliable marker of tissue perfusion. Clinical studies have demonstrated that high P(sl)CO(2) values and, especially, high gradients between P(sl)CO(2) and arterial PCO(2) (DeltaP(sl-a)CO(2)) are associated with impaired microcirculatory blood flow and a worse prognosis in critically ill patients. Although some questions remain to be answered about sublingual capnometry and its utility, this technique could offer new hope for tissue PCO(2) monitoring in clinical practice.

Detection of ischemia by PCO2 before adenosine triphosphate declines in skeletal muscle

OBJECTIVE

Ischemia is a serious problem in clinical medicine, and effective methods are needed to detect ischemia before the injury becomes irreversible. In experimental studies on several organs, PCO2 was found to increase rapidly after the onset of supply-dependent anaerobic metabolism. A shortcoming of these studies was that PCO2 was not correlated with tissue concentrations of lactate and the energy status in the cell. Thus, in this study we have measured tissue concentrations of lactate, phosphocreatine, and adenosine triphosphate. We hypothesized that during ischemic conditions, PCO2 reflects lactate generation in the cell and not exhausted energy stores per se. If this is the case, PCO2 can be used to detect ischemia before the energy stores are depleted. Consequently, therapy can be instituted at a time when the organ is salvageable.

DESIGN

Prospective laboratory study.

SETTING

University research laboratory.

SUBJECTS

Seven pigs.

INTERVENTIONS

In a porcine model, gluteal skeletal muscles with no-flow ischemia were examined. PCO2 was measured both in situ and in vitro at increasing periods of time. Concomitantly, tissue lactate, adenosine triphosphate, and phosphocreatine were analyzed.

MEASUREMENTS AND MAIN RESULTS

Tissue surface CO2 tension (PtCO2) increased rapidly after onset of ischemia. From a baseline of 63 +/- 3 torr (8.4 +/- 1.2 kPa) under aerobic conditions, it increased to 157 +/- 6 torr (21 +/- 2.2 kPa) after 30 mins of ischemia and 386 +/- 9 torr (51.5 +/- 3 kPa) at 120 mins. The rapid increase of PtCO2 correlated well with increasing values of lactate (r2 >.9) in the tissue. Adenosine triphosphate was essentially unchanged for 45 mins after onset of ischemia, after which it declined. Phosphocreatine decreased earlier than adenosine triphosphate in accordance with the notion that high-energy phosphate groups are transferred from phosphocreatine to adenosine triphosphate.

CONCLUSION

In this porcine model of skeletal muscle ischemia, PtCO2 correlates well with tissue lactate and increases long before the energy stores of phosphocreatine and most notably adenosine triphosphate are severely reduced. Thus, PtCO2 could be monitored to detect and treat earlier stages of ischemia.

Tissue Carbon Dioxide Tension:

Free flap surgical procedures are technically challenging, and anastomosis failure may lead to arterial or venous occlusion and flap necrosis. To improve myocutaneous flap survival rates, more reliable methods to detect ischemia are needed. On the basis of theoretical considerations, carbon dioxide tension, reflecting intracellular acidosis, may be suitable indicators of early ischemia. It was hypothesized that tissue carbon dioxide tension increased rapidly when metabolism became anaerobic and would be correlated with acute venoarterial differences in lactate levels, potassium levels, and acid-base parameters. Because metabolic disturbances have been observed to be less pronounced in flaps with venous occlusion, it was hypothesized that tissue carbon dioxide tension and venoarterial differences in lactate and potassium levels and acid-base parameters would increase less during venous occlusion than during arterial occlusion. In 14 pigs, latissimus dorsi myocutaneous flaps were surgically isolated, exposed to acute ischemia for 150 minutes with complete arterial occlusion (seven subjects) or venous occlusion (seven subjects), and reperfused for 30 minutes. After arterial occlusion, pedicle blood flow decreased immediately to less than 10 percent of baseline flow. Blood flow decreased more slowly after venous occlusion but within 3 minutes reached almost the same low levels as observed during arterial occlusion. Venous oxygen saturation decreased from approximately 70 percent to approximately 20 percent, whereas oxygen uptake was almost arrested. Tissue carbon dioxide tension increased to two times baseline values in both groups (p < 0.01). The venoarterial differences in carbon dioxide tension, pH, base excess, glucose levels, lactate levels, and potassium levels increased significantly (p < 0.01). Tissue carbon dioxide tension measured during the occlusion period were closely correlated with venoarterial differences in pH, base excess, glucose levels, lactate levels, and potassium levels (median r2, 0.67 to 0.92). After termination of arterial or venous occlusion, more pronounced hyperemia was observed in the arterial occlusion group than in the venous occlusion group (p < 0.05). Oxygen uptake (p < 0.05) and venoarterial differences in lactate and potassium levels (p < 0.05) were significantly more pronounced in the arterial occlusion group. In the venous occlusion group, with less pronounced hyperemia, venoarterial differences in acid-base parameters remained significantly different from baseline values before occlusion (p < 0.01). The data indicate that tissue carbon dioxide tension can be used to detect anaerobic metabolism, caused by arterial or venous occlusion, in myocutaneous flaps. The correlations between carbon dioxide tension and venoarterial differences in acid-base parameters were excellent. Because carbon dioxide tension can be measured continuously in real time, such measurements are more likely to represent a clinically useful parameter than are venoarterial differences.

Gastric mucosal acidosis and cytokine release in patients with septic shock

OBJECTIVE

It has been postulated that in critically ill patients, splanchnic hypoperfusion may lead to cytokine release into the systemic circulation. The presence of cytokines could trigger an inflammatory response and cause multiple organ dysfunction syndrome. Although experimental studies support this hypothesis, humans studies remain controversial. The aim of the study was to determine the relationship between splanchnic hypoperfusion and cytokine release during septic shock.

DESIGN

Human prospective study.

SETTING

Medical intensive care unit at a university hospital.

PATIENTS

A total of 30 patients with mean arterial pressure of <60 mm Hg after volume loading with either oliguria or hyperlactatemia.

MEASUREMENTS

Gastric intramucosal measurements as an indicator of splanchnic hypoperfusion and blood samples were obtained at admission to the medical intensive care unit and repeated during 48 hrs. Cytokine (tumor necrosis factor-alpha and interleukin-6) values were evaluated by enzyme-linked immunoassays at the following periods: at the time of admission and 2, 4, 8, 12, 24, 36, and 48 hrs later.

MAIN RESULTS

High levels of interleukin-6 and tumor necrosis factor-alpha were observed at admission in survivors and nonsurvivors, without significant difference. At 48 hrs, cytokine levels were significantly higher in patients who died compared with the survivors (tumor necrosis factor: 163 +/- 16 for nonsurvivors vs. 34 +/- 9 ng/mL for survivors; interleukin-6: 2814 +/- 485 for nonsurvivors vs. 469 +/- 107 ng/mL for survivors). At 48 hrs, the PCO2 gap was significantly higher in the nonsurvivors compared with survivors (25.87 +/- 2.73 vs. 11.35 +/- 2.25 mm Hg), despite systemic hemodynamic variables in the normal range. A positive relationship was demonstrated between plasma levels of tumor necrosis factor-alpha and interleukin-6 and the PCO2 gap throughout the study. The PCO2 gap was not correlated with hemodynamic variables.

CONCLUSIONS

Our data suggest a relationship between gastric mucosal acidosis, as assessed by PCO2 gap, and cytokine levels in critically ill patients with septic shock. Gut injury may be a contributor of the inflammatory response in patients with septic shock.

Detection of organ ischemia during hemorrhagic shock

BACKGROUND

In a porcine hemorrhagic shock model we aimed to determine: (a) whether blood flow to the intestine and kidney was more reduced than cardiac output; (b) whether parameters of anaerobic metabolism correlated with regional blood flow; and (c) whether metabolic parameters in intestine, kidney and skeletal muscles detected a compromised metabolic state at an earlier stage than did systemic parameters.

METHODS

In an animal research laboratory at a university hospital six domestic pigs were subjected to volume-controlled hemorrhage. Every 30 min samples of blood were withdrawn. Systemic and regional hemodynamic parameters and tissue levels of PCO2 were monitored. Whole body and organ-specific oxygen consumption (VO2) and veno-arterial (VA) differences of lactate, glucose, potassium (K+), PCO2, H+ and base excess (BE) were calculated every 30 min.

RESULTS

With progressive hemorrhage, intestinal blood flow decreased to the same extent as cardiac output, whereas the reduction in renal blood flow was more pronounced. We found a concomitant reduction in VO2 (onset of supply dependent metabolism) in intestine, kidney and skeletal muscles. In muscular tissue PCO2 increased to levels three times higher than baseline, while renal and intestinal PCO2 increased eightfold. Supply dependency was associated with a concomitant increase in VA CO2 and VA H+. Also, VA lactate increased, mostly in intestine and least in skeletal muscle. Intestinal and renal VA K+ increased, while muscular VA K+ decreased. Arterial lactate and H+ increased considerably, whereas arterial BE decreased.

CONCLUSION

With progressive hemorrhage, renal blood flow, but not intestinal and skeletal muscle blood flow, was reduced more than cardiac output. Supply dependent oxygen metabolism (VO2) and organ acidosis occurred simultaneously in the three organs, despite differences in blood flow reductions. Organ ischemia coincided with a pronounced change in arterial lactate and systemic acid base parameters.

Early antioxidant therapy with Tempol during hemorrhagic shock increases survival in rats

BACKGROUND

Hemorrhagic shock (HS) is associated with the generation of reactive oxygen species, which may contribute to delayed multiple organ system failure and death. Previous studies have shown that the antioxidant Tempol improved physiologic variables, although not necessarily outcome, in septic shock and HS. We hypothesized that the combination of free Tempol with polynitroxylated albumin (PNA)-bound Tempol (which prolongs half-life and decreases toxicity) improves outcome after HS in rats.

METHODS

In study 1, HS was induced by blood withdrawal of 3 mL/100 g over 15 minutes. Mean arterial pressure was maintained at 40 mm Hg with either infusion of normal saline or withdrawal of blood from 20 to 90 minutes. Resuscitation (90-270 minutes) was with infusion of shed blood. Observation was to 72 hours. At HS 45 min, albumin (ALB) (n = 10) or PNA + Tempol (n = 10) was infused slowly (1 mL/100 g/h) until 120 minutes. Study 2 was the same as study 1 (n = 6 per group), but terminated at 150 minutes. Study 3 was the same as study 1, but started with ALB or PNA + Tempol (n = 7 per group) at 20 minutes. The primary outcome variable in studies 1 and 3 was survival, whereas the primary outcome variables in study 2 were antioxidant reserve (ability of the serum or tissue homogenate to scavenge peroxyl radicals produced by 2,2'-azobis [2-aminodipropane]-dihydrochloride) in serum and small intestine, and low-molecular-weight thiols in tissues (liver, small intestine, and kidney).

RESULTS

In study 1, 72-hour survival was 1 of 10 (ALB group) versus 2 of 10 (PNA + Tempol group). At 90 minutes, pH was lower in the ALB group versus the PNA + Tempol group (p = 0.02) and remained low. Arterial lactate increased to 8.9 +/- 3.2 (mean +/- SD) versus 6.5 +/- 1.8 mmol/L (p = 0.04) and base excess was -9.6 +/- 4.3 versus -5.2 +/- 3.2 mmol/L (p = 0.01) (ALB vs. PNA + Tempol groups, respectively). In study 2, antioxidant reserve in serum was lower in the ALB group versus the PNA + Tempol group (p = 0.002). There were no differences between groups in antioxidant reserve in the small intestine or low-molecular-weight thiols in liver, kidney, and small intestine. In study 3, 72-hour survival was zero of seven (ALB group) versus five of seven (PNA + Tempol group) (p = 0.02). Heart rate and systolic blood pressure during late HS were higher in the ALB group in studies 1 and 3 (p < 0.05).

CONCLUSION

When infused early in HS, PNA + Tempol can increase survival. When given late, it significantly improves acid-base and serum antioxidant status, without an effect on survival. Additional studies will be required to determine whether early resuscitation with PNA + Tempol attenuates reactive oxygen species-mediated injury as the mechanism for preventing the progression toward multiple organ failure and death after HS. The results suggest that antioxidant therapy with Tempol deserves further study as a potential adjunct in the initial resuscitation from HS.

Effects of mild hypothermia on survival and serum cytokines in uncontrolled hemorrhagic shock in rats

Previous studies have suggested benefit of mild hypothermia during hemorrhagic shock (HS). This finding needs additional confirmation and investigation into possible mechanisms. Proinflammatory cytokines are mediators of multiple organ failure following traumatic hemorrhagic shock and resuscitation. We hypothesized that mild hypothermia would improve survival from HS and may affect the pro- and anti-inflammatory cytokine response in a rat model of uncontrolled HS. Under light halothane anesthesia, uncontrolled HS was induced by blood withdrawal of 3 mL/100 g over 15 min followed by tail amputation. Hypotensive (limited) fluid resuscitation (to prevent mean arterial pressure [MAP] from decreasing below 40 mmHg) with blood was started at 30 min and continued to 90 min. After hemostasis and resuscitation with initially shed blood and Ringer's solution, the rats were observed for 72 h. The animals were randomized into two HS groups (n = 10 each): normothermia (38 degrees C +/- 0.5 degrees C) and mild hypothermia (34 degrees C +/- 0.5 degrees C) from HS 30 min until resuscitation time (RT) 60 min; and a sham group (n = 3). Venous blood samples were taken at baseline, RT 60 min, and days 1, 2, and 3. Serum interleukin (IL)-1beta, IL-6, IL-10, and tumor necrosis factor (TNF)-alpha concentrations were quantified by ELISA. Values are expressed as median and interquartile range. Survival time by life table analysis was greater in the hypothermia group (P = 0.04). Survival rates to 72 h were 1 of 10 vs. 6 of 10 in the normothermia vs. hypothermia groups, respectively (P = 0.057). All cytokine concentrations were significantly increased from baseline at RT 60 min in both HS groups, but not in the shams. At RT 60 min, in the normothermia vs. hypothermia groups, respectively, IL-1beta levels were 185 (119-252) vs. 96 (57-135) pg/mL (P = 0.15); IL-6 levels were 2242 (1903-3777) vs. 1746 (585-2480) pg/mL (P = 0.20); TNF-alpha levels were 97 (81-156) vs. 394 (280-406) pg/mL (P= 0.02); and IL-10 levels were 1.7 (0-13.3) vs. 15.8 (1.9-23.0) pg/mL (P = 0.09). IL-10 remained increased until day 3 in the hypothermia group. High IL-1beta levels (>100 pg/mL) at RT 60 min were associated with death before 72 h (odds ratio 66, C.I. 3.5-1255). We conclude that mild hypothermia improves survival time after uncontrolled HS. Uncontrolled HS induces a robust proinflammatory cytokine response. The unexpected increase in TNF-alpha with hypothermia deserves further investigation.

Small intestine intramucosal PCO(2) and microvascular blood flow during hypoxic and ischemic hypoxia

OBJECTIVE

To determine whether small intestine intramucosal PCO(2) and mucosal blood flow changes would be different between ischemic and hypoxic hypoxia.

DESIGN

Randomized animal experiment.

SETTING

Research laboratory.

SUBJECTS

Anesthetized, mechanically ventilated, and surgically instrumented pigs.

INTERVENTIONS

Systemic oxygen delivery was lowered in a stepwise manner to decrease it beyond critical oxygen delivery by lowering either FIO(2) or blood volume.

MEASUREMENTS AND MAIN RESULTS

In hypoxic hypoxia pigs (n = 6), arterial oxygen concentration and oxygen delivery decreases were achieved by progressively reducing arterial PO(2) while cardiac index remained unchanged. In ischemic hypoxia pigs (n = 5), oxygen delivery reduction was achieved by progressively reducing cardiac index while arterial PO(2) remained unchanged. In control pigs, oxygen delivery remained unchanged. The lowest oxygen delivery measured in both hypoxia and ischemia experiments was 3.60 +/- 0.26 vs. 2.93 +/- 0.77 mL x kg(-1) x min(-1), respectively (p =.23). At the lowest oxygen delivery level, differences between ischemic hypoxia and hypoxic hypoxia experiments were observed for arterial lactate concentration (468 +/- 308 vs. 1070 +/- 218 mmol/L, respectively; p =.03), mixed venous arterial PCO(2) difference (10 +/- 7 vs. 4 +/- 2 torr, respectively; p =.04), and small intestine mucosal blood flow (6.2 +/- 2.1 vs. 15.7 +/- 7.4 perfusion units, respectively; p =.02). Small intestine intramucosal-arterial difference was higher in ischemic hypoxia than in hypoxic hypoxia (52 +/- 15 vs. 31 +/- 12 torr, respectively; p =.03).

CONCLUSION

Small intestine intramucosal PCO(2) increases may indicate systemic oxygen uptake supply limitation in ischemic and hypoxic hypoxia related to conditions of mucosal flow stagnation and CO(2) generation.

PCO(2) in the large intestine of mice, rats, guinea pigs, and dogs and effects of the dietary substrate

PCO(2) in the lumen and serosa of cecum and colon was measured in rats, guinea pigs, and dogs to examine the relationship between serosal PCO(2) and the incidence of intestinal necrotic lesions after administration of gas-carrier contrast agents in rodents. The effects of the dietary substrate were tested in a group of mice maintained on a diet based on glucose as the only carbohydrate source. The anesthetic used was a fentanyl-fluanison-midazolam mixture (rodents) and pentobarbital (dogs). PCO(2) was measured in vivo and postmortem, and the kinetics of the postmortem serosal PCO(2) [transmural CO(2) flux (J(CO(2)))] was calculated. PCO(2) in the cecal serosa and lumen, respectively, was 64 +/- 4 and 392 +/- 18 Torr in rats, 67 +/- 3 and 276 +/- 17 Torr in guinea pigs, and 73 +/- 6 and 137 +/- 7 Torr in mice on glucose-based diet. In the colon serosa and lumen of dogs, PCO(2) was 30 +/- 6 and 523 +/- 67 Torr, respectively. Serosal PCO(2) increased rapidly after death in rats and slower in guinea pigs and mice, and the slowest change was observed in dogs. Compared with dogs, serosal PCO(2) and J(CO(2)) of rats and guinea pigs were significantly higher. Serosal PCO(2) of guinea pigs was similar to that of rats, whereas the J(CO(2)) of guinea pigs was significantly lower. These data suggest a causal relationship between the ability of the cecal and colonic wall to act as a barrier to CO(2) diffusion and the presence of characteristic gas-carrier contrast agent-induced intestinal lesions in mice and rats and their absence in guinea pigs, dogs, and other species.

Intraperitoneal, but not enteric, adenosine administration improves survival after volume-controlled hemorrhagic shock in rats

OBJECTIVE

To circumvent the potential adverse systemic side effects of adenosine, this study explored the potential benefit of intraperitoneal or enteric adenosine on survival and inflammatory responses after volume-controlled hemorrhagic shock.

DESIGN

Prospective, randomized, and blinded. A three-phase, volume-controlled hemorrhagic shock model was used: hemorrhagic shock phase (120 mins), resuscitation phase (60 mins), and observation phase (72 hrs). Three groups were compared: controls, intraperitoneal adenosine, and enteric adenosine.

SETTING

Animal research facility.

SUBJECTS

Male Sprague-Dawley rats.

INTERVENTIONS

Starting at 20 mins of hemorrhagic shock and continuing through the resuscitation phase, all three groups received both intraperitoneal lavage and repeated bolus injections into the ileum of vehicle (normal saline) or adenosine. In the intraperitoneal adenosine group (n = 10), adenosine solution (0.1 mM) was used for intraperitoneal lavage. In the enteric adenosine group (n = 10), adenosine (1.0 mM) was injected into the ileum. Blood cytokine concentrations and leukocyte infiltration in lungs and liver were studied in 12 separate rats (control and intraperitoneal adenosine, n = 6 each) with the same hemorrhagic shock model at resuscitation time 1 hr or 4 hrs.

MEASUREMENTS AND MAIN RESULTS

Mean arterial pressure and heart rate were similar between the three groups during hemorrhagic shock and resuscitation. Potassium, lactate, and blood urea nitrogen concentrations were lower and arterial pH was higher in the intraperitoneal and enteric adenosine groups compared with the control group (both p <.05). Survival time to 72 hrs was longer in the intraperitoneal adenosine group than in the control group(p <.05). Neither plasma interleukin-1beta, interleukin-6, interleukin-10, and tumor necrosis factor-alpha concentrations nor leukocyte infiltration in the lungs and liver was different between the control and intraperitoneal adenosine groups.

CONCLUSIONS

The administration of adenosine via the intraperitoneal route improves survival time after severe volume-controlled hemorrhagic shock in rats without worsening hypotension or bradycardia. This beneficial effect may not be attributable to effects of adenosine on the inflammatory response.

Dopamine-1 receptor stimulation attenuates the vasoconstrictive response to gut ischemia

The effects of fenoldopam, a dopamine-1 (DA-1) receptor agonist, were studied in two groups of anesthetized dogs before and after induction of splanchnic ischemia by way of hemorrhage. During the first portion of the experiment, both groups received fenoldopam (1.5 microg x kg(-1) x min(-1)) for 45 min followed by a 45-min washout. During the second portion, hemorrhage (10 ml/kg) was induced, followed by no intervention in group I (controls) and restarting of the fenoldopam infusion in group II. Prehemorrhage, fenoldopam increased composite portal blood flow by 33% (P < 0.01). After hemorrhage-induced splanchnic ischemia, fenoldopam restored portal vein blood flow to near baseline, maintained the splanchnic fraction of cardiac output, and attenuated the rise in gut mucosal PCO(2). DA-1 receptor stimulation increased portal blood flow and redistributed blood flow away from the serosal layer in favor of the mucosa during basal conditions and after hemorrhage, suggesting a more concentrated distribution of splanchnic DA-1 receptors within the mucosal layer vasculature. Fenoldopam maintained splanchnic blood flow during hypoperfusion and attenuated the splanchnic vasoconstrictive response to hemorrhage.

Application of fiberoptic sensors for the study of hepatic dysoxia in swine hemorrhagic shock

OBJECTIVES

To determine whether the simultaneous measurement of tissue pH, Pco2, and Po2 with a multiple-parameter fiberoptic sensor can be used to indicate the onset of hepatic dysoxia, to determine critical values, and to assess their use in predicting negative outcomes.

DESIGN

Prospective animal study.

SETTING

University research laboratory.

SUBJECTS

Fourteen Yorkshire swine.

INTERVENTIONS

Hemorrhagic shock (n = 11) was induced over 15 mins to lower systolic blood pressure to 40 mm Hg and was maintained for 30, 60, or 90 mins. Resuscitation was achieved with shed blood and warm saline to maintain mean pressure >60 mm Hg for 120 mins. Sham animals (n = 3) were subjected to 90 mins of sham shock, followed by a 120-min recovery period.

MEASUREMENTS AND MAIN RESULTS

The multiple-parameter sensor continuously measured tissue pH, Pco2, and Po2. pH and Pco2, indicators of anaerobic metabolism, were plotted against tissue Po2. All shocked animals, but no sham animals, showed a biphasic relationship between Po2 and both pH and Pco2. Curves were fit to both an exponential and a dual-line linear function to determine critical values for Po2, pH, and Pco2. The length of time the animal was dysoxic was evaluated as a predictor of negative outcome. Critical values determined from the exponential models were more sensitive indicators of negative outcome than values determined from the linear model and more sensitive than arterial lactate and tonometric intramucosal pH and Pco2.

CONCLUSIONS

The multiple-parameter sensor offers the unique opportunity to study solid as well as hollow organ dysoxia through the simultaneous measurement of interstitial pH, Pco2, and Po2 in a small tissue region. The gradual transition from sufficient oxygen availability to dysoxia as a result of hemorrhage was better described by an exponential equation. The length of time that pH was below or Pco2 was above the critical value determined from the exponential model was predictive of a negative outcome.

Mild hypothermia increases survival from severe pressure-controlled hemorrhagic shock in rats

BACKGROUND

In previous studies, mild hypothermia (34 degrees C) during uncontrolled hemorrhagic shock (HS) increased survival. Hypothermia also increased mean arterial pressure (MAP), which may have contributed to its beneficial effect. We hypothesized that hypothermia would improve survival in a pressure-controlled HS model and that prolonged hypothermia would further improve survival.

METHODS

Thirty rats were prepared under light nitrous oxide/halothane anesthesia with spontaneous breathing. The rats underwent HS with an initial blood withdrawal of 2 mL/100 g over 10 minutes and pressure-controlled HS at a MAP of 40 mm Hg over 90 minutes (without anticoagulation), followed by return of shed blood and additional lactated Ringer's solution to achieve normotension. Hemodynamic monitoring and anesthesia were continued to 1 hour, temperature control to 12 hours, and observation without anesthesia to 72 hours. After HS of 15 minutes, 10 rats each were randomized to group 1, with normothermia (38 degrees C) throughout; group 2, with brief mild hypothermia (34 degrees C during HS 15-90 minutes plus 30 minutes after reperfusion); and group 3, with prolonged mild hypothermia (same as group 2, then 35 degrees C [possible without shivering] from 30 minutes after reperfusion to 12 hours).

RESULTS

MAP during HS and initial resuscitation was the same in all three groups, but was higher in the hypothermia groups 2 and 3, compared with the normothermia group 1, at 45 and 60 minutes after reperfusion. Group 1 required less blood withdrawal to maintain MAP 40 mm Hg during HS and more lactated Ringer's solution for resuscitation. At end of HS, lactate levels were higher in group 1 than in groups 2 and 3 (p < 0.02). Temperatures were according to protocol. Survival to 72 hours was achieved in group 1 by 3 of 10 rats, in group 2 by 7 of 10 rats (p = 0.18 vs. group 1), and in group 3 by 9 of 10 rats (p = 0.02 vs. group 1, p = 0.58 vs. group 2). Survival time was longer in group 2 (p = 0.09) and group 3 (p = 0.007) compared with group 1.

CONCLUSION

Brief hypothermia had physiologic benefit and a trend toward improved survival. Prolonged mild hypothermia significantly increased survival after severe HS even with controlled MAP. Extending the duration of hypothermia beyond the acute phases of shock and resuscitation may be needed to ensure improved outcome after prolonged HS.

Gastric intramucosal PCO2 and pH variability in ventilated critically ill patients

OBJECTIVE

Gastric intramucosal PCO2 (PiCO2) and pH (pHi) are currently used as indices of the adequacy of splanchnic perfusion and as end points to guide therapeutic intervention. However, little is known about their spontaneous variability over time. The present study was designed to define the magnitude of spontaneous variability of PiCO2 and pHi in sedated medical intensive care unit (ICU) patients using an automated recirculating air tonometer and to test whether high-level positive end-expiratory pressure (PEEP) or inverse inspiratory/expiratory (I:E) ratio ventilation resulted in a greater variability than low PEEP with conventional I:E ratio ventilation.

DESIGN

Prospective study.

SETTING

Medical ICU in a tertiary medical center.

PATIENTS

Twenty-three acute respiratory failure patients.

INTERVENTIONS

After being sedated, patients were randomized to undergo pressure control ventilation at the following three settings: A, high PEEP (15 cm H2O) with conventional I:E ratio (1:2), and B, low PEEP (5 cm H2O) with inverse I:E ratio (2:1) alternately, and then C, low PEEP (5 cm H2O) with conventional I:E ratio (1:2). Each ventilation setting period lasted 1 hr.

MEASUREMENTS AND MAIN RESULTS

The PiCO2 and pHi were measured at baseline (time 0), and at 15, 30, 45, and 60 mins thereafter. The corresponding coefficients of variation (CVs) of PiCO2 for overall pooled group and settings A, B, and C were 4.0%, 4.4%, 3.4%, and 4.2%, respectively. The corresponding CVs of pHi for overall pooled group and settings A, B, and C were 0.36%, 0.37%, 0.33%, and 0.4%, respectively. Analysis of variance showed no significant difference in the CVs of PiCO2 or pHi between the three settings. The 95% confidence interval is approximately +/-8% variability for PiCO2 and +/-0.7% variability for pHi.

CONCLUSIONS

In critically ill medical ICU patients with stable hemodynamics, the spontaneous variability of PiCO2 or pHi are not substantial. High PEEP (15 cm H2O) and inverse ratio ventilation (2:1), which does not change the cardiac output or hemodynamics, does not contribute to increased spontaneous variability in PiCO2 or pHi.

Metabolic component of intestinal PCO(2) during dysoxia

The adequacy of intestinal perfusion during shock and resuscitation might be estimated from intestinal tissue acid-base balance. We examined this idea from the perspective of conventional blood acid-base physicochemistry. As the O(2) supply diminishes with failing blood flow, tissue acid-base changes are first "respiratory, " with CO(2) coming from combustion of fuel and stagnating in the decreasing blood flow. When the O(2) supply decreases to critical, the changes become "metabolic" due to lactic acid. In blood, the respiratory vs. metabolic distinction is conventionally made using the buffer base principle, in which buffer base is the sum of HCO(3)(-) and noncarbonate buffer anion (A(-)). During purely respiratory acidosis, buffer base stays constant because HCO(3)(-) cannot buffer its own progenitor, carbonic acid, so that the rise of HCO(3)(-) equals the fall of A(-). During anaerobic "metabolism," however, lactate's H(+) is buffered by both A(-) and HCO(3)(-), causing buffer base to decrease. We quantified the partitioning of lactate's H(+) between HCO(3)(-) and A(-) buffer in anoxic intestine by compressing intestinal segments of anesthetized swine into a steel pipe and measuring PCO(2) and lactate at 5- to 10-min intervals. Their rises followed first-order kinetics, yielding k = 0. 031 min(-1) and half time = approximately 22 min. PCO(2) vs. lactate relations were linear. Over 3 h, lactate increased by 31 +/- 3 mmol/l tissue fluid (mM) and PCO(2) by approximately 17 mM, meaning that one-half of lactate's H(+) was buffered by tissue HCO(3)(-) and one-half by A(-). The data were consistent with a lumped pK(a) value near 6.1 and total A(-) concentration of approximately 30 mmol/kg. We conclude that the respiratory vs. metabolic distinction could be made in tissue by estimating tissue buffer base from measured pH and PCO(2).

Gastric mucosal end-tidal PCO2 difference as a continuous indicator of splanchnic perfusion

Gastric mucosal and arterial blood PCO2 must be known to assess mucosal perfusion by means of gastric tonometry. As end-tidal PCO2 (PE'CO2) is a function of arterial PCO2, the gradient between PE'CO2 and gastric mucosal PCO2 may reflect mucosal perfusion. We studied the agreement between two methods to monitor gut perfusion. We measured the difference between gastric mucosal PCO2 (air tonometry) and PE'CO2 (= DPCO2gas) and the difference between gastric mucosal PCO2 (saline tonometry) and arterial blood PCO2 (= DPCO2sal) in 20 patients with or without lung injury. DPCO2gas was greater than DPCO2sal but changes in DPCO2gas reflected changes in DPCO2sal. The bias between DPCO2gas and DPCO2sal was 0.85 kPa and precision 1.25 kPa. The disagreement between DPCO2gas and DPCO2sal increased with increasing dead space. We propose that the disagreement between the two methods studied may not be clinically important and that DPCO2gas may be a method for continuous estimation of splanchnic perfusion.

Intraperitoneal tonometry for detection of regional enteric ischaemia

BACKGROUND

The objective of this study was to test the hypothesis that intraperitoneal tonometry can be a specific monitor for ischaemia in the small intestine.

METHODS

Twelve pigs were anaesthetized and mechanically ventilated. The celiac artery (CA), the superior mesenteric artery (SMA) and the inferior mesenteric artery (IMA) were identified. Tonometry catheters were positioned intraperitoneally at three different locations where blood supply varied. One at a time of the mesenteric arteries was occluded, producing regional ischaemia in different splanchnic organs.

RESULTS

Regional PCO2 (Pr CO2) increased significantly in the intestinal region, in the small intestine, only during the SMA clamping. In the epigastric region, i.e. in the space between the liver and the stomach, PrCO2 increased significantly only during CA clamping.

CONCLUSION

Intraperitoneal tonometry in the intestinal region can be a specific monitor of ischaemia in the small intestine.

INVESTIGATION

The care and handling of the animals was in accordance with legislation by the Swedish Board of Agriculture.

Effects of increased oxygen breathing in a volume controlled hemorrhagic shock outcome model in rats

It is believed that victims of traumatic hemorrhagic shock (HS) benefit from breathing 100% O(2). Supplying bottled O(2) for military and civilian first aid is difficult and expensive. We tested the hypothesis that increased FiO(2) both during severe volume-controlled HS and after resuscitation in rats would: (1) increase blood pressure; (2) mitigate visceral dysoxia and thereby prevent post-shock multiple organ failure; and (3) increase survival time and rate. Thirty rats, under light anesthesia with halothane (0. 5% throughout), with spontaneous breathing of air, underwent blood withdrawal of 3 ml/100 g over 15 min. After HS phase I of 60 min, resuscitation phase II of 180 min with normotensive intravenous fluid resuscitation (shed blood plus lactated Ringer's solution), was followed by an observation phase III to 72 h and necropsy. Rats were randomly divided into three groups of ten rats each: group 1 with FiO(2) 0.21 (air) throughout; group 2 with FiO(2) 0.5; and group 3 with FiO(2) 1.0, from HS 15 min to the end of phase II. Visceral dysoxia was monitored during phases I and II in terms of liver and gut surface PCO(2) increase. The main outcome variables were survival time and rate. PaO(2) values at the end of HS averaged 88 mmHg with FiO(2) 0.21; 217 with FiO(2) 0.5; and 348 with FiO(2) 1. 0 (P<0.001). During HS phase I, FiO(2) 0.5 increased mean arterial pressure (MAP) (NS) and kept arterial lactate lower (P<0.05), compared with FiO(2) 0.21 or 1.0. During phase II, FiO(2) 0.5 and 1. 0 increased MAP compared with FiO(2) 0.21 (P<0.01). Heart rate was transiently slower during phases I and II in oxygen groups 2 and 3, compared with air group 1 (P<0.05). During HS, FiO(2) 0.5 and 1.0 mitigated visceral dysoxia (tissue PCO(2) rise) transiently, compared with FiO(2) 0.21 (P<0.05). Survival time (by life table analysis) was longer after FiO(2) 0.5 than after FiO(2) 0.21 (P<0. 05) or 1.0 (NS), without a significant difference between FiO(2) 0. 21 and 1.0. Survival rate to 72 h was achieved by two of ten rats in FiO(2) 0.21 group 1, by four of ten rats in FiO(2) 0.5 group 2 (NS); and by four of ten rats of FiO(2) 1.0 group 3 (NS). In late deaths macroscopic necroses of the small intestine were less frequent in FiO(2) 0.5 group 2. We conclude that in rats, in the absence of hypoxemia, increasing FiO(2) from 0.21 to 0.5 or 1.0 does not increase the chance to achieve long-term survival. Breathing FiO(2) 0.5, however, might increase survival time in untreated HS, as it can mitigate hypotension, lactacidemia and visceral dysoxia.

Mild or moderate hypothermia, but not increased oxygen breathing, increases long-term survival after uncontrolled hemorrhagic shock in rats

OBJECTIVE

To test the hypotheses that, for uncontrolled hemorrhagic shock (UHS) in rats, mild hypothermia, compared with normothermia, would increase long-term survival as well as moderate hypothermia, oxygen breathing would increase survival further, and hypothermia and oxygen would mitigate visceral ischemia (dysoxia) during UHS.

DESIGN

Prospective, randomized study.

SETTING

Animal research laboratory.

SUBJECTS

A total of 54 male Sprague-Dawley rats.

INTERVENTIONS

Under light anesthesia and spontaneous breathing, rats underwent UHS phase I of 75 mins, with initial withdrawal of 3 mL/100 g of blood over 15 mins, followed by UHS via tail amputation and limited fluid resuscitation to maintain mean arterial pressure at > or =40 mm Hg; resuscitation phase II of 60 mins (from 75 mins to 135 mins) with hemostasis and aggressive fluid resuscitation to normalize hemodynamics; and observation phase III to 72 hrs. Rats were randomly divided into nine groups (n = 6 each) with three rectal temperature levels (38 degrees C [normothermia] vs. 34 degrees C [mild hypothermia] vs. 30 degrees C [moderate hypothermia]) by surface cooling; each with 3 FIO2 levels (0.25 vs. 0.5 vs. 1.0).

MEASUREMENTS AND MAIN RESULTS

Hypothermia increased blood pressure compared with normothermia. Increased FIO2 had no effect on blood pressure. Additional blood loss from the tail cut was small, with no differences among groups. Hypothermia and FIO2 of 0.5 decreased visceral hypoxia, as measured by the difference between visceral (liver and jejunum) surface Pco2 and PaCO2 during UHS. Compared with normothermia, mild hypothermia increased the survival time and rate as well as moderate hypothermia (p < .01 by life table), without a significant difference between mild and moderate hypothermia. Increased FIO2 had no effect on survival time or rate.

CONCLUSIONS

After severe UHS and resuscitation in rats, mild hypothermia during UHS, compared with normothermia, increases blood pressure, survival time and 72-hr survival rate as well as moderate hypothermia. Mild hypothermia is clinically more feasible and safer than moderate hypothermia. Increased FIO2 seems to have no significant effect on outcome.

Systemic hemodynamics, gastric intramucosal PCO2 changes, and outcome in critically ill burn patients

OBJECTIVES

To define the hemodynamic and gastric intramucosal PCO2 (PiCO2) changes during the first 48 hrs after burn trauma and to analyze their relationship with outcome.

DESIGN

Prospective, observational study in a cohort of consecutively admitted critically ill burn patients.

SETTING

Intensive care burn unit in a university hospital.

PATIENTS

Forty-two patients with burns covering >20% of body surface area or inhalation injury.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Patients were monitored with an oximetric pulmonary arterial catheter and a gastric tonometer to measure PiCO2. The difference between arterial and gastric mucosal PCO2 (P[i-a]CO2) was considered indicative of gastric mucosal hypoxia. Hemodynamic and PiCO2 measurements were performed during the first 48 hrs after admission. Patients suffered burns covering 36.1% +/- 14.3% (mean +/- SD) and 45.3% +/- 21.9% of body surface area (survivors and nonsurvivors, respectively). All patients were successfully resuscitated by conventional standards. Nonsurvivors (n = 16) died a median of 17 days after admission. In univariate analysis, the presence of shock during the resuscitation phase, age, mixed venous pH, P[i-a]CO2, right atrial pressure, pulmonary arterial pressure, pulmonary arterial occlusion pressure, cardiac index, systemic and pulmonary vascular resistance, left ventricular stroke work index, mixed venous oxygen saturation, and systemic oxygen delivery, consumption, and extraction ratio, measured over the first 12 hrs after admission, were significantly (p < .05) different between survivors and nonsurvivors. These differences disappeared after 12 hrs after admission. Multivariate analysis identified age, percentage body surface area burned, and oxygen delivery index (6 hrs after admission) as factors independently associated with a poor outcome. P[i-a]CO2 (12 hrs after admission) was significantly greater in patients with than in those without inhalation injury (17 +/- 13 torr [2.26 +/- 1.73 kPa] vs. 6 +/- 10 torr [0.79 +/- 1.33 kPa]; p = .005). Patients with a P[i-a]CO2 difference (6 hrs after admission) > or =10 torr (1.33 kPa) had a mortality rate of 56% vs. 25% of those patients with <10 torr (p = .044).

CONCLUSIONS

Our data indicate that there are hemodynamic and biochemical changes that occur early after burn trauma that are associated with prognosis after an apparently successful resuscitation. Particularly, a hemodynamic profile characterized by systemic acidosis, low systemic blood flow, and systemic and pulmonary vasoconstriction early after trauma is associated with a poor outcome. Additionally, intestinal mucosal acidosis occurs after burn trauma, is influenced by inhalation injury, and is a variable related to outcome.

Accuracy of mucosal pH and mucosal-arterial carbon dioxide tension for detecting mesenteric hypoperfusion in acute canine endotoxemia

OBJECTIVE

To determine the level of mucosal-arterial Pco2 (Pco2 gap) that is both sensitive and specific for the detection of mesenteric hypoperfusion as defined by either a >50% reduction in portal blood flow or release of lactate by the gut.

DESIGN

Animal experiment.

SUBJECTS

Seven anesthetized, intubated, mechanically ventilated, and surgically instrumented mongrel dogs.

INTERVENTION

Escherichia coli endotoxin (1 mg/kg) given intravenously for 5 mins.

MEASUREMENTS AND MAIN RESULTS

Tonometric Pco2, arterial blood gases, arterial and portal venous lactates, and portal and systemic hemodynamic variables were measured. Mucosal pH (pHi) was calculated according to the manufacturers' instructions. From these data, receiver operating characteristics were calculated. Although animals were resuscitated to maintain a constant cardiac output, portal flow decreased from 350+/-101 to 152+/-75 mL/min (p<.01) and the gut released lactate into the portal circulation in all animals. Pco2 gap increased from 13.1+/-3.9 to 40.2+/-39.2 torr (p<.01) and was inversely correlated with portal blood flow (r2 = .20; p<.05). For detection of a >50% reduction in portal blood flow, a Pco2 gap of 20 torr yielded a maximum accuracy of 67% (sensitivity, 55%; specificity, 73%) and was less accurate than a pHi of 7.20, which yielded a maximum accuracy of 76% (sensitivity, 90%; specificity, 70%), although this difference was not significant (p = .24). There was also a correlation between pHi and portal blood flow (r2 = .31; p<.01). For detection of lactate release by the gut, a Pco2 gap of 20 torr was also 67% accurate (sensitivity, 53%; specificity, 78%), whereas a pHi of 7.10 achieved an accuracy of 64% (sensitivity, 40%; specificity, 83%), which was not significantly different.

CONCLUSION

Pco2 gap measurements are neither sensitive nor specific for mesenteric hypoperfusion with regard to total gut blood flow reductions of >50% or the release of lactate into the portal circulation.

Therapeutic hypothermia in traumatology

Despite its proven clinical application for protection-preservation of the brain and heart during cardiac surgery, hypothermia research has fallen in and out of favor many times since its inception. Since the 1980s, there has been renewed research and clinical interest in therapeutic hypothermia for resuscitation of the brain after cardiac arrest or TBI and for preservation-resuscitation of extracerebral organs, particularly the abdominal viscera in low-flow states such as HS. Although some of the fears regarding the side effects of hypothermia are warranted, others are not. Without further laboratory and clinical studies, the significance of these effects cannot be determined and ways to overcome these problems cannot be developed. Currently, at the turn of the century, there are significant data demonstrating the benefit of mild-to-moderate hypothermia in animals and humans after cardiac arrest or TBI and in animals during and after HS. The clinical implications of uncontrolled versus controlled hypothermia in trauma patients and the best way to assure poikilothermia for cooling without shivering are still unclear. It is time to consider a prospective trial of therapeutic, controlled hypothermia for patients during traumatic HS and resuscitation. The authors believe that the new millennium will witness remarkable advantages of the use of controlled hypothermia in trauma. Starting in the prehospital phase, mild hypothermia will be induced in hypovolemic patients, which will not only decrease the immediate mortality rate but perhaps also will protect cells and reduce the likelihood of secondary inflammatory response syndrome, multiple organ failure, and late deaths. The most futuristic applications will be hypothermic strategies to achieve prolonged suspended animation for delayed resuscitation in traumatic exsanguination cardiac arrest.

Accuracy of intramucosal pH calculated from arterial bicarbonate and the Henderson-Hasselbalch equation: assessment using simulated ischemia

OBJECTIVES

To determine the accuracy of intramucosal pH (pHi) calculated using arterial bicarbonate instead of mucosal capillary bicarbonate in the Henderson-Hasselbalch equation.

DESIGN

Simulation of progressive ischemia in mucosal capillary blood.

SETTING

University research laboratory.

SUBJECTS

Normal human blood diluted with plasma.

INTERVENTIONS

Three venous blood specimens were heparinized and diluted to a mean hemoglobin concentration of 5.0 (+/-0.9) g/dL by addition of plasma (2:1, vol:vol). Mucosal capillary aerobic flow stagnation was simulated by multiple exposures of each cooled specimen to a gas mixture containing 90% nitrogen and 10% CO2. When PCO2 measured at 37 degrees C (98.6 degrees F) was approximately 120 torr (16 kPa), the assigned anaerobic threshold, subsequent anaerobic flow stagnation was simulated by mixing the hypercapnic specimens in sealed syringes with five to six successive small aliquots (<100 microL) of lactic acid (10 g/L).

MEASUREMENTS AND MAIN RESULTS

The relationship between Pco2 and pH in the specimens was compared with the relationship between the same PCO2 values and pHi calculated by substituting bicarbonate concentrations of 22 and 26 mmol/L in the Henderson-Hasselbalch equation. As PCO2 rose from 50 torr (8 kPa), conventionally calculated pHi increasingly underestimated simulated mucosal capillary pH, with bias >0.1 pH unit at the simulated anaerobic threshold of 120 torr (16 kPa). As PCO2 rose further the values converged, becoming equivalent at PCO2 approximately 150 torr (20 kPa). From PCO2 > or =200 torr (26.7 kPa), conventional pHi progressively overestimated simulated mucosal pH. The difference was >0.3 pH units at PCO2 = 250 torr (33.3 kPa).

CONCLUSIONS

In the mucosal PCO2 range usually encountered clinically, the arterial bicarbonate substitution causes underestimation of mucosal capillary pH. With moderate mucosal capillary lactic acidosis the error becomes small, and in severe regional ischemia there is significant overestimation of mucosal capillary pH.

Earlier prediction of anastomotic insufficiency after thoracic esophagectomy by intramucosal pH

OBJECTIVES

To assess the value of using intramucosal pH (pHi) measurements to evaluate the viability of the gastric tube after thoracic esophagectomy, and to determine whether these measurements may be used for early prediction of anastomotic insufficiency.

DESIGN

Prospective, observational study.

SETTING

University hospital in Japan.

PATIENTS

Thirty-nine patients who had undergone thoracic esophagectomy as a treatment for esophageal cancer.

INTERVENTIONS

The blood flow within the gastric tube was measured using a laser Doppler flowmeter during surgery. Periodic measurement of the pHi within the gastric tube (gastric pHi) began during surgery and continued until the second postoperative day. In 30 patients, the pHi within the rectum (rectal pHi) was measured simultaneously with the gastric pHi. The patients were divided into two groups: those patients who experienced anastomotic insufficiency constituted the leakage(+) group (n = 13); those patients who did not experience these complications were designated the leakage(-) group (n = 26).

MEASUREMENTS AND MAIN RESULTS

The gastric pHi values correlated significantly with simultaneous measurements of the blood flow at the anastomotic site (p < .01). The postoperative gastric pHi values increased gradually in the leakage(-) group but stopped increasing after surgery in the leakage(+) group. The rectal pHi values increased gradually after surgery in both groups. Furthermore, there was a significant difference between the two groups when their gastric pHi values were subtracted from their rectal pHi values from the morning of the first postoperative day until the morning of the second postoperative day (p < .05).

CONCLUSIONS

The gastric pHi values well reflected the viability of the gastric tube, especially when combined with the rectal pHi values. By measuring pHi, we can more accurately predict the risk of anastomotic insufficiency earlier after surgery and therefore give those patients who need it additional care to improve the viability of the gastric tube.

Mild or moderate hypothermia but not increased oxygen breathing prolongs survival during lethal uncontrolled hemorrhagic shock in rats, with monitoring of visceral dysoxia

OBJECTIVE

To test the hypotheses that during lethal uncontrolled hemorrhagic shock (UHS) in rats compared with normothermia and room air breathing: a) mild hypothermia would prolong survival time as well as moderate hypothermia; b) oxygen breathing would prolong survival further; and c) hypothermia and oxygen would mitigate visceral ischemia (dysoxia) during UHS.

DESIGN

Prospective, randomized, controlled laboratory animal study.

SETTING

Animal research facility.

SUBJECTS

Male Sprague-Dawley rats.

INTERVENTION

Fifty-four rats were lightly anesthetized with halothane during spontaneous breathing. UHS was induced by blood withdrawal of 3 mL/100 g over 15 mins, followed by 75% tail amputation with topical application of heparin. Five minutes after tail cut, rats were randomly divided into nine groups (6 rats each) with three rectal temperature levels (38 degrees C [100.4 degrees F; normothermia] vs. 34 degrees C [93.2 degrees F; mild hypothermia] vs. 30 degrees C [86 degrees F; moderate hypothermia]) by surface cooling; each with 3 FIO2 levels (0.25 vs. 0.5 vs. 1.0). Rats were observed without fluid resuscitation until death (apnea and pulselessness). Visceral ischemia was monitored by observing liver and gut surface PCO2.

MEASUREMENTS AND MAIN RESULTS

Mean survival time, which was 51 mins in the control group with normothermia and FIO2 of 0.25, was more than doubled with hypothermia, to 119 mins in the combined mild hypothermia groups (p < .05) and to 132 mins in the combined moderate hypothermia groups (p < .05; NS for moderate vs. mild hypothermia). FIO2 had no statistically significant effect on survival time. Increases in visceral surface PCO2 correlated with hypotension (r2 = .22 for intestine and .40 for liver). Transiently, increased FIO2, not hypothermia, mitigated visceral ischemia.

CONCLUSIONS

Both mild and moderate hypothermia prolonged survival time during untreated, lethal UHS in rats. Increased FIO2 had no effect on survival. The effects of hypothermia and increased FIO2 during UHS on viscera, the ability to be resuscitated, and outcome should be explored further.

Gas supersaturation in the cecal wall of mice due to bacterial CO2 production

PCO2 in the lumen and serosa of cecum and jejunum was measured in mice. The anesthetic used was a fentanyl-fluanisone-midazolam mixture. PCO2 was recorded in vivo and postmortem. PCO2 was 409 +/- 32 Torr (55 +/- 4 kPa) in the cecal lumen and 199 +/- 22 Torr (27 +/- 3 kPa) on the serosa in normal mice. Irrigation of the cecum resulted in serosal and luminal PCO2 levels of 65-75 Torr. Cecal PCO2 was significantly lower in germ-free mice (65 +/- 5 Torr). Cecal PCO2 increased significantly after introduction of normal bacterial flora into germ-free mice. Introduction of bacterial monocultures into germ-free mice had no effect. After the deaths of the mice, cecal PCO2 increased rapidly in normal mice. The intestinal bacteria produced the majority of the cecal PCO2, and the use of tonometry in intestinal segments with a high bacterial activity should be interpreted with caution. We propose that serosal PCO2 levels >150-190 Torr (20-25 kPa) in the cecum of mice with a normal circulation may represent a state of gas supersaturation in the cecal wall.

Effect of gastric feeding on intragastric P(CO2) tonometry in healthy volunteers

PURPOSE

The tonometric detection of a high intragastric regional P(CO2) (PrCO2) reflecting an elevated intramucosal P(CO2) can be helpful to diagnose mucosal ischemia, if acid secretion is suppressed to avoid intragastric CO2 production through buffering of acid by bicarbonate in the stomach. It is recommended to perform tonometry in the fasting state, but this may hamper feeding of the critically ill. On the other hand, postfeeding tonometry could serve as a diagnostic stress test because feeding increases mucosal blood flow demand, provided that the meal itself does not hamper diffusion of CO2 from mucosa to tonometer balloon and does not generate intragastric CO2, independently from intramucosal P(CO2). We therefore studied the effect of a standard meal on intragastric PrCO2 tonometry in healthy volunteers with suppression of meal-stimulated gastric acid secretion and, presumably, with an adequate mucosal blood flow reserve.

MATERIAL AND METHODS

The gastric juice pH and tonometric PrCO2 were measured in 14 human volunteers, after gastric acid secretion suppression by either ranitidine (100-mg bolus, followed by 25 mg/h i.v., n = 7) or by ranitidine plus pirenzepine (10-mg bolus, followed by 3 mg/h i.v., n=7) to suppress any residual meal-stimulated gastric acid secretion, before and at 30-minute intervals until 120 minutes after oral ingestion of a standard liquid test meal (Pulmocare [Abbott, the Netherlands]; 500 mL, 750 kcal, P(CO2) 5 mm Hg, pH 7.50).

RESULTS

The gastric juice pH, which was >4.0 in all individuals throughout the study, and the PrCO2 did not depend on the regimen for gastric acid secretion suppression, and therefore the data were pooled. The PrCO2 (median [range]) after feeding was 69% (56% to 170%) of baseline (42 [37-51] mm Hg) from 0 to 30 minutes (P < .001), 85% (72% to 167%) of baseline from 30 to 60 (P < .05), 97% (57% to 193%) from 60 to 90 minutes, and 112% (97% to 189%) of baseline from 90 to 120 minutes with a rise above baseline in 10 of 14 patients. In vitro, the liquid test meal generated CO2 after adding bicarbonate but not after hydrochloric acid.

CONCLUSION

We recommend intragastric tonometry to be performed in the fasting state and discourage tonometry after feeding as a stress test, because a single test meal changes tonometric PrCO2 in a time-dependent manner until 2 hours after gastric feeding of healthy volunteers. The fall in PrCO2 directly after feeding can be attributed to dilution, whereas a rise above baseline in some patients may have been caused, as supported by CO2 production after adding bicarbonate to the test meal in vitro, by CO2 production through buffering of meal-derived acid by gastric bicarbonate, in the absence of stimulated gastric acid secretion by feeding.

Human PaCO2 and standard base excess compensation for acid-base imbalance

OBJECTIVES

Renal and respiratory acid-base regulation systems interact with each other, one compensating (partially) for a primary defect of the other. Most investigators striving to typify compensations for abnormal acid-base balance have reported their findings in terms of arterial pH, PaCO2, and/or HCO3-. However, pH and HCO3- are both altered by both respiratory and metabolic changes. We sought to simplify these relations by expressing them in terms of standard base excess (SBE in mM), which quantifies the metabolic balance and is independent of PaCO2.

DESIGN

Meta-analysis.

SETTING

Historical synthesis developed via the Internet.

PATIENTS

Arterial pH, PaCO2, and/or HCO3- data sets were obtained from 21 published reports of patients considered to have purely acute or chronic metabolic or respiratory acid-base problems.

INTERVENTIONS

We used the same data to compute the typical compensatory responses to imbalances of SBE and PaCO2. Relations were expressed as difference (delta) from normal values for PaCO2 (40 torr [5.3 kPa]) and SBE (0 mM).

MEASUREMENTS AND MAIN RESULTS

The data of patient compensatory changes conformed to the following equations, as well as to the traditional PaCO2 vs. HCO3- or H+ vs. PaCO2 equations: Metabolic change responding to change in PaCO2: Acute deltaSBE = 0 x deltaPaCO2, hence: SBE = 0, Chronic deltaSBE = 0.4 x deltaPaCO2. Respiratory change responding to change in SBE: Acidosis deltaPaCO2 = 1.0 x deltaSBE, Alkalosis deltaPaCO2 = 0.6 x deltaSBE.

CONCLUSION

Data reported by many investigators over the past 35 yrs on typical, expected, or "normal" human compensation for acid-base imbalance may be expressed in terms of the independent variables: PaCO2 (respiratory) and SBE (metabolic).

Mitochondrial redox state as a potential detector of liver dysoxia in vivo

Dysoxia can be defined as ATP flux decreasing in proportion to O2 availability with preserved ATP demand. Hepatic venous beta-hydroxybutyrate-to-acetoacetate ratio (beta-OHB/AcAc) estimates liver mitochondrial NADH/NAD and may detect the onset of dysoxia. During partial dysoxia (as opposed to anoxia), however, flow may be adequate in some liver regions, diluting effluent from dysoxic regions, thereby rendering venous beta-OHB/AcAc unreliable. To address this concern, we estimated tissue ATP while gradually reducing liver blood flow of swine to zero in a nuclear magnetic resonance spectrometer. ATP flux decreasing with O2 availability was taken as O2 uptake (VO2) decreasing in proportion to O2 delivery (QO2); and preserved ATP demand was taken as increasing Pi/ATP. VO2, tissue Pi/ATP, and venous beta-OHB/AcAc were plotted against QO2 to identify critical inflection points. Tissue dysoxia required mean QO2 for the group to be critical for both VO2 and for Pi/ATP. Critical QO2 values for VO2 and Pi/ATP of 4.07 +/- 1.07 and 2.39 +/- 1.18 (SE) ml . 100 g-1 . min-1, respectively, were not statistically significantly different but not clearly the same, suggesting the possibility that dysoxia might have commenced after VO2 began decreasing, i.e., that there could have been "O2 conformity." Critical QO2 for venous beta-OHB/AcAc was 2.44 +/- 0.46 ml . 100 g-1 . min-1 (P = NS), nearly the same as that for Pi/ATP, supporting venous beta-OHB/AcAc as a detector of dysoxia. All issues considered, tissue mitochondrial redox state seems to be an appropriate detector of dysoxia in liver.