| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Anaesthesiologische Universitätsklinik Freiburg, Freiburg im Breisgau, Germany;
Klinik und Poliklinik für Anästhesiologie und Intensivtherapie, Universität Rostock, Rostock, Germany;
Institut für Experimentelle und Klinische Pharmakologie und Toxikologie der Universität Freiburg, Freiburg im Breisgau, Germany; and
Chirurgische Forschung, Chirurgische Universitätsklinik Freiburg, Freiburg im Breisgau, Germany
Address correspondence and reprint requests to Dierk A. Vagts, MD, DEAA, EDIC, Klinik und Poliklinik für Anästhesiologie und Intensivtherapie, Universität Rostock, Schillingallee 35, D-18055 Rostock, Germany. Address e-mail to dierk.vagts{at}medizin uni-rostock.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAppendix 1References |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
The second mechanism is the classic arterial autoregulation: increasing MAP leads to vasoconstriction of the hepatic artery and vice versa (5). Extrinsic flow regulation through humoral factors and the rich sympathetic nerve supply of the hepatic artery are not fully understood (5). Nerve-induced vasoconstriction is mediated by 
-adrenergic receptors. However, in the dog, vasoconstriction is maintained, whereas in the cat, vasoconstriction escapes despite continued stimulation (6). The situation in humans is unknown. If hepatic arterial vasoconstriction is dependent on sympathetic nerve activity, sympathicolysis might be able to induce vasodilation and, consequently, an increase in HABF independent of the HABR.
Reduction of sympathetic activity can be induced by epidural anesthesia (EA), which is often combined with general anesthesia to reduce perioperative stress and provide pain therapy. However, the application of this technique often results in a decrease in MAP and CO (7). Changes in hepatic blood flow after the induction of EA can be the consequence of changes in systemic hemodynamics and/or redistribution of blood flow because of sympathicolysis and vasodilation or because of reactive increased sympathetic activity and subsequent vasoconstriction. There are very few studies (8–10) examining the effect of regional hepatic sympathicolysis (T5 to T12) on hepatic perfusion. Greitz et al. (8) investigated the effects of an almost complete block of sympathetic outflow from T1 to L2, which resulted in a marked decrease of MAP and CO, whereas Meissner et al. (9) investigated the effects of a sole block of T1 to T5, which did not induce changes in MAP, and could only estimate HABF. The different spread of thoracic epidural anesthesia (TEA) in these studies had different implications for changes in CO, MAP, and splanchnic sympathetic activity. All previous studies in humans examined the THBF indirectly (10).
In clinical practice, volume loading frequently treats a decrease in MAP. There are no data about the effects of TEA and volume loading on hepatic perfusion and oxy-genation. This study was designed to examine the effects of a regional TEA (T5 to T12) on HABF, PVBF, and liver oxygenation with and without volume loading.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
A 20-gauge epidural catheter (Braun, Melsungen, Germany) was inserted under radiograph control with the tip at T8. Correct position of the catheter was verified with contrast agent (Ultravist 300®; Schering AG, Berlin, Germany).
A pulmonary artery catheter was placed in the right jugular vein via an introducer system, and the right femoral artery was cannulated (COLD® catheter; 4F Oxymetrie-Thermo-Dye-Dilutionssonde; Pulsion Medical Systems, Munich, Germany).
After median laparotomy, the left hepatic and portal veins were cannulated as previously described (3). Ultrasonic perivascular flowprobes (Transonic Systems Inc., Ithaca, NY) of appropriate sizes were placed around the hepatic artery and the portal vein (Fig. 1). Care was taken to preserve the perivascular nerve plexus and to place the flowprobe of the hepatic artery behind the departure of the superior gastroduodenal artery. Intermittently a multiwire surface electrode was placed onto the liver to measure tissue surface PO2.
|
All intravascular catheters were connected to pressure transducers, and signals were recorded online (PO-NE-MAH®; Digital Acquisition Analysis and Archive Systems, Simsbury, CT). Measured variables were heart rate (HR), MAP, CVP, pulmonary capillary wedge pressure (PCWP), HABF, PVBF, and blood gases. Equations for derived variables (systemic vascular resistance [SVR], THBF, hepatic oxygen delivery [DO2TH], and so on) are listed in Appendix 1. The indocyanine green plasma disappearance rate (PDRICG) was determined with the double indicator dilution technique by using the COLD® catheter.
A microprocessor-supported system (Ingenieurbüro Meß- und Datentechnik Mußler, Aachen, Germany) with eight-channel multiwire platinum surface electrodes (Eschweiler, Kiel, Germany) was used for measuring liver surface oxygen partial pressure. During each measurement, approximately 200 individual PO2 values were registered at 10 different electrode locations to receive representative data of PO2 distribution. The mean values of these data (Fig. 3) reflect tissue oxygenation, which is the net result of nutritive blood flow and tissue oxygen consumption. Methods were described in detail previously (3).
|
After baseline readings (t0), the BF animals received 0.75 mL of 0.5% bupivacaine per segment (Carbostesin®; Astra, Wedel, Germany) epidurally to block segments T5 to T12. VL animals additionally received hydroxyethylstarch (HES) 15 mL/kg (HAES 6%, 200/0.5; Fresenius-Klinik] IV within 30 min before the induction of EA. Subsequently, 0.75 mL of 0.5% bupivacaine per segment was injected epidurally. HES infusion in VL animals was continued at 10 mL · kg-1 · h-1. Repeat measurements in both groups were made 60 min (t1h) and 120 min (t2h) after epidural catheter injection.
In the controls, no further interventions were made after instrumentation. Fluid replacement was based on the fluid demand during the stabilization period and continued in amounts of crystalloids, which were necessary to keep filling pressures and hematocrit stable during the stabilization period (10–15 mL/kg). All variables were measured 60 min (t1h), 120 min (t2h), and 180 min (t3h) later. All animals were killed by KCl infusion in deep anesthesia.
Plasma concentrations of epinephrine and norepinephrine were determined in arterial and portal venous blood (11) at t0, t1h, and t2h. Spread of TEA higher than T5 was excluded at the end of the experiment by inducing cardioacceleration with nitroglycerine (NTG) 0.3 mg/kg IV. This was done before and after injection of atropine 30 µg/kg IV. The persistence of reflex cardioacceleration in both situations was interpreted as a sign of unblocked sympathetic innervation of the heart (T1 to T4) (12).
CO, SVR, HABF, hepatic arterial vascular resistance (HAVR), and PVBF were indexed to compensate for differences in body weight. Statistical analysis was performed with JMP® (SAS Institute, Cary, NC). Medians are given throughout with interquartile ranges (25%–75%) (13). Data were analyzed by Friedman’s statistic, Wilcoxon’s signed rank test, Bonferroni’s correction, and Mann-Whitney U-tests. The level of significance was set at P 
0.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
|
BF Group
TEA elicited a decrease in MAP and SVR (Table 1). HR, CO, CVP, and PCWP did not change (Table 1). HABF decreased by 25%; HAVR, PVBF, and portal venous resistance did not change (Table 2). THBF, hepatic venous pressure, and portal venous pressure remained stable.
|
|
O2TH) did not change. Mean surface PO2 values of the liver were approximately 70 mm Hg and were not affected by TEA (Fig. 3). PDRICG, hepatic lactate uptake of the liver, and glucose flux (Table 2) were also not influenced.
VL Group
In the VL group, MAP decreased similarly as in the BF group (Table 1). After volume loading, there was an increase in CO, CVP, and PCWP. SVR decreased as in the BF group (Table 1). The arterial hemoglobin concentration decreased by 22% after 1 h.
HABF decreased despite the increase in CO. HAVR, PVBF, and portal venous resistance did not change. Because of the decrease in HABF and the decrease portal venous oxygen content, DO2TH decreased. O2TH did not change compared with baseline values. There was no decrease in oxygen surface partial pressure of the liver (Fig. 3). TEA did not affect PDRICG, hepatic lactate uptake, or glucose flux (Table 2).
In three control animals (Table 3), differences between baseline values and those obtained during the following 3 h did not exceed 10%. HRs increased by 49 ± 22 bpm (mean ± SD) after NTG, 39 ± 17 bpm after atropine, and 21 ± 7 bpm after NTG after atropine.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
During surgical procedures, especially laparotomy, the liver may be at risk for perioperative dysfunction secondary to hypoperfusion (14) due to intraabdominal surgical stimulation (15), perioperative blood loss, clamping of splanchnic arterial vessels, and endotoxemia (16). In many of these situations, sympathetic activity is increased, resulting in vasoconstriction of splanchnic vessels (17).
TEA may reduce regional vasoconstriction because of decreased sympathetic activity resulting in regional relaxation of precapillary sphincters and of smooth muscles of metarterioles. The resulting decrease in regional vascular resistance might be a factor for redistribution of regional blood flow. However, TEA may also induce a decrease in systemic perfusion pressure, with the potential risk of regional flow reduction.
In our experiments, TEA induced a decrease in SVR without changes in CO. This resulted in a 30% reduction of MAP, which is in agreement with previous studies (18) in anesthetized, ventilated animals.
Despite maintained PVBF, HABF decreased because of a reduction of MAP. The calculated HAVR did not change. With respect to the HABR, this result is a puzzling finding. Maintained PVBF should not affect HABF. The MAP-dependent reduction of HABF suggests that the mechanism of arterial hepatic autoregulation (5) is very weak or even does not exist under regional sympathicolysis. Further explanations for the lack of hepatic artery vasodilation might be that the hepatic artery was maximally vasodilated at baseline—which seems unlikely, because the HABF can be tripled because of HABR—or that the baseline sympathetic activity of the hepatic artery is very small, indicating that sympathetic stimulation as an extrinsic factor plays a role only for vasoconstriction, but not for vasodilation (6). However, the reduction of 25% in HABF (equaling only approximately 5% of THBF) was not clinically relevant because PVBF, which accounts for 80% of THBF, did not change. In conclusion, the maintained THBF during unchanged CO and reduced MAP support the thesis that THBF is mainly flow and not pressure dependent.
Kennedy et al. (10) described a 17% decrease of MAP, a maintained CO, and a 23% decrease in THBF after lumbar EA up to T5. These differing results might be explained with indirectly estimated THBF in contrast to direct measurement, with greater extension of the epidural block, and with awake humans in contrast to anesthetized animals. Meissner et al. (9) did not find an effect of restricted high TEA (T1 to T4) on hepatic perfusion. However, because they used microspheres, they were able to estimate only the HABF. Greitz et al. (8) found the same effect of EA on HABF as we did: despite a reduction of PVBF, there was no HABR.
DO2TH and O2TH did not change during the experiments. This suggests that DO2TH was not affected by changes of oxygen uptake in preportal organs during TEA and that the reduction of regional sympathetic activity did not reduce the metabolic activity of the liver, estimated by the surrogate O2. Unchanged mean hepatic surface PO2 values suggest that the microcirculation was not disturbed, despite the reduction of MAP.
Because THBF was unaffected by TEA in our experiments, PDRICG was a good marker for hepatic cell function. Indocyanine green is removed from the blood almost exclusively by the liver. Its plasma disappearance rate is dependent only on sinusoidal perfusion, membrane transport, and hepatic secretory capacity. In our experiments, we did not find a decrease of PDRICG. These results are in accordance with findings of PDRICG after reduction of MAP by 20% induced by lumbar EA in humans (19). Lactate uptake, an energy-demanding process and a good marker for hepatic cellular function, was also unaffected.
Clinically, volume loading is used to compensate for epidural-induced MAP decreases. We compared animals with BF application with volume loading with 6% HES. The amount used per kilogram of body weight was in accordance with clinical practice in humans. The resulting increase in PCWP was pronounced, but there were no signs of cardiac failure, such as pulmonary edema or a relevant increase of CVP. After volume loading prior to and during TEA, CO was increased. The increase in CO was due to hypervolemic hemodilution, which is in agreement with previous work regarding the effects of hemodilution (20). Despite volume loading, the MAP still decreased by 16%. As seen by other investigators, in TEA it was not possible to maintain MAP by volume loading, especially during general anesthesia (21).
Nearly all demonstrated and discussed effects of EA on hepatic blood flow and oxygenation were qualitatively and quantitatively comparable to the results of the BF group. These results are in accordance with earlier findings, when MAP was maintained within 15% of baseline values by volume loading (22) after spinal anesthesia. Thus, volume loading before and during EA did partly attenuate the decrease in MAP, but this was not reflected by any beneficial effect on hepatic microcirculation and oxygenation. DO2TH even decreased in the VL group because of hemodilution and a decrease in portal venous oxygen content. Nevertheless, we did not find an influence on liver surface PO2.
Verification of the spread of TEA, aiming for a block of T5 to T12, in animals during general anesthesia is very difficult (23). Some groups have used a modified pinprick technique, postmortem analysis of injected dye, or capillary PO2 measurement. However, even direct measurement of the activity in splanchnic fibers to different organs exhibits different patterns of activity (24). We controlled the position of our catheter radiologically. To exclude a spread higher than T4, we used two pharmacological tests: the HR response to hypotension was evaluated in animals with and without blockade of cardiac vagal transmission with atropine. Atropine decreases parasympathetic activity, therefore eliciting tachycardia by increasing sympathetic effects on the heart. NTG as a short-acting drug was used to elicit reflex tachycardia due to pronounced hypotension if cardiac nerves were not blocked. Cardioacceleration was maintained in both situations, indicating intact function of cardiac sympathetic nerves. This, in turn, indicates that spinal segments higher than T4 (the origin of preganglionic cardiac sympathetic neurons) were not blocked by TEA in our study.
Sympathicolysis itself was proven by reduced arterial and portal venous norepinephrine and epinephrine plasma concentrations compared with baseline values (12). In control animals, we did not find a reduction of catecholamine plasma levels. TEA blocks preganglionic sympathetic transmission and decreases plasma norepinephrine concentrations. Because the liver is supplied autonomically from the same segments as the adrenal glands, reduction of epinephrine plasma levels is proof of sympathicolysis of liver and supplying vessels. The extent of reduction of catecholamine concentrations is dependent on the extension and intensity of the sympathetic nervous blockade and the baseline activity of the sympathetic nervous system. Nevertheless, findings in the literature regarding catecholamines and TEA are not unanimous. Stevens et al. (25) found little change in plasma epinephrine and norepinephrine levels in young, healthy volunteers, even with an extensive epidural block to C8. With a segmental TEA from T4 to T11, we elicited effects due to blocking spinal nerve roots at the epidural space but also blocked the adrenal glands.
TEA combined with general anesthesia neither impaired THBF, despite a decrease in MAP, nor improved THBF by potential flow redistribution due to vasodilation. HABF was decreased, but liver oxygenation was not affected. Lactate uptake and PDRICG clearance as indicators for hepatocellular function remained unchanged. Volume loading before TEA to maintain MAP did not show any benefit with regard to hepatic perfusion, oxygenation, and function.
|
Appendix 1 |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
|
|
|
|
|
|
where MAP is mean arterial pressure, CVP is central venous pressure, PVP is portal venous pressure, HVP is hepatic venous pressure, CO is cardiac output, HABF is hepatic arterial blood flow, PVBF is portal venous blood flow, and bw is body weight.
Equations Used for Calculation of Oxygen Supply/Uptake
|
|
|
|
|
|
where DO2TH is total hepatic oxygen delivery, DO2HA is hepatic arterial oxygen delivery, DO2PV is portal venous oxygen delivery, CO2A is systemic arterial oxygen content, CO2PV is portal venous oxygen content, CO2HV is hepatic venous oxygen content, and O2TH is total hepatic oxygen uptake, SO2 is hemoglobin oxygen saturation.
|
|
|
|
|
|
where DlacL is hepatic lactate uptake; ClacHA is hepatic arterial lactate concentration, ClacVH is hepatic venous lactate concentration, ClacPV is portal venous lactate concentration, and ClacA is arterial lactate concentration.
|
|
|
|
where FgluL is hepatic glucose flux; CgluHA is hepatic arterial glucose concentration, CgluVH is hepatic venous glucose concentration, and CgluPV is portal venous glucose concentration.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
