Acta anaesth. scand. 1977, 2 1, 457-469
The Different Responses of the Hepatic Arterial Bed to Hypovolaemia and to Halothane Anaesthesia MAGNA ANDREEN, LARSIRESTEDT and BENGTZETTERSTROM The Department of Anaesthesia and the Thoracic Surgery Research Laboratory, Karolinska sjukhuset, Stockholm, Sweden
Ten dogs were subjected to a period of hypovolaemia (bleeding volume: 2% of body weight) and to a period of halothane anaesthesia (end-tidal halothane concentration: 1%). Mean arterial blood pressure decreased to 79% of control value during hypovolaemia and to 58% of control value during halothane anaesthesia. Mean total peripheral and preportal vascular resistances increased during hypovolaemia and were unchanged during halothane. Mean hepatic arterial and portal venous blood flows decreased to 82% and 55% of control values, respectively, during hypovolaemia, and to 41% and 56% of control value, respectively, during exposure to halothane. Mean hepatic arterial resistance was unchanged during hypovolaemia, but increased during halothane. Mean hepatic oxygen consumption did not change significantly during hypovolaemia, but decreased during halothane anaesthesia, in spite of a n increased extraction of oxygen from both the hepatic arterial and the portal venous blood. Possible mechanisms which may maintain oxygen supply to the liver by increasing the hepatic arterial fraction of total liver blood flow when portal venous blood flow is reduced are discussed. I t is concluded that this mechanism is upset or inhibited during halothane anaesthesia.
Receiued I 4 January, accepted for publication 3 February 1977
Halothane anaesthesia may depress myocardial contractility and cardiac output (SEVERINCHAUS & CULLEN1958). Regional blood flow is influenced by changes in both cardiac output and regional vascular tone. The response to halothane varies in different vascular beds (WESTERMARK 1969, AMORY et al. 1971). Thus, during halothane anaesthesia, portal venous and hepatic arterial blood flow proved to decrease in the dog (THULIN et al. 1975). The decrease in the hepatic arterial blood flow was more pronounced than in the portal venous blood flow, due to a rise in resistance to flow in the arterial vascular bed. These results do not agree with the concept of an intrinsic mechanism in the liver, which would regulate the flow in the hepatic artery so as to compensate for impairments in
portal blood supply (BURTON-OPITZ 191 1, SCHWIEGK1932, SANCETTA1953, SCHENR et al. 1962, COHN& KOUNTZ1963, FISCHER & TAKACS 1964, TERNBERC & BUTCHER 1965, HANSON& JOHNSON 1966, GREENWAY et al. 1967b). The present study was undertaken in order to examine further the influence of halothane anaesthesia on liver blood flow and liver oxygen consumption, and to compare these effects with those obtained during a moderate hypovolaemic hypotension leading to approximately the same decrease in portal venous flow rate as the exposure to halothane. I t would then be possible to determine whether the hepatic arterial vascu1a.r bed reacts differently in response to a decreased portal flow rate due to halothane anaesthesia or to hypovolaemia.
458
M. ANDREEN, L. IRESTEDT AND B. ZETTERSTROM
MATERIAL AND METHODS Ten mongrel dogs, weight 19-35 kg, which had fasted overnight, were anaesthetized with thiopental (25 mg/kg b.w.) intravenously. Anaesthesia was maintained with 70% N,O in O2and supplementary doses of thiopental (5 mg/kg b.w.) when needed. The animals were endotracheally intubated and intermittent positive pressure ventilation was provided by a n Engstrom respirator (LKB Medical, Stockholm, Sweden) a t a frequency of 20/min. Ventilation was adjusted according to arterial Pco2 in order to achieve normocapnia. Surgical preparation. The abdomen was opened through a right subcostal incision. Great care was taken to ensure good haemostasis. The gastroduodenal artery was ligated close to the junction between the proper hepatic arteries and the common hepatic artery to allow measurement of proper hepatic arterial blood flow. Dissection of the common hepatic artery and the portal vein was made without severing perivascular nerves. Catheters for collecting samples of blood were placed in the right femoral artery, the portal vein, one hepatic vein and the pulmonary artery. The portal catheter was inserted via a small mesenteric tributary and advanced into the common portal vein close to the liver. The hepatic venous catheter was introduced through the external jugular vein. The correct location of this catheter was verified at autopsy. Pulmonary arterial blood was collected through a floating catheter inserted into the pulmonary artery. The position of this catheter was verified by pressure tracing. BloodJows. Calibrated flow probes were placed on the common hepatic artery and the portal vein, and blood flows in these vessels were measured with square-wave electromagnetic flow meters (Nycotron AS., Drammen, Norway), Cardiocoutput. Cardiac output was measured by thermodilution with a cardiac output computer (Type 3750, Cardio Vascular Instruments Ltd., Welwyn Garden City, U.K.) (ANDKEEN 1974). In the present study, saline was used instead of glucose as the thermal indicator. Blood pressures. Arterial blood pressure was measured in the abdominal aorta through a catheter inserted via the left femoral artery. Portal venous and hepatic venous blood pressures were measured through the blood-collecting catheters. Pulmonary arterial blood pressure was measured through the thermistor catheter. The catheters were connected to transducers. Mean blood flows and mean blood pressures were continuously recorded on a Grass Polygraph (Grass Instrument Co., Quincy, Mass., U.S.A.) Haemoglobin concentration and oxygen saturation. Arterial,
portal venous, hepatic venous and pulmonary arterial blood samples were collected simultaneously. Oxygen content was calculated according to the formula ( H b (g/l) x 1.34~ So,%/lOO+P~, (kPa) x 0.23) by duplicate determinations of haemoglobin concentration and oxygen saturation on a Co-oximeter (IL 182, Lexington, Mass., U.S.A.). Blood gases and acid-base slate. Arterial Pol was determined with an oxygen electrode (Radiometer, Copenhagen, Denmark). Arterial Pco2 and p H were determined according to ASTRUP (1956) with a pH-electrode (Radiometer, Copenhagen, Denmark). Base excess was calculated according to SICGAARD ANDERSEN & ENCEL(1960). Temperature. Blood temperature was measured via thc thermistor in the pulmonary artery. End-tidal halothane concentration. End-tidal halothane concentration was measured continuously with a mass spectrometer (MGA 200, Centronic, Croydon, U.K.) which was calibrated with a Vapor vaporizer (Dragerwerk, Lubeck, West Germany) (HILL1963). The mass spectrometer sampling catheter was connected to the endotracheal tubc. Experimental procedure. After preparation of the animal had been completed, a n interval of 30 min was allowed to elapse to establish a circulatory steady state. No further thiopental was given during this period. Each animal was exposed to one period of halothane anaesthesia and one period of hypovolaemia. Five animals had their halothane period first, and, after an interval of at least 60 min during which no halothane was administered, they were exposed to a period of hypovolaemia. Five animals started with a hypovolaemic period, and, after retransfusion of the shed blood, an interval of a t least 60 min was allowed to elapse before they were exposed to halothane. Blood samples for measurements of arterial, portal venous, hepatic venous and pulmonary arterial oxygen content were collected immediately before and a t the end of each period. No thiopental was given during the halothane or the hypovolaemic periods or 30 min before any period. Halothane period. Halothane was added to the nitrous oxide-oxygen mixture in a concentration of 1.5-2.0% by the Vapor vaporizer. The animals were a n a n thetized for 25 min, so that arterial blood pressure decreased to about 60% of control values, which was expected to cause a n equally large decrease in portal venous blood f l ~ w(THULIN et al. 1975). Hypouolaeinic period. The left femoral arterial catheter was connected to a n open, hcparinized glass bottle. The animals were bled for 15-20 min until a blood
RESPONSES OF HEPATIC ARTERIAL BED
volume corresponding to approximately 2% of b.w. was removed, which was expected to cause a fall in portal venous blood flow of approximately the same order of magnitude as the halothane anaesthesia (HINSHAW et al. 1961). When the blood had been removed, the connection between the arterial catheter and the reservoir was closed and the hypovolaemic level was maintained for another 5-10 min, so that the total hypovolaemic period lasted 25 min. Before and at the end of each period, cardiac output was determined by triplicate determinations. During the periods, cardiac output was repeatedly determined once or twice every 5 minutes. Arterial blood gases and acid-base state were checked during the preparation of the animal and before each experimental period. Arterial Pc0, was maintained between 35 and 45 mmHg (4.5 and 6.0 kPa) (mean 36 mmHg = 4.8 kPa) and base excess between + 3 and -3 mmol/l (mean value -1 mmol/l). Metabolic acidosis was prevented and, when necessary, corrected with 0.6 M sodium bicarbonate. Blood temperature was kept between 36" and 37.5"C (mean value 36.4"C) by means of a heat pad. Mean decrease in blood temperature during the experiment was 0.5" C. Isotonic saline and Ringer's solution were infused intravenously throughout the experiment to cover fluid losses by evaporation and surgical bleeding. Approximate infusion rate was 0.4-0.5 l/h. Average haematocrit of all animals was 41 vol % at the beginning and 39 vol % at the end of the experiment. The difference was not significant. Calculations Pressures are expressed in mmHg and kPa ( I mmHg = 0.133 kPa). Vascular resistance. Vascular resistances were calculated according to the general formula (R = AP/Q), where R denotes resistance, A P blood pressure gradient in kPa, and blood flow in ml/min. Caval venous pressure was assumed to be 0 kPa.
0
Oxygen consumption. Total liver oxygen uptake was determined as the sum of hepatic arterial oxygen uptake and portal venous oxygen uptake. Hepatic arterial oxygen uptake was calculated as arteriohepatic venous oxygen content difference multiplied by hepatic arterial blood flow. Portal venous oxygen uptake was calculated as portal-hepatic venous oxygen content difference times portal venous blood flow. The values were expressed in per cent of total liver oxygen uptake. Whole body oxygen uptake was calculated as the product of arterio-pulmonary arterial oxygen content difference and cardiac output. Estimation of mean arterial halothane concentration. Mean end-tidal halothane concentration was multiplied by the halothane blood/gas partition coefficient (2.3)
459
(LARSON et al. 1962) to reach mean arterial concentration in vol %. This fraction was turned into g halothane/litre blood by dividing by Avogrado's number (22.414) and multiplying by the molerular weight for halothane (197.39) (SECHER 1971).
(x
fs.d.) of Statistics. Means and standard deviation controls and experimental values were calculated. Student's t-test for paired values was used to establish the statistical significance of the observed changes. The following probability (P)levels of significance were used: P
The Different Responses of the Hepatic Arterial Bed to Hypovolaemia and to Halothane Anaesthesia MAGNA ANDREEN, LARSIRESTEDT and BENGTZETTERSTROM The Department of Anaesthesia and the Thoracic Surgery Research Laboratory, Karolinska sjukhuset, Stockholm, Sweden
Ten dogs were subjected to a period of hypovolaemia (bleeding volume: 2% of body weight) and to a period of halothane anaesthesia (end-tidal halothane concentration: 1%). Mean arterial blood pressure decreased to 79% of control value during hypovolaemia and to 58% of control value during halothane anaesthesia. Mean total peripheral and preportal vascular resistances increased during hypovolaemia and were unchanged during halothane. Mean hepatic arterial and portal venous blood flows decreased to 82% and 55% of control values, respectively, during hypovolaemia, and to 41% and 56% of control value, respectively, during exposure to halothane. Mean hepatic arterial resistance was unchanged during hypovolaemia, but increased during halothane. Mean hepatic oxygen consumption did not change significantly during hypovolaemia, but decreased during halothane anaesthesia, in spite of a n increased extraction of oxygen from both the hepatic arterial and the portal venous blood. Possible mechanisms which may maintain oxygen supply to the liver by increasing the hepatic arterial fraction of total liver blood flow when portal venous blood flow is reduced are discussed. I t is concluded that this mechanism is upset or inhibited during halothane anaesthesia.
Receiued I 4 January, accepted for publication 3 February 1977
Halothane anaesthesia may depress myocardial contractility and cardiac output (SEVERINCHAUS & CULLEN1958). Regional blood flow is influenced by changes in both cardiac output and regional vascular tone. The response to halothane varies in different vascular beds (WESTERMARK 1969, AMORY et al. 1971). Thus, during halothane anaesthesia, portal venous and hepatic arterial blood flow proved to decrease in the dog (THULIN et al. 1975). The decrease in the hepatic arterial blood flow was more pronounced than in the portal venous blood flow, due to a rise in resistance to flow in the arterial vascular bed. These results do not agree with the concept of an intrinsic mechanism in the liver, which would regulate the flow in the hepatic artery so as to compensate for impairments in
portal blood supply (BURTON-OPITZ 191 1, SCHWIEGK1932, SANCETTA1953, SCHENR et al. 1962, COHN& KOUNTZ1963, FISCHER & TAKACS 1964, TERNBERC & BUTCHER 1965, HANSON& JOHNSON 1966, GREENWAY et al. 1967b). The present study was undertaken in order to examine further the influence of halothane anaesthesia on liver blood flow and liver oxygen consumption, and to compare these effects with those obtained during a moderate hypovolaemic hypotension leading to approximately the same decrease in portal venous flow rate as the exposure to halothane. I t would then be possible to determine whether the hepatic arterial vascu1a.r bed reacts differently in response to a decreased portal flow rate due to halothane anaesthesia or to hypovolaemia.
458
M. ANDREEN, L. IRESTEDT AND B. ZETTERSTROM
MATERIAL AND METHODS Ten mongrel dogs, weight 19-35 kg, which had fasted overnight, were anaesthetized with thiopental (25 mg/kg b.w.) intravenously. Anaesthesia was maintained with 70% N,O in O2and supplementary doses of thiopental (5 mg/kg b.w.) when needed. The animals were endotracheally intubated and intermittent positive pressure ventilation was provided by a n Engstrom respirator (LKB Medical, Stockholm, Sweden) a t a frequency of 20/min. Ventilation was adjusted according to arterial Pco2 in order to achieve normocapnia. Surgical preparation. The abdomen was opened through a right subcostal incision. Great care was taken to ensure good haemostasis. The gastroduodenal artery was ligated close to the junction between the proper hepatic arteries and the common hepatic artery to allow measurement of proper hepatic arterial blood flow. Dissection of the common hepatic artery and the portal vein was made without severing perivascular nerves. Catheters for collecting samples of blood were placed in the right femoral artery, the portal vein, one hepatic vein and the pulmonary artery. The portal catheter was inserted via a small mesenteric tributary and advanced into the common portal vein close to the liver. The hepatic venous catheter was introduced through the external jugular vein. The correct location of this catheter was verified at autopsy. Pulmonary arterial blood was collected through a floating catheter inserted into the pulmonary artery. The position of this catheter was verified by pressure tracing. BloodJows. Calibrated flow probes were placed on the common hepatic artery and the portal vein, and blood flows in these vessels were measured with square-wave electromagnetic flow meters (Nycotron AS., Drammen, Norway), Cardiocoutput. Cardiac output was measured by thermodilution with a cardiac output computer (Type 3750, Cardio Vascular Instruments Ltd., Welwyn Garden City, U.K.) (ANDKEEN 1974). In the present study, saline was used instead of glucose as the thermal indicator. Blood pressures. Arterial blood pressure was measured in the abdominal aorta through a catheter inserted via the left femoral artery. Portal venous and hepatic venous blood pressures were measured through the blood-collecting catheters. Pulmonary arterial blood pressure was measured through the thermistor catheter. The catheters were connected to transducers. Mean blood flows and mean blood pressures were continuously recorded on a Grass Polygraph (Grass Instrument Co., Quincy, Mass., U.S.A.) Haemoglobin concentration and oxygen saturation. Arterial,
portal venous, hepatic venous and pulmonary arterial blood samples were collected simultaneously. Oxygen content was calculated according to the formula ( H b (g/l) x 1.34~ So,%/lOO+P~, (kPa) x 0.23) by duplicate determinations of haemoglobin concentration and oxygen saturation on a Co-oximeter (IL 182, Lexington, Mass., U.S.A.). Blood gases and acid-base slate. Arterial Pol was determined with an oxygen electrode (Radiometer, Copenhagen, Denmark). Arterial Pco2 and p H were determined according to ASTRUP (1956) with a pH-electrode (Radiometer, Copenhagen, Denmark). Base excess was calculated according to SICGAARD ANDERSEN & ENCEL(1960). Temperature. Blood temperature was measured via thc thermistor in the pulmonary artery. End-tidal halothane concentration. End-tidal halothane concentration was measured continuously with a mass spectrometer (MGA 200, Centronic, Croydon, U.K.) which was calibrated with a Vapor vaporizer (Dragerwerk, Lubeck, West Germany) (HILL1963). The mass spectrometer sampling catheter was connected to the endotracheal tubc. Experimental procedure. After preparation of the animal had been completed, a n interval of 30 min was allowed to elapse to establish a circulatory steady state. No further thiopental was given during this period. Each animal was exposed to one period of halothane anaesthesia and one period of hypovolaemia. Five animals had their halothane period first, and, after an interval of at least 60 min during which no halothane was administered, they were exposed to a period of hypovolaemia. Five animals started with a hypovolaemic period, and, after retransfusion of the shed blood, an interval of a t least 60 min was allowed to elapse before they were exposed to halothane. Blood samples for measurements of arterial, portal venous, hepatic venous and pulmonary arterial oxygen content were collected immediately before and a t the end of each period. No thiopental was given during the halothane or the hypovolaemic periods or 30 min before any period. Halothane period. Halothane was added to the nitrous oxide-oxygen mixture in a concentration of 1.5-2.0% by the Vapor vaporizer. The animals were a n a n thetized for 25 min, so that arterial blood pressure decreased to about 60% of control values, which was expected to cause a n equally large decrease in portal venous blood f l ~ w(THULIN et al. 1975). Hypouolaeinic period. The left femoral arterial catheter was connected to a n open, hcparinized glass bottle. The animals were bled for 15-20 min until a blood
RESPONSES OF HEPATIC ARTERIAL BED
volume corresponding to approximately 2% of b.w. was removed, which was expected to cause a fall in portal venous blood flow of approximately the same order of magnitude as the halothane anaesthesia (HINSHAW et al. 1961). When the blood had been removed, the connection between the arterial catheter and the reservoir was closed and the hypovolaemic level was maintained for another 5-10 min, so that the total hypovolaemic period lasted 25 min. Before and at the end of each period, cardiac output was determined by triplicate determinations. During the periods, cardiac output was repeatedly determined once or twice every 5 minutes. Arterial blood gases and acid-base state were checked during the preparation of the animal and before each experimental period. Arterial Pc0, was maintained between 35 and 45 mmHg (4.5 and 6.0 kPa) (mean 36 mmHg = 4.8 kPa) and base excess between + 3 and -3 mmol/l (mean value -1 mmol/l). Metabolic acidosis was prevented and, when necessary, corrected with 0.6 M sodium bicarbonate. Blood temperature was kept between 36" and 37.5"C (mean value 36.4"C) by means of a heat pad. Mean decrease in blood temperature during the experiment was 0.5" C. Isotonic saline and Ringer's solution were infused intravenously throughout the experiment to cover fluid losses by evaporation and surgical bleeding. Approximate infusion rate was 0.4-0.5 l/h. Average haematocrit of all animals was 41 vol % at the beginning and 39 vol % at the end of the experiment. The difference was not significant. Calculations Pressures are expressed in mmHg and kPa ( I mmHg = 0.133 kPa). Vascular resistance. Vascular resistances were calculated according to the general formula (R = AP/Q), where R denotes resistance, A P blood pressure gradient in kPa, and blood flow in ml/min. Caval venous pressure was assumed to be 0 kPa.
0
Oxygen consumption. Total liver oxygen uptake was determined as the sum of hepatic arterial oxygen uptake and portal venous oxygen uptake. Hepatic arterial oxygen uptake was calculated as arteriohepatic venous oxygen content difference multiplied by hepatic arterial blood flow. Portal venous oxygen uptake was calculated as portal-hepatic venous oxygen content difference times portal venous blood flow. The values were expressed in per cent of total liver oxygen uptake. Whole body oxygen uptake was calculated as the product of arterio-pulmonary arterial oxygen content difference and cardiac output. Estimation of mean arterial halothane concentration. Mean end-tidal halothane concentration was multiplied by the halothane blood/gas partition coefficient (2.3)
459
(LARSON et al. 1962) to reach mean arterial concentration in vol %. This fraction was turned into g halothane/litre blood by dividing by Avogrado's number (22.414) and multiplying by the molerular weight for halothane (197.39) (SECHER 1971).
(x
fs.d.) of Statistics. Means and standard deviation controls and experimental values were calculated. Student's t-test for paired values was used to establish the statistical significance of the observed changes. The following probability (P)levels of significance were used: P
![Responses of the sympathetically-innervated hepatic arterial vascular bed of the dog to intra-arterial injections of dopamine [proceedings].](/assets/img/hold.png)
