Organ shortage is a major obstacle to meeting the increasing demand for transplants. One strategy for expanding the donor pool is to procure and use organs from marginal donors. Organs procured from hemodynamically unstable cadaveric donors exposed to extended periods of high-dose dopamine (DA) or norepinephrine (NE) are usually considered to be inferior.1 Furthermore, at this time, there is no consensus with respect to the benefits of one pressor over the other, in part because few studies have studied the effects of catecholamines on hepatosplanchnic circulation and tissue oxygenation in clinically relevant models of brain death under controlled laboratory conditions. To fill this knowledge gap, we tested the hypothesis that splanchnic and systemic hemodynamics are improved with equi-effective pressor doses of DA versus NE in a well-described swine model.
We tested the hypothesis that hepatosplanchnic and systemic hemodynamics are improved with equi-effective doses of dopamine (DA) versus norepinephrine (NE) in a brain-dead swine model. Pigs (n = 18) were anesthetized and ventilated. Brain death was induced by epidural balloon inflation, hypoventilation, and hypoxia. After 30 minutes, mechanical ventilation was restored without anesthesia. During 60 and until 480 minutes, half received DA (10 μg/kg/minute) and half received NE (0.1 μg/kg/minute) titrated to a mean arterial pressure (MAP) > 60 mm Hg with supplemental fluid to maintain a central venous pressure > 8 mm Hg. Hemodynamics, hepatic laser Doppler blood flow, and hepatic and gastric tissue oxygenation with near-infrared spectroscopy were continuously monitored. Serial blood samples were analyzed for blood gases and electrolytes, coagulation changes, and serum chemistries. Balloon inflation caused brain death and autonomic storm, and 8 of 18 were nonsurvivors. After 30 minutes, the MAP, mixed venous O2 saturation, and partial pressure of arterial oxygen values decreased to 37 ± 2 mm Hg, 38 ± 4, and 49 ± 8 mm Hg, respectively. Serum lactate increased to 5.4 ± 0.7 mM. Among survivors (n = 10), MAP stabilized with either pressor. Urine output was maintained (>1 mL/kg/hour), but creatinine increased >30% with respect to the baseline. Tachyphylaxis developed with NE but not with DA (P < 0.05). Cardiac index was higher with DA versus NE (P < 0.05). There were no differences in stroke volume, metabolic indices, or liver blood flow. Liver tissue O2 was higher with DA versus NE at 8 hours (P < 0.05). Coagulation tests and liver enzymes were similar with NE versus DA (P > 0.05). In conclusion, after brain death, cardiac index and hepatic oxygenation were significantly improved with equi-effective doses of DA versus NE. Liver Transpl 14:1287–1293, 2008. © 2008 AASLD.
MATERIALS AND METHODS
Animals were housed in a facility that was approved by our Institutional Association for Assessment and Accreditation of Laboratory Animal Care with veterinarians available at all times. All procedures were performed according to the National Institutes of Health guidelines for the use of laboratory animals and were pre-approved. Animals were anesthetized during all surgical interventions, and all efforts were made to minimize the number of animals involved and to alleviate their pain and distress.
Farm-raised, crossbred, fasted swine of both sexes (43 ± 1 kg, n = 18) were sedated with an intramuscular injection of 30 mg/kg ketamine and 3.5 mg/kg xylazine. Anesthesia was induced with 2% isoflurane and then maintained with intravenous infusions of 10 mg/kg/hour ketamine, 0.5 mg/kg/hour xylazine, and 50 μg/kg/hour fentanyl. After orotracheal intubation, mechanical ventilation (model 754 impact portable adult ventilator, Impact Systems, West Caldwell, NJ) was instituted with tidal volumes of 10 mL/kg and 8 to 16 breaths/minute to maintain an end tidal carbon dioxide concentration of 40 ± 5 mm Hg and a fraction of inspired oxygen (FiO2) of 0.4. Pulse oximetry (pulse oximeter, Nellcor, Hayward, CA) and electrocardiogram were continuously monitored.
Catheters were placed via cutdown in the femoral artery for continuous arterial blood pressure monitoring (Agilent model 66 bedside hemodynamic monitor, Hewlett-Packard, Palo Alto, CA) and in the bilateral external jugular veins for fluid and drug administration. Additional catheters were placed in the pulmonary artery via the right internal jugular vein to measure mixed venous oxygen saturation, pressures, and cardiac output (Abbott Critical Care Systems, Abbott Laboratories, North Chicago, IL) and in the urinary bladder to measure urine output. Next, the stomach and liver were exposed through a midline laparotomy. Two laser Doppler blood flow probes (BPM 402, Vasomedics, Eden Prairie, MN) were sutured to the liver surface to measure microcirculatory blood flow, as we previously described.2 In addition, near-infrared spectroscopy probes were sutured to the liver surface and to the anterior gastric wall (InSpectra tissue spectrometer, Hutchinson Technologies, Inc., Hutchinson, MN) to measure tissue oxygen saturation, as we previously described.3, 4 The abdomen was then closed with towel clips. During these procedures, 1 L of crystalloid was administered to maintain vascular volume.
Finally, a 12-F balloon-tipped catheter (all-silicone Foley catheter, Bardex, Covington, GA) was introduced into the epidural space through a 1-cm craniotomy in the right frontoparietal region. After instrumentation, there was a 1-hour recovery period.
Baseline data were collected for 30 minutes, then the FiO2 was reduced to 0.2, the respiratory rate was reduced to 4/minute, the epidural balloon catheter was inflated with 10 mL of saline over 5 minutes, and the anesthetic infusion was stopped. After 30 minutes, resuscitation was initiated; by FiO2 being increased to 1.0, the respiratory rate was increased to maintain the end tidal carbon dioxide concentration between 35 and 45 mm Hg, the positive end expiratory pressure was set to 5 cm H20, and at least 1 L of crystalloid was administered. In those that responded to initial ventilator management and fluid resuscitation, pressor therapy was initiated after 30 minutes (60 minutes after balloon inflation). Half received DA (maximum dose, 10 μg/kg/minute) and the other half received NE (maximum dose, 0.1 μg/kg/minute) titrated to a target mean arterial pressure (MAP) > 60 mm Hg. Supplemental crystalloid was administered to maintain a target central venous pressure > 8 mm Hg. Data were collected for 480 minutes. The experiment was terminated by stopping of pressor infusion, which caused immediate cardiovascular collapse.
MAP, central venous pressure, pulmonary artery pressure, pulmonary capillary wedge pressure, mixed venous oxygen saturation, cardiac output, laser Doppler liver flow, and gastric and hepatic tissue oxygen saturation were measured continuously. At 30-minute intervals, arterial blood samples were drawn for analysis of blood gases, lactate, and electrolytes (Stat Profile Ultra, Nova, Waltham, MA). Urine output was measured hourly.
At baseline and 1, 3, and 8 hours after resuscitation, blood coagulation parameters were evaluated with thromboelastography (TEG hemostasis analyzer, Haemoscope, Niles, IL), as we previously described.4 Briefly, paired 2-mL whole blood samples were drawn from the femoral artery catheters at 5 mL/minute into 10-mL syringes at baseline and 3 time points after injury. A 360-μL aliquot from each sample was analyzed in duplicate at precisely 2 minutes on side-by-side bench-top instruments. Native whole blood samples were used. The 2 thromboelastography measurements were averaged to obtain 1 value at each time point.5
At baseline and 1, 3, and 8 hours after resuscitation, blood samples were drawn, and aspartate aminotransferase, alanine aminotransferase, blood urea nitrogen, total bilirubin, and creatinine were measured in the hospital pathology laboratory according to standard procedures.
There were 2 treatment groups; the pressors were administered in a randomized, blinded fashion. Data are expressed as mean ± standard error of the mean. The Student t test and linear mixed models were used to analyze the data. The linear mixed model was used to perform a repeated measures analysis of variance with planned comparisons between DA and NE at each time point. SAS 9.1 (SAS Institute, Inc., Cary, NC) was used for all analyses. A 2-tailed P value of <0.05 was considered significant.
Epidural balloon inflation provoked an immediate autonomic storm characterized by profound but transient tachycardia and hypertension (Table 1). Within 2 minutes, the heart rate and MAP doubled, but by 30 minutes, every animal that survived was in a moribund condition: liver blood flow and tissue O2 saturation were each reduced by half.
|Baseline||After 2 to 5 Minutes||After 30 Minutes|
|MAP, mm Hg||83 ± 1||160 ± 11||39 ± 2|
|Heart rate, beats/minute||71 ± 3||186 ± 11||99 ± 10|
|Cardiac index, mL/minute/kg||114 ± 6||131 ± 16||94 ± 12|
|SvO2, %||72 ± 1||66 ± 4||43 ± 4|
|Liver tissue O2, %||90 ± 1||67 ± 7||57 ± 5|
|Liver BF, mL/minute/100 g||23 ± 2||37 ± 8||9 ± 2|
|Gastric tissue O2, %||72 ± 2||57 ± 4||47 ± 4|
|Arterial pH||7.46 ± 0.02||7.23 ± 0.02|
|PaCO2, mm Hg||44 ± 3||71 ± 3|
|PaO2, mm Hg||182 ± 29||51 ± 13|
|Glucose, mg %||102 ± 7||230 ± 28|
|Lactate, mM||1.01 ± 0.23||5.54 ± 0.65|
After 30 minutes, 8 of 18 were either already dead or unresponsive to fluid. The remaining 10 animals all required pressors to maintain MAP and were brain-dead as characterized by bilateral dilated fixed pupils and lack of spontaneous respiratory effort in the absence of anesthesia.
Figure 1 shows there was no significant treatment-related difference for systemic pressures for any value at any time. Toward the end of the observation period, maintenance of the target MAP in the NE group required progressive increases in the infusion rate.
The left panel of Fig. 2 shows that the crystalloid required to maintain filling pressure averaged about 3.5 mL/kg/hour and that urine output averaged about 2 mL/kg/hour; these values were not different between treatments. The right panel shows that tachyphylaxis developed to NE but not to DA. The P values for the repeated measures analysis of variance main effects for dose and time as well as the interaction of dose and time were all below 0.05. The significant interaction indicates that the dosing rates of DA and NE were different over time. Once a threshold dose for DA had been established, the infusion rate remained relatively constant over the 480-minute observation period. In contrast, the infusion rate for NE was lower than that for DA from 70 to 250 minutes, from 250 minutes on. The difference in rates was supported by the significantly lower maximum dose percentage for NE compared to that for DA from 250 to 430 minutes.
Figure 3 shows the metabolic response to brain death and the effect of the 2 pressors. Inflation of the epidural balloon provoked an initial 2-fold increase in plasma glucose and a 5-fold increase in lactate. Over the next 480 minutes, there was a gradual recovery in glucose, and there was a partial recovery in lactate, but this was followed by a slow, steady rise. Plasma K+ progressively increased by about 1 mM by the end of the observation. The values with DA trended higher on each graph, but these apparent differences were not significant with respect to NE. Other electrolytes and hemotocrit remained within normal limits, so those data are not shown.
Figure 4 shows the effect of the 2 pressors on regional oxygenation. The left panel shows that the mixed venous oxygen saturation was virtually identical between treatments. The center panel shows that gastric tissue oxygenation was about 10% higher with DA versus NE at each time point after treatment, but the changes were variable in individual animals and did not reach statistical significance. The right panel shows that liver tissue oxygenation was 10% to 20% higher for all 8 hours with DA versus NE and reached statistical significance at 6 to 8 hours (P < 0.05). The changes in microcirculatory blood flow, as measured by laser Doppler, decreased by more than half within 30 minutes after brain death (see also Table 1) and generally showed a partial recovery with resuscitation. However, the values were highly variable between liver regions within and between animals, and the signal often became sporadic and unreliable. There was no obvious treatment effect, so those data are not shown.
Table 2 shows the changes in liver enzymes and serum chemistries after brain death. By 8 hours after brain death, there was no change in serum alanine aminotransferase, but there was a 50% to 100% increase in aspartate aminotransferase and total bilirubin, a 30% to 40% increase in plasma creatinine, and a 20% to 30% increase in blood urea nitrogen. Thromboelastogram coagulation parameters were within normal limits at each time point in each treatment group, so those data are not shown. Figure 5 shows that cardiac output was 20% to 50% higher with DA versus NE for at least 4 hours but was driven by tachycardia (for both P < 0.05). Stroke volume progressively fell during the observation period and did not significantly differ between treatment groups.
|Baseline||65 ± 15||58 ± 19|
|Brain death||63 ± 14||82 ± 25|
|3 hours||81 ± 21||87 ± 16|
|8 hours||89 ± 11||123 ± 35|
|Baseline||51 ± 6||60 ± 5|
|Brain death||51 ± 4||63 ± 6|
|3 hours||54 ± 5||63 ± 5|
|8 hours||53 ± 3||63 ± 5|
|Total bilirubin, mg %|
|Baseline||0.18 ± 0.04||0.18 ± 0.02|
|Brain death||0.22 ± 0.04||0.26 ± 0.04|
|3 hours||0.35 ± 0.03||0.34 ± 0.07|
|8 hours||0.30 ± 0.05||0.40 ± 0.11|
|BUN, mg %|
|Baseline||9.8 ± 1.9||8.6 ± 1.4|
|Brain death||10.0 ± 1.9||8.6 ± 1.1|
|3 hours||11.8 ± 1.4||8.6 ± 1.2|
|8 hours||12.2 ± 1.7||10.0 ± 1.6|
|Creatinine, mg %|
|Baseline||1.52 ± 0.16||1.70 ± 0.23|
|Brain death||1.56 ± 0.14||1.88 ± 0.20|
|3 hours||1.60 ± 0.11||1.70 ± 0.14|
|8 hours||2.16 ± 0.17||2.32 ± 0.21|
This study was designed to elucidate the effects of DA versus NE on the hepatosplanchnic circulation in the setting of hemodynamic instability induced by sudden brain death in swine. We found that DA significantly increased the cardiac index, heart rate, and hepatic tissue oxygenation in comparison with NE.
Rapid epidural balloon inflation produced a biphasic hemodynamic response observed in other studies.6–8 The serum lactate level, which is regarded as a reliable marker of anaerobic metabolism, was not different between groups. We were unable to demonstrate a difference between drugs with respect to renal and liver function tests or gastric tissue oxygenation. Our findings contradict those of Nakatani et al.,9 who concluded that DA has adverse effects on hepatic mitochondrial function,9 as well as other studies suggesting deleterious effects of DA in brain-dead animals.10–11 Our findings are consistent with those of Dougel et al.,12 who demonstrated a favorable effect of DA in reversing the hemodynamic perturbations observed in brain death in a dog model. We showed that NE produced tachyphylaxis in this animal model. However, further study is required to investigate the reasons for this phenomenon.
Our study has a number of limitations. We did not assess solid organ tissue oxygen consumption or blood flow, and the sample size was small. We did not measure the lactate/pyruvate ratio, which may be a more sensitive indicator of the metabolic state of the tissues and could have shed light on our findings.
In summary, these data showed that for an 8-hour period after brain death, DA, in comparison with NE, produced a higher cardiac index and heart rate, better hepatic tissue oxygenation, and no tachyphylaxis. However, further research is required to better characterize the benefits of DA, as well as other commonly used pressors, in the setting of the hemodynamically unstable cadaver donor.
We are grateful to Barbara Gallea (Hutchinson Technology, Hutchinson, MN) for providing the near-infrared probes and technical assistance with the near-infrared spectroscopy system. We also thank Mr. George Beck of Impact Instrumentation (West Caldwell, NJ) for providing the ventilators.