Clinical Science ( 1990) 78, 193- I97
193
Effects of haemorrhage and volume expansion on portal-systemic collateral vascular resistance in conscious portal hypertensive rats ABRAHAM KOSHY, TATSUYA SEKIYAMA, JEAN-MICHEL CEREDA, ANTOINE HADENGUE, CATHERINE GIROD AND DIDIER LEBREC Unit6 de Recherches de Physiopathologie Hkpatique (INSERM U-24). Hbpital Beaujon, Clichy, France
(Received 2 June/ll September 1989; accepted 2 October 1989)
SUMMARY
INTRODUCTION
1. In order to study the acute effects of blood volume changes on the vascular resistance of portal-systemic collaterals (collateral vascular resistance), a model of total portal vein occlusion with 100% portal-systemic shunts was developed in the rat. In this model, we determined the haemodynamic effects of haemorrhage (1.8 m1/100 g body weight) or intravenous infusion of a volume expander (1.8 m1/100 g body weight). Cardiac output and regional blood flows were measured by the radioactive microsphere method. 2. Haemorrhage significantly reduced arterial pressure from 1 0 8 + 4 to 9 2 f 4 mmHg ( m e a n f s m ) , cardiac output from 5 6 f 4 to 2 4 f 2 ml min-' 100 g-' body weight, portal pressure from 15.1 f 1.5 to 10.0 f 1.4 mmHg and portal tributary blood flow from 19.9 f 2.3 to 8.3 1.4 ml/min. Consequently, collateral vascular resistance significantly increased from 6.6 f0.9 x 103 to 11.1 f2.0 x 1O3 kPa 1- s. 3. Volume expansion reduced arterial pressure from 98 3 to 90 f3 mmHg, and significantly increased cardiac output from 4 3 f 3 to 55*3 ml min-I 100 g-' body weight, portal pressure from 13.9f0.7 to 16.5 f0.8 mmHg and portal tributary blood flow from 16.4+ 1.3 to 28.2 f3.2 ml/min. Consequently, collateral vascular resistance significantly decreased from 7.0 f0. 5 x 103to 4.9 f0.4 x 10' kPa I-' s. 4. This study shows that in rats with portal hypertension, portal-systemic collateral vascular resistance is modified by alterations in blood volume.
Patients with cirrhosis have higher cardiac output and body fluid volume but lower systemic vascular resistance than normal subjects [l]. Although plasma volume expansion increases [2, 31 and depletion decreases [4, 51 portal pressure in cirrhosis, we are unaware of studies in patients, on the effects of blood volume changes on portal venous and collateral blood flows. Such studies have been hampered by the fact that, in man, accurate measurement of absolute portal-systemic collateral blood flow is not feasible. Even in animals, a model of partially ligated portal vein with incomplete shunting has been used to obtain an approximate estimate of collateral blood flow and vascular resistance [6]. Thus, we developed a model of total portal vein occlusion with 10O0/~portal-systemic shunts which permits accurate measurement of collateral blood flow and hence calculation of vascular resistance in these shunts [7]. It was, therefore, the aim of this study completely to divert the portal tributary blood flow into portal-systemic collaterals and to use this model to ascertain if blood volume changes can alter collateral vascular resistance.
*
+
MATERIALS AND METHODS Animals Sixteen male adult Sprague-Dawley rats (Charles River, Saint-Aubin-lks-Elbeuf, France) were permitted free access to food and water until the evening before the haemodynamic study, when food was withdrawn.
Key words: blood volume, collateral vascular resistance, haemorrhage, portal hypertension, splanchnic circulation.
Portal vein ligation
Correspondence: Dr D. Lebrec, INSERM U-24, HGpital Beaujon, 92 1 I8 Clichy, France.
Total portal vein ligation was performed in two stages since acute portal vein occlusion in animals is fatal [7]. Partial portal vein obstruction was first made in rats
.
,
194
A. Koshy et al.
weighing 160-180 g as previously described [8, 91. The abdomen was opened under light ether anaesthesia and the portal vein ligated over a 0.96 mm polyethylene catheter. The catheter was then removed. A 3-0 silk thread passed through another polyethylene catheter was left around the portal vein to facilitate total ligation withogt major dissection, and the abdomen closed. Three weeks later, the abdomen was reopened under light ether anaesthesia, the polyethylene catheter left in sitir during partial portal vein ligation was removed while retaining the silk thread and the portal vein ligated totally [7]. The abdomen was closed using catgut. All rats survived total ligation of the portal vein. Study protocol Basal haemodynamic studies were performed 2-3 h after total portal vein ligation at least 30 min after reappearance of consciousness when haemodynamics were stable and after at least 15 min of quiet rest. Rats were divided into two groups of eight animals each. Each group underwent basal haemodynamic study immediately followed by haemorrhage or infusion of a volume expander. The volume expander contained 3% (w/v) modified gelatine, 150 mmol of sodium, 5 mmol of potassium and 100 mmol of chloride (Plasmagel, Laboratoires Roger Bellon, Neuilly/Seine, France). Haemorrhage was performed from the femoral artery catheter with a Harvard pump at a rate of 0.75 ml/min for about 8 min into pre-weighed syringe. The actual amount withdrawn was 1.8 m1/100 g body weight. Plasma expander was infused (1.8 m1/100 g body weight) at a rate of 0.75 ml/ min. Haemodynamic measurements were repeated at the tenth min from the start of haemorrhage or infusion. Plasma volume was measured in the basal state by injecting 0.3 ml of 50"/0 (w/v) Evans Blue dye in 0.9% (w/v) NaCl (saline) and withdrawing 0.2 ml of blood after 10 min for spectrophotometric reading of plasma at 610 nm. Packed cell volume was measured in the basal state and immediately after volume expansion, by the capillary method using 0.1 ml of blood. Portal pressure Portal pressure was measured in all rats. A catheter was placed in the trunk of the superior mesenteric vein just before the portal vein and the pressure was recorded in the conscious restrained rat [lo]. Haemodynamic studies All haemodynamic studies were performed in conscious rats placed in restraining cages as previously described [7, 8, 101. Cardiac output and regional blood flows were measured using radioactive microspheres [ 1 11. In brief, the femoral artery, internal jugular vein and left ventricle were catheterized. Rectal temperature was maintained at 37.0 f 0.5"C. A pre-counted aliquot of approximately 60 000 15 pm-diameter I 13Sn-labelled microspheres with a specific activity of 10 mCi/g (New
England Nuclear, Boston, MA, U.S.A.) suspended in 10% (w/v) dextran and ultrasonically agitated, was injected and flushed with 1 ml of saline over 45 s into the left ventricle. Starting 5 s before the microsphere injection, the reference sample was drawn into a Harvard pump driven syringe at a rate of 1 ml/min for 1 min. After haemorrhage or volume expansion, a second injection of 141Ce-labelled microspheres was given and the same technique was used. The animals were then killed with an overdose of pentobarbital. The spleen, stomach, small intestine, colon, mesentery with pancreas, liver and kidneys were counted separately in a y-counter (Compteur Gamma G 4000; Kontron, Montigny-Le-Bretonneux, France) at energy settings of 280-1000 and 70-210 keV for "'Sn and IJ1Ce,respectively. The error due to the spillover of 1 1 3 S ~ into the I4"Ce channel was corrected using "?3n and IJ1Ce standards. We have previously shown that the microsphere method is reproducible in portal hypertensive rats [ lo]. Adequate microsphere mixing was assured by < 10% difference between the two kidneys. One animal having inadequate mixing was rejected. Cardiac output, regional blood flows and vascular resistances were calculated as follows. Cardiac output (ml/min) was calculated as [radioactivity injected (c.p.m.)/reference sample radioactivity (c.p.m.)]x 1.0. Regional organ blood flow was calculated as [organ radioactivity (c.p.m.)/radioactivity injected (c.p.m.)]x cardiac output (ml/min). Portal tributary blood flow was taken as the sum of spleen, stomach, small intestine, colon and mesentery with pancreas blood flows, and represents portal systemic shunt blood flow since shunting is total. Liver represents hepatic arterial blood flow. Splanchnic blood flow is the sum of tributary blood flow and hepatic arterial blood flow. Total vascular resistance was determined as (arterial pressure x 80)/cardiac output. Preliminary studies showed that atrial pressure is zero in rats; this pressure changed by less than 1 mmHg after volume depletion or expansion. Portal tributary vascular resistance was calculated as [(arterial pressure - portal pressure) x 80]/portal tributary blood flow, and portal systemic collateral vascular resistance as (portal pressure x 80)/portal tributary blood flow. Statistical analysis Results are expressed as means f SEM. Student's paired f-test was used to determine differences between paired haemodynamic measurements and analysis of variance was used for unpaired data. 1' values
193
Effects of haemorrhage and volume expansion on portal-systemic collateral vascular resistance in conscious portal hypertensive rats ABRAHAM KOSHY, TATSUYA SEKIYAMA, JEAN-MICHEL CEREDA, ANTOINE HADENGUE, CATHERINE GIROD AND DIDIER LEBREC Unit6 de Recherches de Physiopathologie Hkpatique (INSERM U-24). Hbpital Beaujon, Clichy, France
(Received 2 June/ll September 1989; accepted 2 October 1989)
SUMMARY
INTRODUCTION
1. In order to study the acute effects of blood volume changes on the vascular resistance of portal-systemic collaterals (collateral vascular resistance), a model of total portal vein occlusion with 100% portal-systemic shunts was developed in the rat. In this model, we determined the haemodynamic effects of haemorrhage (1.8 m1/100 g body weight) or intravenous infusion of a volume expander (1.8 m1/100 g body weight). Cardiac output and regional blood flows were measured by the radioactive microsphere method. 2. Haemorrhage significantly reduced arterial pressure from 1 0 8 + 4 to 9 2 f 4 mmHg ( m e a n f s m ) , cardiac output from 5 6 f 4 to 2 4 f 2 ml min-' 100 g-' body weight, portal pressure from 15.1 f 1.5 to 10.0 f 1.4 mmHg and portal tributary blood flow from 19.9 f 2.3 to 8.3 1.4 ml/min. Consequently, collateral vascular resistance significantly increased from 6.6 f0.9 x 103 to 11.1 f2.0 x 1O3 kPa 1- s. 3. Volume expansion reduced arterial pressure from 98 3 to 90 f3 mmHg, and significantly increased cardiac output from 4 3 f 3 to 55*3 ml min-I 100 g-' body weight, portal pressure from 13.9f0.7 to 16.5 f0.8 mmHg and portal tributary blood flow from 16.4+ 1.3 to 28.2 f3.2 ml/min. Consequently, collateral vascular resistance significantly decreased from 7.0 f0. 5 x 103to 4.9 f0.4 x 10' kPa I-' s. 4. This study shows that in rats with portal hypertension, portal-systemic collateral vascular resistance is modified by alterations in blood volume.
Patients with cirrhosis have higher cardiac output and body fluid volume but lower systemic vascular resistance than normal subjects [l]. Although plasma volume expansion increases [2, 31 and depletion decreases [4, 51 portal pressure in cirrhosis, we are unaware of studies in patients, on the effects of blood volume changes on portal venous and collateral blood flows. Such studies have been hampered by the fact that, in man, accurate measurement of absolute portal-systemic collateral blood flow is not feasible. Even in animals, a model of partially ligated portal vein with incomplete shunting has been used to obtain an approximate estimate of collateral blood flow and vascular resistance [6]. Thus, we developed a model of total portal vein occlusion with 10O0/~portal-systemic shunts which permits accurate measurement of collateral blood flow and hence calculation of vascular resistance in these shunts [7]. It was, therefore, the aim of this study completely to divert the portal tributary blood flow into portal-systemic collaterals and to use this model to ascertain if blood volume changes can alter collateral vascular resistance.
*
+
MATERIALS AND METHODS Animals Sixteen male adult Sprague-Dawley rats (Charles River, Saint-Aubin-lks-Elbeuf, France) were permitted free access to food and water until the evening before the haemodynamic study, when food was withdrawn.
Key words: blood volume, collateral vascular resistance, haemorrhage, portal hypertension, splanchnic circulation.
Portal vein ligation
Correspondence: Dr D. Lebrec, INSERM U-24, HGpital Beaujon, 92 1 I8 Clichy, France.
Total portal vein ligation was performed in two stages since acute portal vein occlusion in animals is fatal [7]. Partial portal vein obstruction was first made in rats
.
,
194
A. Koshy et al.
weighing 160-180 g as previously described [8, 91. The abdomen was opened under light ether anaesthesia and the portal vein ligated over a 0.96 mm polyethylene catheter. The catheter was then removed. A 3-0 silk thread passed through another polyethylene catheter was left around the portal vein to facilitate total ligation withogt major dissection, and the abdomen closed. Three weeks later, the abdomen was reopened under light ether anaesthesia, the polyethylene catheter left in sitir during partial portal vein ligation was removed while retaining the silk thread and the portal vein ligated totally [7]. The abdomen was closed using catgut. All rats survived total ligation of the portal vein. Study protocol Basal haemodynamic studies were performed 2-3 h after total portal vein ligation at least 30 min after reappearance of consciousness when haemodynamics were stable and after at least 15 min of quiet rest. Rats were divided into two groups of eight animals each. Each group underwent basal haemodynamic study immediately followed by haemorrhage or infusion of a volume expander. The volume expander contained 3% (w/v) modified gelatine, 150 mmol of sodium, 5 mmol of potassium and 100 mmol of chloride (Plasmagel, Laboratoires Roger Bellon, Neuilly/Seine, France). Haemorrhage was performed from the femoral artery catheter with a Harvard pump at a rate of 0.75 ml/min for about 8 min into pre-weighed syringe. The actual amount withdrawn was 1.8 m1/100 g body weight. Plasma expander was infused (1.8 m1/100 g body weight) at a rate of 0.75 ml/ min. Haemodynamic measurements were repeated at the tenth min from the start of haemorrhage or infusion. Plasma volume was measured in the basal state by injecting 0.3 ml of 50"/0 (w/v) Evans Blue dye in 0.9% (w/v) NaCl (saline) and withdrawing 0.2 ml of blood after 10 min for spectrophotometric reading of plasma at 610 nm. Packed cell volume was measured in the basal state and immediately after volume expansion, by the capillary method using 0.1 ml of blood. Portal pressure Portal pressure was measured in all rats. A catheter was placed in the trunk of the superior mesenteric vein just before the portal vein and the pressure was recorded in the conscious restrained rat [lo]. Haemodynamic studies All haemodynamic studies were performed in conscious rats placed in restraining cages as previously described [7, 8, 101. Cardiac output and regional blood flows were measured using radioactive microspheres [ 1 11. In brief, the femoral artery, internal jugular vein and left ventricle were catheterized. Rectal temperature was maintained at 37.0 f 0.5"C. A pre-counted aliquot of approximately 60 000 15 pm-diameter I 13Sn-labelled microspheres with a specific activity of 10 mCi/g (New
England Nuclear, Boston, MA, U.S.A.) suspended in 10% (w/v) dextran and ultrasonically agitated, was injected and flushed with 1 ml of saline over 45 s into the left ventricle. Starting 5 s before the microsphere injection, the reference sample was drawn into a Harvard pump driven syringe at a rate of 1 ml/min for 1 min. After haemorrhage or volume expansion, a second injection of 141Ce-labelled microspheres was given and the same technique was used. The animals were then killed with an overdose of pentobarbital. The spleen, stomach, small intestine, colon, mesentery with pancreas, liver and kidneys were counted separately in a y-counter (Compteur Gamma G 4000; Kontron, Montigny-Le-Bretonneux, France) at energy settings of 280-1000 and 70-210 keV for "'Sn and IJ1Ce,respectively. The error due to the spillover of 1 1 3 S ~ into the I4"Ce channel was corrected using "?3n and IJ1Ce standards. We have previously shown that the microsphere method is reproducible in portal hypertensive rats [ lo]. Adequate microsphere mixing was assured by < 10% difference between the two kidneys. One animal having inadequate mixing was rejected. Cardiac output, regional blood flows and vascular resistances were calculated as follows. Cardiac output (ml/min) was calculated as [radioactivity injected (c.p.m.)/reference sample radioactivity (c.p.m.)]x 1.0. Regional organ blood flow was calculated as [organ radioactivity (c.p.m.)/radioactivity injected (c.p.m.)]x cardiac output (ml/min). Portal tributary blood flow was taken as the sum of spleen, stomach, small intestine, colon and mesentery with pancreas blood flows, and represents portal systemic shunt blood flow since shunting is total. Liver represents hepatic arterial blood flow. Splanchnic blood flow is the sum of tributary blood flow and hepatic arterial blood flow. Total vascular resistance was determined as (arterial pressure x 80)/cardiac output. Preliminary studies showed that atrial pressure is zero in rats; this pressure changed by less than 1 mmHg after volume depletion or expansion. Portal tributary vascular resistance was calculated as [(arterial pressure - portal pressure) x 80]/portal tributary blood flow, and portal systemic collateral vascular resistance as (portal pressure x 80)/portal tributary blood flow. Statistical analysis Results are expressed as means f SEM. Student's paired f-test was used to determine differences between paired haemodynamic measurements and analysis of variance was used for unpaired data. 1' values
