| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
MS 8690 Received 5 September 1998; accepted after revision 12 January 1999.
 |
ABSTRACT |
|---|
 TopAbstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The liver represents only a small percentage of body weight; however, it receives about one quarter of the cardiac output. The predominant blood flow is from the portal vein, which is controlled by the vascular resistance of the splanchnic organs (Greenway & Lautt, 1989). The portal venous resistance is small and influences portal venous pressure rather than flow (Lautt & Greenway, 1987). Hepatic arterial flow is unrelated to the metabolic activity of the liver under normo-volaemic conditions, maybe as a consequence of the large portal flow (Lautt, 1980). For example, during alcohol infusion hepatic metabolism massively increases, the hepatic arterial bed does not dilate but oxygen extraction increases (Bledfeldt et al. 1985). An inverse relation between the arterial and portal blood flow has been described (Burton-Opitz, 1910). This relationship is termed the 'arterial buffer response' as the arterial circulation is thought to provide some compensation for a reduction in portal flow but not vice versa (Lautt, 1981). The evidence in favour of the hepatic arterial buffer response is restricted to the effects of pharmacological agents on the relative contribution of the hepatic arterial and portal venous flows to hepatic flow in unstressed animals, as well as to experiments in which the portal venous flow was restricted (Greenway & Lautt, 1989). Thus, it is not known whether the arterial buffer response is important in determining the hepatic blood flow, or hepatic oxygenation and metabolism during haemorrhage.
In order to examine whether the arterial flow can maintain hepatic oxygenation and metabolism during haemorrhage, we monitored the portal venous and hepatic arterial flow in the pig. In addition, hepatic oxygen uptake, tissue oxygenation, lactate metabolism, and noradrenaline spill-over were followed during progressive haemorrhage. Hepatic variables were compared with those of the systemic and renal circulation.
|
METHODS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Eighteen pigs (sows - Danish Landrace : Yorkshire, 1 : 1; boars - Duroc; weight, 34·5 ± 3·7 kg) were studied in accordance with the guidelines laid down by the Danish Ministry of Justice. They were premedicated with midazolam (15 mg I.M.) and anaesthesia was achieved with etorfin (0·31 mg I.M.) and acepromazin (1 mg I.M.). After intubation, they were ventilated with 66 % N2O in oxygen and 1 % halothane until a central venous access was established, after which halothane was discontinued. Anaesthesia was then maintained with I.V. methohexithone (100 mg h-1) and fentanyl (0·75 mg h-1). Vecuronium (30 mg h-1) was given as a muscle relaxant and the animals were ventilated with oxygen and air to an inspiratory fraction of 0·4. The anaesthetic regime was assessed by the anaesthesiologist and considered adequate when no cardiovascular responses were noted in response to surgery, and this level of anaesthesia was unchanged during the whole experiment. Animals that were alive after a 40 % blood loss were killed with an intravenous injection of potassium. The post-mortem weight of the liver and the kidney were noted.
Mean arterial pressure (MAP) was determined from a catheter in the right carotid artery connected to a transducer placed at heart level. This line was continually flushed with saline (3 ml min-1) (Baxter, Uden, Holland). Heart rate (HR) and the electrocardiogram (Danica, Copenhagen, Denmark) were obtained from precordial electrodes (Medicotest, Ølstykke, Denmark). Cardiac output (CO), pulmonary arterial pressure (PAP), pulmonary arterial wedge pressure (PAwP), and mixed venous oxygen saturation (SvO2) were measured through a pulmonary artery catheter (7·5F; Baxter, Uden, Holland) and displayed on a monitor. For measurement of CO 10 ml ice-cold saline was used as a bolus for the thermo-dilution estimate. Central venous pressure (CVP) was determined from the right atrium (8F, baby-feeding tube; Unoplast, Roedovre, Denmark). Catheters were placed in the hepatic vein via the external jugular vein (6F; Cook, Bjaeverskov, Denmark), the renal vein via the femoral vein (6F; Cook), and the portal vein via the splenic vein (8F, baby-feeding tube; Unoplast). These catheters were used for blood sampling and measurement of pressures (HvP, RvP and PvP, as displayed on another Danica monitor). A mid-line incision and minimal dissection in the free edge of the lesser omentum permitted application of flow probes around the portal vein and the hepatic artery, while branches to extrahepatic tissue were ligated. The peritoneum over the right kidney was incised and a flow probe was placed around the renal artery. A catheter was placed in the portal vein via a branch to the splenic vein and the abdomen was closed.
Flow in the hepatic artery (HaBF; probe 5 mm), portal vein (PBF; probe 8 mm), and renal artery (RBF; probe 4 mm) were assessed by ultrasound transit time (Medistim, Norway). Hepatic flow (HBF) was calculated as the sum of HaBF and PBF. Hepatic arterial vascular resistance was calculated as (MAP - PvP)/HaBF, portal venous resistance as (PvP - HvP)/HBF, vascular resistance of the splanchnic area as (MAP - PvP)/PBF, and renal vascular resistance as (MAP - CVP)/RBF. Pressures and blood flows were measured continuously and vascular resistances were calculated off-line from these parameters.
Haematocrit (Hct) was measured by centrifugation. An ABL 625 apparatus (Radiometer, Copenhagen, Denmark) was used to measure arterial and venous blood gas variables, pH, standard base excess, potassium, sodium, glucose and lactate. Blood for catecholamine analysis was collected into chilled glass tubes containing ethyleneglycol-bis (
-aminoethylether)-N,N'-tetraacetic acid and reduced glutathione and placed on ice. Samples were centrifuged promptly at 4°C with plasma stored at -80°C. Catecholamine concentrations were determined by a single-isotope radioenzymatic method (Ben-Jonathan & Porter, 1976) using HPLC.
Hepatic noradrenaline uptake was calculated using the following relationship: Hepatic uptake = (1 - Hct) × (PBF × NApv + NAa × HaBF - HBF × NAhv). Splanchnic noradrenaline spillover was calculated as: Spl NAspill = (1 - Hct) × PBF × (NApv - NAa) and renal noradrenaline spillover was calculated as: Ren NAspill = (1 - Hct) × RBF × (NArv - NAa). NAa, NApv, NArv, NAhv are arterial, portal, renal and hepatic venous concentrations of noradrenaline, respectively.
Oxygen uptake from the hepatic arterial system was calculated as (HaVO2 - HvVO2) × HaBF; oxygen uptake from the portal system was calculated as (PvVO2 - HvVO2) × PBF and the hepatic uptake was assumed to be the sum of these two values. Hepatic oxygenation was measured with near-infrared spectroscopy at four wavelengths with the emitter and sensor spaced 3·5 cm apart on the surface of the right lobe of the liver (Niro-500, Hamamatsu Phototonics KK, Japan). Changes in the concentration of oxygenated haemoglobin (
HbO2) and deoxygenated haemoglobin (Hb) were calculated from the algorithm of Wray et al. (1988). Chromophore concentrations were set to zero in the steady state before any blood was drained. Although the exact sample volume assessed by near-infrared light is not known, the method has been validated against flow measurements for the brain and skeletal muscle during hypovolaemic shock in humans (Madsen & Secher, 1998).
Renal oxygen consumption was calculated as the arterio-venous oxygen difference multiplied by RBF. Surface oxygenation in the kidney was followed by continuous light spectroscopy (INVOS, Somanetics, USA). Haemoglobin oxygen saturation was estimated from the ratio of two wavelength absorbances, taken to reflect deoxygenated haemoglobin and the sum of deoxygenated and oxygenated haemoglobin.
Hemacel and isotonic saline were administered intravenously until a maximal SvO2 was established. This manoeuvre was necessary in order to ensure that CO and oxygen delivery to the tissue were not volume dependent during baseline conditions (Ejlersen et al. 1995). From each of the five instrumented vessels (the hepatic artery, pulmonary artery, portal vein, hepatic vein and renal vein) 0·5 ml of blood was taken for measurements by the ABL-625 and 5 ml for catecholamine analysis. Ten per cent of the estimated blood volume (EBV) was drained via the central venous catheter. One half-hour later cardiovascular variables were determined and blood samples were obtained. Similar haemorrhage and recording procedures were repeated until circulatory collapse (i.e. the inability to maintain a stable blood pressure, e.g. due to cardiac arrhythmia). The EBV was 6·5 % of body weight.
In eight of the pigs, blood flow was recorded by the ultrasound transit time technique and compared during steady-state conditions with hepatic flow determined by infusion of Indocyanine Green (ICG; Cardiogreen, Becton Dickinson, USA; 0·12 µmol min-1). The plasma concentration of the dye was determined by HPLC (Ott et al. 1993). Five samples were taken simultaneously from the artery and a hepatic vein at 5 min intervals after at least 90 min of infusion. Hepatic plasma flow was calculated using Fick's principle with correction for deviations from the steady state as previously described (Ott et al. 1993). The Spearman rank correlation was r2 = 0·86, P < 0·001 (Fig. 1). Complete removal and re-application of the probes evaluated reproducibility of the values obtained by the transit time technique. The coefficients of variation were 4·6 % for portal flow, 3·2 % for the hepatic artery and 4·9 % for the renal artery.
 |
View larger version [in this window] [in a new window] |
|
|
Hepatic blood flow measured by ultrasound transit time and by Indocyanine Green (ICG) clearance. | ||
Results are presented as means ± S.E.M. and evaluated by repeated analysis of variance including all pigs from baseline to 30 % haemorrhage, but only 13 pigs during 40 % haemorrhage. The analysis of variance was followed by the Student-Newman-Keuls post hoc test using 0·05 as the significance level.
|
RESULTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Before circulatory collapse, it was possible to drain 30 % of EBV in all pigs studied. Furthermore, in 13 pigs an additional 10 % could be removed. Unless stated otherwise, the baseline and observations at 30 % haemorrhage are reported. Hypovolaemia increased heart rate from 88 ± 4·1 to 169 ± 10 beats min-1 (P < 0·001, Fig. 2). Blood pressure was reduced from 104 ± 5 to 35 ± 3 mmHg, pulmonary wedge pressure from 7·1 ± 0·5 to 4·7 ± 0·4 mmHg, central venous pressure from 4·4 ± 0·6 to 1·2 ± 0·3 mmHg, and cardiac output from 4·4 ± 0·3 to 1·3 ± 0·2 l min-1, while pulmonary arterial pressure (20·7 ± 0·6 mmHg) did not change significantly.
 |
View larger version [in this window] [in a new window] |
|
|
Values are means ± S.E.M. Heart rate (HR), mean arterial pressure (MAP), pulmonary artery wedge pressure (PAwP), pulmonary artery pressure (PAP), central venous pressure (CVP), cardiac output (CO), portal venous pressure (PvP), and hepatic venous pressure (HvP). P values represent the ANOVA test. * Significantly different from baseline, P < 0·05. | ||
Total hepatic blood flow decreased from 152 ± 8 to 35 ± 6 ml min-1 (100 g)-1. Simultaneously, portal venous flow fell from 130 ± 8 to 21 ± 4 ml min-1 (100 g)-1 (P < 0·001). By contrast, decreases in arterial hepatic flow were modest until more than 30 % of EBV was drained (23 ± 3 to 14 ± 2 ml min-1 (100 g)-1; Fig. 3). Of the total cardiac output, 28 ± 2 % passed through the liver. This relationship was not affected by blood loss. This finding reflects the fact that the proportion of cardiac output that passed through the hepatic artery increased from 4 ± 0 to 9 ± 1 % (at 30 % bleeding, with no further change after additional bleeding, P < 0·001), while the proportion that passed through the portal vein decreased from 24 ± 2 to 14 ± 2 % (P < 0·001). Portal venous pressure was 8 ± 1 mmHg and was not significantly affected by haemorrhage (Fig. 2), while hepatic venous pressure was decreased when more than 20 % EBV was removed (4·6 ± 0·7 to 1·8 ± 0·6 mmHg). Renal blood flow diminished progressively from 285 ± 22 to 44 ± 10 ml min-1 (100 g)-1.
 |
View larger version [in this window] [in a new window] |
|
|
Values are means ± S.E.M. Total hepatic blood flow (HBF), portal venous blood flow (PBF), hepatic arterial blood flow (HaBF), and renal blood flow (RBF). P values represent the ANOVA test. * Significantly different from baseline, P < 0·05. | ||
During progressive haemorrhage, hepatic arterial vascular resistance decreased from 43 ± 4 to 28 ± 4 × 103 dyn s cm-5 and further to 25 ± 3 × 103 dyn s cm-5 after 40 % bleeding (Fig. 4), whereas portal vascular resistance increased from 0·2 ± 0·1 to 1·6 ± 0·8 × 103 dyn s cm-5. Splanchnic vascular resistance increased from 8 ± 1 to 24 ± 4 dyn s cm-5 and renal vascular resistance was elevated from 16 ± 2 to 52 ± 10 dyn s cm-5. Thus, the hepatic arterial vascular bed was selectively dilated during haemorrhage.
 |
View larger version [in this window] [in a new window] |
|
|
Values are means ± S.E.M. Portal venous resistance (PVR), hepatic arterial vascular resistance (HaVR), splanchnic vascular resistance (SplVR), and renal vascular resistance (RVR). P values represent the ANOVA test. * Significantly different from baseline, P < 0·05. | ||
Baseline hepatic oxygen consumption was 20 ± 5 ml min-1 and remained stable until 40 % of EBV was drained (Fig. 5). The amount of oxygen derived from the two vascular systems of the liver, however, changed markedly. Before bleeding, 32 ± 4 % of hepatic oxygen uptake was derived from the arterial system; this increased to 67 ± 5 % after 30 % bleeding and to 74 ± 6 % after a 40 % blood loss. Neither splanchnic nor renal oxygen consumption changed significantly, as the decreases in blood flow to these two regions was compensated by increased oxygen extraction. Hepatic tissue oxygenation, expressed as HbO2, did not change significantly, whereas renal tissue oxygenation was reduced when more than 10 % of EBV was drained (Fig. 5).
 |
View larger version [in this window] [in a new window] |
|
|
Values are means ± S.E.M. Systemic oxygen consumption (VO2), hepatic oxygen consumption, hepatic arterial oxygen consumption (HaVO2), portal oxygen consumption, splanchnic oxygen consumption, renal oxygen consumption, and renal and hepatic tissue oxygenation by near-infrared spectroscopy (NIR). P values represent the ANOVA test. * Significantly different from baseline, P < 0·05. | ||
The arterial, portal, hepatic and renal venous concentrations of sodium and glucose did not change significantly during bleeding (Table 1). The arterial and venous concentrations of potassium increased when blood loss corresponded to 40 % EBV. Furthermore, the pH of the portal, hepatic and renal venous blood was reduced at 40 % EBV with no change in arterial values. Standard base excess was decreased in all vascular beds with a 40 % blood loss. Blood lactate remained stable until blood loss reached more than 30 %. At this stage blood lactate increased from 1·1 ± 0·2 to 3·9 ± 0·8 mmol l-1 in the artery (baseline to 40 %). The corresponding venous values were 1·1 ± 0·2 to 5·3 ± 0·9 mmol l-1 in the portal vein, 0·8 ± 0·1 to 5·3 ± 0·8 mmol l-1 in the hepatic vein, and 0·9 ± 0·2 to 3·0 ± 0·6 mmol l-1 in the renal vein (Fig. 6). Hepatic lactate uptake decreased from 289 ± 55 to 77 ± 73 µmol min-1 after drainage of 30 % of EBV, and was absent when 40 % of EBV was drained. Renal lactate uptake was reduced from 120 ± 55 to 44 ± 9 µmol min-1 (P < 0·01).
Table 1. Blood biochemical variables during haemorrhage
| Blood volume depletion (%) | ||||||
| 0 | 10 | 20 | 30 | 40 | ||
| Sodium (mmol l-1) |
Artery | 139 ± 2 | 139 ± 2 | 138 ± 2 | 139 ± 2 | 137 ± 2 |
| Portal vein | 140 ± 2 | 140 ± 2 | 140 ± 2 | 138 ± 2 | 139 ± 2 | |
| Hepatic vein | 140 ± 1 | 141 ± 3 | 139 ± 2 | 140 ± 2 | 138 ± 3 | |
| Renal vein | 140 ± 2 | 139 ± 2 | 139 ± 2 | 139 ± 2 | 138 ± 3 | |
| Potassium (mmol l-1) |
Artery  |
3·6 ± 0·4 | 3·8 ± 0·4 | 3·9 ± 0·5 | 4·2 ± 0·5 | 4·8 ± 0·6 * |
| Portal vein |
3·8 ± 0·4 | 4·0 ± 0·4 | 4·2 ± 0·5 | 4·5 ± 0·6 | 5·7 ± 1·2 * | |
| Hepatic vein |
3·7 ± 0·4 | 4·0 ± 0·5 | 4·2 ± 0·4 | 4·7 ± 0·8 | 6·6 ± 1 * | |
| Renal vein |
3·6 ± 0·4 | 3·7 ± 0·5 | 4·0 ± 0·5 | 4·6 ± 0·6 | 5·5 ± 1 * | |
| pH | Artery | 7·39 ± 0·08 | 7·39 ± 0·09 | 7·38 ± 0·08 | 7·35 ± 0·08 | 7·32 ± 0·06 |
| Portal vein |
7·34 ± 0·08 | 7·33 ± 0·08 | 7·30 ± 0·08 | 7·25 ± 0·11 | 7·06 ± 0·13 * | |
| Hepatic vein |
7·34 ± 0·09 | 7·33 ± 0·09 | 7·28 ± 0·08 | 7·24 ± 0·11 | 7·05 ± 0·15 * | |
| Renal vein |
7·38 ± 0·08 | 7·38 ± 0·09 | 7·36 ± 0·10 | 7·34 ± 0·09 | 7·18 ± 0·12 * | |
| Standard base excess (mmol l-1) |
Artery |
1·9 ± 2·1 | 1·6 ± 1·6 | 1·4 ± 1·4 | 0·5 ± 2·1 | -4·8 ± 3·1 * |
| Portal vein |
3·5 ± 1·9 | 3·4 ± 1·4 | 3·2 ± 1·4 | 2·1 ± 1·8 | -5·1 ± 3·2 * | |
| Hepatic vein |
3·7 ± 2·2 | 3·9 ± 1·7 | 3·9 ± 1·1 | 3·1 ± 1·9 | -4·6 ± 4·1 * | |
| Renal vein |
3·6 ± 1·9 | 3·9 ± 1·4 | 3·7 ± 1·5 | 2·8 ± 1·2 | -1·6 ± 2·6 * | |
| Glucose (mmol l-1) |
Artery | 5·5 ± 0·4 | 4·5 ± 0·3 | 5·5 ± 0·5 | 4·2 ± 0·5 | 6·4 ± 0·6 |
| Portal vein | 4·8 ± 0·4 | 4·1 ± 0·5 | 4·2 ± 0·3 | 4·2 ± 0·4 | 4·4 ± 0·6 | |
| Hepatic vein | 4·5 ± 0·6 | 5·1 ± 0·5 | 5·2 ± 0·5 | 4·2 ± 0·5 | 7·5 ± 0·7 | |
| Renal vein | 4·7 ± 0·4 | 5·2 ± 0·5 | 4·7 ± 0·5 | 5·2 ± 0·6 | 6·4 ± 0·6 | |
| Noradrenaline (nmol l-1) |
Artery |
1·61 ± 0·21 | 2·01 ± 0·19 | 2·63 ± 0·33 | 3·07 ± 0·45 * | 3·25 ± 0·50 * |
| Portal vein |
1·73 ± 0·33 | 2·23 ± 0·42 | 3·11 ± 0·57 | 4·11 ± 0·67 * | 5·30 ± 0·71 * | |
| Hepatic vein |
1·64 ± 0·20 | 2·01 ± 0·31 | 2·68 ± 0·40 | 3·41 ± 0·44 * | 4·21 ± 0·52 * | |
| Renal vein |
1·77 ± 0·39 | 2·31 ± 0·83 | 3·46 ± 0·79 | 5·58 ± 0·89 * | 7·12 ± 0·94 * | |
| Adrenaline (nmol l-1) |
Artery | 0·63 ± 0·10 | 0·68 ± 0·14 | 0·73 ± 0·16 | 0·81 ± 0·17 | 1·01 ± 0·18 |
| Portal vein | 0·58 ± 0·11 | 0·62 ± 0·18 | 0·70 ± 0·13 | 0·78 ± 0·15 | 0·98 ± 0·18 | |
| Hepatic vein | 0·58 ± 0·10 | 0·64 ± 0·12 | 0·69 ± 0·13 | 0·76 ± 0·14 | 0·85 ± 0·14 | |
| Renal vein | 0·57 ± 0·16 | 0·63 ± 0·15 | 0·68 ± 0·16 | 0·79 ± 0·18 | 0·96 ± 0·16 | |
Significantly changed by bleeding, P < 0·05. * Significantly (P < 0·05) different from baseline.
 |
View larger version [in this window] [in a new window] |
|
|
Values are means ± S.E.M. Arterial (A), portal venous (PV), hepatic venous (HV), and renal venous (RV) lactate concentrations and hepatic and renal lactate uptake. P values represent the ANOVA test. * Significantly different from baseline, P < 0·05. | ||
The arterial and venous portal, hepatic and renal concentrations of noradrenaline were elevated during haemorrhage (Table 1). The most pronounced increases were in portal venous blood (from 1·7 ± 0·3 to 5·3 ± 0·4 nmol l-1) and renal venous blood (from 1·8 ± 0·7 to 7·1 ± 0·9 nmol l-1). The increases in hepatic venous and arterial concentrations were from 1·6 ± 0·2 to 4·2 ± 0·5 nmol l-1 and 1·6 ± 0·2 to 3·2 ± 0·3 nmol l-1, respectively.
From the vascular bed draining to the portal vein, noradrenaline spillover was elevated from 40 ± 12 to 118 ± 29 pmol min-1. Similarly, the renal spillover was increased from 15 ± 4 to 62 ± 21 pmol min-1 (P < 0·01; Fig. 7). Hepatic uptake increased from 32 ± 12 to 69 ± 24 pmol min-1. However, additional haemorrhaging decreased hepatic uptake to 26 ± 10 pmol min-1.
 |
View larger version [in this window] [in a new window] |
|
|
Values are means ± S.E.M. Hepatic noradrenaline uptake (Hep NA uptake), splanchnic noradrenaline spillover (Spl NAspill) and renal noradrenaline spillover (Ren NAspill) during progressive hypovolaemia. P values represent the ANOVA test. * Significantly different from baseline, P < 0·05. | ||
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The results of the present study indicate that as a consequence of progressive haemorrhage in the pig, the hepatic arterial vascular system was selectively dilated and hepatic arterial blood flow represented an increasing part of cardiac output. The hepatic arterial flow was maintained until just prior to circulatory collapse and the fall in the portal venous blood flow resulted in a pronounced reduction in total hepatic blood flow. In consequence, hepatic oxygen consumption was maintained until just prior to circulatory collapse and the hepatic surface tissue oxygenation did not change. A maintained hepatic arterial flow was required to preserve hepatic oxygenation because the portal blood flow not only decreased, but the portal venous blood became increasingly desaturated with oxygen (from 88 to 24 %) because of increased oxygen extraction in the splanchnic area.
Our findings on noradrenaline showed that spillover increased consistent with sympathetic activation and vasoconstriction in the vascular supply to the splanchnic area and the kidney. Indeed, the renal tissue oxygenation was reduced even in response to modest bleeding. The liver metabolizes catecholamines and during mild hypovolaemia the uptake of noradrenaline was increased (Fig. 7). However, the elimination was reduced during severe hypovolaemia and at that stage the blood levels increased sharply (Table 1). Concurrently, lactate elimination from both the liver and the kidney became low or undetectable with subsequent elevations of the plasma concentration.
The hepatic uptake of a substance depends on both the liver cell function and the hepatic blood flow. According to the sinusoidal perfusion model, the hepatic extraction of a substance is expressed as 1 - e-PS/HBF, where PS is a constant for each substance (Keiding, 1991). This model describes normal liver physiology provided that all sinusoids are recruited. Under this assumption a reduction of the hepatic blood flow will be followed by a more modest reduction in the hepatic uptake of a substance, because the extraction fraction increases. In the present study the hepatic uptake of lactate decreased by 73 % after drainage of 30 % EBV and was eliminated after a 40 % blood loss, while the hepatic blood flow was reduced by only 62 % and 77 %, respectively. In contrast to the normal physiological situation, the hepatic lactate uptake was reduced more than the hepatic blood flow. This suggests that the intrinsic ability of the liver to take up lactate was reduced by haemorrhage. Ischaemia could not account for this observation because the hepatic oxygen uptake and the tissue oxygenation were preserved (Fig. 5). One possibility is that uptake of lactate was affected by progressive derecruitment (collapse) of the liver sinusoids, which would lead to deviation from the sinusoidal perfusion model in the observed direction.
The noradrenaline data are consistent with the idea that the liver sinusoids collapse in response to haemorrhage. At baseline, the liver extracted noradrenaline from the portal blood and maintained the hepatic venous noradrenaline concentration close to the arterial concentration. The baseline uptake of 31 pmol min-1 was increased to 56 pmol min-1 at 10 % blood loss and 69 pmol min-1 at 20 % blood loss. In these situations, the hepatic noradrenaline uptake increased more than the portal noradrenaline concentration even though the hepatic blood flow was reduced. This indicates that the intrinsic ability for hepatic uptake of noradrenaline was upregulated during the blood loss. However, after a 30 and 40 % blood loss the hepatic noradrenaline uptake declined to 55 and 26 nmol min-1, respectively, as would be expected if partial sinusoidal collapse took place. Our observations, therefore, suggest sinusoidal 'collapse' after a 30-40 % blood loss. A similar phenomenon has been reported in the cat, where a reduction of hepatic blood flow was associated with a derecruitment of 50 % of the sinusoids (Greenway, 1989).
Central pressures were reduced in parallel with the amount of blood withdrawn except for pulmonary arterial pressure which did not change significantly. This was not due to the type of anaesthesia chosen as the pulmonary artery pressure was also stable during bleeding in pigs under halothane anaesthesia (Krantz et al. 1997). The portal venous pressure remained constant even after a 77 % reduction of the hepatic blood flow. This was a consequence of an increase in the vascular resistance of the portal venous system. These findings are in accordance with the demonstration of hepatic venous sphincters, which are located in the larger hepatic veins (Lautt et al. 1986) and constrict in response to noradrenaline, angiotensin and sympathetic nerve activation (Lautt et al. 1987; Lautt & Legare, 1987). These sphincters ensure the maintenance of the porto-systemic pressure gradient. Reduction of the tone in these vessels allows for the portal venous flow to increase without a rise in the portal venous pressure, even to the extent that resistance may approach zero (Lautt & Greenway, 1987). Our study indicates that these mechanisms are active even during severe haemorrhagic stress.
Several intrinsic as well as extrinsic factors may influence hepatic arterial blood flow. Intrinsic regulation balances the arterial and portal flow, so that a reduction of portal venous flow is followed by an increase in arterial hepatic flow or an increase in portal flow is followed by a reduction of hepatic arterial flow, but there is no regulation of portal venous flow as a consequence of changes in hepatic arterial flow (Lautt, 1981). The mechanism behind this 'arterial buffer response' is not known. Lautt (1981) suggested as an intrinsic mechanism that a reduction of portal venous flow leads to a reduced washout of adenosine, assumed to be produced at a constant rate in the hepatic sinusoids. As adenosine dilates the hepatic artery, changes in adenosine washout by changes in portal venous flow would explain the arterial response to the changes in portal venous flow. Extrinsic regulation of liver blood flow is very complex. The liver receives nerve fibres from both vagi, the right phrenic nerve, and sympathetic ganglia T7-T10 (Biolac-Sage et al. 1990) with a large variation among species (Ballet, 1990; Lautt, 1990). Stimulation of 
-receptors constricts the hepatic artery (Greenway et al. 1967; Richardson & Withrington, 1977), whereas -receptor stimulation dilates the hepatic artery (Greenway & Lawson, 1967). The general response to sympathetic stimulation is constriction of the hepatic artery (Greenway & Lawson, 1967), and cholinergic nerves are of little importance (Greenway et al. 1967). Our results could be explained if haemorrhage leads to a specific decrease in sympathetic action towards the hepatic arterial circulation.
Many hormones affect hepatic blood flow. Catecholamines reduce the hepatic arterial flow and also increase the portal pressure with no change in portal blood flow (Ross & Kurrasch, 1969; Richardson & Withrington, 1978). Therefore, circulating catecholamine levels may contribute to the preserved portal pressure but not to the selective reduction of the hepatic vascular resistance and flow during haemorrhage. Whether other humoral mechanisms maintain the hepatic arterial flow during severe haemorrhage cannot be explained from this study.
In conclusion we found that during haemorrhage and subsequently reduced cardiac output and portal blood flow, one or more mechanisms selectively dilated the hepatic arterial vascular bed. This maintained hepatic cellular oxygenation and hepatic metabolism until just prior to circulatory collapse. The previously described hepatic arterial buffer response to changes in portal blood flow is one explanation of our findings and our study demonstrates that this compensatory mechanism is capable of maintaining hepatic perfusion and oxygenation even during severe haemorrhage.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| Ballet, F. (1990). Hepatic circulation: Potential for therapeutic intervention. Pharmacology and Therapeutics 47, 281-328 | [Medline] |
| Ben-Jonathan, N. & Porter, J. C. (1976). A sensitive radioenzymatic assay for dopamine, norepinephrine and epinephrine in plasma and tissue. Endocrinology 98, 1497-1507 | [Abstract] |
| Biolac-Sage, P., Lafon, M. E., Saric, J. & Balabaud, C. (1990). Nerves and perisinusoidal cells in human liver. Journal of Hepatology 10, 105-112. | [Medline] |
| Bledtfeldt, J. E., Riley, E. M. & Groszmann, R. J. (1985). Compensatory mechanism in response to an elevated hepatic consumption in chronic eternol-fed rats. American Journal of Physiology 248, G507-511 | [Medline] |
| Burton-Opitz, R. (1910). The vascularity of the liver. I. The flow in the hepatic artery. Journal of Experimental Physiology 3, 297-313. | |
| Ejlersen, E., Skak, C., Møller, K., Pott, F. & Secher, N. H. (1995). Central cardiovascular variables at a maximal mixed venous oxygen saturation in severe hepatic failure. Transplantation Proceedings 27, 3506-3507. | [Medline] |
| Greenway, C. V. (1989). Derecruitment in cat liver: extension of undistributed parallel tube model to effect of low hepatic blood flow on ethanol uptake. Canadian Journal of Physiology and Pharmacology 67, 1225-1231 | [Medline] |
| Greenway, C. V. & Lautt, W. W. (1989). Hepatic circulation. In Handbook of Physiology, section 6, The Gastrointestinal System, vol. 1, Motility and Circulation, part 2, ed. Schultz, S. G., pp. 1519-1564. American Physiological Society, Bethesda, MD, USA. | |
| Greenway, C. V. & Lawson, A. E. (1967). -Adrenergic receptors in the hepatic arterial bed of the anesthetized cat. Canadian Journal of Physiology and Pharmacology 47, 415-419. |
|
| Greenway, C. V., Lawson, A. E. & Mellander, S. (1967). The effect of stimulation of the hepatic nerves, infusion of noradrenaline and occlusion of the carotid arteries in the anesthetized cat. Canadian Journal of Physiology 192, 21-41. | |
| Keiding, S. (1991). Dynamic aspects of hepatic removal of circulating substances. In Oxford Textbook of Clinical Hepatology, ed. McIntyre, N., Benhamou, J., Bircher, J., Rizzetto, M. & Rodes, J., pp. 78-87. Oxford Publishers, Oxford. | |
| Krantz, T., Sztuk, F., Swiatek, F., Jakobsen, J. & Secher, N. H. (1997). Central cardiovascular and oxygen variables during haemorrhage in the pig. Acta Anaesthesiologica Scandinavica 41, 719-724 | [Medline] |
| Lautt, W. W. (1980). Control of the hepatic arterial blood flow: Independence from liver metabolic activity. American Journal of Physiology 239, H559-564 | [Medline] |
| Lautt, W. W. (1981). Role and control of the hepatic artery. In Hepatic Circulation in Health and Disease, ed. Laut,W. W., pp. 203-226. Raven Press, New York. | |
| Lautt, W. W. (1990). Hepatic nerves: A review of their function and effects. Canadian Journal of Physiology and Pharmacology 58, 105-123. | |
| Lautt, W. W. & Greenway, C. V. (1987). Conceptual review of the hepatic vascular bed. Hepatology 7, 952-963 | [Medline] |
| Lautt, W. W., Greenway, C. V. & Legare, D. J. (1986). Localization of intrahepatic portal vascular resistance. American Journal of Physiology 14, G375-381. | |
| Lautt, W. W., Greenway, C. V. & Legare, D. J. (1987). Effect of hepatic nerves, norepinephrine, angiotensin, elevated central venous pressure on postsinusiodal resistance sites and intrahepatic pressure. Microvascular Research 33, 50-61 | [Medline] |
| Lautt, W. W. & Legare, D. J. (1987). Effect of histamine, noradrenaline and nerves on vascular resistance and pressures in dog liver. American Journal of Physiology 252, G472-478 | [Medline] |
| Madsen, P. L. & Secher, N. H. (1998). Near-infrared oxymetry of the brain. Progress in Neurobiology (in the Press). | |
| Ott, P., Keiding, S. & Bass, L. (1993). Plasma elimination of indocyanine green in the intact pig after bolus injection and during constant infusion: comparison of spectrophotometry and high- pressure liquid chromatography for concentration analysis. Hepatology 18, 1504-1515 | [Abstract] |
| Richardson, P. D. I. & Withrington, P. G. (1977). The role of -adrenoreceptor in the responses of the hepatic arterial vascular bed of the dog to phenylephrine, isoprenaline, noradrenaline and adrenaline. British Journal of Pharmacology 60, 239-249 |
[Medline] |
| Richardson, P. D. I. & Withrington, P. G. (1978). Pressure-flow relationships and effects of noradrenaline and isoprenaline on hepatic arterial and portal venous vascular beds in the dog. The Journal of Physiology 282, 451-470 | [Abstract] |
| Ross, G. & Kurrasch, M. (1969). Adrenergic responses of the hepatic circulation. American Journal of Physiology 16, 1380-1385. | |
| Wray, S., Cope, M., Delpy, D. T., Wyatt, J. S. & Reynolds, E. O. (1988). Characterization of the near infrared absorption spectra of cytochrome aa3 and haemoglobin for the non-invasive monitoring of cerebral oxygenation. Biochimica et Biophysica Acta 933, 184-192 | [Medline] |
This study was supported by The Boel Foundation. We thank Søren Haagen Nielsen, Charlotte Sick Nielsen, Letty Klarskov and Mette Olesen for excellent technical assistance.
Corresponding author
A. Rasmussen: Department of Surgical Gastroenterology and Transplantation C, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen Ø, Denmark.
Email: rh01976{at}rh.dk
This article has been cited by other articles:
|
|
 |
|
  S. M. Jakob Splanchnic Blood Flow in Low-Flow States Anesth. Analg., April 1, 2003; 96(4): 1129 - 1138. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. B. Nielsen, J. O. Clemmesen, C. Skak, P. Ott, and N. H. Secher Attenuated hepatosplanchnic uptake of lactate during intense exercise in humans J Appl Physiol, April 1, 2002; 92(4): 1677 - 1683. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Jakob, J. J. Tenhunen, S. Laitinen, A. Heino, E. Alhava, and J. Takala Effects of systemic arterial hypoperfusion on splanchnic hemodynamics and hepatic arterial buffer response in pigs Am J Physiol Gastrointest Liver Physiol, May 1, 2001; 280(5): G819 - G827. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
