Splanchnic Haemodynamics and Vasoactive Agents in Experimental Cirrhosis

It is well known that portal hypertension is associated with a hyperdynamic systemic circulatory state. This study measures systemic and splanchnic haemodynamics in an experimental rat model of hepatic cirrhosis. It also investigates the association between haemodynamic changes in cirrhotic animals and circulating levels of the vasoactive hormones glucagon and vasoactive intestinal polypeptide (VIP). Splanchnic blood flow was significantly increased in the cirrhotic group (13.2 ± 1.3 vs. 9.2 ± 1.6 ml/min, P < 0.05). Circulating levels of glucagon and VIP were two and five fold increased respectively in cirrhotic animals compared to controls. There was a strong correlation between portal pressure and glucagon levels in the cirrhotic group (r = 0.85). Raised splanchnic blood flow is partly responsible for elevated portal pressure in this model and this rise may be humorally mediated.


ertension is
ssociated with hyperdynamic systemic and splanchnic circulatory states 1'2.It has also been demon- strated that raised splanchnic blood flow plays an important role in the maintenance of a chronically elevated portal pressure in experimental portal hyper- tension3''.Several theories have been proposed to explain the splanchnic hyperaemia observed in portal hypertension including decreased sensitivity of splanchnic arterioles to noradrenaline 5 and the opening of intestinal arteriovenous shunts6.It has also been sug- gested that circulating vasoactive agents, normally catabolised by the liver, are present in increased quan- tities in hepatic cirrhosis and may mediate these haemodynamic changes7.Glucagon, a potent intesti- nal vasodilator, is present in increased concentration in the plasma of patients and animals with portal hypertension and has been implicated in the intestinal hyperaemia associated with this conditiona-l.The present study defines the systemic and splanchnic haemodynamics changes in an animal model ofexperi- mental cirrhosis, and assesses their relationships with circulating concentration of putative splanchnic vaso- dilators (glucagon and vasoactive intestinal peptide).


MATERIALS AND METHODS

Male Sprague-Dawley rats

antin and Kingman Ltd.
K) were housed in a controlled environment with a 12-hour light: dark cycle.They were fed standard rat diet but were fasted overnight in wire bottom cages prior to measurement of splanchnic haemodynamics.


Production of Experimental Cirrhosis

Hepatic cirrhosis was produced using oral administra- tion of carbon tetrachloride (CC14) as previously de- scribed11.Animals weighing 150 +__ 20 gm were given phenobarbitol (150/100ml) in their drinking water to induce the enzyme cytochrome P450 which has been shown to increase rat liver sensitivity to CCI,12.This hepatotoxin was given once weekly using an orogastric tube under light ether anaesthesia.The initial dose was 0.15ml and each subsequent weekly dose was cal- culated according to the animals weight loss following the previous dose.Carbon tetrachloride was stopped as soon as ascites developed and experiments were performed 3 weeks later.Control animals of similar weight also received phenobarbitol by weekly gavage under ether anaesthesia and were weighed daily; how- ever they did not receive carbon tetrachloride.We have previously found 11 that 40% of animals receiving CC14 as described above will develop ascites and micro- nodular hepatic cirrhosis (Figure 1).A total of 15 ani- mals were gavaged With CC14 and the six animals which developed ascites in this group had haemo- dynamic experiments performed.


Surgical Preparation,

Systemic and splanchn

haemodynamics were me
s- ured using a reference microsphere sample tech- nique13.Each animal had a tracheostomy performed and was ventilated using a Harvard rat ventilator (model 68) with a 2:1 nitrous oxide--oxygen mixture in 0.5 % halothane.The femoral artery was cannulated on both sides and mean arterial pressure recorded using a strain gauge transducer connected to a Gould 8000 semi-dual recorder.Arterial blood gas concentra- tions were analysed (ABL, Radiometer) and only ani- mals with a P02 over 90 mmHg, a PC02 in the range 35-42 mmHg and a pH greater than 7.35 were included in the study.Rectal temperature was monitored and kept at 37__+0.5C by means of a heat lamp.The abdomen was opened using a midline incision and a cannula inserted into the ileocolic branch of the portal vein to measure portal venous pressure.The abdomen was closed and portal pressure measured a minimum of 30 minutes later and only when a respir- atory pattern was present on portal venous tracing.Mean arterial and portal pressure measurements were taken simultaneously.Finally a cannula was inserted into the left ventricle via the right carotid artery for injection of radio-labelled microspheres.Insertion was performed under continuous pressure monitoring and correct positioning was verified both by the presence of a ventricular pattern on pressure recording and by checking the position of the tip of the cannula at the end of the experiment.


Haemodynamic Measurements

Once the animal w

haemodynamically stable a
d pressure measurements had been recorded, approxi- mately 100,000 Ruthenium-labelled microspheres were injected into the left ventricle over 25 seconds 13.The microspheres were suspended in 10% Dextran with a drop of 0.01% "Tween" 80 added to prevent clump- ing.The amount of radioactivity in the injection syringe (volume 0.2ml) was measured in a gamma- scintillation counter (Packard 5000) and the syringe was vortexed for 30 seconds prior to injection.Arterial pressure was recorded continuously during infusion and a change in pressure of more than 10% over basal value led to automatic exclusion.The femoral artery cannula was connected to

periphe
al 2 ml syringe, which was attached to a with- drawal pump (Sage Instruments).Wi hdrawal was commenced 10 seconds before intra-ventricular injec- tion of microspheres and stopped 60 seconds later.The reference sample in the syringe therefore served as an "artificial organ" with a known blood flow (1 ml/min).

The radioactivity in the syringe was counted in a gamma-scintillation counter.The degree of portasy- stemic shunting was measured by injecting 100,000 Tn-labelled microspheres into the ileocolic branch of the portal vein as previously described14.The animal was killed 2 minutes later and the lungs, liver, splanchnic organs and kidneys were removed and placed in vials for measurement of radioactivity.


Calculations

Cardiac output (ml/min) Injected radioactivity (dpm)x Reference organ flow (ml/min)

Reference organ radioactivity (dp

ood flow (ml/min) Organ radioactivity
dpm)x Reference organ flow (ml/min)

Reference organ radioactivity (dpm)

Portal venous inflow was the sum of flows to the stomach, spleen, small and large bowel, and mesentery.

Vascular resistance (mm Hg/ml/min) Change in pressure across vascular bed (mm Hg) Blood flow to vascular bed (ml/min) Portasystemic Shunting (%)

Lung radioactivity (dpm)

x 100

Liver + lung radioactivity (dpm)

Uniform mixing of microspheres in blood was checked by comparing radioactivity in both kidneys and animals with a greater than 10% difference were excluded.

Measurement of Glucagon and VIP At the end of the experiment blood was tak

from the inferior vena
ava for measurement of glucagon and vasoactive intestinal polypeptide levels.Each sample was placed in a heparinised tube and frozen.A radio- immunoassay technique was used to measure glucagon and vasoactive intestinal polypeptide 15,16.


Statistical Analysis

Results are presented as mean __+ standard error of the mean (mean + SEM).The statistical tests used are the Mann-Whitney ind Pearsons correlation coefficient.


RESULTS


Systemic and splanchnic haemodynamics

Portal venous pressure was significantly higher in cirrhotic animals compared to

ontrols (Ta
le 1).

Portasystemic shunting was not detectable in controls but 23.4% portal blood was diverted into the systemic circulation in cirrhotic animals.

Portal hypertension was associated with a rise in cardiac output and reduced total portal resistance (Table 1).There was a significant rise in splanchnic blood flow which was associated with a reduction in splanchnic arteriolar resistance.Total portal resistance was 50% greater in cirrhotic animals but this difference compared to controls was not statistically significant.


Hormonal Concentrations

There was a five fold increase in the level of vasoactive intestinal polypeptide in the cirrhotic group compared to controls (Table 2).However there was no correla- tion between portal pressure and levels of vasoactive intestinal polypeptide in cirrhotic animals.Although glucagon levels were not significantly increased in cirrhotic animals (Table 2) there was a strong correla- tion (r=0.85) between portal venous pressure and glucagon levels (Figure 2).There was no correlation between glucagon levels and the magnitude of porta- systemic shunting.Total peripheral (mmHg/ml/min) 1.1 +/-0.1 1.7 +/-0.2*resistance Splanchnic inflow (ml/min) 13.2 1.3 9.2 + 1.6" Splanchnic arteriolar (mmHg/ml/min) 8.0 + 1.3 15.1 +/-3.1"resistance Total portal resistance (mmHg/ml/min) 1.2 + 0.1 0.8 0.1"

Values are mean +/-SEM; * p < 0.01


DISCUSSION

This study shows that experimental cirrhosis is asso- ciated with hyperdynamic systemic and splanchnic circulatory states.These results are in agreement with previous observations that raised splanchnic inflow is the primary factor in the maintenance of an elevated portal pressure in experimental portal hyperten- sion 1'2'3.It has been suggested that this rise in splan- chnic inflow may be hormonally mediatedT.Glucagon and vasoactive intestinal polypeptide are known to increase splanchnic blood flow.These hormones are produced by splanchnic organs and are normally catabolised by the liver.In portal hypertension how- ever, glucagon and vasoactive intestinal polypeptide may escape into the systemic circulation via portasy- stemic collateral vessels.

The second aim of the present study therefore was to measure the levels of glucagon and vasoactive intes- tinal peptide in portal hypertension and to establish whether they were associated with the rise in portal pressure seen in experimental cirrhosis.The results show that both glucagon and va oactive intestinal polypeptide levels are increased in experimental cirr- hosis.They also demonstrate a strong correlation be- tween portal pressure and glucagon levels suggesting that this agent may be responsible in part for the rise in portal pressure seen in cirrhotic animals.In contrast there was no association between circulating levels of vasoactive intestinal polypeptide and portal venous pressure.It is not clear why these hormones are raised in the model of cirrhosis.One possible mechanism is that they pass directly into the systemic circulation in portasystemic collateral vessels and therefore escape catabolism in the liver.There was however no correlation between circulating levels of glucagon and the magnitude of portasystemic shunting seen in animals with cir hosis.Portasystemic collateral vessels are only partly responsible however for physiological shunting in portal hypertension.Reduced catabolism of sub- stances normally degraded by the liver will also give rise to physiological shunting in hepatic cirrhosis.It is possible that the rise in glucagon levels in portal hypertensive animals in the present study may be due in part to reduced catabolism by a cirrhotic liver.Pancreatic hypersecretion of glucagon is known to occur in ex- perimental portal hypertension and this could have also contributed to the raised glucagon levels observed in the present study17.

There is now strong experimental evidence to sug- gest that humoral factors mediate the splanchnic hyperaemia seen in chronic prehepatic portal hypertension.It has been shown that cross-perfusion of the mesenteric circulation of normal rats with arterial blood from rats with portal hypertension results in a 30% increase in intestinal blood flow.Figure 2 There was a strong correlation between glucagon levels and portal pressure measurements in cirrhotic animals.

Recent studies have concentrated on the role of glucagon as a possible mediator of this response as it reduces intestinal arteriolar resistance and levels are known to be elevated in the plasma of patients and animals with portal hypertensions'9.This hypothesis was tested by Benoit and colleagues 1 who showed that administration of glucagon antiserum caused a 30% reduction in portal venous inflow in rats with portal hypertension.

The present study differs from

revious studies in
hat it measures splanchnic haemodynamics, glucagon and vasoactive intestinal polypeptide levels, and portal venous pressure in experimental cirrhosis.The results support the contention of Benoit et al. that glucagon plays a role in the maintenance of a raised portal pressure in portal hypertension 7,t.Given that resis- tance to portal blood flow also plays a role in elevating portal pressure, it is likely that there is a synergistic effect between increased portal venous inflow and resistance to flow resulting in portal hypertension.Raised portal pressure in hepatic cirrhosis could there- fore be explained by the following sequence of events.

As cirrhosis develops, portal pressure rises and beyond a critical pressure portasystemic collateral vessels open.Humoral agents such as glucagon are released into the systemic circulation through these shunts giving rise to a reduction in splanchnic arteriolar resis- tance.The resulting increase in splanchnic blood flow then produces a rise in portal venous pressure.The evidence supporting this h pothesis however is limited almost exclusively to animal studies.Similar experi- ments in humans are difficult to perform because of the lack of established non-invasive techniques to measure splanchnic blood flow, magnitude of portasystemic shunting and portal venous pressure.The present study suggests that glucagon may play a role in the maintenance of raised portal pressure but this hypothesis needs to be tested in a clinical setting.


INVITED COMMENTARY

Several years ago, we proposed a relationship between portosystemic shunting, glucagon and the hemo- dynamic consequences of chronic portal hypertension.The essence of our hypothesis was that portosystemic shunting led to increased circulating levels of glucagon which began to act as a dilator of the splanchnic vasculature.As a result, the hyperdynamic splanchnic circulation developed.In the years since this initial report, several studies have examined the role of hu- moral factors in portal hypertensive conditions.The work of Geraghty et al., joins a growing body of literature supporting the role ofglucagon as a mediator of the hemodynamic derangements in chronic portal hypertension.A major finding of the present study was a correlation between portal pressure and glucagon levels in cirrhotic rats.The authors suggested that glucagon may mediate t e increased portal pressure.There are two possible mechanisms by which glueagon could raise portal pressure: 1) increased portal venous inflow and 2) increased portal vascular resistance.

Previous reports from our laboratory have clearly shown that glucagon is capable of increasing portal venous inflow in chronic portal hypertension.Further- more, we have suggested that the increased portal inflow can accou