De Kock, Marc MD*; Laterre, Pierre-Francois MD*; Andruetto, Patricia MD†; Vanderessen, Lionel MD*; Dekrom, Sylvie MD*; Vanderick, Bernard MD*; Lavand’homme, Patricia MD*
Combined epidural/general anesthesia is used for patients undergoing major abdominal surgery (1). This technique, however, may cause arterial hypotension that may be corrected by intravascular fluid administration and by use of vasoconstrictors. Fluid administration, for example, may decrease the hematocrit and artificially increase the requirements of packed red blood cells. During recovery from anesthesia and surgery, excessive intravascular fluid volume may occur. Vasoconstrictor administration increases both cardiac and metabolic work because these drugs act by recruiting catecholaminergic receptors. For these reasons, there is clearly a need for alternative solutions.
Arginine vasopressin is a potent endogenous vasoconstrictor that increases blood pressure by direct action on specific receptors located on vascular smooth muscles. In contrast with the catecholaminergic vasoconstrictors, this is obtained without tachycardia and exaggerated metabolic cost. Moreover, experimental evidence supports the concept that the endogenous vasopressin system is the naturally occurring compensatory mechanism to restore arterial blood pressure when spinal sympathetic outflow is reduced by high epidural anesthesia (2). Our rationale was to test the efficacy and safety of ornipressin to counteract hypotension consecutive to epidural anesthesia. Ornipressin is a derivative of arginine vasopressin acting specifically on the V1 subtype receptors. In other words, it is a derivative that mimics the effects of arginine vasopressin on vascular smooth muscle; however, it has no antidiuretic effect. Ornipressin was compared with norepinephrine and dopamine, two catecholamines currently given to these types of patients. Gastric tonometry was used to evaluate the specific effect of this vasoconstrictor on splanchnic circulation.
After receiving institutional ethics committee approval and informed consent, we conducted a double-blinded, randomized study on 60 patients between 18 and 55 yr of age scheduled for extensive intestinal resection or second stage anastomosis. Exclusion criteria were cardiovascular or pulmonary diseases, chronic use of cardiovascular medications, renal or hepatic dysfunction, or a history of allergic reaction to any of the study drugs. On the night before surgery, all patients received 2 mg of lormetazepam and 150 mg of ranitidine orally. The same drugs were given 45 min before the procedure (lormetazepam orally and ranitidine IM).
A power analysis based on our pilot study under the present protocol revealed that 15 patients would provide a power greater than 0.8 (α = 0.05) for detection of differences (50% increase) in time to restore baseline arterial blood pressure (3). An epidural catheter was inserted in all patients at the T7-8 vertebral interspace by using the loss-of-resistance technique. The epidural catheter was threaded 5 cm into the epidural space. General anesthesia was induced by using approximately 2 mg/kg propofol IV titrated until loss of consciousness and 0.5 mg/kg atracurium. Thereafter, IV boluses of 0.1 μg/kg sufentanil and 1 mg/kg lidocaine were administered before tracheal intubation. Anesthesia was maintained with propofol infusion of 3 mg · kg−1 · h−1. Additional boluses of 0.5 mg/kg propofol were given to maintain a bispectral index between 55 and 65. Mechanical ventilation was adjusted to maintain an end-expiratory CO2 concentration of 35–36 mm Hg. During the induction of anesthesia, all of the patients received a fluid load of approximately 7 mL/kg crystalloids. During anesthesia administration, a basal fluid infusion of ±5 mL · kg−1 · h−1 was administered. Fluid supplementation (colloids) was given to compensate blood loss.
At 30 min after the induction of anesthesia and 20 min before surgical incision, patients received an epidural bolus of 7–8 mL of bupivacaine 0.5% with 2 μg/kg clonidine and 0.05 μg/kg sufentanil. It was immediately followed by an infusion of 5 mL of bupivacaine 0.06% with 0.5 μg · kg−1 · h−1 clonidine and 0.1 μg/h of sufentanil. This epidural infusion was discontinued in the recovery ward when a patient-controlled epidural analgesia programmed to deliver 5 mL/h of bupivacaine 0.06% with 3.5 μg/h of clonidine and 0.05 μg/h of sufentanil was started. Additional boluses of 5 mL of the solution could be delivered on demand every 50 min during the first three postoperative days.
In addition to routine monitoring (body temperature, oxygen saturation, twitch depression, and bispectral index), intraoperative monitoring included a 5-lead electrocardiograph with automated S-T segment analysis, an intraarterial catheter for systemic blood pressure monitoring, and a central venous catheter for venous pressure monitoring. To evaluate the effect of the different vasopressors on splanchnic perfusion, all patients had a gastric sump tonometer introduced nasogastrically, which provided the normal function of a nasogastric tube and facilitated measurement of the intracellular gastric mucosal Pco2 (Pico2). Measurement of Pico2 was achieved every 10 min with a tonometer (Tonocap TC-200; Datex-Engstrom, Helsinki, Finland) that allows simultaneous assessment of end-expiratory CO2 concentration. We performed Pico2 measurements in all of the patients from the induction of anesthesia until 2 h after the recovery. Oxygen consumption (V̇o2) was continuously recorded during anesthesia by using an analyser (Deltatrac™ MBM-100; Datex-Engstrom). This noninvasive monitoring of V̇o2 was performed to demonstrate that the Pico2 measurements were done in stable metabolic conditions.
Definition of hypotension requiring vasopressor therapy : After epidural injection, a decrease of 20% or more of baseline systolic arterial pressure (recorded after the induction during stable anesthesia) in a patient with a central venous pressure > 12 mm Hg determined the administration of one of the following vasopressors:
Group 1: dopamine started at 2 μg · kg−1 · min−1.
Group 2: norepinephrine started at 0.04 μg ·kg−1 · min−1.
Group 3: ornipressin (Por 8) started at 1 IU/h.
Patients were allocated to one of the treatment groups according to a computer-generated randomization list. The vasopressors were administered by a central line, where a continuous flow of 20 mL/h saline was achieved by using a calibrated infusion pump. Each vasopressor was started at the rate of 2 mL/h and infused by using a second calibrated infusion pump. In the case of no response after 5 min, the flow rate was doubled and consequently, so was the dose of vasopressor. In the case of a poor or intermediate response after an additional 5 min, the flow of vasopressor was increased by 2 mL. Return to baseline values (recorded after the induction during stable anesthesia) was considered as an adequate response. The time elapsed between vasopressor initiation and return to baseline value was used as a criterion to evaluate the efficacy of the studied drugs. Patients were allocated to one of the different vasopressors according to a computer-generated randomization list. The first 15 patients presenting without a decrease in systolic arterial blood pressure were used as control subjects (Group 4). Patients developing hypotension with central venous pressure < 12 mm Hg received first, fluid supplementation. In the case of return to baseline systolic arterial blood pressure, patients otherwise allocated to receive one of the studied vasopressors were used as control subjects. After 90 min, the vasopressor was discontinued and time to reach the 20% decrease of baseline value was recorded. When this criterion was met, the vasopressor was reintroduced. The vasopressor was discontinued after patients’ recovery in all cases.
Study solutions were prepared by an anesthesiologist not involved in patient care, and both the patient and the anesthesiologist who delivered anesthesia were blinded to the study solutions. Samples were retrieved to analyze arterial blood gas and lactate before, during, and at the end of vasoconstrictor administration. Patient temperature was maintained by using a warming blanket, fluid, and a forced-air warmer.
Levels of thermoanalgesia were assessed by using an ether swab 1 h after the recovery. The postoperative follow-up of patients was recorded during the first 3 mo, and all of the complications were noted.
Comparisons of independent variables and time elapsed before restoring baseline arterial blood pressure were based on analysis of variance. Observed proportions were compared by using χ2 and Fisher’s exact test with Yates correction when appropriate. Variables over time were evaluated statistically by using two-way analysis of variance with repeated measures. Post hoc comparisons were made by using Tukey’s test (CSS Statitica; Statsoft, Tulsa, OK). Homogeneity of the variance of the data was assessed by using the Levene’s test. A P < 0.05 was considered significant.
Demographic data of the patients are summarized in Table 1. Epidural catheters were placed easily and successfully on the first attempt in all patients. The surgical procedures were simple, and no patient had to receive a blood transfusion during surgery or during the first postoperative days.
The 15 control group patients (Group 4, patients without hypotension) were recruited before the other groups were completed. The results of further patients presenting without hypotension after the 15 initial subjects (seven patients) were not considered in this study.
All of the vasopressors tested were able to restore baseline systolic arterial blood pressure (Fig. 1). It took 11 ± 3 min in Group 1 (dopamine) versus 7 ± 3 and 8 ± 2 min in Groups 2 and 3 (norepinephrine and ornipressin, respectively) (P < 0.05, dopamine vs norepinephrine and ornipressin). In Group 1, the majority of patients (9 of 15) required 6 μg · kg−1 · min−1 dopamine to restore baseline arterial blood pressure (2 μg · kg · min was sufficient in one patient and 4 μg · kg−1 · min−1 was adequate in four patients). In Group 2, 13 patients required 0.04 μg · kg−1 · min−1 norepinephrine, and 2 patients needed 0.08 μg · kg−1 · min−1 norepinephrine to restore baseline. In Group 3, 2 IU/h of ornipressin was sufficient to restore baseline in the majority of the patients (11 of 15). Only one patient required 1 IU/h of ornipressin, and two needed 3 IU/h of ornipressin to restore baseline. Criteria for vasopressor reintroduction was a decrease of 20% or more of the baseline systolic arterial pressure recorded after the induction during stable anesthesia and before epidural injection. After discontinuation of the vasopressors, time elapsed to reach the criteria for vasopressor reintroduction was 15 ± 4 min in patients of Group 1 (dopamine), 15 ± 4 min in patients of Group 2 (norepinephrine), and 26 ± 8 min in patients of Group 3 (ornipressin) (P < 0.05). After recovery, all of the patients were weaned from the vasopressor infusions during the first 30 min.
Dopamine (Group 1) significantly increased heart rate when compared with norepinephrine (Group 2) and ornipressin (Group 3) (P < 0.05). Ornipressin did not induce significant bradycardia (Fig. 1). No S-T segment changes indicating cardiac ischemia were recorded in any patient.
During anesthesia administration and surgery, Pico2 significantly increased in patients who received ornipressin (Group 3, P < 0.05, Fig. 2). This occurred despite the absence of any modification of the end-expiratory CO2 concentrations or major modification of the alveolar-arterial CO2 gradient (Table 2). An increase in Pico2 was also noted in patients of the norepinephrine group; however, the increase did not reach statistical significance. Moreover, during the administration of anesthesia, when vasopressors were discontinued, a significant decrease in Pico2 was observed in Group 3 (P < 0.05) (Fig. 2). At the recovery, when the vasopressors were stopped, no intergroup difference was noted in the intracellular gastric/end-expiratory Pco2 gradient (Fig. 3). This gradient, however, was significantly modified during this period (P < 0.05).
During anesthesia administration, arterial lactate significantly increased in Groups 1 (dopamine) and 2 (norepinephrine). Lactate concentrations, however, remained in the normal range. Measurements of oxygen consumption during anesthesia are presented in Figure 2. The administration of dopamine significantly increased intraoperative oxygen consumption (P < 0.05).
Perioperative fluid administration is presented in Table 3. None of the patients required packed red blood cells or albumin.
At recovery, an adequate level of thermoanalgesia (±T5) was measured in all of the patients by using an ether swab. There was no intergroup difference. Postoperative analgesia provided by continuous epidural infusion was satisfactory in all patients.
Only seven patients required postoperative care. A fistula was detected in four patients (one in Group 1, one in Group 2, and two in Group 4), and a second surgical intervention was required in only one of these four patients. Two patients (one in Group 2 and one in Group 3) suffered septic complications originating in the urinary tract. One patient in Group 1 presented with late (>48 h) postoperative atrial fibrillation. At 3 mo after the surgical procedure, no anastomotic stricture was recorded for any patient.
Our results show that ornipressin restores arterial blood pressure in patients undergoing intestinal resection for inflammatory bowel disease under combined general and thoracic epidural anesthesia. This observation may not appear particularly important because the potent vasoconstrictor effects of the vasopressin derivatives are evident in various clinical situations (4,5). More interesting, however, is the dosage required to obtain this effect. Relatively small doses are sufficient when compared with the dose used to prevent the recurrence of bleeding esophageal varices (from 4 to 16 IU/h). In fact, this observation supports the experimental work done by Peters et al. (2). By using a model of conscious dogs, these authors demonstrated that the endogenous vasopressin system is the naturally occurring compensatory mechanism to restore arterial blood pressure when spinal sympathetic outflow is markedly reduced by high epidural anesthesia. Exogenous vasopressin may be used to treat hypotension caused by high epidural block because opioids and α2-adrenoceptor agonists, two classes of drugs currently used during the administration of anesthesia and epidural analgesia, inhibit endogenous vasopressin release (6). Consequently, the administration of a vasopressin analog may mimic the physiologic adaptation to sympathetic block induced by epidural anesthesia.
Moreover, a clinical observation by Landry et al. (7) highlights the efficacy and relative safety of exogenous small-dose (±2 IU/h) vasopressin administration in critically ill patients. These authors reported a significant increase in arterial blood pressure and urine output in five patients suffering vasodilatory septic shock refractory to large-doses of norepinephrine and epinephrine. This was obtained without any evidence of cardiac or splanchnic ischemia or a significant decrease in cardiac output. This sensitivity to vasopressin in septic patients with blunted vasopressor response to norepinephrine is in agreement with experimental work in the rat (8). Another clinical situation where specific sensitivity to small-dose exogenous vasopressin has been described is primary autonomic failure (9).
These observations prompted us to test the efficacy of a small-dose vasopressin derivative in a situation where the sympathetic system is impaired, i.e., the sympathetic block induced by epidural anesthesia, particularly when local anesthetics are combined with the α2-adrenoceptor agonist clonidine. This substance, administered by the epidural route, induces direct local and systemic sympatholytic effects (10). In our population, as observed by Landry et al. (7), none of the 15 patients presented with any sign of cardiac ischemia.
Another interesting observation from our study is the demonstration of the specific effect of ornipressin on splanchnic perfusion by using the gastric tonometry technique. The tonometer consists of a gas-permeable silicone balloon at one end of a gas-impermeable sampling tube. When inserted into the stomach, the balloon is automatically filled with air, and the CO2 in the gastric mucosa (Pico2) diffuses into, and equilibrates with, the air-filled balloon. Every 10 minutes, the gas contained in the balloon is analyzed and compared with the expiratory Pco2. Any increase in Pico2 associated with an increase of the gradient between Pico2 and the end-expiratory CO2 concentration or CO2 gap, indicates splanchnic hypoperfusion.
In critically ill patients, gastric intramucosal pH, a variable derived from gastric tonometry, was reported to be a sensitive prognostic indicator of outcome, and failure to correct this variable significantly worsened the patient prognosis (11). In healthy subjects, the value of gastric tonometry as an early monitor of hypovolemia has been demonstrated by Hamilton-Davies et al. (12). A progressive blood loss of 25% of the blood volume in humans resulted in a significant increase in Pico2 before a decrease in stroke volume measured by suprasternal Doppler, and changes in invasive arterial blood pressure, heart rate, or lactate levels. This observation clearly shows the accuracy of gastrointestinal tonometry in detecting reduced splanchnic perfusion. Interestingly, it has been shown that a 15% decrease in total blood volume results in a reduction of approximately 40% in splanchnic blood volume with no change in any of the commonly measured cardiovascular variables, such as blood pressure or heart rate (13). In our study, at the start of the Pico2 measurements, absolutely no gradient was detected, which indicates that none of the patients was hypovolemic. According to the values recorded in the control group, epidural anesthesia did not induce any important change of this variable. In contrast, small-dose ornipressin significantly increased Pico2. Nevertheless, the increase in CO2 gradient remained in the upper limit of the normal range and was rapidly reversible after discontinuation of the drug. This may explain the mechanism by which ornipressin restores arterial blood pressure in the clinical situations studied. Ornipressin, at the dose used, may induce a specific vasoconstriction of the splanchnic capacitance vessels that are dilated by epidural and general anesthesia. Vasopressin derivatives reduce bleeding in acute hemorrhagic gastritis by specific V1 receptor-mediated vasoconstriction of the gastric vascular bed.(14). A significant increase in intramucosal Pico2 and in CO2 gap may be caused by either failure to clear CO2 caused by gastric mucosal hypoperfusion or locally increased production of CO2 via anaerobic mechanisms.
We suggest that ornipressin caused splanchnic vasoconstriction, thus reducing CO2 clearance in the gastrointestinal mucosa because the patients were metabolically stable without lactic acidosis. Interestingly, norepinephrine a nonspecific arterial vasoconstrictor also tended to increase Pico2; however, this tendency did not reach statistical significance. After anesthesia and discontinuation of the vasoconstrictors, during the first 1.5 h of the recovery period, the CO2 gap significantly increased in all groups. This transient increase may be caused by the relative hypovolemia after the discontinuation of vasoconstrictors or metabolic disturbances associated with emergence of anesthesia or cold stress.
In the current study, ornipressin was compared with norepinephrine, an α-adrenergic agonist usually administered in case of important vasodilation, and to dopamine, a mixed β- and α-adrenergic agonist. Dopamine also stimulates dopamine receptors in the splanchnic circulation, causing vasodilation. During anesthesia, when the vasopressors were discontinued, arterial blood pressure decreased much slower in patients who received ornipressin. This significant difference is related to the half-life of the study drugs.
In conclusion, small-dose ornipressin restores arterial blood pressure in patients receiving combined general and thoracic epidural anesthesia because of splanchnic vasoconstriction, without major ischemic side effects, tachycardia, or lactate production. Nevertheless, our results cannot ascertain the safety of small-dose ornipressin in patients suffering from coronary artery disease.
1. Liu SS, Carpenter RL, Mackey DC, et al. Effects of perioperative analgesic technique on rate of recovery after colon surgery. Anesthesiology 1995; 83:757–65.
2. Peters J, Schlaghecke R, Thouet H, Arndt J. Endogenous vasopressin supports blood pressure and prevents severe hypotension during epidural anesthesia in conscious dogs. Anesthesiology 1990; 73:694–702.
3. Fisher LD, van Belle G. Sample sizes for observational studies. In: Fisher LD, van Belle G, eds. Biostatistics: a methodology for health science. 1st ed. New York, NY: Wiley-Interscience Publishers, 1993: 644–54.
4. Eyraud D, Brabant S, Nathalie D, et al. Treatment of intraoperative refractory hypotension with terlipressin in patients chronically treated with an antagonist of the renin-angiotensin system. Anesth Analg 1999; 88:980–4.
5. Overand PT, Teply JF. Vasopressin for the treatment of refractory hypotension after cardiopulmonary bypass. Anesth Analg 1998; 86:1207–9.
6. Jackson EK. Vasopressin and other agents affecting the renal conservation of water. In: Hardman JG, Limbird LE, eds. Goodman & Gilman’s the pharmacological basis of therapeutics. 9th ed. New York: McGraw-Hill, 1996: 721.
7. Landry DW, Levin HR, Gallant EM, et al. Vasopressin pressor hypersensitivity in vasodilatory septic shock. Crit Care Med 1997; 25:1279–82.
8. Baker CH, Sutton ET, Zhou Z, et al. Microvascular vasopressin effects during endotoxin shock in the rat. Circ Shock 1990; 30:81–95.
9. Wagner HN Jr, Braunwald E. The pressor effect of the antidiuretic principle of the posterior pituitary in orthostatic hypotension. J Clin Invest 1956; 35:1412–18.
10. Eisenach JC, De Kock M, Klimsha W. α2-Adrenergic agonist for regional anesthesia. Anesthesiology 1996; 85:655–74.
11. Doglio GR, Pusajo JF, Egurrola MA. Gastric mucosa pH as a prognostic index of mortality in critically ill patients. Crit Care Med 1991; 19:1226–33.
12. Hamilton-Davies C, Mythen MG, Salmon JB, et al. Comparison of commonly used clinical indicators of hypovolemia with gastrointestinal tonometry. Intensive Care Med 1997; 23:276–81.
13. Price HL, Deutsch S, Marshall BE. Hemodynamic and metabolic effects of hemorrhage in man with particular reference to the splanchnic circulation. Circ Res 1966; 18:469–74.
14. Peterson W. Gastrointestinal bleeding. In: Sleisenger MH, Fordtran JS, eds. Gastrointestinal disease. Philadelphia: WB Saunders, 1988: 397–449.