Total occlusion (or embolization) of proper hepatic artery (HA) is generally well tolerated when performed in patients with normal hepatic function (usually for massive traumatic liver bleeding) but risk for hepatic failure is high in patients with poor liver function and limited hepatic reserve. Selective HA has been used safely in patients with compromised liver function (e.g. cirrhosis) usually for hepatocellular carcinoma in patients who are not surgical candidates. I am not aware of any clinical studies involving total or selective HA embolization to correct the arterio-portal shunts that develop in patients with cirrhosis or portal hypertension.

Thank you.

I do not mean clinical or experimental studies either, just to discuss a vague concept of a possible open arterio-portal shunts' impact on natural history of the liver cirrhosis. In fact, it was my students who suggested the idea yesterday, and I did not know what to answer.

Sorry I misunderstood the question. No doubt the development of arterio-portal shunting exacerbates any variceal bleeding episode due to the high arterial pressures and blood flows. It would be interesting if we could identify subpopulations of portal hypertension patients who have no arterio-portal shunting and those that do and follow natural clinical history to see if progression to variceal bleeding episodes is more rapid in the group with arterio-portal shunting. But I don't know how those subgroups could be identified. Your student is quite perceptive - my credit to the mentor!

It was my fault, so I beg your pardon.

The mentor was learning Physics long time ago so that forgot it thoroughly, but my studentS (not one) do not. They asked why the VENOUS blood from the esophageal varixes is not venous in colour, but is crimzon, or cherry-red like an arterial blood, and why nobody has noticed seemingly this contradiction and explained it?

And they suggested their own version as follows:

1) The Elias's paradox: according to the famous and reprinted everywhere concept by H.Elias, the lobular branch of the hepatic artery with the biliary duct run in a tender trabecula of connective tissue...and opens into central sinusoid of the app. same caliber [H.Elias. A re-examination of the structure of the mammalian liver. Am.J.Anat.1949; 85(3): 379-456. Plate 8 on the page 443 and explanation on the previous page].

The blood pressure in a hepatic artery is about 100 mm Hg and that one in a sunusoid is about 5 mm Hg. Why the blood does not escape from the artery to the portal system causing the portal hypertension development under the normal conditions?

So, they rendered this picture:

I answered, not at once though, that possible explanations for this phenomenon were suggested by E.H.Bloch. The termination of hepatic arterioles and the functional unit of the liver as determined by microscopy of the living organ. Ann. NY Acad of Sciences, p.78-87, where he found the special arterio-portal sphincters in the liver (Fig.1 of the article cited). However, further studies failed to prove the sphincters existence or that one of the so-called 'intralobular arteriole', for example,

Matsumoto T, Komori R, Magara T, Ui T, Kawakami M, Tokuda T, Takasaki S, et al. A study on the normal structure of the human liver, with special reference to its angioarchitecture. Jikeikai Med J 1979;26: 1-40.

As a result, Ekataksin & Wake, 1997, supposed that the hepatic artery branches supply the non-hepatocytes liver cells only and their blood do not intermingle with that of the sinusoids' untill it turned into the post-capillar venous (and of low pressure) blood [Ekataksin W, Wake K. New concepts in biliary and vascular anatomy of the liver. Prog Liv Dis 1997;15:1-30]. It sounds logical, but remained unproved still, to the best of knowledge.

Taking apart Ekataksin concept for a moment, my students suggested then an explanation of their own based on liquids laws. I saw the next picture:

As the Bernoulli law tells:

Delta P is directly proportional to the square of the D, where P - is the pressure in a given vessel, D - the vessel's diameter, and minus the friction force that may be neglected. So, in case the diameter of a vessel increased as twice, the resulting pressure decrease will be 4-fold. So, to solve the Elias's paradox we just need to find the square root from 20 times of the difference between the hepatic artery and sinusoid pressure: app. 4.5 We found that to solve this problem and equilibrate the pressure difference, the hepatic artery branch should be connected with 4-5 sinusoidal vessels of the same diameter, like on this picture:

Of course, there are other explanations may be taken. For example, S.Takasaki &H.Hano. Three-dimensional observations of the human hepatic artery. J.Hepat 2001; 34: 455-66, provided images with the portal (lobular) vein is 4-5-fold as wider comparing with the so-called axial artery (Fig.1-4) that meet the same hydrodynemic criteria.

However, the students' hypothesis is interesting by that it may suggest futher development of the liver microcirculation, namely, in case the aforementioned proportion becomes altered to the arterial side, the arterio-portal escape should be going on, as rendered on the pictured. And this version is the only or so it seems at any rate that could explain the bleeding with the scarlet blood from esophageal varixes.

The short answer is no – there is no sense to interrupt the hepatic arterial inflow in liver cirrhosis as the other major inflow arteries contributing to portal venous blood pull are still there. It just can worsen the course of disease due the sharp reduction of oxygen supplementation of the liver. The portal hypertension, irrespective of the cause, raised due to increase of intrahepatic vascular resistance. A well known systemic and splanchnic reaction to portal hypertension is the hyperdynamic circulation with precapillary arterio-venous shunting of various degrees (hepato-pulmonary syndrome is a good example). The amount of the shunt inversely correlates with tissue oxygen extraction (saturation in lungs). So it is not surprising that the blood coming from variceal bleeding seems little bit more reddish. And the regulatory mechanisms are much more complex than just fluid mechanics can explain. However the initial problem in cirrhosis with portal hypertension remain the increased outflow vascular resistance.

I can recommend one of the good reviews on the subject:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2999290/pdf/WJH-2-208.pdf

Another reason the blood coming from variceal bleeding appears more "red" than typical controlled venupuncture lab collection is that the bleeding varix has physically "punctured" through the surface of stomach or esophageal mucosal lining causing tissue disruption including small arteries and arterioles in the submucosal tissues. The blood from clinical bleeding varices is likely s mixuture of predominantly portal venous circulation but also small amounts of arterial and venous blood.

Dr.Julianov,

Thank you for the link, it is quite plausible and, by a lucky chance, has been familiar to me already. I am aware as well that the regulatory mechanisms involved are complex and remain unclear to some, or preciser, to the major extent.

However, there is evident lack of the properly hydrodynamic studies in the liver cirrhosis animal models (what is difficult from both practical and theoretical points of view) with profit of the pharmacological and biochemical that leads to lots of contradictions between various studies results. One may tell that is normal for the ongoing scientific research, but the disintegrated condition is fruitless without an ingretavie concepts, the simplier the better.

I should repete again, that thai idea is not mine, but of my students' of the 4th year, the pediatrician phaculty. However I stand on sceptical positions about it and about possibilities of various treatment modalities to succeed in liver cirrhosis as a whole, this idea seems to me interesting in terms that future researchers will develop a novel, non-routine approach to this end-stage disease of the liver management.

The hyperdynamic circulation described in the portal hypertension syndrome is hardly appropriate for the 'reddish' colour of the esophageal bleedings, and what is why for:

The hyperdynamic condition is relative one to that that commonly known in cases of the hypovolemic shock with the only difference that the shunting should be by far less on its degree in the former. The shock lasts hours, the portal hypertension does several years as usual. However, the venous blood is of usual 'venous' colour in shock.

To the best of my knowledge, there is no data about the O2 portal blood saturation;

To the best of my knowledge, nobody has counted a ratio between the sinusoids and axial arteries' number in the cirrhosis (or other causes of the P.Hypertension).

To the best of my knowledge, nobody performed the partial or total hepatic artery ligation in experimental animals to determine the effect of that action.

At last, to cite the review provided by Dr.Julianov:

"...Portal hypertensive rats treated with signalling inhibitors of VEGF and platelet

derived growth factor (PDGF) significantly reduce their porto-systemic collaterization[33]. Also, the use, in experimental rat models of portal hypertension, of Sorafenib, a potent inhibitor of proangiogenic VEGF receptor-2 and PDGF receptor-β, induced an important decrease in splanchnic neovascularisation and in the extent of porto-systemic collaterals, along with a marked attenuation of hyperdynamic splanchnic and systemic circulations[34]."

This piece does not contradict to the idea discussed here about the failed intrahepatic (or extrahepatic) arterio-sinusoid ratio in the liver cirrhosis complicated with the portal hypertension.

Thank you Dr. Borodach, for your comment and questions. Actually most of them have their answers as they are important for liver surgery. Oxygen saturation in hepatic artery is as usual (nearly 95%) and saturation of portal blood is greater than other veins - during the fasting state is nearly 85%. So the liver receives its oxygen supply equally from portal and arterial inflow vessels. There are a lot of clinical data about the consequences of partial or total hepatic artery ligation and hepatic surgeons take in consideration them In everyday practice. Clinical data shows that interruption of blood flow of the proper hepatic artery in healthy liver carries 35-50% risk of death due to liver failure. What figures to expect in friable diseased liver? However you are absolutely right for animal experiments – the architecture and regulation of hepatic blood flow in humans is not completely comparable to them.

Thank you for the post, Dr.Julianov,

At first, those oxygen saturation data can be applid to the healthy liver only to the best of my knowledge, and even in this case many questions remained unclear.

For example, what is happining with the hepatocytes' oxygen supply during feeding and what is the role of the hepatic artery inflow for this period, and how it may be altered if feeding regimen changed (obesity, etc.) I could not find in literature.

As for "clinical data show(s) that interruption of blood flow of the proper hepatic artery in healthy liver carries 35-50% of death" (the upper figure may be too high, from my point of view, and concerns possibly to the hepatic artery's injury and hemorrhagic shock)... So, the data are about the healthy liver, and the cirrhosis is absolutely another story, so that the results of the proper hepatic artery ligation in cirrhotics may differ, to worse or better.

I am agreed that the liver remains the most enigmatic organ after the brain, the organ that went through seemingly 500,000,000 year long evolutional path and was invented under those conditions far beyond any extant creature to survive. But this was a lyrical digression, sorry.

The mysterious point is that the eosophageal varices bleedings are with light, arterial-like blood. As that from the hemorrhoids, for example. It is strange enough, is not it?

Dear Dr. Borodach,

Portal vein oxygen saturation does not depend so much on the liver as the 02 extraction is related to the metabolism of the other splanchnic organs. And during feeding portal venous blood saturation normally can drops down to 65%. So called hepatic artery buffer response (HABR) play an important role in keeping inflow/saturation to the liver. And the HABR is so independent that it works even in transplanted livers. Regarding hepatic artery or its branches ligation – almost all of the relevant clinical data comes from planned or incidental surgical ligation during elective surgery, not in trauma or other hemodinamically uncontrolled conditions. All those questions are well explored and you can find the answers easily in the recent authoritative liver surgery/transplantation textbooks as well those focused primarily on hepatic circulation in health and disease. If you keep in mind 85% saturation of portal blood in fasting, and add the well known up to 30% aterio-portal shunt in portal hypertension with well developed porto-systemic collaterals (as in bleeding varices), the color of the blood does not seems to be so red I think. It is however a very good question.

Dr.Julianov provided here the most profound and insightful posts as his usual way is.

There is little to add is left, I think.

1) It is a pity that liver morphologists do not join this discussion.

2) The HABR does work when the liver is more or less healthy, even if being transplanted. The point is that even at an early stage of the liver fibrosis the HABR is distorted by prolonged pathologically compensatory arterial inflow [PMCID: PMC3259325, a section 'Imaging vascular changes'; etc.]. So, one could speculate that that shift may imitate the condition of the prolonged or continuous metabolic state of the ‘faulse satiety’. Indeed, the progressive muscle depletion and the total body weight loss are characteristic for advanced stages of liver fibrosis/cirrhosis.

3) On the other hand, this arteriolization of the hepatic circulation is hardly curable with medication. So, there might exist a hope on the vascular interventional therapy, very frail and elusive as it is. It would be too risky due to the increasing porto-hepatic shunting that leads to a hypoxic environment of the hepatic tissue in advanced cirrhosis. Perhaps, an experimental study should concern to a certain early stage of the liver fibrosis?

There are several of clinical indications for embolisation of pathology via the hepatic artery. Intrahepatic bleeding from the hepatic artery (from trauma or secondary to a ruptured hepatocellular carcinoma) is one example where targeted coil embolisation may be of benefit. Chemoemblisation of liver tumours via the hepatic artery is also commonly performed (to gain control).

However there is only one setting where embolisation via the hepatic artery may be of benefit in the setting of portal hypertension. This is where there is an arterioportal fistula leading to secondary portal hypertension (may be due to trauma, tumour or other pathology).

In the setting of cirrhosis the portal hypertension is being driven by the loss of hepatocyte mass combined with nodular regeneration as well as fibrosis. Embolising the hepatic artery is not able to address the underlying pathology which is altering the haemodynamics, nor can it address the portal hypertension in this setting.

As for variceal haemorrhage most of the time the blood appears darker to visual inspection than for fully oxygenated arterial blood as explained by Alexander Julianov

Dear Dr. Verrah,

I agree completely with your comments. However please note that the question posted by Dr. Borodach have two clear points:

1. Are there indications for interruption of flow through PROPER hepatic artery in cirrhosis with PH?

2. Possible explanation for a CASE of bright red bleeding from esophageal varices?

I just propose answers to these questions. We accept the latter in clinical setting (pts with PH and bright red GI bleeding – not so common as you mentioned) only when endoscopy ruled out source of arterial bleeding (erosions, ulcer etc.) as it is common in cirrhotics, but this is a separate discussion.

Thank you for the comprehensive answer, Dr.Verran.

What you have mentioned about the tumour and the arterio-portal fistula embolization via the hepatic artery is absolutely true, and we performed those interventions in Novosibirsk for HCC, and for other locations as kidney and brain (the arterio-portal fistula has not been met to us yet).

As for the oesophageal varices blood brightness, I think there may be even better way to compare it with that of a given patient's venous blood from the cubital vein:

to perform the direct measurement of their O2 saturation what does not seem as a problem. I hope my students will make it up if they do not abandon this matter after the Students' Scientific Congress in Novosibirsk State Medical University that is going to start in this April.

Backing to the 'bright red' colour I should speek up of one article in Russian as well, but I cannot find it (it was published at some local congress collection). So, iis authors reported that they found cavernous tissue proliferaton on the level of the cardia and lower eosophagus in the cirrhotic patients in post mortem examination.

As commonly known, the cavernous vascular tissue is characteristic for all the sphincters' submucosa layer studied: the anal sphincter, the sphincter of Oddi, etc. This tissue is supposed to serve as a 'padding' for a better hermetization of a certain opening. It (the tissue) consists of the regulating arterio-venous anastomosis that can lead under pathological conditions to the formation of large cysts, or 'lakes' containing the mixed or even completely arterial blood as we can see in case of the quite common hemorrhoidal bleeding.

About the cavernous tissue in the papilla of Vater, you can see at:

1) 234) Födisch H.J. Feingewebliche Studien fur Orthologie und Pathologie der Papilla Vateri / H.J.Födisch. - Stuttgardt: G. Thieme, 1972. - 55 s.

2) Borodach A.V.& A.L.Popov, Some Pathogenetic Aspects of the Surgical Treatment of the Complicated Gall Stone Disease. Novosibirsk: The Siberian University Publishing House, 2008. -Р.188.,

etc.

Another group of my students suggests another possible solution for the Elias paradox. They believe that the abundant absorption of the serum from the presinusoids equilibrates the arterio-portal difference. I do not know. Everybody measures the liver blood inflow, and nobody seemingly does as to its outflow under the normal conditions.

It may be possible to address this so called physiological paradox by animal experiments. Under normal conditions ouflow must equal inflow otherwise there would be a problem (such as when the liver becomes congested in the setting of right heart failure).

Dr.Verran,

Thank you.

You highlighted here two important and fundamental questions those cannot be underestimated:

1) how large maybe a gap between an experimental modelling and its real physiological or pathological prototype in its whole integrity with the human organism.

2) as to the adequacy of our views on the liver microcirculation to reality.

Unfortunately, both the formers are not more than a certain degree of approximation.

On the other hand, the arterial inflow and venous outflow is hardly equal to each other at each moment in terms of the precise hydrodynamics because of the lymphatic outflow existence as well, and this very process was meant by the picture above. How much is the volume of the hepatic lymph produced under the normal conditions and how much is its influence on the liver lobule microcirculation and the pressure difference that is the question addressed to wiser colleagues. I do not claim that there is not any datum, but only that I failed to find it (or them if multiple) in the available literature.

Dear Prof. Borodach:

I had read a clinic article about the risk of hepatic failure vs. ligation of proper hepatic artery, and the risk was 5% in a normal liver. In terms of liver cirrhosis, I think this risk will increase in proportion to the severity of the cirrhosis. In my personal experience as a surgeon, I dare not to challenge this issue if cirrhosis is not in early stage.

There is also a phenomenon of splanchnic hyperdynamic state in cirrhosis. A-V shuntings appear from splanchnic arteries to portal venous system. My teacher and I had sometimes purposely ligated left hepatic artery and splenic artery and aimed to decrease the portal pressure for prophylaxis during hepatic resections of hepatocellular carcinoma on severe cirrhotic patients with obvious esophageal varices or with a history of EV bleeding. Theoretically we thought it was right. 

Just for your reference.

Yu-Chung Chang

Dear Prof. Chang,

Thank you for your opinion. I am sure, it is very valuable for all the participants and, moreover, for me.

May I ask, what results of that ligation did you obtain in terms of both the liver function and morphology?

According to Ekataksin et al, the arterial blood supplyes mainly (or even entirely) the connective tissue component of the liver parenchyma, so that the increasing of the arterial inflow may provoke the fibrotic tissue overgrowth. So, is it possible the opposite process, in other words, would  some cutting of the arterial blood inflow dicrease the amount of the fibrosis of the liver? Did you find in the liver specimens any signs of the diminishing or degradating of the fibrotic tissue after the partial elimination of the arterial hepatic inflow?

On the other hand, I can understand that the cirrhosis is the terminal stage of severest liver damage, and there is hardly any hope on some 'miraculous' remedy or procedure to make it better save the transplantation.

Sincerely yours,

Andrey Borodach

Dear Prof. Borodach:

I  can't answer or comment your questions. we just focused to decrease portal flow (pressure) from A-V shunting. We did not follow the liver parenchyma status after ligation of left heptic or splenic artery.