Anesth Analg 2003;97:1573-1577
© 2003 International Anesthesia Research Society
CARDIOVASCULAR ANESTHESIA
The Effects of Desflurane and Propofol on Portosystemic Pressure in Patients with Portal Hypertension
M. Susan Mandell, MD PhD*,
Janette Durham, MD
,
David Kumpe, MD,
James F. Trotter, MD
,
Gregory T. Everson, MD, and
Claus U. Niemann, MD
Departments of *Anesthesiology, Radiology, and Hepatology, University of Colorado Health Sciences Center, Denver; and Department of Anesthesia and Perioperative Care, University of California, San Francisco
Address correspondence and reprint requests to M. Susan Mandell, MD, PhD, Department of Anesthesiology, University of Colorado Health Sciences Center, Campus B-113, 4200 E. Ninth Ave., Denver, Colorado 80262. Address e-mail to Susan.Mandell{at}UCHSC.edu
 |
Abstract
|
|---|
Physicians perform hepatic venous pressure measurements to guide medical therapy aimed at reducing portal hypertension. These measurements are frequently performed during general anesthesia. Since most general anesthetic drugs reduce liver blood flow, it is likely that hepatic venous pressures will be altered. We therefore examined the effects of two frequently used anesthetic drugs on hepatic venous pressure in a prospective randomized study to determine if pressure measurements taken during general anesthesia were similar to awake values. We studied 21 patients with hepatitis C, excluding patients with hepatofugal flow and portal vein thrombosis. All patients had free and wedged hepatic venous pressures measured awake with sedation and after anesthesia with either propofol or desflurane. Desflurane significantly increased free hepatic venous pressure (11.9 ± 4.4 to 23.5 ± 4.1 mm Hg; P < 0.05) and decreased hepatic venous pressure gradient (21.6 ± 7.4 to 14.7 ± 5.2 mm Hg; P < 0.05), whereas propofol did not change these variables. We conclude that desflurane, but not propofol, alters hepatic venous pressure measurements from the awake state, significantly increasing free hepatic venous pressure and decreasing the hepatic venous pressure gradient, an indirect measure of portosystemic pressure. Changes in the hepatic venous pressure gradient must be interpreted with caution during desflurane general anesthesia.
IMPLICATIONS: Desflurane reduces the blood pressure difference between the portal and systemic circulations. This can cause errors in assessment of the success of medical therapy of portal hypertension. Propofol has less effect on the difference between the portal and systemic circulation.
|
Introduction
|
|---|
Complications of portal hypertension are controlled and prevented by decreasing the pressure gradient between the portal and systemic venous circulations (1,2). When patients fail medical therapy, physicians often insert a transjugular intrahepatic portosystemic shunt (TIPS) to treat complications of portal hypertension (3,4). Since portal pressure is not measured directly during medical therapy or TIPS insertion, the hepatic venous pressure gradient (HVPG) is used as an indirect measure of the portosystemic gradient (5,6). The HVPG is the measured difference between the wedged hepatic venous pressure (WHVP), an estimate of portal venous pressure, and the free hepatic venous pressure (FHVP), an estimate of systemic venous pressure (6). The WHVP is a reliable surrogate of portal pressure in normal and most cirrhotic patients, so that the HVPG serves as an accurate measure of the portosystemic pressure gradient (1,6).
The sedative midazolam provides comfort during measurement of hepatic venous pressure but is limited, because doses larger than 0.02 mg/kg increase the FHVP and erroneously reduce the HPVG (7). Many chronically ill patients do not tolerate the discomfort associated with pressure measurement or TIPS insertion. Because movement and agitation during the procedure is a safety hazard for the patient and the radiologist, general anesthesia is sometimes required. However, all inhaled anesthetic drugs decrease total hepatic blood flow in normal study subjects (8) by a dose-dependent reduction in vascular resistance of the portal and splanchnic circulation (9). In contrast, the sedative and anesthetic drug propofol may increase hepatic blood flow in experimental study subjects (10,11). The mathematical relationship between pressure and flow suggests that these drugs may also affect hepatic venous pressure. We therefore designed this study to determine if two frequently used general anesthetic drugs, desflurane and propofol, affect hepatic venous pressure.
|
Methods
|
|---|
IRB approval was obtained. After informed consent, we prospectively studied 24 liver transplant candidates with biopsy-proven cirrhosis scheduled to undergo a TIPS procedure for refractory ascites or recurrent upper gastrointestinal bleeding. None of the patients had clinically significant encephalopathy.
All procedures were nonemergent in patients who required routine large volume paracentesis, despite aggressive pharmacologic therapy for ascites or patients who experienced recurrent gastrointestinal bleeding after multiple variceal banding. All patients had a diagnosis of hepatitis C without additional cause of liver disease. Patients with portal vein thrombosis and hepatofugal blood flow diagnosed by abdominal Doppler ultrasound were not entered into the study.
All patients had monitoring with 5-lead electrocardiogram, noninvasive blood pressure, axillary temperature, and end-tidal CO2 through either nasal cannula or endotracheal tube and pulse oximetry.
All study subjects received 3–5 mL/kg of 5% albumin before the study followed by a crystalloid infusion at 50 mL/h. Patients received analgesia with 1–2 µg/kg of fentanyl while the radiologist inserted a right internal jugular venous catheter.
All studies were performed in the Interventional Radiology Unit. Under ultrasound guidance, the right internal jugular vein was punctured and a 10F sheath placed using a Seldinger technique. A 5F Multi-purpose catheter (Cook, Bloomington, IN) was advanced through the right atrium into the inferior vena cava and into a hepatic vein under fluoroscopic control. This catheter was exchanged for an 11.5 mm occlusion balloon that was used to measure FHVP and WHVP at the same site in the hepatic vein. Wedged hepatic venography confirmed the accuracy of the WHVP. Pressure signals were obtained from a precalibrated pressure transducer with an external reference set to the midaxillary line, connected to a multichannel Hewlett-Packard recorder (Model M1094B, Hanover, MA). All pressures, measured in duplicate, and results were reported as the mean of the duplicate measurement.
Patients were randomized using a sequential number list to receive either desflurane or propofol after awake measurement of the FHVP and WHVP. The protocol blinded the radiologist to the anesthetic designation. It was not possible to blind the anesthesiologist to the choice of general anesthesia because the modes of administration of the two anesthetic drugs differed.
General anesthesia was induced in a rapid sequence with etomidate 0.2–0.6 mg/kg in the desflurane group, 1.5–2 mg/kg of propofol in the propofol group, and 1 mg/kg of succinylcholine. Neuromuscular relaxation with cisatracurium was monitored by peripheral nerve stimulation. We administered desflurane to achieve a stable end-expired concentration of 5%–6% or propofol 100–140 µg · kg-1 · min-1 for at least 15 min before continuing the study. All subjects were ventilated using a tidal volume of 7–10 mL/kg without positive end-expiratory airway pressure to maintain an end-tidal CO2 of approximately 35 mm Hg and airway pressures between 20–25 cm H2O. We monitored continuous arterial blood pressure from a femoral access sheath placed for TIPS insertion in addition to noninvasive monitors. TIPS insertion proceeded after measurement of all hepatic venous pressure measurements.
Excluded from study were patients whose heart rates or blood pressures varied more than 20% from their awake resting values or required vasopressors for blood pressure support during general anesthesia. Three patients were excluded from the study; one because of the presence of a veno-venous communication that prevented hepatic venous occlusion and measurement of the wedged pressure and two desflurane study patients because of persistent hypotension that required the use of vasopressors to improve blood pressure. The data from the remaining 21 patients were analyzed. Age, sex, medical indication for TIPS, and Child-Pugh score were recorded for each patient at the time of study. Variables collected for analysis included: arterial systolic and diastolic blood pressures, heart rate, oxygen saturation, FHVP, WHVP, and the calculated HVPG (WHVP - FHVP).
We conducted a power analysis to determine the number of patients required to detect a difference of 30% in the mean with a SD of 30% within groups. The analysis revealed that 10 patients per study group were required to reach a power >0.8 with an 
of 0.05. All data are presented as mean ± SD. Variables within groups were compared with a two-tailed paired t-test. Analyses of variables between groups were performed using a two-tailed unpaired t-test. A confidence interval of 95% was chosen for all tests. Statistical significance was accepted when P was smaller than 0.05. All calculations were performed using JMP software (version 4.0; SAS Institute, Cary, NC).
|
Results
|
|---|
Of 21 patients studied, 11 received desflurane and 10 propofol to maintain general anesthesia. All patient data were included in the final analysis. There were no significant differences in demographic traits between study groups. The average age in both groups was 51 yr (51.1 ± 5.6 and 51 ± 5.5 yr in the desflurane and propofol groups, respectively), whereas 70% and 60% of study subjects were men in the desflurane and propofol study groups, respectively. TIPS were indicated for recurrent upper gastrointestinal bleeding in seven desflurane and six propofol subjects. The remaining patients in each group had refractory ascites. Child-Pugh scores were similar in both groups (10.4 ± 1.6 in desflurane versus 10 ± 1.4 in propofol).
In the desflurane group, systemic variables measured during awake and general anesthesia were not statistically different (Table 1). However, the FHVP significantly increased from 11.9 ± 4.4 mm Hg to 23.5 ± 4.1 mm Hg (P < 0.05), and the HVPG decreased significantly from 21.6 ± 7.4 to 14.7 ± 5.2 mm Hg (P < 0.05). There was no difference from the awake to the anesthetized state in systemic variables, FHVP, WHVP, and HVPG in the propofol group.
We found no significant differences between the desflurane and propofol study groups in awake systemic and hepatic venous pressure variables (Table 1 and 2). In contrast, the FHVP (23.5 ± 4.1 mm Hg for desflurane versus 13.7 ± 4.6 mm Hg for propofol; P < 0.05) and WHVP (38.2 ± 3.8 mm Hg for desflurane versus in 31.6 ± 5.2 mm Hg for propofol; P < 0.05) were significantly greater in the desflurane compared with propofol study groups. The HVPG was not significantly different between the study groups.
|
Discussion
|
|---|
Desflurane, but not propofol, significantly increased FHVP from the awake state in all subjects studied. An increase in the FHVP reduced the HVPG, a measure of portosystemic gradient. These studies suggest that hepatic venous pressures approximate the awake state better during general anesthesia with propofol.
We compared the effects of the two frequently used anesthetic drugs, desflurane and propofol, on hepatic venous pressures. Because anesthetic drugs cause changes in hepatic blood flow that are dose-dependent (12), we chose doses of both anesthetic drugs to maintain comparable levels of general anesthesia (13). To minimize variability in our study, we excluded patients with conditions that interfered with the accurate estimation of portal venous pressures by balloon occlusion. The WHVP can underestimate true portal pressure in patients with defects outside the sinusoids of the liver, such as primary biliary cirrhosis (14). Overestimation of portal pressures by balloon occlusion also caused us to exclude patients with portal vein thrombosis and hepatofugal portal venous flow on Doppler analysis (15–17). Because anatomical defects in the hepatic sinusoids of patients with hepatitis C generally cause portal hypertension, there is a good correlation between WHVP and portal pressure (15).
We administered albumin to minimize the decrease in blood pressure caused by anesthetic drugs in our study subjects (18), avoiding the use of vasopressors, which can independently alter portal pressure (19). Whereas some investigators argue that rapid volume administration may independently increase portal pressure (20), all of our study subjects received similar treatment that did not produce differences between the study groups. Significant differences in hepatic venous pressures occurred only after general anesthesia and thus are likely caused by effects of the anesthetic drugs.
Hepatic venous and systemic pressure measurements of the propofol and desflurane awake study groups were similar, suggesting that the selective increase in FHVP after administration of desflurane was drug induced. The decrease in HVPG after administration of desflurane was caused by a significant increase in the FHPV compared with the WHVP (Table 1). The reason(s) desflurane but not propofol increased FHVP is uncertain and not readily explained by our study. However, an isolated increase in FHVP can result from two mechanisms: myocardial depression or an increase in total effective vascular compliance (21).
Desflurane maintains cardiac output by a balance between direct myocardial depression and a decrease in systemic vascular resistance (22). We could not use central venous pressure to determine if changes in cardiac function explained differences in the FHVP caused by desflurane because there are significant and dynamic pressure gradients between the abdominal and thoracic venous pressure in patients with portal hypertension (23). Patients with end-stage liver disease have hyperdynamic circulations characterized by intrinsic changes in cardiac function and systemic vascular resistance (24) and their response to desflurane will therefore likely differ from normal patients. A hypothesis of increased FHVP based upon actions of inhaled anesthetic drugs in normal patients may be flawed, and further studies are required to explore the relationship between cardiac function and FHVP in patients with portal hypertension.
Physicians reduce effective circulating venous volume without changing total circulatory volume to decrease portal hypertension (21). Drugs such as somatostatin and nonspecific ß-blockers such as propranolol decrease total effective venous capacity by producing splanchnic venoconstriction (20,25), causing splanchnic steal (26). The significant increases in FHVP associated with propranolol (21) and somatostatin (25) are similar to our observations with desflurane. Investigators hypothesize that enhanced venous tone increases the FHVP (21). Although desflurane causes splanchnic vasodilation in normal study subjects (27,28), its effects on patients with portal hypertension are unknown. Because patients with portal hypertension already have profound splanchnic vasodilation, it is of value to determine if changes in vascular resistance between the systemic and splanchnic circulation favor a decrease in effective circulating venous volume causing desflurane-induced increased FHVP.
Medical therapy or shunt insertion can decrease the HPVG by 20%, or <12 mm Hg, values associated with a significant decrease in the risk of portal hypertensive complications (29–31). We did not study hepatic venous pressure values after TIPS or medical therapy and are thus unable to determine if similar effects occur after therapeutic intervention under general anesthesia. However, the significant decrease in HPVG during desflurane anesthesia raises concern that the type of general anesthetic drug administered could influence how physicians judge the success of interventions to decompress portal pressure.
In summary, general anesthesia drugs altered hepatic venous pressure measurements from the awake resting state in our study. General anesthesia with desflurane, but not propofol, increases FHVP and consequently reduces the HVPG. Further evaluation is required to identify the specific mechanisms responsible for alterations in venous pressure caused by volatile anesthetics such as desflurane. These observations are important to physicians who provide care for patients with portal hypertension because anesthetic drugs can significantly influence hepatic venous pressure measurements.
|
References
|
|---|
- Villanueva C, Balanzo J, Novella M, et al. Nadolol plus isosorbide mononitrate compared with sclerotherapy for the prevention of variceal rebleeding. N Engl J Med 1996; 334: 1624–9.[Abstract/Free Full Text]
- Rossle M, Ochs A, Gulberg V, et al. Comparison of paracentesis and transjugular intrahepatic portosystemic shunting in patients with ascites. N Engl J Med 2000; 342: 1701–7.[Abstract/Free Full Text]
- Azoulay D, Castaing D, Majno P, et al. Salvage transjugular intrahepatic portosystemic shunt for uncontrolled variceal bleeding in patients with decompensated cirrhosis. J Hepatol 2001; 35: 590–7.[ISI][Medline]
- Escorsell A, Banares R, Garcia-Pagan J, et al. TIPS vs. drug therapy in preventing variceal rebleeding in advanced cirrhosis: a randomized controlled trial. Hepatology 2002; 35: 385–92.[ISI][Medline]
- Chutaputti A. Management of refractory ascites and hepatorenal syndrome. J Gastroenterol 2002; 17: 456–61.
- Groszmann R, Glickman M, Blei A, et al. Wedged and free hepatic venous pressure measured with a balloon catheter. Gastroenterology 1979; 76: 253–7.[Medline]
- Steinaluf A, Garcia-Tsao G, Zakko M, et al. Low-dose midazolam sedation: an option for patients undergoing serial hepatic venous pressure measurements. Hepatology 1999; 29: 1070–3.[ISI][Medline]
- Merin R, Bernard J, Doursout M, et al. Comparison of the effects isoflurane and desflurane on cardiovascular dynamics and regional blood flow in the chronically instrumented dog. Anesthesiology 1991; 74: 568–74.[ISI][Medline]
- Armbruster K, Noldge-Schomberg G, Dressler I, et al. The effects of desflurane on splanchnic hemodynamics and oxygenation in the anesthetized pig. Anesth Analg 1997; 84: 271–7.[Abstract]
- Carmichael F, Crawford M, Khayyam N, Saldivia V. Effects of propofol on splanchnic hemodynamics and liver oxygen consumption in the rat: a dose response study. Anesthesiology 1993; 79: 1051–60.[ISI][Medline]
- Wouters P, Van de Velde M, Marcus M, et al. Hemodynamic changes during induction of anesthesia with eltanolone and propofol in dogs. Anesth Analg 1995; 81: 125–31.[Abstract]
- Frink E, Morgan S, Coetzee A, et al. The effects of sevoflurane, halothane, enflurane and isoflurane on hepatic blood flow and oxygenation in chronically instrumented gray hounds. Anesthesiology 1992; 76: 85–90.[ISI][Medline]
- Apfelbaum J, Lichtor J, Lane B, et al. Awakening, clinical recovery and psychomotor effects after desflurane and propofol anesthesia. Anesth Analg 1996; 83: 721–5.[Abstract]
- Navasa M, Pares A, Bruguera M, et al. Portal hypertension in primary biliary cirrhosis: relationship with histological features. J Hepatol 1987; 5: 292–8.[Medline]
- Perello A, Escorell A, Concepcio B, et al. Wedged hepatic pressure adequately reflects portal pressure in hepatitis C virus-related cirrhosis. Hepatology 1999; 30: 1393–7.[ISI][Medline]
- Boyer T, Triger D, Horisawa M, et al. Direct transhepatic measurement of portal vein pressure using a thin needle: comparison with wedged hepatic vein pressure. Gastroenterology 1977; 72: 584–9.[Medline]
- Rector W, Hoefs J, Hossack K, Everson G. Hepatofugal portal flow in cirrhosis: observations on hepatic hemodynamics and the nature of arterioportal communications. Hepatology 1988; 8: 16–20.[ISI][Medline]
- Nathan N, Vial G, Benrhaiem M, et al. Induction with propofol target-concentration infusion vs. 8% sevoflurane inhalation and alfentanil in hypertensive patients. Anaesthesia 2001; 56: 251–7.[Medline]
- Rothe C, Maass-Moreno R. Hepatic venular resistance responses to norepinephrine, isoproterenol, adenosine, histamine, and Ach in rabbits. Am J Physiol 1998; 274: H777–85.[Medline]
- Vlavianos P, Mac Mathuna P, Williams R, Westaby D. Splanchnic and systemic haemodynamic response to volume changes in patients with cirrhosis and portal hypertension. Clin Sci 1999; 96: 457–81.
- Andreu V, Perello A, Moitinho E, et al. Total effective vascular compliance in patients with cirrhosis: effects of propranolol. J Hepatol 2002; 36: 356–62.[Medline]
- De Hert S, Van der Linden P, ten Broecke P, et al. Effects of desflurane and sevoflurane on length-dependent regulation of myocardial function in coronary surgery patients. Anesthesiology 2001; 95: 357–63.[ISI][Medline]
- Groszmann R, Glickman M, Blei A, et al. Wedge and free hepatic venous pressure measured with a balloon catheter. Gastroenterology 1979; 76: 253–8.
- Ma Z, Lee S. Cirrhotic cardiomyopathy: getting to the heart of the matter. Hepatology 1996; 24: 451–9.[ISI][Medline]
- Hsia H, Lin H, Lee F, et al. Changes of hepatic and systemic haemodynamics following somatostatin administration in patients with hepatitis B-related cirrhosis. Gastroenterol Hepatol 1993; 8: 15–20.
- Newby D, Hayes P. Hyperdynamic circulation in liver cirrhosis: not peripheral vasodilation but splanchnic steal. Q J Med 2002; 95: 827–30.
- O’Riordan J, O’Beirne A, Young Y, Bellamy M. Effects of desflurane and isoflurane on splanchnic microcirculation during major surgery. Br J Anaesth 1997; 78: 95–6.[Abstract/Free Full Text]
- Sundeman H, Aneman A, Broome M, et al. Effects of desflurane on the pig intestinal circulation during hypotension. Acta Anaesthesiol Scand 1999; 43: 1069–77.[Medline]
- Haskal Z, Martin L, Cardella J, et al. Quality improvement guidelines for transjugular intrahepatic portosystemic shunts: standards of practice. J Vasc Interv Radiol 2001; 12: 131–6.[Medline]
- Feu F, Garcia-Pagan J, Bosch J, et al. Relation between portal pressure response to pharmacotherapy and risk of recurrent variceal haemorrhage in patients with cirrhosis. Lancet 1995; 346: 1056–9.[Medline]
- Vorobioff J, Groszmann R, Picabea E, et al. Prognostic value of hepatic venous pressure gradient measurements in alcoholic cirrhosis: a 10 year prospective study. Gastroenterology 1996; 111: 701–9.[Medline]
Accepted for publication July 23, 2003.
Top
Abstract
Introduction
