INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sepsis and its related syndromes develop frequently in hospitalized patients, with an associated mortality of 10 to 20%. In the presence of circulatory failure, this figure rises to over 60% and may account for up to 200,000 deaths per annum in the United States alone (1). Most patients succumb to a multiple organ dysfunction syndrome (MODS) rather than hypotension per se (2), but the reasons for this are not clear. Sepsis is known to disrupt microcirculatory flow and nutrient exchange, and an impaired response to endogenous and exogenous pressor agents is often reported (3). Intravascular leukaggregation, abnormal red blood cell deformability, increased microvascular permeability, interstitial protein loss, and tissue edema are frequently observed (4). These changes are promoted by proinflammatory mediators and are partly modulated by the endothelium. It is hypothesized that endothelial injury exacerbates maldistribution of regional blood flow and leads to cellular hypoxia and vital organ dysfunction.

In patients with sepsis, intravenous fluids and inotrope administration are widely used to support the systemic circulation. Although these interventions may reverse latent intravascular hypovolemia and improve myocardial contractility, there are few data regarding their effects on tissue oxygen tension (Pt_O2), and they have not been associated with any survival benefit in patients with septic shock. Clinically, a hyperdynamic circulatory response to fluid resuscitation is observed. Second, peripheral vasodilatation and systemic hypotension are characteristic features of severe sepsis and despite reflex increases in cardiac output are associated with impairment of left ventricular function (5). The endogenous vasodilator nitric oxide (NO), formed from the substrate L-arginine by NO synthase (NOS), is reported to play a major role in the cardiovascular response to sepsis (6). Potential beneficial effects of NOS inhibition in animal models and clinical sepsis have been the subject of considerable speculation (7, 8), but there are no studies investigating the effects of generalized NOS inhibition on peripheral tissue oxygenation. Thus, any improvements in systemic hemodynamic indices may be complicated by impaired tissue nutrition, in addition to other theoretical disadvantages (9).

We have demonstrated recently that significant tissue hypoxia and abnormal microvascular regulation of tissue oxygenation occur in endotoxemic rats, despite apparently normal microcirculatory perfusion (10). The aims of this study were fourfold: first, to further assess global hemodynamic performance (and thus oxygen delivery), gas exchange, and Pt_O2 in a rodent model of endotoxemia; second, to examine the effect of fluid resuscitation on these parameters in endotoxemic rats; third, to examine the effects of generalized NOS inhibition on these parameters, in the presence and absence of endotoxemia; and finally, to examine the combined effect of fluid resuscitation and generalized NOS inhibition on these parameters in the presence and absence of endotoxemia.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals and Anesthesia

All procedures and protocols were performed in accordance with the Animal (Scientific Procedures) Act, 1986, and approved by the Home Office (UK) Inspectorate. Male Wistar rats (275-300 g) received Salmonella enteritidis (Sigma Chemical, Poole, UK) endotoxin (20 mg · kg¹, intraperitoneally) or saline vehicle (3 ml · kg¹, intraperitoneally) 3.5 h before anesthesia with pentobarbitone (60 mg · kg¹, intraperitoneally). Animals were placed supine on an electric heating pad, maintaining body temperature between 36 and 37° C. A cannula was inserted into the left femoral artery and attached to a pressure transducer (Hewlett-Packard, Bracknell, UK) for continuous mean arterial blood pressure (MAP) monitoring and withdrawal of blood samples. A 1-Fr thermistor microprobe (Columbus Instruments, OH) was inserted into the aortic arch via the right carotid artery.

A tracheostomy was performed, and the animals ventilated using a mixture of 21% O₂, 5% CO₂, and balance N₂ using a volume-cycled, small animal ventilator (Harvard Apparatus, Chatham, Kent, UK). A cannula was inserted into the right atrium via the right jugular vein, and attached to a microinjector (Columbus Instruments, Columbus, OH) allowing 150 µl injection of room-temperature saline. Thermistor cardiac output (CO) was measured (Cardiotherm 500 AC-R cardiac output computer; Columbus Instruments), and the mean of two consecutive measurements within 10% of each other recorded. The left jugular vein was cannulated to permit volume resuscitation in appropriate animals; 4 ml · kg¹ · body wt¹ bolus of 0.9% saline being administered after placement of the cannulae, followed by an infusion of 16 ml · h¹ · kg¹ · body wt¹. Intermittent boluses of intraperitoneal pentobarbitone were given to supplement anesthesia as required.

Microelectrodes

Techniques for measuring Pt_O2 using gold disk microelectrodes are described elsewhere (10). Briefly, gold disk electrodes (125 µM) embedded in 25-gauge hypodermic needles were constructed in our laboratory. Current-voltage curves were constructed over the range 0 to 1.2 V to determine appropriate polarizing voltages. Electrodes were calibrated before and after each experiment with air-equilibrated and nitrogen-sparged saline. Electrodes were polished between experiments with alumina to remove adsorbed protein, and were ultrasonicated for 15 min in deionized water before recalibration. Electrodes were polarized using a two-channel potentiostat (EMS Instruments, Oxford, UK). The resulting raw current was sampled at 30 Hz with a personal computer (Dell Latitude LX 4100; Dell, Bracknell, Berkshire, UK), using a data acquisition board (Data Shuttle; Strawberry Tree, Sunnyvale, CA), and processed by data acquisition software (Workbench for Windows; Strawberry Tree).

Experimental Protocols

The microelectrodes were carefully inserted into the rectus femoralis muscle to a depth of ~ 3 mm. Initial polarization of the microelectrodes was followed by an equilibration period of 20 min. Fraction of inspired oxygen (FI_O2) was altered every 10 min in the sequence; baseline (0.21)  0.5  0.21  0.1  0.21  0.95 while continuous measurements of Pt_O2 and MAP were made. The two 0.21 FI_O2 sequences after baseline (0.21 FI_O2) were included as an equilibration period during which Pt_O2 could return to baseline (i.e., 0.21 FI_O2) values. CO measurements and arterial blood gases (MODLE) were measured at the midpoint of ventilation with 0.21 (baseline), 0.5, 0.1, and 0.95 FI_O2.

Animals were divided at random into the following groups:

Experiment 1: Effects of endotoxemia. Sham-treated animals (Group 1, n = 5) and endotoxemic animals (Group 2, n = 5).

Experiment 2: Effect of volume resuscitation. Sham-resuscitated animals (Group 3, n = 5) and endotoxemic-resuscitated animals (Group 4, n = 5). Resuscitation commenced 10 min prior to baseline recording.

Experiment 3: Effect of N^G-nitro-L-arginine methyl ester (L-NAME) (1 mg · kg¹, intravenous bolus). Sham-L-NAME animals (Group 5, n = 5) and endotoxemic-L-NAME animals (Group 6, n = 6). L-NAME was administered 10 min prior to baseline recording.

Experiment 4: Effect of combined L-NAME (1 mg · kg¹, intravenous bolus) and volume resuscitation. Sham-resuscitated-L-NAME animals (Group 7, n = 5) and endotoxemic-resuscitated-L-NAME animals (Group 8, n = 6). L-NAME was administered simultaneously with volume resuscitation 10 min prior to baseline recording.

Data Analysis

All results are expressed as mean ± SEM. Multiple group comparisons were analyzed using a one-way nonparametric analysis of variance (ANOVA) followed by Student-Newman-Keuls post-test to isolate differences. Differences between individual group means were tested by a Student's unpaired t test. Values of p  0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experiment 1: Effect of Endotoxemia

No significant differences in MAP or Pa_CO2 were observed at baseline, nor throughout the experimental period (Table 1 and Figure 1A). CO was similarly unchanged at baseline, although a significant decrease was observed in endotoxin-treated animals (Table 1, Group 2) at 0.1 FI_O2 (Figure 1B). A significant metabolic acidosis was observed in the lipopolysaccharide (LPS)-treated animals and this persisted throughout the experimental period (Figure 1C). Although Pa_O2 was normal at baseline in endotoxemic animals, mean muscle Pt_O2 was significantly depressed compared with sham-treated rats (Table 1 and Figures 1D and 1E). Significant impairment of the muscle Pt_O2 response to hyperoxia was observed in endotoxemic animals at 0.5 and 0.95 FI_O2 (Figure 1D).

View larger version (15K): [in this window] [in a new window]   Figure 1.   Changes in (A) mean arterial pressure, (B) cardiac output, (C ) blood pH, (D) tissue oxygen tension, (E ) PaO2, (F ) PaCO2 in Group 1 (sham, open squares), Group 2 (endotoxemic, closed squares), and Group 4 (endotoxemic-resuscitated, closed circles) animals after changes in FIO2. Values expressed as mean ± SEM. p < 0.05 compared with Group 1; p < 0.05 compared with Group 2.

Experiment 2: Effect of Volume Resuscitation

Volume-resuscitated rats (Group 3) displayed no differences in CO, blood pH, Pa_O2, Pa_CO2, or muscle Pt_O2 compared with sham-treated animals (Group 1) at baseline (Table 1) or during the experiment (data not shown). MAP was unchanged at baseline, but a significant rise was observed at 0.95 FI_O2 (140 ± 2.1 versus 119 ± 2.7 mm Hg; p < 0.01). In volume-resuscitated LPS-treated rats (Group 4), no significant differences in Pt_O2, MAP, CO, Pa_O2, or pH were observed at baseline compared with sham-treated rats (Table 1 and Figures 1A-1E), although Pa_CO2 was lower at baseline and at 0.1 FI_O2 (Figure 1F). Pt_O2 was unchanged at baseline and throughout the experiment (Table 1 and Figure 1D). A significant difference in Pt_O2 between volume-resuscitated and nonvolume-resuscitated endotoxemic-treated rats was observed at 0.95 FI_O2 (Figure 1D). Similarly, no significant differences in CO nor pH were observed at baseline nor throughout the experiment compared with sham-treated rats (Table 1 and Figures 1B and 1C). MAP was also unaltered at baseline, but was significantly different from LPS-treated rats at 0.1 FI_O2 (Table 1 and Figure 1A). Pa_O2 was unchanged throughout (Figure 1E).

Experiment 3: Effect of L-NAME

L-NAME alone (Group 5) caused a significant increase in MAP, and a significant fall in CO at baseline compared with sham-treated rats (Table 1 and Figures 2A and 2B). This effect was sustained throughout the experiment, with a significant fall in CO at all FI_O2 (Figure 2B), coupled with a significant increase in MAP at 0.95 FI_O2 (Figure 2A). Pt_O2 was unchanged at baseline, but was significantly reduced at 0.5 and 0.95 FI_O2 (Table 1 and Figure 2D). Blood pH and Pa_CO2 were unchanged at baseline and throughout the experiment (Table 1 and Figures 2C and 2F). Blood Pa_O2 was significantly reduced at baseline and 0.95 FI_O2 (Table 1 and Figure 1E).

View larger version (16K): [in this window] [in a new window]   Figure 2.   Changes in (A) mean arterial pressure, (B) cardiac output, (C ) blood pH, (D) tissue oxygen tension, (E ) PaO2, (F ) PaCO2 in the response of Group 1 (sham, open squares), Group 5 (sham-L-NAME, closed diamonds), and Group 6 (endotoxemic-L-NAME, closed circles) animals after changes in FIO2. Values expressed as mean ± SEM. p < 0.05 compared with Group 1.

In L-NAME-treated LPS animals (Group 6), no significant differences in MAP nor CO were observed at baseline (Figures 2A and 2B). MAP did not change throughout the experiment, whereas CO was significantly reduced at 0.95 FI_O2 only (Figures 2A and 2B). Pt_O2 was unchanged at baseline, but was significantly reduced at 0.5 and 0.95 FI_O2 (Table 1 and Figure 2D). Blood pH was also significantly reduced at baseline and 0.1 FI_O2 (Table 1 and Figure 2C). Pa_CO2 was significantly reduced at baseline, but was unchanged throughout the experiment (Figure 2F), whereas Pa_O2 was unchanged throughout (Figure 2E).

Experiment 4: Effect of Combined L-NAME and Volume Resuscitation

Sham-resuscitated-L-NAME rats (Group 7) displayed no significant differences in CO or MAP at baseline compared with sham control rats (Group 1), whereas a significant difference in MAP was observed at 0.1 FI_O2 only (Figures 3A and 3B). Moreover, Pt_O2 (Figure 3D), Pa_O2, and Pa_CO2 (Figures 3E and 3F) were unchanged both at baseline, and throughout the experiment. By contrast, blood pH was significantly lower at baseline and throughout the experiment (Figure 3C). In endotoxin-treated, volume-resuscitated, L-NAME rats (Group 8), no significant differences in MAP were observed compared with sham control rats (Group 1), whereas a significant fall in CO was detected at 0.1 FI_O2 only (Figures 3A and 3B). Moreover, both Pa_O2 and Pa_CO2 were unchanged throughout (Figures 3E and 3F). Pt_O2 was unchanged at baseline, but significantly reduced at 0.5 and 0.95 FI_O2 (Figure 3D) compared with sham control rats (Group 1). Blood pH was significantly reduced at baseline and throughout the experiment (Figure 3C).

View larger version (8K): [in this window] [in a new window]   View larger version (8K): [in this window] [in a new window]   Figure 3.   Changes in (A) mean arterial pressure, (B) cardiac output, (C ) blood pH, (D) tissue oxygen tension, (E ) PaO2, (F ) PaCO2 in Group 1 (sham, open squares), Group 7 (sham-resuscitated-L-NAME, closed triangles), and Group 8 (endotoxemic-resuscitated-L-NAME, closed circles) after changes in FIO2. Values expressed as mean ± SEM. p < 0.05 compared with Group 1.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have previously shown significant tissue hypoxia and abnormal microvascular regulation in endotoxemic rats, even in the presence of normal microcirculatory perfusion (10). Using the same model, we have now examined the relationship between tissue oxygenation and systemic hemodynamic changes by also measuring cardiac output; and have investigated the effects of volume resuscitation and NOS inhibition on this relationship. The data presented are in accord with our previous study in that significant tissue hypoxia was observed in endotoxemic muscle compared with sham-treated animals (10). Pt_O2 was similarly resistant to increasing FI_O2. These peripheral changes were accompanied by a decrease in blood pH, although MAP and CO were maintained. However, it has been previously shown that persistent metabolic acidosis may not represent tissue hypoxia, and thus the fall in blood pH observed may not reflect altered O₂ delivery (11). Previous studies investigating the state of tissue oxygenation in endotoxemic and other models of sepsis have produced varying results (10). In porcine hypotension, impaired oxygenation is observed in skeletal muscle only in the latter stages of the condition (12), whereas in rabbit muscle, the effects are extremely rapid in onset (13). Conversely, another study demonstrated elevated Pt_O2 in the endotoxemic rat bladder epithelium, indicating that changes in tissue oxygenation may be species and/ or organ-specific (14). Recently, muscle Pt_O2 in patients with septic shock has been shown to be lower than in control subjects (15). By contrast, others have found that mean Pt_O2 in patients with sepsis/septic shock was significantly elevated compared with those with "limited infection" or cardiogenic shock (16). However, neither investigator considered the effects of the administration of inotropic drugs, which are known to elevate Pt_O2 (17).

Patients with sepsis frequently require volume resuscitation, although few data exist regarding its effects on endotoxemia-induced tissue hypoxia. In rats with peritonitis, volume resuscitation reverses the observed fall in both Pt_O2 and cardiac output (18), although experiments with dogs suggest the reversibility is organ-specific (19). Our current data demonstrate that volume resuscitation can reverse the effects of endotoxemia on Pt_O2, without affecting hemodynamics. Our model displayed a normodynamic circulation, suggesting that fluid deficit was limited to the peripheral microcirculation only. Another possibility was that the hypoxia observed in the endotoxemic animals represented a redistribution of blood flow from the muscle to more vital organs. This seems unlikely however, as we have previously demonstrated in this model that nutritive muscle blood flow appears to be normal in endotoxemic animals, indicating possible impaired cellular oxygen extraction (10). Further studies will be needed to examine changes in flow at the level of the microcirculation, and the mechanism of action by which volume resuscitation improves tissue oxygenation in this model. Irrespective of this, and coupled with the partial reversal of the acidemia, our data emphasize the clinical importance of adequate fluid resuscitation in sepsis.

The role of NO in modulating the vascular response to sepsis is recognized, but poorly understood. Under physiological circumstances, NO is produced by a constitutive isoform of NOS, and appears to be essential in the maintenance of vascular tone and the regulation of organ blood flow (20). Bacterial LPS and several mediators of the immunogenic response to gram-positive and gram-negative organisms stimulate the production of the inducible form of NOS (iNOS) in a number of tissues, including macrophages, endothelium, and vascular smooth muscle (6). Immunostimulation results in iNOS induction and the largely unregulated release of NO. Many studies have shown increased expression of iNOS and increased levels of NO in both animal models and patients with sepsis (7, 21, 22).

Beneficial effects of generalized NOS inhibition in sepsis have been reported. NOS inhibition normalized cardiac index and systemic vascular resistance, MAP, and decreased heart rate without impairing oxygen consumption in unanesthetized sheep with hyperdynamic sepsis (23). In murine endotoxemia, neither NOS inhibition nor antibiotics alone altered mortality, but coadministration of the two treatments significantly improved survival (24). NOS inhibition increases blood pressure and systemic vascular resistance in patients with severe septic shock (8) and does not limit skeletal muscle O₂ uptake in the hindlimb, despite decreased blood flow (25). By contrast, several studies have demonstrated a potentially detrimental effect of generalized NOS inhibition in sepsis. Thus, in porcine endotoxemia, L-NAME reduced venous return and increased left ventricular afterload, which could compromise cardiovascular function (26). Furthermore, NOS inhibition has decreased survival time in a murine model of sepsis (27). L-NAME has also been shown to impair tissue oxygen extraction in the canine mesenteric circulation (28). Further, in the same model rendered endotoxemic, NOS inhibition did not improve critical oxygen extraction ratios, despite improved hemodynamic performance (28). However, there are no data regarding the effects of generalized NOS inhibition on skeletal muscle tissue oxygenation. Our study demonstrates that L-NAME alone causes significant tissue hypoxia, possibly due to the accompanying fall in cardiac output and arterial oxygen tension. Alternatively, L-NAME may have increased shunt. However, this is unlikely, as we have previously demonstrated that both selective and nonselective NOS inhibition augment hypoxic pulmonary vasoconstriction in this model (29, 30). Not surprisingly, L-NAME had no beneficial effect on Pt_O2 in endotoxemic animals. One possibility is that the tissue hypoxia observed in endotoxemic animals was caused by impaired oxygen uptake, as discussed earlier, whereas the tissue hypoxia induced by L-NAME alone was caused by redistribution of blood flow. Moreover, administration of L-NAME together with volume resuscitation attenuated the beneficial effects of resuscitation alone on the endotoxemic-induced changes in tissue oxygenation described previously. By contrast, recent experiments have demonstrated that specific inhibition of iNOS, rather than generalized inhibition, improves mortality in animal models of endotoxemia (31, 32). These, together with our observations raise important questions about the potential clinical benefits of generalized NOS inhibition. Further experiments using specific iNOS inhibitors will be required to fully understand the role of NO in sepsis.

In summary, this study indicates that significant skeletal muscle hypoxia is a complication of experimental endotoxemia in the rat. In this model, impaired tissue microvascular control was present, and was not improved by increasing FI_O2. These changes were reversed by volume resuscitation, emphasizing the potential clinical importance of this maneuver. NOS inhibition alone induced a significant fall in Pt_O2 and did not improve tissue oxygenation in endotoxemia. Lastly, coadministration of volume resuscitation and NOS inhibition did not improve tissue oxygenation, indicating that the beneficial effects of fluid resuscitation described above were nullified by L-NAME. Further investigation into the effects of generalized NOS inhibition on tissue oxygenation is required to assess the overall clinical importance of this approach to treating patients with sepsis.