INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Inadequate splanchnic blood flow (spl) is regarded to play a pivotal role in the development of multiple organ failure (MOF) associated with septic shock (1). Recently, we showed that exogenous -stimulation is crucial for the maintenance of spl and oxygen exchange (6). Furthermore, the synthetic -mimetic dobutamine induced an increase in spl (7, 8), thereby disclosing pathologic regional O₂ uptake/supply dependency (8). Dobutamine, however, mainly increases spl in close correlation to cardiac index (8, 9). Moreover, it did not prevent alterations of hepatic ultrastructure in hyperdynamic porcine septic shock induced by fecal peritonitis (10).

By contrast, the combined dopamine and -adrenergic agonist dopexamine not only increased blood flow to the splanchnic area in patients with congestive heart failure (11) but also improved parameters of intestinal and hepatic oxygenation and function in the critically ill (12, 13). Moreover, it improved gastric mucosal capillary oxygenation in human septic shock (14). Finally, in the above-mentioned porcine shock model dopexamine only maintained hepatic ultrastructure (10), and in rats improving oxygen delivery with dopexamine attenuated the endotoxin-induced alterations of the microcirculation (15).

Therefore, the goal of this study was to investigate the hypothesis whether dopexamine improves hepatosplanchnic blood flow and thereby parameters of regional energy balance in patients with hyperdynamic septic shock.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The study was approved by the ethics committee of the Ulm University Medical School. Twelve patients with septic shock were studied, all requiring noradrenaline to maintain mean arterial pressure > 60 mm Hg (noradrenaline infusion rate  0.04 µg · kg¹ · min¹). The patients fulfilled the following criteria: (1) age between 18 to 70 yr; (2) cardiac index  3.0 L/min/m²; (3) temperature  38° C or  36° C; (4) leukocytes  4,000 or  12,000 10⁹/L.

All studies were accomplished during volume-controlled mechanical ventilation (Servo 900 C; Siemens AG, Solna, Sweden), and the patients were deeply sedated and given analgesic intravenously with continuous midazolam (Dormicum; Hoffmann LaRoche AG, Basel, Switzerland) and fentanyl (Janssen, Neuss, Germany) and relaxed with cisatracurium (Nimbex; Glaxo Wellcome, Bad Oldesloe, Germany). During the protocol the patients were not fed enterally, all intravenous fluids were kept at maintenance infusion rates, and no red blood cells were given. Turning or other nursing procedures were prohibited in order to avoid manipulation-induced variations of global or splanchnic oxygen exchange.

In addition to routine monitoring (radial and pulmonary artery catheters), a balloon-tipped catheter was inserted into one hepatic vein. For this purpose the hepatic vein was cannulated via the right internal jugular vein, and the correct position of the catheter verified before and after the study by fluoroscopy using a small amount of contrast dye. A nasogastric tube (TRIP NGS Catheter; Tonometrics, Inc., Worcester, MA) was inserted in the stomach (16). Correct position was confirmed by X-ray.

Regional Hemodynamic Measurements

The total hepatosplanchnic blood flow was estimated using a primed, continuous infusion of indocyanine green (ICG), based on the Fick principle and hepatic venous sampling (17). The method was previously evaluated in intensive care patients and described in detail (17, 18).

Briefly, after a prime injection of ICG (12 mg; Cardiogreen, Becton-Dickinson Microbiology Systems, Cockeysville, MD), the dye was infused at a constant rate (0.5 mg/min) into a peripheral vein. At 20, 25, and 30 min of infusion, blood samples were taken from the hepatic venous an the arterial catheter for the analysis of ICG levels and subsequent estimation of hepatosplanchnic blood flow. The mean of the three blood flow values was taken for calculations. The arterial and hepatic venous ICG concentrations were in a steady-state plateau at each measurement as indicated by the coefficient of variation of 1 ± 3% (mean ± SD) for the arterial and the hepatic vein dye concentrations for the consecutive samples taken at 20, 25, and 30 min of each measurement. The mean ICG extraction of our patients was 35 ± 18% (mean ± SD), exceeding the limit of 10% in 35 of 36 measurements, which is required for valid application of this method (19). During the second baseline only in Patient 11 the extraction was 8%.

The balloon-tipped catheter positioned in a hepatic vein enabled us to continuously determine the hepatic venous pressure (Phv) and intermittently, by inflating the balloon, the hepatic venous occlusion pressure (Phvo). Phvo is regarded to yield a close estimate of the portal venous pressure (20, 21). According to the following formulas we then calculated the resistances across the splanchnic vascular bed. Total splanchnic resistance = (  Phv)/spl; prehepatic splanchnic resistance = (  Phvo)/spl; hepatic resistance = (Phvo  Phv)/spl where = mean systemic arterial pressure.

Regional oxygen delivery (DO₂) was calculated as a product of spl and arterial oxygen content (Ca_O2). Regional oxygen extraction was calculated as arterial-hepatic venous oxygen content difference/Ca_O2, and the regional O₂ as the product of spl times the arterial-hepatic venous oxygen content difference.

The gastric mucosal PCO₂ was measured semicontinuously (10-min intervals) via the nasogastric tube with a Tonocap (Tonocap TC 200; Baxter, München, Germany) using air to inflate the balloon.

Other Measurements

Systemic and pulmonary vascular pressures were continuously measured via an arterial catheter and a Swan-Ganz catheter inserted in the pulmonary artery via the vena subclavia or vena jugularis interna. Systemic oxygen consumption (O₂) was measured continuously from the inspired and expired respiratory gases by open-circuit indirect calorimetry (Deltatrac; Datex-Engstroem, Helsinki, Finland) (22). The oxygen extraction was calculated as the arterial-mixed venous oxygen content difference divided by Ca_O2. Cardiac output () was measured continuously (Vigilance. Baxter, München, Germany) or by thermodilution in triplicate using 10 ml of room temperature saline after 60 and 90 min of each phase of the study.

Blood gases were sampled at the 30 min timepoint of ICG sampling and blood gases and hemoglobin O₂ saturations were measured using a clinical blood gas analyzer (Nova Stat Profile M; Nova Biomedical, Rödermark, Germany). Oxygen contents were calculated as: Hb · SO₂ · 1.36 ± PO₂ · 0.03.

Protocol

Before the baseline measurement of gas exchange, systemic and regional blood flow, global hemodynamics, and gas exchange were monitored for a 60-min period to assure a stable condition. During this period, no changes were made in the ventilator settings or any other treatment. During the following 30 min, the hepatosplanchnic blood flow was measured using the ICG infusion method (17).

After the baseline measurement a dopexamine infusion was started, and the infusion rate was incrementally adjusted until a 25% (20 to 30%) increase in cardiac index had been obtained. To obtain this goal, 1-4 µg · kg¹ · min¹ were required. After 60 min of stabilization, the regional blood flow measurement was repeated during a 30-min period, and a second set of data was collected. Then the dopexamine infusion was withdrawn. After another 90 min of stabilization, the third set of regional blood flow measurement and collection of data was performed. The total duration of the study, hence, was 280 to 300 min depending on the time to titrate the infusion rate of dopexamine.

Statistical Methods

Data are presented as median and 25th-75th percentiles. After exclusion of normal distribution the differences between the measurements were analyzed by the Friedman rank sign analysis of variance and a subsequent Wilcoxon-Wilcox rank sign test for multiple comparisons. Statistical significance was considered at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Global Hemodynamics

The increased cardiac index was due to both a rise in heart rate (90/100-87 versus 116/128-110 min, p < 0.01) and, to a smaller extent, to an increased stroke volume index (45/49-40 versus 51/69-48 ml/m², p < 0.01) Neither central venous pressure (Pcv) nor pulmonary vascular pressures were affected by dopexamine (Table 1). The increased cardiac index resulted in increased systemic DO₂ (537/592-465 versus 636/795-540 ml/ min/m², p < 0.01) despite a decrease in Pa_O2 (12.8/14.4-10.7 versus 11.4/12.3-10.1 kPa, p < 0.01) whereas mean systemic O₂ remained unaffected (Table 2). In five patients we could find an increase in body temperature by 0.5 to 0.8° C during dopexamine infusion. The changes in temperature were not correlated to changes in O₂.

Regional Hemodynamics

The increased cardiac index resulted in a significant increase in spl (0.86/1.23-0.66 versus 0.96/1.42-0.85 L/min/m², p < 0.01) (Table 1). The fractional contribution of spl to cardiac index (21/28-13 versus 19/28-12%, p < 0.05) decreased in all but one patient (Figure 1). Neither mean Phv nor Phvo was influenced by dopexamine (Table 1). Although the driving pressures of venous return were not significantly affected, both the total and prehepatic resistances significantly decreased without significant effect on hepatic resistance (Table 1). Splanchnic DO₂ increased in all patients (Figure 2, upper panel) but, similar to systemic O₂, mean splanchnic O₂ was not significantly affected (Table 2) although there was profound interindividual variation (Figure 2, lower panel).

View larger version (24K): [in this window] [in a new window]   Figure 1.   Fractional contribution of spl to cardiac index.

View larger version (21K): [in this window] [in a new window]   Figure 2.   (Upper panel  ) Individual changes in splanchnic oxygen delivery (splDO2) after a dopexamine-induced increase in cardiac index. (Lower panel  ) Individual changes in splanchnic oxygen consumption (spl O2) after a dopexamine-induced increase in cardiac index.

Neither the mean gastric mucosal-arterial PCO₂ gradient nor the gastric mucosal-hepatic venous PCO₂ gradient was significantly altered (Table 3) although again there was substantial interindividual variability (Figure 3). One patient had to be excluded from the analysis of PCO₂ gradients for technical reasons.

View larger version (20K): [in this window] [in a new window]   Figure 3.   (Upper panel  ) Individual changes in gastric mucosal-arterial PCO2 gradients after a dopexamine-induced increase in cardiac index. (Lower panel  ) Individual changes in gastric mucosal-hepatic venous PCO2 gradients after a dopexamine-induced increase in cardiac index.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study investigated the hypothesis whether dopexamine may improve hepatosplanchnic hemodynamics, oxygen transport, and parameters of regional energy balance in patients with hyperdynamic septic shock requiring noradrenaline to maintain mean arterial blood pressure. The key findings are as follows: the dopexamine-induced increase in resulted in a concomitant increase in spl; the fractional contribution of spl to decreased; and dopexamine did not consistently influence regional oxygen exchange nor the tonometric PCO₂ gradient.

In good agreement to previous studies (11, 23, 24) dopexamine consistently increased both and spl. In most studies such an increase in spl mostly paralleled the increase in systemic blood flow (12, 13). It is noteworthy that the fractional contribution of the regional blood flow to total was even reduced during dopexamine infusion in our patients (Figure 1). Although this redistribution of blood flow was statistically significant, this finding probably lacks clinical significance: the decrease in percent contribution of spl to was only modest (from 21/28-13 to 19/28-12) and the absolute regional blood flow nevertheless increased as a result of the pronounced rise in . It should be noted, however, that Cain and Curtis (24) had also found a decreased fractional blood flow to the gut 3 h after the start of a combined dextran and dopexamine infusion in endotoxic dogs. We can only speculate about the discrepancy between our findings and the previous human studies, but both the underlying disease and the treatment may have assumed importance in this context: In contrast to the investigations mentioned previously, our patients received ongoing noradrenaline intravenously to maintain for 24 to 48 h prior to the study. Modulation of the -adrenergic receptor density and affinity resulting from both the noradrenaline treatment and the underlying sepsis per se (25) may have altered hemodynamic response to the dopexamine infusion.

Although we could not demonstrate any preferential redistribution of to the splanchnic region, we found clear effects of dopexamine within the splanchnic region. Measuring Phvo and Phv enabled us to calculate the prehepatic and hepatic resistance. Despite inconsistent changes in the driving pressures of venous return, the splanchnic vascular resistance significantly decreased obviously partly as a result of a significant decrease in prehepatic resistance (Table 1). Mean hepatic resistance, by contrast, was not affected (Table 1). The main effect of dopexamine within the splanchnic region, hence, seems to be the prehepatic area, a finding which is consistent with the pharmacologic profile of dopexamine (28) as a combined ₂- and dopamine receptor agonist as well as the particular distribution of ₂-receptors within the splanchnic region (29).

In all patients, dopexamine increased both systemic and splanchnic DO₂, while neither mean systemic nor splanchnic O₂ significantly changed. The findings of our study, therefore, do not allow disclosure of a pathologic splanchnic oxygen uptake/supply relationship. Nevertheless, it should be noted that three patients exhibited a steep increase in splanchnic O₂ concomitant with the rise in splanchnic DO₂ (Figure 2). In contrast to previous data (8) the small number of patients in our study together with a potential mathematical coupling of shared variables, namely splanchnic DO₂ and O₂ (30), however, precludes any general conclusion with respect to regional DO₂/O₂ relationships.

Dopexamine infusion did not significantly change the mean gastric mucosal-arterial PCO₂ gradient (Figure 3). Even in those patients in whom the PCO₂ gradient exceeded the threshold that Schlichtig and Bowles (31) had determined as being compatible with aerobic metabolism, e.g., a gradient of 25 mm Hg, the effect of dopexamine did not produce a consistent response. The gastric mucosal-hepatic venous PCO₂ gradient did not show a homogenous response either (Figure 3), and variations were mostly caused by changes in the gastric mucosal PCO₂. Finally, the changes in the PCO₂ gradients were not related to variations in splanchnic O₂ (Figure 4). Such a dissociation between changes in PCO₂ gradients on the one hand and regional blood flow, oxygen exchange, and/or intracapillary oxygenation on the other hand has been previously reported during dopexamine infusion by other investigators (14, 32): Temmesfeld-Wollbrück and coworkers (14) could not find a clear change in CO₂ gradients despite a significant increase in gastric mucosal intracapillary O₂ saturation and relative hemoglobin concentration as assessed by remission spectrophotometry. Uusaro and coworkers (32) did not find improved PCO₂ gradients either when spl was increased by dopexamine in patients after coronary artery bypass graft surgery.

View larger version (12K): [in this window] [in a new window]   Figure 4.   Correlation between splanchnic oxygen consumption (spl O2) and gastric mucosal-arterial PCO2 gradients after dopexamine infusion.

In conclusion, in patients with hyperdynamic septic shock who had stabilized with volume infusion and noradrenaline, infusing dopexamine to further increase resulted in increased spl. A preferential effect, however, on the splanchnic circulation with increased fractional contribution of the regional blood flow to could not be detected.