·11
Fig 1
Fatal rupture of an oesophageal
Tarix

·THE PATHOLOGY AND
MANAGEMENT OF
PORTAL HYPERTENSION
By
R MILNES WALLFR
MS, FRCS
Professor of Surgery Umeersity of Bristol
LOVDO
EDILARD ARTOLD (PUBLISHERS) LTD

·© R. Milnes Walker 1959
First published 1959
Made and printed to Great Britain by
Il'illiam Clouses and Sons, Limited, London and Beccles

·CONTENTS Page INTRODUCTION 1 M 7 3 Chapter ANATOVY AND PHYSIOLOGY OF THE PORTAL SYSTEM Gross anatomy Portal systemic anastomosis The stomach and oesophagus The falciform ligament intra hepatic circulation Physiology Experimental portal hypertension Measurement of portal pressure Intra splenic pressure Ilepatic vein pressure Normal portal pressure Liver blood flow 7 7 S 9 9 21 PATHOLOGY OF PORTAL HYPERTENSION Extra-hepatıc obstruction Intra-hepatic obstruction Congenital fibrosis of the liver The pattern of fibrosis The capsule of the iner Hepatic rein obstruction Mechanism of ascites formation 18 21 2. 26 26 30 32 40 43 NII EFFECTS OF PORTAL HYPERTENSION The collateral circulation Cruselhier Baumgarten syndrome 'The spleen V CLINICAL ASPECTS OF PORTAL HYPERTENSION The oesophageal or gastric rafices The spleen eins on the abdominal wall 16 V VANACENTENT OF PORTAL HYPERTENSION Diagnosis Venography 49 49 30 ·IV CO TETS Chapter VI TREATMENT OF PORTAL HYPERTENSION Treatment of ascites Page 60 60 60 61 63 VII PORTACAVAL AVASTOVOSIS Indications Selection of patients Technique 66 66 66 67 VIII SPLENOREVAL ANASTOVOSIS 77 IY DIRECT OPERATIONS ON THE VARICES 82 X PORTAL SYSTESTIC NEUROPATHY The symptoms Treatment 88 89 91 XI PORTAL ITYPERTENSIONY CHILDREN 93 100 100 XII RESULTS OF TREATMENT Portal systemic anastomosis Transection operations and direct operations on the varices Hepatıc artery ligation Oesophago gastric resection Arterialisation of the portal vein Injection of the varices 104 105 105 10% IIO 111 INDEX ·INTRODUCTION The upsurge of interest in vascular surgery which has occurred in the last twenty years has cast its mantle over the portal circulation Progress in the knowledge of this circulation has been hindered by its inaccessibility to the ordinary methods of examination and investigation used in the systemic blood vessels During the last few years the use of hepatic sein catheterisation, splenic puncture, portal venography and Inver needle biopsy have contributed greatly to our knowledge of the portal circulation and its pathology It has not been possible to reproduce easily in animals the abnormal conditions found in man, so that much of the research into this circulation has had to be clinical research, carried out by physicians, surgeons and radiologists This volume is based on the experience of over two hundred patients scen during the last ten years, most of these patrents have had surgical operations, and thus there has been an oppor. tunity to study the pathology of the living, a study which is densed to many workers who only see their patients from the outside or when they have passed the terminal stages of their disease Venous shunting operations for portal hypertension have now been in routine use for rather more than a decade, and their early technical problems have been overcome, there have been no major changes in the management of this condition during the last five years or so, and thus the time seemed appropriate to review the present position in the light of the experience which has been gained There are still many problems to be worked out both as regards the pathology and the treatment In the majority of cases little is known about the cause of the disease in the hver or the mechanical factors which lead to the hypertension, and the problem of the management of those patients in whom a shunt operation is either inadu sable or impossible remains as great as ever It is the purpose of this book to serve as a guide to the clinicians who have to treat these cases, and at the same time to attempt to stimulate further investigation, both by experimental methods and by clinical er perience, which will throw light on the many unsolved problems and thus in turn assist in the prevention and management of portal hypertension During the last few ycars a voluminous literature of the subject has appeared but most of it records personal experience, often with very inadequate data both as regards numbers of patients and length of time for which they have been studied 'I he clinician thus has great difficulty in assessing the present position amongst this maze of reports, and in this Yolume I have tried to set out briefly how the matter stands at the ·Y1 INTRODUCTION moment, realising full well that great changes may take place in the future concerning the management of these patients No attempt has been made here to review the whole series of my own patients, but a number of short case histories are recorded to illustrate special points I have not recapitulated the historical aspect of the subject as this has already been admirably reviewed by Child (1954) Tuo monographs describing series of cases have appeared during the last year by C-A Ekman (1957) giving the experience of 82 patients in the Unnersity Hospital at Lund, and by A H Hunt (1958) who recounts an experience of 250 patients at St Bartholomer's Hospital, London In addition the extensive clinical experience of Blakemore in New York and of Linton in Boston is described in a number of papers Those who are interested in the experimental aspect of hepatic cirrhosis are referred to the compre hens e review by Kobak (1956) Mly thanks are due to a great many people who have assisted in providing the experience on which this work is based First of all come the patients, who have nearly all been most helpful in keeping in touch so that the follow up of their treatment has been almost complete Next come the general practitioners, physicians and surgeons sho have referred their patients and placed confidence in the team who are interested in this work at the Bristol Royal Infirmary These are so numerous that it is im possible to mention them by name, but particular gratitude is due to a few physicians who have amongst them contributed a high proportion of the patients on whose study this work is based To my assistants, anaesthetists and nursing staff I take this opportunity of expressing my deep debt of grattude for their untiting help, any successful results of surgical treatment are due to their work more than to the actual technical operation, the radiologists and pathologists form indispensable members of the team, for without their co operation very little clinical research is possible Practically all the photographs in this book have of the Tinn ersity ity to ed by courtesy of the editor of the Journal of Phy stology It is a tragedy that patients who are otherwise fit still die of haemorthage from oesophageal or gastric varices An example of this is shown in the illustration opposite the title page Many of these patients can be saved for years of active and useful life by timely surgical treatment, and if the work on which this monograph is based has saved the lives of a few of them, it will have at least some justification Bristol, 1958 steps of the 1 1-L ·INTRODUCTION REFERENCES CHILD, C G (1954) The llepatic Circulation and Portal Hypertension Philadelphia and London Saunders KVA, C-A (1957) Portal Hypertension Diagnosis and Surgical Treat- Acta Chir scand, Suppl 222, P 143 UNT, A II (1958) A Contnbution to the Study of Portal Hypertension Edinburgh and London Lungstone KOBAK M (1956) Experimental Surgery of Hepatic Cirrhosis" Int Abstr Surg , 102, 521-5++ ment" P 230

·No text on this page.

·CHAPTER 1 ANATOMY AND PHYSIOLOGY OF THE PORTAL SYSTEM Gross anatomy The portal venous system outside the lwer consists of three main vessels and their tributaries The main vessels are the superior mesenteric vein and the splenic vein, which unite behind the neck of the pancreas to form the third main vessel, the portal vein After a course of 5 or 6 cm in an oblique direction upwards and to the right, this vein du des in the hilum of the lever into two main branches Occasionally a third branch comes off at the same level, an important point of the branches are being ligatured when making a portacaval anastomosis Venography has shown that the direction of the portal vein is much more oblique than is usually shown in anatomical text-books, and it may lie almost horizontal The distribution of the branches in the liver has been fully described by Couinaud (1953) and by Gans (1955) and affects the surgery of the liver rather than the extra-hepatic portal venous system The tributaries of the main vessels are variable The two most important are the left gastric vein and the inferior mesenteric vein, for these vessels are the main channels by which there is an anastomosis wsth the systemic veins, draining as they do the tuo ends of the intra abdominal part of the alimentary canal The left gastric sein is normally the largest sein which drains the stomach In about two-fifths of all people it enters the upper border of the splenic sein within 3 cm of the termination of the latter, in a similar proportion it joins the right border of the portal vein and in the remaining fifth it comes in just at the junction of the splenic and superior mesenteric veins The site of the termination of the inferior mesenteric vein bears similar proportions, in two-fifths into the splenic vein behind the pancreas, in two-fifths into the upper part of the superior mesenteric iein, and in the remaining fifth it joins the angle between these veins at their junction (Fig 2) Portal-systemic anastomosis From the point of view of pathology the anastomoses between the portal and systemic veins are important Normally these anastomoses are insignificant, but if there is obstruction to the flow of portal blood along its normal course they assume great importance ·2 ANATOMY OF THE PORTAL SYSTEM The stomach and oesophagus The anastomoses at the cardiac end of the stomach nuch communicate with the oesophageal reins need particular attention Butler (1931) has made a very careful study of these communications He has shown that in the lower part of the oesophagus there are sub epithelial and sub mucous enous pleruses separated by the muscularis mucosae From the sub mucous plexus seins perforate the muscularcoat and unite to form extrinsic verns and these perforating veins are prouded with values which direct the blood outwards The two plexuses in the oesophagus communicate lelt gast ic ven Splenic ven Superior mesentere Info or mesenteric ved ven Fig 2 The main seins of the Portal System The common variations in the terminations of the left gastric and inferior mesenteric seins are shoun by broken i nes The boundary between the areas of drstribution of the night and left branches of the portal vein in the laer corresponds to a line drawn verticaly from the gall bladder fossa freely with corresponding plexuses in the stomach and those communica tsons which lie in the sub mucous layer may contain salles which direct the blood from the oesophagus to the stomach There are communications betwees the anterior and posterior tributaries of the left gastric sein and the corresponding veins which accompany the vagus nerses terminating in the azygos ken on the right side and the hemeazygos and left posterior bronctual ieins on the left Vost of the extrinsic sejas join the azygos tein and its tributaries but a fer drain into intercostal veins and com muntcate freely with the periaortic senous plexus Small oesophageal veins join the superior and inferior phurentic teins Butler found that India ink injected into the superior mesenteric vein of the foetus found its say into the sub epithelial plexus at the loner end of the oesophagus but ·ANATOMY OF THF PORTAL SYSTEM 3 some of it also passed by the left gastric vein to the azygos and posterior bronchial eins At the other end of the alimentary canal there are communications between the radicals of the inferior mesenteric vein in the sub-mucosa of the rectum and tributaries of both the middle and inferior rectal seins which drain into the systemic veins The falciform ligament Other important collateral channels occur in the falciform ligament Butler (1952) found large veins present in this site in two cases, in both a thick-walled branch came off the left branch of the portal vein and after running for some way in the falciform ligament it was continued as a dilated thin-walled vessel which drained into the veins of the anterior abdominal wall, and was in fact a greatly enlarged vein of Burow branches of the umbilical vein, are normally quite small, but are constant both in the foetus and in the adult Clinical observation confirms these These veins, 1 wall some little way above the umbilicus Of the paired umbilical seins in the embryo, it is usually the left one alone that persists, so that normally the umbilical scin comes off the left branch of the portal sein, but occasionally, as was well shown in one of our tenograms, it joins with the right branch of the portal vein Clsen here there are many small communications between the portal and systemic icons In the retroperitoneal tissues there are communications to the lumbar and to the azygos and hemiazygos seins, to the adrenal and renal veins and to the lower intercostal veins particularly by way of vessels in the lieno-renal fold of peritoneum In addition communications are to be found in the ligaments of the liver leading to the diaphragm, and draining mainly to the pericardio-phrenic veins The portal vein itself is relatively thich-walled containing much more smooth muscle than any systemic vein of similar size (Fig 3) Intra-hepatic circulation In order to appreciate the pathology of some of the conditions which influence the portal circulation it is necessary to study the minute anatomy of the liver lobules, and much light has been thrown on this recently, particularly by the work of Knisely et al (1948) Branches of the hepatic artery and the portal sem drside m the portal tracts and enter separate sinusoids at the periphery of the Irser tobulc, but these sinuses unite and by the time the blood from these two sources rcaches the parts of the sinusoids near the centre of the lobule it is intimately mised The arterioles entening the lobules are much smaller than the corresponding ·4 ANATOMY OF THE PORTAL SYSTEM portal venules, and it is probable that the resistance in these arterioles leads to an equalisation of pressure in the sinusoids which are supplied by blood from both sources (Fig 4) Injection studies have shown that in the portal tracts there are communications between branches of the hepatic arteries and the portal veins, their numbers vary in different animals but they probably become enlarged in cases of portal hypertension No definite evidence has been found of communications in the liver between the hepatic arteries and portal veins on the one hand and the hepatic veins on the other in the substance of the liver except through the sinusoids Where there is (a) x 60 severe fibrosis such communications probably form so that blood can flow through the liver without going through the sinusoids In some animals the orifices of the hepatic veins as they enter the in ferior vena cava are surrounded by smooth muscle fibres which exert a sphincteric action on the veins and may thus control the outflow of blood from the liver Such a mechanism has not been demonstrated in man Physiology The main function of the portal circulation is to convey some of the products of digestion from the alimentary canal to the liver The liver ·PHYSIOLOGY OF TICE PORTAL SYSTEM takes up glucose and stores it as gh cogen Some amino acids are carried to the liver and there undergo changes resulting in their detoxication before they are released into the systemic circulation. There is ven httle factual knowledge of the control of this circulation The inflon JS dependent on the Row of blood in the spianchnic branches of the aorta, which in its turn is under the control of the sympathetic nervous system It is possible that there is also some control of its outflou from the Inver „Daniel and Prichard (1951) have shown by radiographic means that in 5 दि Bila duet Hepatic story Portal rain Fig + Reconstruction of a luser tobule The sinusoids recente blood from both the hepatic artery and the portal ein, in the portal tracts the arteries are not nearly 50 straight as the veins Central (hepatic) voin rats there are considerable variations in the blood flow through the incr Sometimes the portal blood flos is confined to short ride sinusoids situated centrally in the inter, at other ames the flow is evenly distributed right out to the periphery of the organ and the rate of flow 18 much slovet 'l he sigmficance of these changes is not known and they require further Investigation (Fig ) The pressure gradient between the portal and hepatic ieins is small, in health not more than 10 mm Hg, so that cry litle obstruction to the fou through the liver all tend to bring about a rise in the portal senous pressure.

·6 PHYSIOLOGY OF THE PORTAL SYSTEM R (c) 140. (b) Fig5 The portal blood flow in the rat Studies made by injection of an opaque medium into the portal circulation In the upper series all parts of the liver were filled In the lower series only the central area of the liver was filled and the blood passed through more quickly (Daniel R M and Pritchard M M I) Experimental portal hypertension Attempts to produce portal hypertension in animals have in the main been unsuccessful Sudden occlusion of the portal vein is rapidly fatal in cats dogs and rabbits but if a renous anastomosis is made to prevent portal congestion recovery ensues However, the macaca monkey will survive portal sein ligation but the portal pressure falis to normal within sıx to ten days (Child et al, 1950) Slow occlusion of the vein may be obtained ·PIIYSIOLOGY OF THE PORTAL SYSTEM 7 by wrapping cellophane around it but in animals the collateral circulation which forms is adequate to maintain the portal pressure within normal limits Measurement of portal pressure In man it is no simple matter to measure the portal enous pressure and in fact it was only just over twenty years ago (Rousselot 1936) that it was first recorded It can only be done either at a laparotomy or by indirect means At a laparotomy the simplest method is the use of a mano meter filled with salıne connected by a three way tap to a needle or canula inserted into one of the veins of the portal system 17 one of the smaller cins is used a needle is Irable to damage the wall and more reliable read ings are obtained with the use of a small canula This can be placed in the rightgastro epiplosc lein or in a son of the mesentery of the small intestine or the great omentum which may be Igatured afterwards When a recorder with a pressure head of the transducer or strain gauge type is arulable tracings can be obtained with a needle in either the portal or splenic vein The fact that the abdomen is open when the measurements are made probably does not affect the riad ng appreciably but nar Fig 6 A sal ne manometer used for malls intra abdominal pressure recording the portal icnous pressure is transmitted to the cins and will thus influence the portal venous pressure When making a reading It is important to ensure that nothing is obstructing the icin Ietween the point of reading and the liver and there should be a defimte excursion of the fluid corresponding with respiratory movement (lig 6) Intra-splenic pressure ltkinson and Sherloch (199+) have shou n tkat the intra spk nic pressure 13 a reasonally good indication of the portal venous pressure I needle ·8 PHYSIOLOGY OF THE PORTAL SYSTEV is inserted through the parietes into the spleen and, when blood drips from the needle, it is connected to a manometer The measurement can conveniently be made at the same time as a trans-splenic portal senogram These writers found levels ranging from 3 to 17 mm Hg in 14 patients without evidence of portal hypertension, but in cases of portal hypertension the range was from 17 to 35 mm Hg The intra splenic pressure exceeded the portal venous pressure estimated by hepatic vein catheterisation by 2 to 6 mm Hg Hepatic vein pressure Another indirect means of measurement of the portal pressure ss by means of hepatic vein catheterisation and wedging a catheter with a terminal opening in a_tributary of a hepatic sein (Fig 7) When this is done it is assumed, in cases of intra hepatic ob. struction, that the pressure immediately proxunal to the catheter «ll build up to the portal pressure In cases of extra hepatic ob struction this build up of pressure will not occur, so that as a means of measuring the pressure this is only useful when there is disease of the liver without extrahepatic obstruction This Fig 7 A cardiac catheter wedged in a method is occasionally of tributary of a hepatic vein in order to record value in distinguishing bethe portal venous pressure tween intra and extrahepatic obstruction but otherwise has little place in routine clinical practice, of hepatic sein catheterisation is being done as a means of measuring the liver blood flow, In 24 patients a pressure recording may be made at the same time without liser disease a mean pressure of 68_mm_Hg was found but in un patients with cirrhosis of the Inter, levels varied from 156 to 352 mm Hg with a mean of 22 0 mm Hg (Paton et al, 1953) Allison (1951) made measurements of the pressure in oesophageal varices by direct puncture using a needle through an oesophagoscope The method is difficult ia practice, and in any case the pressure in ·PHYSIOLOGY OF THE PORTAL SYSTEM 9 these vessels is not a true indication of the pressure in the veins nearer the liver. Normal portal pressure A number of measurements of portal venous pressure has been made by means of a saline manometer during operations on patients whose livers and portal systems were to all appearances quite normal In our series the range was from 7 5 to 12 mm Hg taking the level of the portal vein as zero Other workers have found rather wider limits Gray (1951) taking the level of the body of the adjacent vertebra as the base line found figures varying from 10 to 16 mm Hg If the portal vein is temporarily occluded during a laparotomy, the pressure.Itses rapidly to something between 40 and 50 mm Hg and remains at this level while the obstruction is maintained Liver blood flow Bradley and his colleagues (1945) devised a means of measuring the liver blood flow in man_by the bromsulphalein excretion method, this depends on the absorption of this substance by the liver alone Intra- Venous infusion maintains a constant level in the systemio circulation and samples taken from a catheter in a hepatic vein indicate the amount removed by the liver and from this the flow can be calculated More recently, attempts to measure the flow by the use of radio-active isotopes have been made (Riddell et al, 1957) Normally the flow is about 1500 ml per min in adults These investigations record the total liver blood flow, and a satisfactory method has yet to be devised by which the relative hepatic artery and portal vein flows can be determined in man REFERENCES ALLISON, P (1951) "The Measurement of Blood Pressure in Oesophageal Vances Thorax, 6, 325-327 ATKINSON, MI and SHERLOCK, S (1954) "Intrasplenic Pressure as Index of Portal Venous Pressure" Lancet, 1, 1325-1327 BRADLEY, S E, INGELFINGER, FJ, BRADLEY, G P and CURRY, J J (1945) "The Estimation of Hepatic Blood Flow in Man" clin Invest, 24, 890-897 BUTLER, H (1951) "The Veins of the Desophagus" Thorax, 6, 276-296 (1952) Gastro-oesophageal Haemorrhage in Hepatic Cirrhosis" Thorax, 7, 159-166 CHILD, CG, MIL.NES, RF, HOLSWADE, G R and GORE, A L (1950) "Sudden and Complete Occlusion of the Portal Vein in the Macaca Níulatta Monkey" Ann Surg. 132, 475-495 COUINAUD, C (1953) "Étude de la Verne Porte intra-hépatique" Pr med, 61,1434-1438 DANIEL, P M and PRICHARD, M M L (1991) "Variations in the Circulation of the Portal Venous Blood within the Liver " J Phy stol , 114, 521-537. 2-PH ·8 PHYSIOLOGY OF THE PORTAL SYSTEM is inserted through the parietes into the spleen and, when blood drips from the needle, it is connected to a manometer The measurement can conveniently be made at the same time as a trans-splenic portal zenogram These writers found levels ranging from 3 to 17 mm Hg in 14 patients without evidence of portal hypertension, but in cases of portal hypertension the range was from 17 to 35 mm Hg The intra-splenic pressure exceeded the portal venous pressure estimated by hepatic iein catheterisation by 2 to 6 mm Hg Hepatic vein pressure Another indirect means of measurement of the portal pressure is by means of hepatic vein catheterisatron and wedging a catheter with a terminal opening in a_tributary of a hepatic rein (Fig 7) When this is done it is assumed, in cases of intra-hepatic obstruction, that the pressure immediately proximal to the catheter will build up to the portal pressure In cases of extra-hepatic obstruction this build up of pressure will not occur, SO that as a means of measuting the pressure this is only Useful when there is disease of the liver without extrahepatic obstruction This A cardiac catheter wedged in a method is occasionally of tributary of a hepatic vein in order to record value in distinguishing bethe portal venous pressure theen intra and extrahepatic obstruction, but otherwise has little place in routine clinical practice, of hepatig sein catheterisation is being done as a means of measuring the hver blood flow, a pressure recording may be made at the same time In 24 patients without liver disease a mean pressure of 68 mm Hg was found, but in II patients with cirrhosis of the liver, levels varied from 156 to 332 mm Hg with a mean of 22 0 mm Hg (Paton et al, 1953) Allison (1951) made measurements of the pressure in oesophageal varices by direct puncture using a needle through an oesophagoscope The method is difficult in practice, and in any case the pressure ifl Fig 7 ·PHYSIOLOGY OF THE PORTAL SYSTEM 9 these vessels is not a true indication of the pressure in the veins nearer the Iner Normal portal pressure A number of measurements of portal venous pressure has been made by means of a salıne mindmeter during operations on patients whose lovers and portal systems were to all appearances quite normal In our series the range was from 5 to 12 mm Hg taking the level of the portal vein as zero Other workers have found rather wider limits Gray (1952) taking the level of the body of the adjacent vertebra as the base line found figures varying from 10 to 36 mm Hg If the postal vein is temporarily occluded during a laparotomy, the pressure rises rapidly to something between 40 and 50 mm Hg and remains at this level while the obstruction is maintained Liver blood flow Bradley and his colleagues (1945) devised a means of measuring the Inver blood flow in man by the bromsulphalein excretion method, this depends on the absorption of this substance by the Iner alone Intravenous infusion maintains a constant level in the systemic circulation and samples taken from a catheter in a hepatic vein indicate the amount removed by the liver and from this the flow can be calculated More recently, attempts to measure the flow by the use of radio-active isotopes have been made (Riddell et al, 1957) Normally the flow is about 1500 ml per min in adults These investigations record the total liver blood flow, and a satisfactory method has yet to be devised by which the relatie hepatic artery and portal vein flows can be determined in man REFERENCES Allisov, P (1951) "The Measurement of Blood Pressure in Oesophageal Varices" Thorar, 6, 325-327 ATKINSON, M and SHERLOCK, S (1954) "Intrasplenic Pressure as Inder of Portal Venous Pressure" Lancet, 1, 1325-1327 BRADLEY, S E , INGELFINGER, F ), BRADLEY, G P and CLARY, (1945) The Estimation of Hepatic Blood Flow in Man" jdm. Imest, 24, 890-897 BUTLER, HI (1951) “The Veins of the Desaphagus" Thorax, 6, 276-296 (1952) Gastro-oesophageal Haemorrhage in Hepatic Cirrhosis" Thorax, 7, 159-166 CHILD, CG, MILNES, R F HOLSU ADE, G R and GORE, A L (1950) "Sudden and Complete Occlusion of the Portal Vein in the Macaca Mulatta Monkey" 814 Surg. 132, 475-495 Courneo, C (1953) "Etude de la l'eine Porte intra-hépatique Pt med, 61, 1434-1438 DINEL, P M and PRICHARD, NI A L (1991) "Variations in the Circulation of the Portal Venous Blood within the Liver" 7 Physiol, 114, 521-537 2-PH

·10
PHYSIOLOGY OF THE PORTAL SYSTEM
GANS, H (1955) "Introduction to Hepatic Surgery
Elsevier Publishing
Circula
1
A mota dam
I Danske
Videnskab Selskab Biol Skrifter IV, nr_7
PATON, A , REYNOLDS, T B and SHERLOCK, S (1953) "Assessment of Portal
Venous Hypertension by Catheterisation of Hepatic Vein " Lancet, 1,
918-921
RIDDELL, A G, GRIFFITHS, D B, McALISTER, J M and OSBORN, S B (1957)
"The Measurement of Liver Blood Flow with Colloidal Radogold
(198Au)"
Clin Sct, 16, 315-324
ROUSSELOT, L M (1936) "The Role of Congestion (Portal Hypertension) in
So called Banti's Syndrome" Amer med Ass , 107, 1788–1793

·CHAPTER 11 PATHOLOGY OF PORTAL HYPERTENSION The pressure in the portal circulation may rise either as a result of increased inflow through the splanchnic vessels, or of increased resistance to the authow There has been some discussion regarding the significance of increased infow, if the outflow is capable of carrying an extra volume of blood the additional inflow is unlikely to influence the pressure When the resistance to the inhou is diminished, as for example in the case of an arterio-venous fistula in the splanchnic circulation then the portal pressure will inevitably rise Evidence has been produced (Leather, 1957) that in cases of gross splenomegaly without obstruction to the portal blood flow there may be increased portal pressure, but this is based mainly on measurements of intra spletic pressure, and in such cases this may not gue a true indication of the portal venous pressure Increased resistance to the outflow of blood from the portal system is caused by lesions affecting the portal vein (pre- or extra hepatic obstruction), lesions of the liver (intra-hepatic obstruction) or lesions which raise the pressure in the hepatic veins (post-hepatic obstruction) The last group includes thrombosis of the inferior vena cava, tricuspid valve stenosis or incompetence, and constrictive pericarditis Such back pressure leads to irreversible changes in the liver which add an intrahepatic obstruction to the post-hepatic cause, and care has to be taken in distinguishing between the two or the mistake may be made of treating the hepatic lesion without realising that it is not the primary cause of the portal hypertension In these post-hepatic lesions the portal pressure 13 not greatly raised, the spleen is rarely palpable far below the costal margın, and a collateral circulation is not much in evidence In other words, the portal hypertension plays a minor role and as such does not require special treatment There are a few patients in whom there is portal hypertension, who develop a collateral circulation, but in whom no evidence of either in. crewed flow of obstruction to the outflow can be found It's possible that there may be a functional obstruction to the outflow, in the form of spason of the portal vein or its branches in the Inter The view has also been expressed that in some cases there may be obstruction to systemic veins which normally have anastomoses with the portal circulation, eg the azygos vein, and this leads to the diversion of blood by way of these anastomoses into the portal circulation, it seems unlikely that such a mechanism is ever significant ·12 PATHOLOGY OF PORTAL HYPERTENSION In this monograph, portal hypertension is taken to be a rise in the pressure throughout the portal venous system However, it occasionally happens that in cases of obstruction to the splenic vein, the hypertension is confined to the veins of the spleen and the upper part of the greater curvature of the stomach These cases should be called "splenic hypertension" Though they are rare, their recognition is important for they can be cured by removal of the spleen, haematemesis is often the presenting symptom, they may have oesophageal varices, but the collateral circulation tends to cross the stomach and drain into the left gastric vein if that vessel enters the main veins beyond the obstruction The condition can only be diagnosed accurately by venography, both by injection into the spleen and into a tributary of the superior mesenteric vein, which will show the limits of the obstruction Extra-hepatic obstruction Extra-hepatic obstruction of the portal blood flow may be primary or secondary, and may be due to thrombosis in the lumen of the portal vein or to pressure or invasion from without When it is due to pressure or invasion from outside the vein, the treatment is that of the cause and the portal hypertension does not require special attention As an example, patients with malignant lymph nodes in the portal fissure may present with splenomegaly Similarly, malignant tumours of the stomach or pancreas may obstruct the portal or splenic veins and cause a general portal hypertension or a localised splenic hypertension Apart from the conditions which carry an increased risk of intravascular thrombosis, the following causes of portal vein thrombosis are recognised I Congenital 2 Injury 3 Secondary to intra hepatic obstruction 4 Portal pylephlebitis caused by inflammation in the organs drained by the portal system 1 Congenital The majority of patients with extra-hepatic obstruction and normal liver parenchyma are children or adolescents For this reason it has been suggested that the condition is a true congenital abnormality, with failure of proper development of the portal vein This may well be so in rare instances, but the observations of Gibson and Richards (1955) and of Parker and Seal (1955) suggest that much more frequently the cause | 18 an intra-vascular thrombosis of the portal vein It is suggested that this is particularly likely to occur at or soon after birth, either as an extension of infection from the umbilicus, or that the normal process of obliteration of the umbilical vein and ductus venosus extends into the portal vein A few cases are on record in which a definite history of ·PATHOLOGY OF PORTAL HYPERTENSION 13 umbilical sepsis has been obtained, and in the following case, in which an acute suppurative arthritis of the hip occurred during the first seek of life, the source of infection was probably the umbilicus SS, female, had wathin a week of birth, abscesses over her left buttock, left ankle, and on her head, these healed but at the age of 20 months, when she first attempted walking, it was found that she had a pathological dislocation of the left hip, which was treated by traction and a caliper, and later by an osteotomy When just 2 vears old she had her first haematemesis, and r 10 1 A in the oesophagus above the line of section Occasionally, the obstruction in the portal vein is represented by a short stricture, which may well be a true congenital abnormality T removed and an orthodox end-to-side splenorenal anastomosis performed During the five years since then he has remained quite well, but, at the age of z2, some varices are still present in hı3 oesophagus 2 Injury Injury of the portal sein causing obstruction by scarring or by thrombosis is rarc, but cases are on record in which it has been injured during a holecystectomy and the patients have subsequently developed portal hypertension Ligature of the postal vein, however, does not always ghe rise to manifestations of portal hypertension, and L 3 Secondary to intra hepatic obstruction In patients with long-standing 11 L the blood in the portal sein may become almost stationary Both these factors predispose to thrombosis As long as this thrombosis is confined

·14
PATHOLOGY OF PORTAL HYPERTENSION
Fig 8
Venogram by intra splenic injection showing a short
structure of the portal vein
?
Fig 9 Venogram by intra splenic injection
The portal
Tein is narrowed by thrombosis In addition the left branch
of the portal sein was obliterated by thrombosis
A large
left gastric vein leads to a mass of varices around the cardia

·PATHOLOGY OF PORTAL HYPERTENSION 15 to the portal vein it may pass unnoticed clinically provided that the patient has developed an adequate collateral circulation, but if the process extends into and completely obstructs the superior mesenteric vein, infarction and gangrene of the small intestine is an almost inevitable sequel Many patients who die of hepatic cirrhosis are found to have some degree of thrombosis in the portal system Hunt and Whittard (1954) have drawn attention to the importance of this complication (Fig 9) In a series of 8o patients with intra-hepatic obstruction whose portal veins have been explored and a portacaval anastomosis performed, mural thrombt partly occluding the lumen have been found nine times, in spite of the fact that in most cases the presence of a lumen had been demonstrated by preliminary venography Splenectomy appears to be a factor predisposing to thrombosis in the portal vein In all 3 patients in this group who had had a previous splenectomy, mural thrombı vere present in the portal vein and in 1 the lumen was divided into two almost equal halves When the lumen becomes completely obliterated the risk of haemorrhage or portal systemic neuropathy is increased Such a state of affairs renders a portacaval anastomosis impossible, and the risk of this complication is an impressive argument against delaying this operation if the patient has had a severe haemorrhage and his liver function is good Before preliminary routine venography in our series the portal vein was explored in two cases of intra-hepatic obstruction and was found to be completely occluded Jo, male, age 29, had had his spleen removed following three haematemeses four years previously Intermittent bleeding persisted, and on exploring the region of the portal vein it was found to be completely thrombosed, and the pressure in a jejut:3! vein was 32 mm Hg No venous anastomosis was possible so an oesophageal transection was performed, apart from one slight haemorrhage he has remained well for the succeeding three years AW, male, vomited blood for the first time at the age of 16, and shortly after this his spleen was removed He had no further haemorrhage for ten L-1.11 1 +int aftur thudana. operation on his varices These cases demonstrate not only the fact that patients may show no symptomatic evidence of thrombosis of the portal vein but the uselessness of splenectomy in treating this condition They also indicate that a shortterm follow-up is quite alueless ·16 PATHOLOGY OF PORTAL HYPERTENSION 4 Portal phlebitis Before the introduction of modern chemothera peutic and antibiotic agents suppurative portal phlebitis was almost invariably fatal The source of the infection is most frequently the appen dıx, but in a number of patients its origin cannot be located (Shaldon, 1958) as in the following example Nowadays, as a result of treatment, the original infection may be overcome, but it will leave the patient with thrombosis of the portal vein This may be only a mural thrombosis or it may fill the lumen completely, after a matter of months or years the manifestations of portal hypertension will become evident Fig 10 Venogram by intra splenic injection There has been thrombosis of the portal and splenic veins following suppurative portal phlebitis WA, male, age 45, had fourteen months before his first haematemesis, a serious febrile illness which lasted for ten weeks and started with lower rightsided abdominal pain His appendix had been removed fourteen years previously On the thirteenth day of his illness, rigors commenced and on the twenty-second day he had jaundice which lasted four days A laparotomy on the twenty-sixth day revealed turbid intra peritoneal fluid but no source of his infection Following the haematemesis his spleen was just palpable, he had gastric and oesophageal varices, and venography showed thrombosis of both splenic and portal veins (Fig 10) I hule awaiting operation he had another sesere haemorrhage At operation the liver was formal, adhesions were present round the portal sein, and the pressure in a jejunal sein was 30 mm Hg The spleen tras remoied and a gastric transection performed The follox - up is only of six months' duration, but so far he has remained well ·PATHOLOGY OF PORTAL HYPERTENSION 17 Thrombosis of the portal vein leads to great enlargement of small essels which are normally insignificant, but which by-pass the obstruction, sometimes a degree of recanalisation of the vein takes place The result is a meshwork of irregular vessels, of varying size, but sometimes there are larger vascular spaces, giving the region much the appearance of a senous angioma, and the condition has been called "cavernous transformation of the portal vein" The extent of this change will be determined by the extent of the original thrombosis It may be confined to the extra-hepatic part of the portal vein, or it may be carried along many or all of the branches of the vein in the portal tracts of the liver (Tig 11) If Fig 11 A portal tract in the liver of a girl who died of haemorrhage at the age of six, there has been extensive development of collateral venous channels forming the so-called cavernous transformation The liver parenchyma is normal x 30 the thrombosis also affects the splenic vein such a cavernous mass will be found extending along the upper border of the pancreas The umbilical vein is normally a branch of the left branch of the portal vein in the liver, when a collateral circulation develops in the falciform ligament, and dilated veins appear in the subcutaneous tissues around the umbilicus, it is an indication that the obstruction is intra-hepatic In cases of thrombosis of the portal vein in its extra hepatic part, the obstruction is proximal to the junction of the umbilical vein and the left branch of the portal vein, and thus in this type of obstruction a "Caput Medusae" will not form There must be considerable resistance to the flow of blood through the ·18 PATHOLOGY OF PORTAL HYPERTENSION tortuous channels in cases of cavernous transformation, for, although some opaque solution when injected into the portal circulation reaches the Inter the portal pressure is often astonishingly high, frequently being above 30 mm Hg In children with this condition the Iner has certai. characteristics which can be recognised, it is particularly soft in consistency and has a brick red colour rather than the normal plum colour, presumably this is due to the fact that a much higher proportion than normal of its blood supply is arterial, these characteristics have not been seen in adults There still seems to be some doubt whether deprivation of an otherusse normal liver of its portal venous blood supply is detrimental to its function or leads to permanent changes in its parenchyma As a result of the er perience of 25 cases of primary extra hepatic obstruction in which Inver biopsies and biochemical investigations are available, I am very doubtful whether this deprivation of venous blood leads to any disease in the Iner itself Some of these patients have been studied for many years and there has been no evidence of progressive luer damage (see Chapter \I) In fact the only symptom from which these patients suffer apart from blood changes due to hy persplenism is haemorrhage from the varices of their collateral circulation Most of them if untreated will eventually d as a result of haemorrhage Satisfactory treatment is no easy matter bi If the danger of haemorrhage can be overcome, these patients will leat normal active ines Intra-hepatic obstruction Many pathological conditions of the Inter cause obstruction of the blood flow through it Owing to the high pressure gradient between the hepatic artery and the hepatic veins this obstruction needs to be severe before there is any appreciable reduction in the arterial blood flow through the Iner and this obstruction does not influence the general arterial pressure By contrast the low pressure gradient between the portal and hepaa eins results in the flow being very easily obstructed, and as the port circulation is a closed circulation except between capillaries, this obstruc tion leads at an early stage to a rise in pressure in the portal circulation The conditions in the Inter which cause obstruction to the portal blood flow may be temporary or permanent, it is probable that any condition which causes tension inside the liver capsule will hinder the flow of portal blood Thus acute hepatitis or fatty infiltration may lead to a transient rise in portal blood pressure which will subside as the inter returns to normal In 40 patients who had acute hepatitis Reichman and Days (1957) recorded the intrasplenic pressure during the acute illness and again during convalescence and found that in all except 6 cases there was a sig nificant fall in the convalescent period varying from 14 to 64 per cent 5 cases however a higher pressure was found during convalescence thar. during the acute phase of the illness Conditions which set up a fibrous 11 ·19 PATHOLOGY OF PORTAL HYPERTENSION response in the liver give rise to a more permanent portal hypertension, but it has been found that in cases where the inter is extensively replaced by fibrous tissue the portal pressure may be within the normal range or only slightly elevated The reasons for this will be discussed later There are many causes of connective tissue proliferation in the liver which leads to fibrosis In some there has been necrosis of liver cells with collapse of the connective tissue framework, and proliferation of the surviving liver cells, often in irregular nodules The importance of fatty deposits in the liver cells is not clear in many cases of portal hypertension with hepatic fibrosis it is quite impossible to lay the blame on any particular etiological factor Any of the following may be precursors of portal hypertension 1 Virus hepatitis Alost cases of this infection clear up completely and It is not known why a few progress to cirrhosis, but nutritional factors may play a part (Himsworth and Glynn, 194+) Dible (1951) has followed the changes by means of serial liver biopsies It is probable that many cases are due to this cause, but the acute illness is not recognised as hepatitis because evident jaundice has been absent Symptoms of portal hypertension appear between two and ten years after the acute attack ME, female, had an acute illness at the age of 33 lasting nine weeks, with jaundice Nine years later she attended hospital on account of pain in the left loin, and was found to be so anaemic that a blood transfusion of 3 pints was administered Three years later, at the age of 45, she had her first hacmatemesis which required a transfusion of 3 pints of blood, but her next attach eight months later necessitated the administration of 8 pints of blood Her Inver function tests were good, intra splenic pressure was 25 mm Hg and venography showed a patent portal vein When an end to side porta casal anastomosis was made, the pressure in the portal vein before making the shunt was 29 mm Hg 2 Autritional deficiency There is much experimental evidence that a low protein diet causes diffuse hepatic necrosis and post necrotic scarring This may occur in spite of an adequate calorie intake In many countries, particularly in Africa and the Fast, poor nutrition has been blamed for the high incidence of cirrhosis of the liver and portal hypertension but it may be that the poor nutrition increases the liability to virus infection In Southern Rhodesia, hepatic cirrhosis is much the most common cause of haematemesis (Gelfand, 1957) It is possible that the cirrhosis associated with ulcerative colitis or with fibro-cystic disease of the pancreas may be due to deficiency of protein 3 Alcoholic cirrhosis It is not definitely established whether there is a direct toxic effect, or the Inter changes are due to a nutritional deficiency in those who take much alcohol This is, however, an important factor in a proportion of cases, and adversely affects the prognosis after a shunt operation if they continue their alcohol consumption ·20 PATHOLOGY OF PORTAL HYPERTENSION ID, male, age 39, had for the previous twenty years been in the habit of taking 4 pints of beer and s pints of cider a day During the last eight years be had fisc attacks of haematemesis At operation his laser was normal in size bur hard with a nodular surface, and his portal pressure WTS 27 mm Ilg After a portacar al anastomosis, his spieen which had been enlarged to three fingerhrcadths below the costal margin, became impalpable Blowever, hc declined to reduce his alcohol consumption below 2 pints each of beer and cider, and died of Incr faslure two and a half years later 4 llaemochromatosis In this condition the accumulation of iron in the Iner leads to extensisc fibrosis, and thus to portal hypertension MT, male 'I lus patient was boarded out of the Army on account of haemochromatosis in 1940 at the age of 32 In 1949 there were two attacks of sciere haematemesis At a laparotomy the Inter was cirrhotic, the spleen moderately enlarged, and the pancreas brick-red in colour At that time there Was no glycosuria Thc splenic artery was tied and an omentopexy per. formed Iwo and a half years later he was found to have diabetes In 1954 he had another haemorrhage and ascites appeared, and following further haematemesis he died five and a half years after his operation At post mortem there was multilobular cirrhosis with multiple nodules of primary carcinoma of the liver 5 llepato-lenticular degeneration Here there is an excess of copper deposited in the liver, and cvidence of portal hypertension may be found CL , female, first noticed clumsiness of her hands at the age of 15, and soon after this found that she could not pronounce words clearly The following year the diagnosis of hepato lenticular degeneration was made, the cornca showed typical hay ser-flischer rings and her scrum copper was found to be 47 mg/100 ml A year later she had her first haematemesis Iniestigation showcu no gross disturbance of laser function, large varices and a patent portal While in hospital further blecuing occurred and she received 27 pints of blood by transfusion in six days She stas then transferred to the Bristol Royal Infirmary and two days later an end-to side portacaval anastomosis was performu ller portal i cnous pressure 11.19 43 mm Ilg During the ylar since her operation no further blceding has occurred 6 Bilharcia The fixer is snsolved by infection by Schistosoma manson and Schustosoma japonicum The ova obstruct the branches of the portal vein and give rise to a persportal fibrosis 'l his is one of the common causes of portal hypertension in South America AS, male, was born in 1938 in Poland and was taken to Russia in 1941 Ile cached Persia in 1942, when fic suffered from dysentery, in 1943 he was in India and then spent the years from 1943 10 1950 in Rhodesia and 'Tanganyika, where he had malaria lle reached Ingland in 1950 at the age of 12 years and was complaining of lassitude and had persistent ferer A In laparotomy showed cirrhosis of the later duc to Schustasomassa mansons 1952 le had a ses crt haernatemesis for which transfusion was required flis Incr was moderately and his spleen was grossly enlarged Finc mottling was shown radiologically in both lungs After a course of antimony tantrac, a portacas al anastomosis was performed 'Thc portal pressure was 16 mm Hg and fell to 11 mm Hg after the anastomosis The lymph nodes found the Vein ·21 PATHOLOGY OF PORTAL HYPERTENSION portal vein were much enlarged Histology showed lesions containing large giant cells in the lung, lner and a lymph node, in the liver was also porta? infiltration by lymphocytes and plasma cells, but no marked fibrosis A few degenerate parasites were demonstrated in the liver (Figs 12a) During the five years since his operation he has remained very well, and his lover and spleen hare ceased to be palpable 7 Biliary ctrliosts Chronic biliary obstruction due, for example, to gall stones, or injury to the common bile duct, leads to cirrhosis of the liver which in turn causes portal hypertension If the obstruction is in the extra-hepatic passages, surgery should be undertaken for its relief, and operative treatment for the hypertension must be required only very rarely 8 Congenital absence of the bile ducts A particular form of biliary cirrhosis is due to congenital absence of the bile ducts Most of these patients die during the early months of life, but a few survive the initial obstructive jaundice and it appears that the alimentary canal takes on some of the excretory functions of the liver, however, these patients early develop the typical brown complexion of the cirrhotic, and ascites also get oesophageal varices and may have haematemesis The love function is too poor for any surgical treatment to be considered and the usually die of liver failure The ani Congenital fibrosis of the liver There is a rare form of obstruction which occurs in children (Park 1956) It is of the pre-sinusoidal type, the portal tracts are large contain not only an excess of fibrous tissue, but very many small bile ducts, some of which may penetrate, without any sheath, into the inver parenchyma and often contain inspissated bile 'The luwer is enlarged and firm and when exposed at operation has a peculiar appearance, its surface is quite smooth, but it is mottled, many small areas of purple liver parenchyma being separated by a grey or yellow network which represents the fibrous tissue These children have no symptoms except haematemesis, and the changes in the Iner do not appear to be progressite The portal pressure is high, but they do well after a portacaval anastomosis Tour such patients have occurred in this series, 3 girls who had their first haemorrhages at the ages of 3 and 5 and 17 respectively, and a boy who bled for the first time at the age of 6 One of these is described in greater detail LS, female, was found to have an enlarged Iner at the age of 2, when admitted to hospital for an umbilical hernia Haematemesis occurred when she was 3, recurred twice when she was 9 and again at the age of 13 liver was a little enlarged and firm, the spleen ettended three fingerbreadths below the costal margin, but all Iner function tests were normal At the age of 12 a portacaval anastomosis was performed, her portal sein pressure was 24 mm Hg The lner appearance both macro and microscopically was typical as described above (Fig 12b) "The ·Fig 12 Gross and microscopic appearances of the liver in some conditions causing intra hepatic obstruction Photographs and biopsies taken at operation Histology-11atiory stain, * 30 (a) Patient AS, age 14 Schistosoriasis The liver is enlarged, darkcoloured and firm 'The granuloma contains a few guant celisPortal pressure 16 mm Hg (b) Patient LS, age 12 Congenital fibrosis of the liver The Inseris firm, enlarged and has a mottled appearance The portal tracts show numerous bile ducts but portal seins are small and scanty Portal pressure 24 mm Hg (c) Patient W M, age 47 Cirrhosis with coarse regeneration nodules and much fibrous tissue Portal pressure 28 mm Hg (d) Patient LR, age 46 Lier very adherent with thick fibrous capsule and normal parenchyma underneath it Portal pressure 22 mm Hg

·/
NM
lat
Tig 12

·(n
(6)
8
()
F18 13

·Fig 13 Gross and microscopic appearances of the liver in cases of cirrhosis showing the variations in type Photographs and biopsies taken at operation Flistolog) --Vallory stain * 30 (a) The fibrous tissue is confined to the portal tracts and their immediate neighbourhood Portal pressure 36 mm Hg (b) The fibrous tissue is distributed as fine strands or septa linking the portal tracts Portal pressure 25 mm Hg (c) The fibrous tissue is more extensive but the lobular pattern is still recognisable and the surface of the liver is not grossly irregular Portal pressure 21 mm Hg (d) The fibrous tissue is extensne and the loser parenchyma occupies isolated islands in which the normal lobular pattern is unrecognisable there has been much regeneration and the surface presents a more nodular appear ance Portal pressure 27 mm Hg ·24 PATHOLOGY OF PORTAL HYPERTENSION It is now four years since another of these children, described on p gh, had a portacaval anastomosis, and she has remained in excellent health since In about half the cases of cirrhosis of the liver which are referred because of evidence of portal hypertension no definite etiological factor can be assigned It is important, however, to appreciate that in many of these the pathological process in the liver is static and that often cirrhosis is not necessarily a progressive disease leading inevitably to fatal liver failure This applies particularly to those cases which are the result of virus hepatitis The inflammatory process has led to proliferation of connective tissue cells, these become converted into fibrous tissue some regeneration of liver cells takes place, and a liver which can function nor mally remains The only abnormality which endangers the patient is the increased resistance to the flow of blood through the organ The pattern of fibrosis The fibrosis in the liver shows wide variations and owing to the re generation it is not always easy to recognise the particular pattern Occasionally the fibrous tissue is confined to the portal tracts and their immediate neighbourhood, and the Iner parenchyma is normal through out most of the individual lobules, the architecture of the liver being little distorted (Fig 13a) In such cases the obstruction is presinusoidal and functionally the condition will behave like a pre-or extra-hepatic obstruc tion, in this type the liver function is good and will remain unimpaired unless there is a change in the pathology of the liver More commonly the fibrous tissue is distributed as fine strands or septa linking up one portal tract with its neighbours and thus tending to isolate the liver lobules (Fig 13b), when this pattern obtains the liver retains its normal shape and size, the surface will be only slightly less smooth than normal and the consistency only a little more firm Casual palpation of the liver at a lapa rotomy may lead to the pathology being overlooked, though this is a type in which the portal pressure is often very high Some of the cases of this type are undoubtedly a sequel to an attack of virus hepatitis In a few cases this type is progressive and gross fibrosis and liver fatlure supervene, but in the majority the condition is stationary and laser function may remain normal for many years Grosser forms of cirrhosts are commonly the result of nutritional deficiency or of alcoholism, which have led to necrosis and post necrotic scarring (Fig 13c and d) In this type, with much fibrous tissue the portal pressure is not usually so high, but shen large regeneration nodules are seen (Fig 12c) the pressure tends to be higher, possibly because the regeneration nodules distort and compress the portal tracts These cases are often fatal as a result of liver failure, and post-mortem studies by means ·I 4THOLOGY OL PORTAL IIRTENSION of injection casts have shown that there is gross distortion of the portal tracts (laddon et al 195+) There have been a fess cases of portal hy pertension with no evidence of extra hepatic obstruction in which the Iner pathology has shown only the slightest variation from the normal pattern and the suggestion is made that vascular spasm may play a part in leading to the obstruction of 25 wiler histology from a patient with no extra hepatic obstruction and a portal sein pressure of 25 mm Ilg There is no fibrosis in the liver X 100 blood flow through the iner In this group the luser is a little firmer in consistency than normal rather darkly coloured but the surface smooth or only slightly irregular and microscopically there is no increase in fibrous tissue the portal tracts appear normal or show only a little cellular infiltration Three patients u ho fall in this group have done well after portacaval anastomosis The histology of a portal tract in one of thes cases is shown in lig 14 The following is one example 3--IH ·26 PATHOLOGY OF PORTAL HYPERTENSION JC male age 29 had four sei ere haematemeses during six years His portal pressure was 27 mm Hg and fell to 6 mm Hg after a portacayal anastomosis he has remained well for six years since that operation II s liver biopsy showed no excess fibrosis or cellular reaction in the portal areas The capsule of the liver Another pathological type of laser thich is associated with portal hyper tension is one in which the capsule is greatly thickened and may even be calcified in parts there are two examples in this series one a case of haemochromatosis showing much calcification is described on P 44 The other is noted here LR male had at the age of 39 an illness with ascites for which para centesis was performed once and the fluid never recurred Seven years later he had a haematemesis for włuch he was transfused with four pints of blood FIrs spleen was not palpably enlarged but he had extensive oesophageal varices Liver function tests were almost normal At operation the ver was densely adherent to surrounding structures but when these were freed the capsule was thickened and white (Tig 12d) On incision this capsule was 4 mm thick but beneath this the laser substance looked normal "The portal vein was exposed and was healthy except for a small mural thrombus which was removed the pressure in it was 22 mm Hg An end to side portacaval anastomosis was performed The explanation of this thickened capsule is difficult but it may repre sent the end result of extensive peripheral necrosis in the liver somewhat analogous to cortical necrosis in the kidney, Daniel s findings in animals (P5) suggest the idea that sometimes the periphery of the organ is de prised of blood The finding of traces of portal tracts in this fibrous capsule suggests that it does in fact represent a replacement of liver parenchyma by fibrous tissue Hepatrc vein obstruction A particularly interesting form of intra hepatic obstruction to the portal blood flow is seen when the primary lesion is in the hepatıc veins and thus the obstruction is post sinusoidal Two distinct types are seen cho occlusne disease of the Inter which affects the smaller tributaries of the hepatic ieins and Budd Chiari disease in which the larger hepatic seins and particularly their orifices into the inferior vena cava are affected Clinically the two conditions have some resemblance Thus hepato megaly and ascites are prominent features at first the spleen is not greatly enlarged and Iner function is not much impaired If recovery from the acute phase takes place fibrosis which commences at the centre of the lobules occurs in the Inter and signs of portal hypertension become evident increasing enlargement of the spleen and haematemesis from oesophageal varices Death may be due either to Iner failure or haemor rhage from the varices Stuart and Bras (1997) have given a full account of veno occlusne disease as it occurs in Jamaica and suggest that the ·PATHOLOGY OF PORTAL HYPERTENSION 27 ingestion of certain herbs, especially Senecio (ragu ort) and Crotalaria (rattle-bor) may be important aetiological factors The following is a typical case of Budd-Chiass disease JI, male, age 21, noticed strelling of his abdomen which gradually in. crcased, after eight months, pain wluch was never severe was present in the right hypochondrum and he had a httle vomiting and epistaxis On admission to hospital a year after the first symptoms there 1125 no jaundice, his bdomen was greatly distended by ascites, there were numerous spider naert, ut no sacral or ankle ocdema After paracentesis a hard liver edge was palpable and the spleen could just be felt Extensive varices were present in the pesaphagus Laser function tests on these occasions were all within normal limits At a laparotomy the Iner was hard and nodular, the spleen about three times normal size and the portal senous pressure was 19 mm Hg A large Vein *18 present in the falciform hgament Luer biopsy shoxed sex ere centrilobular necrosis and haemorrhage with enlargement of the sinusoida (Fig 15) Following a haematemesis he died in hepatic coma At the autopsy much blood was present in the alimentary canal, the inyer weighed 2070 gm and was nadular, the portal vein was healthy All the orifices of the hepatic veins into the vena cava except one were completely occluded by organised membranes, and the remaining one had an opening only 3 mm in diameter (Fig 16), in the baser the hepatıc veins were dilated In the vena cava was a Irrle arganised thrombosis on the wall, and it must be assumed that the membranes occluding the orifices of the hepatic seins were the result of old thrombs It was surprising that his Iner function »as so bile impaired when there was practically no outlet from the buyer by the hepatic veins It has been shown by Popper and his colleagues (1952) by injection studies of the later in patients who haie died of cirrhosis that the fibrous tissue contains a network of vessels which communicate with both the portal and hepatic reins No such communications were found in normal Iners ft thus appears that where there is extensive fibrosis in the laser intra-hepatic portal-systemsc venous anastomoses deselop, and it is suggested that these may go some way towards reducing the portal blood pressure This might account for the fact that shen the liver is grossly scarred by fibrous tissue the risk of haemorrhage is less than in those patients with only thin incomplete fibrous sepua The appearance of the user when exposed at operation may thus show a wide variation, it may look almost normal, the surface may be finely granular when there is an esen distribution of fine strands of fibrous tissue throughout When the fibrous tissue is less esenly distributed and there are small nodules of regeneration the typical "hob-nasl" appearance is found In a few cascs large nodules may push their way above the surface, the interning capsule being tethered donn by fibrous tissue, and here an extraordinary nadulat appearance is presented (Fig 17) In the majority of these the etiology is obscure but Sery coarse liver nodules have been recorded in association with fibro cystic discase of the pancreas he ·26 PATHOLOGY OF PORTAL HYPERTENSIOV JC, male, age 29, had four severe haematereses during six years Ili portal pressure was 27 mm Hg, and fell to 6 mm Hg after a portacaal anastomosis, he has remained well for six years since that operation Ilus liver biopsy showed no ercess fibrosis or cellular reaction in the portal arcas The capsule of the liver Another pathological type of inter which is associated with portal hypertension is one in which the capsule is greatly thickened and may even be calcified in parts, there are two examples this series, one, a case of haemochromatosis showing much calcification, is described on P 44 The other is noted here LR, male, had at the age of 39 an illness with ascites, for which para centesis was performed once, and the fluid never recurred Sesen years later he had a haematemesis, for which he was transfused with four pints of blood His spleen was not palpahly cnlarged, but he had extensiic ocsophageal Liver function tests were almost normal At operation the per berad uc he varices 1 . ADA us which was removed, the pressure in it sas 22 mm Hg An end-to side portacaval anastomosis was performed The explanation of this thickened capsule is difficult, but it may represent the end result of extensne peripheral necrosis in the Iner, somen hat analogous to cortical necrosis in the kidney, Daniel's findings in animais (P5) suggest the idea that sometimes the periphery of the organ 18 deprived of blood The finding of traces of portal tracts in this fibrous capsule suggests that it does in fact represent a replacement of Inver parenchyma by fibrous tissue Hepatic vein obstruction A particularly interesting form of intra-hepatic obstruction to the portal blood flow is seen when the primary lesson is in the hepatic verns, and thus the obstruction is post-sinusoidal Two distinct types are seen henom Occlusive disease of the Iner which affects the smaller tributanes of the hepatic veins, and Rudd-Chiarı disease in which the larger hepatic sens, and particularly their orifices into the inferior vena cava, are affected Clinically the tuo conditions have some resemblance Thus, hepatomegaly and ascites are prominent features, at first the spleen is not greatly enlarged and liver function is not much impaired It recovery from the acute phase takes place, fibrosis, which commences at the centre of the lobules, occurs in the Inter and signs of portal hypertension become evident, increasing enlargement of the spleen and haematemesis from oesophageal varices Death may be due esther to inser failure or haemosshage from the varices Stuart and Bras (1957) have given a full account of ieno-occlusive disease as it occurs in Jamaica and suggest that the ·PATHOLOGY OF PORTAL FIYPERTENSION 27 ingestion of certain Berbs, especially Senecio (ragsort) and Crotalaria (rattle-box) may be important aetiological factors The fallowing is a typical case of Budd-Chiari disease JH, male, age 21, noticed snelling of his abdomen which gradually increased, atter eight months pain which was never severe was present in the nght kypochondrium and he had a little vomiting and epistaxis On admission to hospital a year after the first symptoms there was no saundice, his abdomes was greatly distended by ascites, there were numerous spider nacit, but no sacral or ankle oedema After paracentesis a hard lner edge was palpable and the spleen could just be felt Extensie varices were present in the oesophagus Liver function tests on these occasions were all within normal limits At a laparotomy the foyer was hard and nodular, the spleen about three times normal size and the postal senous pressure was 19 mm Hg A large vein nas present in the falciform ligament Luer biopsy shou eu senere centrilobular necrosis and haemorrhage with enlargement of the sinusoids (Tig 15) Following a haematemesis he died in hepatic coma At the autopsy, much blood 535 present in the alimentary canal, the int leighed 2070 gm and was nodulat, the portal vein was healthy All the rifices of the heparc tens into the vena cava except one were completely scluded by organised membranes and the remaining one had an opening Hly 3 mm i diameter (Fig 16), in the laer the hepatic veins were dilated the veni cava vas a little organised thrombosis on the wall, and st must be Assumed that the membranes occluding the orifices of the hepatic ieins were the result of old thromb It was surprting that his luer function was so Arctic impaired when there was practically no outlet from the liver by the hepatıc vens It has been shown by Popper and his colleagues (1952) by injection studies of the layer in patients uho have died of arrhosis that the fibrous tissue contains a network of vessels which communicate with both the portal and hepatic eins No such communications were found in normal Iners It thus appears that where there is extenske fibrosis in the iner intra hepatic portal-systemase venous 31astomoses develop, and it is suggested that these may go some way ton 3rds reducing the portal blood pressure This might account for the fact that when the liver is grossly scarred by fibrous tissue the risk of haemorrhage is less than in those patients with only thin incomplete fibrous septa The appearance of the liver when exposed operation may thus show a wide variation, it may look almost normal, the surface may be finely granular when there is an eten distribution of fine strands of fibrous tissue throughout When the fibrous tissue is less evenly distributed and there are small nodules of regeneration the typical "hob-nail" appearance is found In a few cases large nodules may push their way above the surface, the intersching capsule being tethered doun by fibrous tissue, and here an extraordmary nodular appearance is presented (Fig 17) In the mszonty of these the ctrology is obscure, but very coarse liset nodules have been recorded an association Kith fibro-estic disease of the pancreas ·28 PATHOLOGY OF PORTAL HYPERTENSION (6) Budd Chiari syndrome Luer hops shous centrilobular necrosis and haemorrhage (a) x 30, () x 100 Fig 15 There are other findings at operation, the significance of which is not entirely clear, when ascites is present, and sometimes in other cases, oedema may be found between the folds of the lesser omentum and per haps extending to the adjacent duodenum and retroperitoneal tissues other cases, soft vascular enlarged lymph nodes are present around the portal sein while in others enlarged k mphatic sessels may be seen in this situation or on the surface of the liver or diaphragm (Fig 18) These Sessels normally contain clear lymph hut occasionally chyle flous when

·Fg 16 Budd Chiari syndrome The inferior
ena cara has been opened from behind all the
orifices of the hepatic seins are occluded except
one which is reduced to 3 mm diameter
Mit
Fug 17 Inner with vers coarse regeneration nodwes
in a boy aged 12 l'ortal pressure 31 mm Ilg 'This
pattert died of her failure five scars after a pintu
canal anastomosis

·30 PATHOLOGY OF PORTAL HYPERTENSION Fig 18 A large lymphatic tessel on the pleural surface of the diaphragm in a case of cirrhosis without ascites one is divided in the free margin of the lesser omentum, suggesting that in such a case the flow of lymph was upwards, though not necessarily into the liver Mechanism of ascites formation Clinical experience shows that portal hypertension per se does not cause ascites Cases of primary extra hepatic obstruction, many of whom have portal pressures greater than 30 mm Hg, have no ascites Ascites is, however, common in patients with intra-hepatic obstruction Evdence is gradually accumulating (11adden et al, 1954) to show that ascites is due to obstruction of the liver outflow This may occur in cardiac cirrhosis, in which centrilobular fibrosis is found, it also occurs in the Budd-Chiari syndrome and in veno-occlusive disease of the liver, in both there being obstruction to the hepatic veins In these conditions the obstruction to the blood flow is post sinusoidal, and ascites is a prominent feature As a contrast, in those cases in which the fibrosis is confined to the periportal areas and the pressure is high, ascites is as a rule absent There are intermediate cases, in which the architecture of the laser is quite altered by regeneration, and in which the obstruction is both preand post sinusoidal, and then ascites is associated with a raised portal pressure However, when there is very extensne fibrosis intra hepatic portal systemie shunts may help to bring down the portal pressure, and this would account for the fact that cases in which ascites is a prominent feature do not often have severe haemorrhages from oesophageal varices ·PATHOLOCY OF PORTAL HYPERTENSION There are undoubtedly other factors in the formation of ascites and changes in the plasma proteins which occur in laser disease certainly play a part in patients who have had haemorrhages from oesophageal varices and who also have ascites the latter symptom may not reappear aſter a successful portacaval anastomosis which suggests that when there is defective hver function the ascites which is already present may be aggra ated by the raised portal pressure REFERENCES CHC15 ind ONEILL Portal Vein Jsset 94 31 45 DIBL lion Mecrosis and Fibrosis in the Liver n med i 833 841 GEIFAND M (1957) Personal Commun carton Guso) B and RICILARD R. 1 (1955) Cavernous Transformation of the Torral Tein lath Bact 70 81--96 1119SWORTH II P and GLI LE (1974) Toxpathic and Trophopathic Hepatitis Lancet 1 457 465 TILT Å 11 and 1 HITTARD B R. (1954) Thrombosis of the Portal Wein in Cirrhoss Hepatis Lancet 1 281 284 I EATHER HIV (1957) Personal Communication PARKER R A and SEAL RMC (1955) Caucrnous Transformation of the Portal Wein Pat? Bac! 70 97 103 PARKER RG (1956) Fibrosis of the Liver as a Congenital Abnormalny 7 Path Rack 71 359 368 Porrer 11 ELI4S II and PETTY D E (1952) Vascular Pattern of the Curthotic Lies imer din Path 22 -17-729 RECIA S and Dus HD (1957) The Splenic Ipproach to the Portal Circulation Gastroenterology 33 boy-615 5111DON C (1938) Portal !)aema Brut Surg 45 357 360 STUART h I and BRA G (1957) eno occluse Disease of the Liner Quart y llcd z6 291 315 ·CHAPTER III EFFECTS OF PORTAL HYPERTENSION Portal hipertension produces anatomical and functional changes in the splanchnic system these are mainly associated with the development o the collateral circulation and the congestive splenomegaly In long standing cases the visceral peritoneum loses its normal lustre and presents an opaque grey appearance I'he dilated seins of the portal system show changes similar to varicose veins in the legs in places the wall is thinned and mural thrombi are likely to be present such veins tear easily and do not hold sutures well In cases of intra hepatic obstruction other findings may be present the lymph nodes in the hilum of the Inter and along the portal vein may be enlarged and sometimes there is oedema of the neighbouring structures particularly the gall bladder the lesser omentum and the adjacent retroperitoneal tissues These are probably the result of the liver discase rather than being directly due to the portat hypertension and 20) The collateral circulation This occurs where the portal and systemic circulations join (Figs 19 The oesophagus and gastric fundus The most important site is the region of the oesophagus and the fundus of the stomach (Fig 2ra b and c) here the varices may be very extensive often extending in th- sub mucosa of the oesophagus up to the level of the arch of the aorta and even as high as the junction of the pharyny with the oesophagus (Tug 22) their actual arrangement shorts considerable variation there may be one or two large trunks or a network of irregular channels which lie all round the circumference then sened through an oesophagoscope intense purple congestion of the whole mucosa of the lower part may be seen but in some cases a few large vessels stand out projecting into the lumen either as isolated swellıngs or running for some length upwards producing exag gerated longitudinal folds as in Fig 23 usually the blue colour of the con tained blood can be seen through the wall but sometimes the overhy mg epithelium is thickened and the varices appear grey When a thoracotomy is carried out dijated seins are usually seen on the outer surface of the oesophagus and the points where they penetrate the muscular coat may be clearly visible At the same time it may be noted that there are drlared Teins on the upper surface of the diaphragm, in addition the pericardio ·CITECTS OF TORTAL HYPERTENSION 33 phrenic scins and one or two of the lower intercostal eins may be en larged (Tig 59) I lie retroperitoneal fissues In the retroperitoneal tissues the most constant collateral veins are found in the neighbourhood of the spleen thus the vasa brevia are often tremendous carrying blood up from the spleen towards the oesophageal varices in the lieno renal fold of peri toneum the collaterals usually take the form of many small channels each not more than one millimetre in diameter but sometimes a single large Fig 19 The collateral circulation in extra hepatic obstructs in due to throm hosis of the portal vein Fig 20 rhe collateral circulation in intra hepatic obstruction Vessel 18 found running from the lower pole of the spleen into the retro peritoneal tissues and such a vessel will sometimes show on a venogram (I lg 22) When performing a splenorenal anastomosis the Icft adrenal vein is often found to be dilated and varicose These retroperitoneal seins drun into the renal veins (Tig 25) the lover marcoaliansandth the literate szigor and an ins OT more than one occasion when approaching the oesophagus by a left thoracotomy the hemiazygos sein fuus een found to be more than one cm in diameter und either this sein or the azygos sein may shou up as paravertebral shadone an o filme of the chart 11 -6

·34
EFFECTS OF PORTAL ILYPERTENSION
de
me
(6)

·TFFECTS OF PORTAL HYPERTENSION 35 Fig 21 Varices in the gastric fundus showing (a) as fill ng defects when the stomach contains barium and (6) the cins outlined by trans splenic lenography A case of extra hepatic obstruction with thrombosis of the splenic lein (c) The venogram by injection into a jejunal vein shows that the portal vein is patent

·36
EFFECTS OF PORTAL HYPERTENSION
(a)
17
-
(0)
Fig 22 Oesophageal varices extending as high as
the cervical oesophagus in a case of intra hepatic
obstruction (same patient as Fig 13a)

·FFFFCTS OF PORTAL IIYPERTENSIO
37
Fig 23 Oesophageal varices presenting
us a few large verstis running up the
oesophagus
Fig 24 Tenogram in a case of extra hepatic oh
struction a large collateral sessel incircles the lower
pole of the spleen and drains to the chest wall buy the
liem renal ligiment

·Fig 25 Trans splenic venogram in a case of extra-hepatic obstruction, showing retroperitoneal collateral vessels leading to the left renal vein and the inferior vena cava The filling of the latter shows through the gastric translucency, the stomach being displaced by the large spleen Fig 26 (a) Large herniazygos sein showing as a parasertebral shadon The patient had had an end-to-side spleno-renal anastomosis ten years before using a vitallium tube, which can be seen overlying the twelfth rib (b) Tomogram of the same patient showing the shados of the enlargeu azygos vein above the left branchus ·EFFECTS OF TORTAL JIYPERTE\S101 39 Following an operation collaterals develop through adhesions to the scar (Tig 27) The anal canal In the anal canal internal haemorrhoids are often found associated with partal hypertension but to what extent this is cause and effect it is difficult to say, the majority of cascs of cirrhosis of the liver occur in patients who are getting past middle age and this is the age When haemorrhoids are common in children who have portal hypertension haemorrhoids do not occur but some congestion of the rectal and anal mucosa may be present It is doubtful whether serious bleeding from haemorrhoids 1s any more common in patients with portal hypertension than in other patients who have haemorr hoids and are in the same age group The folatformligament Mention has already been made of the col lateral circulation in the falciform ligament, here again it may tahe Fig 27 The same patient as shown in Tug 26 infra red photograph Dilated the foon of a single large channel subcutaneous veins around the scar of (Fig 28) but more often it con the operation sists of a mass of tiny vessels "hose appearance can be readily recognised during 2 peritoneo scopy I rom the aldominal attachment of the falciform ligament these vessels pierce the linea alba and thus give rise to the dilated veins in the subcutaneous tissues of the abdominal wall forming the 'Caput Vicdusac These vessels may be troublesome when making an upper abdominal incision Tihen there are lige ieins in the falciform ligament and one of them pierces the linea alba through a narrou opening a sub cutaneous varix mas form at this point and the non through this will set up currents which cause a palpable thnll and a murmur which can be heard «ith the stethoscope the condition being described as the Crusell hier Raumgarten syndrome it is not necessary to postulate an abnormal anatomical arrangement for such a sein simply represents a persistent cin of Burow or possibly a persistent umbilical ein pitent throughout its course > ·40 EFFECTS 01 PORTAL HYPERTENSION Fig 28 Trans splenic venogram in a case of intra hepatic obstruction showing a large vein coming from the left branch of the portal vein and traversing the falciform ligament Cruveilhier Baumgarten syndrome WJ, male, first noticed some varicose veins near his left groin at the age of 14 and these gradually spread upwards When aged 26 he had some melaena, haematemeses started the following year and were repeated almost every year until he reached the age of 37 His liver was just palpable, the spleen was enlarged to three fingers and present in his oesophagu The dilated tortuous rein in the abdominal wall extended from the pubes to a point just above the xiphisternum Fie cm above the umbilicus was a varix 4 cm in diameter in which varices 1 TA ..barva bromsulphalein test showed 16 per cent retention after 45 minutes An end to side portacayal anastomosis was performed" The large sein in the falciform ligament was a branch of the left portal vein There was some mural thrombosis in the main portal sein and the more recent part of this clot was removed The pressure in the portal sein was 15 mm IIg while that in the sein in the falciform ligament was 14 mm Hg and fell to 3 mm Ilg ·FFFECTS OF PORTAL HYPERTENSION on compression of the portal vein Hefore the anastomosis the pressure in the inferior vena cava svas 7 mm Hg and that in the portal 1e1n fell to this level after the anastomosis was open Since the operation he states that he has felt better than he had done for years there has heen no more haematemesis and he has for five years worked consistently as a selder in the docks The vein in the antenor abdominal wall thrombosed a few days after his operation and soon after became in 1131hic (Tug 30) 11g 29 I arga subcutaneous arices in a patient who had the Cruscolher Baumgarten sindrome 1 1 Some of the largest collatcral channels are seen in the omentum which has become adherent to some structure drained by the systemic sistem rhaicin in the omentum which was adherent to the lateral *m from a patient in whom a large cand tu un osan Occasionally varices de selop in other sites, llius record the case of a patient with cirrhosis of the liver who had severe bleeding froni a karty in the caecum which necessitated an emergency Tight hemicolectom but the patient died of In er failure I atal intraperitoneal haemorrhage has also been recorded by Els Griffiths and VacIntyre (1955)

·42
EFFECTS OF PORTAL HYPERTENSION
ww
Tig 30 The same patient as in Fig 29 a few
months after a portacaval anastomosis
Fig 31
Photograph taken at operation showing a scin 2 cm d ameter
in the omentum which was adherent to the lateral abdominal wal!
the patient had both intra and extra hepatic obstruction

·EFFECTS OF PORTAL JIYPERTENSION It is difficult to gne an explanation of the fact that in some patients one group of collaterals predominates and in others another group It has cavner mental, L 1 2. Ioh han nn netan become insignificant, it seems that very much the same takes place in cases of portal hypertension for one patient may have very large oesophageal varices and little dilated collateral circulation elsew here, while another will have a large group of collaterals in another site and few if any Fig 32 Trans splenic sen ygram showing a large collateral vein in the omentum which was adherent to the wall of the left side of the pelvis in spite of this anastomoss the pressure in the portal vein was 20 mm IIg and the patient had many serious haemor rhages 1 he portal rein is narroned by murat thrombus dilated cins in the oesophagus The same applies to the splenic vein in some cases particularly those with marked splenic enlargement this vein will be greatly dilated and somewhat tortuous but in other cases this vein may not be dilated at all The spleen Che spleen is always enlarged in cases of portal hypertension, and it is unusual for at not to be palpable Bowescr, there is little or no relationship between its size and the portal blood pressurt That its enlargement is parth, the result of congestion is shown by the fact that it will become considerably reduced in size within a few hours of a portacaval anastomosis But long standing congestion causes structural changes in ·# EFFECTS OF PORTAL HYPERTENSION the spleen and these tahe a long time to resolve after the hypertension has been reheved In addition, the work of Cameron and de Sarım (1939) suggests that in disease of the liver there may also be a blood bome factor which causes enlargement of the spleen, for they found in rats that when the spleen was transplanted into the abdominal wall and was entirely cut off from the portal circulation, it would still show some enlargement when cirrhosis of the Iner was induced by means of carbon tetrachloride ily persplenism The enlarged spleen can give rise to changes in the blood, thus there may be anaemia, leucopenia or thrombocytopenis It is sometimes difficult to determine whether the anaemia is due to hacmor- Thage or to an effect of the spleen on the bone marros, and probably there is often a combination of both factors The thrombocytopenia may lead to nose-bleeding, bruising or petechiae of which the patient may complain Regarding the white cells, a leucopenia of about two to three thousand is common but this does not seem to have serious clinical significance After a portacaval anastomosis, as the spleen becomes reduced in size, the blood Picture tends to return to normal, but sometimes the blood changes continue to be troublesome, only once in our series (a case of haemochromatosis) was it necessary, after the establishment of a portacaval anastomosis, to perform a splenectomy because of a persistent haemolytic anaemia MF, female, at the age of 45, had three severe haematemeses within the space of a month necessitating transfusion of 12 pints of blood or plasma. cha น.น 1 Her leucocytes tell at one stage to 600 per cu mm and platelets were scanty After a portacaval anastomosis, at which it was confirmed that the calcrfica t101 was entirely in the thickened capsule, the blood picture improved to normal Horleser, a year later she des eloped a severe haemolytic anaemia, which was only checked by splenectomy At this operation the typical red brown colour of the pancreas found in cases of haemochromatosis was noted Following this she was very fit for three months, then she became mentally confused, loving her musculat power, and died after a further three months REFERENCES CAMFROG R and de SARAY, GS" (1939) "A Method for Permanently Dissociating the Spleen from the Portal Circulation (The Varsupralised Spleen) and its Use : the Study of Experimental Lixer Cirrhosis Path Bact , 48, 41-47 FlLIS, H, GRIFFITHS, P and MACINTYRE, A (1958) "Ifaemoperitoneum." Brut Surg , 45, 606-610 LEVY, JS, HARDIN, I H, SIIIPP, II and KEELING, J 1 (1957) "Yances of Caecum as an Unusual Cause of Gastrointestinal Bleeding Gastroenterology, 33, 637-640 LONGLAND, C. (1953) "The Collateral Circulation of the limb" R Coil Surg, 13, 161-176

·LFFECTS OF PORTAL HYPERTENSION
5
(a)
FIG 33
(6)
Calcification in the capul of the Inter in a case
of haemochromatinis
(a) the trans Sple nic senngrim which in additron ta
the calcification shosts large varies around the
cards and an unusually crtical portal in
(h) lateral sex showtog the calcification which was
shun at operation to be mafined to the capsuk

·CHAPTER IV CLINICAL ASPECTS OF PORTAL HYPERTENSION The only clinical findings due directly to portal hypertension are those associated with the collateral circulation, or the effects of the enlarged spleen Patients with intra hepatic obstruction may have in addition clinical features due to their liver disease The oesophageal or gastric varices When these extend a long way up the oesophagus the patient may feel a sense of fullness in the throat, but the serious aspect is the danger of haemorrhage from these varices Occasionally it manifests itself as a slow ooze, so that the patient is unaware of it, but suffers from anaemia and occult blood is found in the stools More often the bleeding is sudden and severe, in Great Britain such bleeding accounts for less than ten per cent of cases of serious haematemesis, the great majority being due to peptic ulceration but there are parts of the world such as Central Africa where this state of affairs is reversed, and in the great majority of cases of haematemesis the bleeding comes from oesophageal varices However, in children, even in Britain, portal hypertension is the most common cause of haematemesis The actual precipitating factor may be peptic ulceration of the oesophageal mucosa over the varıx, but trauma by food particles has also been blamed, and there may be other factors such as a sudden rise in the intra-abdominal or portal pressure hrook (1956) has shown that the Valsalva manoeuvre can increase the occluded hepatic vein pressure to as much as 100 mm Hg in patients with or without cirrhosis and considers sudden abdominal strain is a likely cause of rupture of varices However, the intervals between one haemorrhage and the next can be very variable, anything from a day or two to many years It is this unpredictability which makes the outlook in the cases so uncertain Some figures show (Patek et al, 1948) that of patients with cirrhosis of the liver, more than half are dead within a year of their first haemorrhage, but some of these die of liver failure and not all of them succumb to another attack of bleeding Contrary to the experience of some u riters, we have found it very unusual for a patient to die as a result of his first haemorrhage, but the immediate prognosis for a patient bleeding from oesophageal varices us more serious than in cases of bleeding from peptic ulceration Usually the blood 18 vomited without any warning at all, though the patient some times notices a salt taste in the mouth before the blood appears In those ·CLINICAL ASPECTS 47 cases in which the blood all passes downwards, the patient may feel faint, and the appearance of black stools follows A second severe haemorrhage within a few days is particularly serious Following a haemorrhage the patient may pass into a state of liver failure, or may suffer from neuropathic symptoms and even pass into coma as a result of the absorption of protein products resulting from the digestion of the blood in the intestine (see Chapter X) The spleen The spleen is always enlarged to some extent in cases of portal hypertension It is not always palpable, for it may not extend below the costa! margin, or the presence of ascites may prevent its being felt In many cases it is enlarged as much as four fingerbreadths belon the costal margin, but it is unusual to find it much larger than this As already mentioned, the size of the spleen bears little or no relationship to the elevation of portal pressure, but it is proportionately more enlarged in young patients than in those in the older age group As a rule it is free of adhesions, but sometimes, probably as a result of infarction adhesions to the parietal peri toneum are present and in that case they will be very vascular As a result of haemorrhage the patient may be anaemic, but this may be in part due to haemolysis, the effect of hypersplenism, there may be petechiae or echy moses, and nose bleeding is a not infrequent symptom Some patients merely complain of lassitude Veins on the abdominal wall The presence of dusted veins in the subcutaneous tissues of the anterior abdominal wall which tend to show a radiating pattern from the umbilicus has already been mentioned as a sign of intra hepatic obstruction (Fig 29) 1 hypertension, there may be evidence of associated disease, particularly of the liver Thus ascites, palmar erythema spider naevi (Fig 34). gynaecomastia and testicular atrophy are all evidence of liver disease The patient may have the characteristic brown complexion of the cirrhotic, or there may be definite jaundice This last is a sign that there is either some biliary obstruction or that the hepatitis is in an active phase, and adds greatly to the danger of any treatment for the hypertension, in fact it is wise to postpone any active treatment of the latter until at least a few months after the jaundice has subsided, unless delay is dangerous for other reasons The liver itself is sometimes palpable and this is an indication that intra-hepatic obstruction is present When the liver can be felt in the epigastrium it does not necessarily mean that it is enlarged, for a liver of ·CHAPTER IV CLINICAL ASPECTS OF PORTAL HYPERTENSION The only chnical findings due directly to portal hypertension are those associated with the collateral circulation, or the effects of the enlarged spleen Patients with intra hepatic obstruction may have in addition clinical features due to their liver disease The oesophageal or gastric varices When these extend a long way up the oesophagus the patient may feel a sense of fullness in the throat, but the serious aspect is the danger of haemorrhage from these varices Occasionally it manifests itself as a slow coze, so that the patient is unaware of it, but suffers from anaemi and occult blood is found in the stools More often the bleeding is sudden and severe, in Great Britain such bleeding accounts for less than ten per cent of cases of serious haematemesis the great majority being due to peptic ulceration, but there are parts of the world, such as Central Africa where this state of affairs is reversed and in the great majority of cases of haematemesis the bleeding comes from oesophageal varices However, in children, even in Britain, portal hypertension is the most common cause of haematemesis The actual precipitating factor may be peptic ulcera tion of the oesophageal mucosa over the varix, but trauma by food particles has also been blamed, and there may be other factors such as a sudden rise in the intra-abdominal or portal pressure Krook (1956) has shown that the Walsala manoeuvre can increase the occluded hepatic vein pressure to as much as 100 mm Hg in patients with or without cirrhosis and conseders sudden abdominal strain is a lihely cause of rupture of varices Honeer, the inter als betu een one haemorrhage and the next can be very variable anything from a day or two to many years It is this unpredictability which makes the outlook in the cases so uncertain Some figures show (Pateh et al, 1948) that of patients with cirrhosis of the iner, more than half are dead within a year of their first haemorrhage, but some of these die of liver failure and not all of them succumb to another attack u phare found it the -Hero $ $ U more serious than in cases of bleeding from peptic ulceration Usually the blood is vomited without any warning at all, though the patient some times notices a salt taste in the mouth before the blood appears In those ·47 In many tension CLINICAL ASPECTS cases in which the blood all passes dos owards the patient may feel faint and the appearance of black stools follows A second ses ere haemorrhage within a few days is particularly serious Tollowing a haemorshage the patient may pass into a state of liver failure or may suffer from neuropathic symptoms and even pass into coma as a result of the absorption of protein products resulting from the digestion of the blood m the intestine (see Chapter V) The spleen The spleen is always enlarged to some extent in cases of portal hyper It is not always palpable for it may not entend below the costal margin or the presence of ascites may prevent its being felt cases it is enlarged as much as four fingerbreadths below the costal margin but it is unusual to find it much larger than this As already mentioned the size of the spleen bears little or no relationship to the elevation of portal pressure but it is proportionately more enlarged in young patients than in those in the older age group As a rule it is free of adhesions but some times probably as a result of infarction adhesions to the parietal peri toneum are present and in that case they will be very vascular As a result of haemorrhage the patient may be anaemic but this may be in part due to haemolysis the effect of hypersplenism there may be petechrae or echymoses and nose bleeding is a not infrequent symptom Some patients merely complain of lassitude Veins on the abdominal wall The presence of dilated cins in the subcutaneous tissues of the anterior abdommal wall which tend to show a radiating pattern from the umbilicus has already been mentioned as a sign of intra hepatic obstruction (Tug 29) Similar i eins may be noted in the region of a scar of a presious operation to the deep aspect of which there are adhes ons (Fig 27) In addition to these signs which are the direct result of the portal hypertension there may be evidence of associated disease particularly of the liver Thus ascites palmar erythema spider nacvi (Fig 34) gynaecomastia and testicular atrophy are all evidence of Iner disease The patient may have the characri ristic bronn complexion of the cirrhotic or there may be definite jaundice This last is a sgn that there is either some bulary obstruction or that the hepatitis is in an active phase and adds greatly to the danger any treatment for the hypertension in fact it is wise to postpone any active treatment of the latter unul at least a few months after the jaundice has subsided unless delay is dangerous for other reasons The lover itself is sometimes palpable and this is an indicanon that intra hepatic obstruction is present When the laser can be felt in the epigastrum it does not necessarily mean that it is enlarged for a lover of

·48
CLINICAL ASPECTS
normal size which is unduly hard will be palpable in the hepatic con
ditions which cause portal hypertension the liver may in fact be smaller
than normal or considerably enlarged but in the majority of Cases it is of
about average size
When it is much enlarged this again is usually 293
indicanon that there is still some active hepatitis
Ascites unless attributable to some quite independent cause is evidence
that the obstruction is intra hepatic though there may of course be
secondary portal vein thrombosis I have only seen one patient with
th
F g jt lnusually extensne sp der naevi on the shoulder of a pat ent
alcohol cc trhoss of the 1 ver
ascites extra hepanic obstruction and no disease of the liver but in th s
case there were probably other factors which gave rise to the intra
peritoneal fluid and it is my bel ef that extra hepatic obstruct on hy itself
never causes ascites
data Ved
REFERE\CES
LROOM H (1956)
C rculatory Stud es in Liver Cirrhoss
Scand Suppl 318 60
PATEK A I Post] RATNOFF OD VAIN H and Herlalu RW
(1948)
Detar) Treatment of Cirrhos s of the Li er
met med
Iss 138 543 548

·CHAPTER V MANAGEMENT OF PORTAL HYPERTENSION In the management of a patient with portal hypertension, the first steps are the establishment of the diagnosis, the decision that the hypertension is the cause of the patient's symptoms, and whether it is sufficiently severe to endanger his life or seriously to impair his health In cases of intrahepatic obstruction it is important to obtain a clear distinction between those symptoms due directly to the disease of the Iner, and those which are a consequence of the hypertension Diagnosis The great majority of patients who have portal hypertension which requires treatment present as a result of haemorrhage from the alimentary canal usually haematemesis Sometimes they go to the doctor because of malarse, and occasionally the patient himself will have noticed the swelling due to his enlarged spleen The presence of these findings will lead the doctor to suspect portal hypertension, but in cases of sudden haemorrhage, the distinction from peptic ulcer may be difficult, and a number of cases are only diagnosed when an emergency laparotomy for haematemesis is performed However, in the great majority the finding of the enlarged spieen will be the most significant sign There may be a history of past jaundice and other signs of liver disease After the clinical examination a barium swallow should be ordered to demonstrate the presence of varices, but the radiologist should also look for any other abnormality such as a peptic ulcer in the stomach When varices are present in the oesophagus or stomach they show as rounded filling defects projecting into the lumen (Tug 22), but if too much barium 18 swallowed at a time it may compress and obliterate the varices and their presence may be missed There are, however, a few patients in whom the varices fail to shost as a result of this radiological examination, and if the diagnosis is seriously in doubt these patients should be examined with an oesophagoscope, by which means it is unlikely that any varices will be overlooked The varices shon up as rounded prominences in the oesophageal mucosa, sometimes up to 1 cm in diameter, and tending to run longitudinally, normally the blue colour of their contained blood shows through the thin mucosa, but sometimes they have an opaque grey colour where thickening of the epithelium has occurred over them Tanner (1958) considers that they can be seen better with a gastroscope as it is ·50 MANAGEMENT OF PORTAL HYPERTENSION withdrawn up the oesophagus than with the oesophagoscope, gentle inflation being applied at the time For practical purposes the presence of varices may be taken as definite evidence of an appreciable degree of portal hypertension, and the reverse also holds good As mentioned earlier, in some patients the collateral circulation may be much more des eloped in one site than another, but as the most serious consequences of the hypertension result from the varices about the cardia, the absence of varices in the oesophagus suggests that there is not a sufficient degree of hypertension to require treatment While these investigations are being undertaken, certain pathological tests should be made A full blood count is required, which should include differential leucocyte and platelet counts This will indicate any anaemia as well as the effects of hypersplenism In difficult cases a bonemarrow examination may be necessary Liver function tests are also required Laboratories differ in the tests that they carry out as a routine, but it is essential to have estimations of the serum bilirubin and alkaline phosphatase, and a differential protein estimation, either chemically or by electrophoresis, with the absolute value of the serum albumin is required At least one of the flocculation tests should be done, preferably the thymol turbidity, as these gre some indication of activity of Iner disease In some clinics much weight is attached to the bromsulphalein excretion test, and this is of value in patients whose liver function is im paired and the question of a major operation is being considered Peritoneoscopy (Walker, 1943) and liver puncture biopsy are sometimes helpful in making the diagnosis, the former has its greatest use if there is a possibility of malignant disease in the liver, either primary or secon dary, and can be carried out under local anaesthesia Liver biopsy by needle puncture is occasionally indicated if there is doubt concerning the nature of any pathological process in that organ It should only be done of the prothrombin time is normal As there may be large vessels ir the falciform ligament the approach through the epigastrium should bu avoided in these cases The method is not without risk (Terry, 1952) and should only be employed if the decision for or against surgical treatment depends on hnonledge of the pathology of the liver It is possible to obtain a liver biopsy under vision with the peritoneoscope but by either method the specimen may be too small to gnea conclusive diagnosis Venography Before any surgical treatment is undertaken it is most important to have a knowledge of the state of the veins of the portal system This will indicate the site of any obstruction, and in fact senography may in some cases be the only way of distinguishing between intra- and extra-hepatic obstruction (Walker, 1957) (Steiner et al, 1957) ·MANAGEMENT OF PORTAL HYPERTENSION 52 Except in patients who have had the spleen removed, this should be done in the first instance by percutaneous splenic puncture (\Valker et al, 1953) Cor adults only local anaesthesia is required but in children a general anaesthetic should be given The puncture is made either below the left costal margin when the spleen is much enlarged, or else through the intercostal space, usually the ninth, which is in line with the aper of the spleen If the spleen is not palpable the injection should be made over the area of splenic dullness, but it is our practice only to carry out this examination of the spleen is impalpable in those patients in whom the information cannot be obtained in any other way A 17-S WG needle is used, and as a rule the change in consistency can be felt as the needle enters the spleen The patient is instructed to take only very shallow breaths while the needle is in place Before the injection 18 made it is usually possible to withdraws blood into the syringe, but we have not found this in every case For an adult, 30 ml of a 70 per cent solution of one of the soluble opaque rodine compounds is used, and a proportionately smaller dose for children The injection is made as rapidly as possible, films are exposed as the injection is completed and after an interval of three seconds, though with special casette changers more films can be exposed, but too will give all the information required The needle is withdrawn as soon as the injection is completed While the needle 19 in place before the injection is given, the opportunity may be taken to record the intrasplenic pressure, using a saline manometer we have had rather inconstant readings, and on one occasion the high pressure indicated that the needle was in an artery It is possible that with a continuous recorder more accurate readings of the intra-splenic pressure can be obtained Further enographic studies can be made only at a laparotomy, and in fact of the patient has had the spleen removed this is the only way cases of intra-hepatic obstruction of a percutaneous venogram shows that the splenic and portal seins are unobstructed the case is clearly one of intra-hepatic obstruction, and an anastomosis to esther the portal or the splenic vein will provide a suitable by pass When, however, none of the injection reaches the portal sein further study may be necessary, for the injected fund may be dn erted because the vein is obstructed by thrombus, or because the flow of blood is reversed To demonstrate the patency of seins at laparotomy the best site for the injection of the opaque medium is into one of the veras of the mesentery of the small intestine, and at the same time the portal pressure can be measured No antempt at a venous anastomosis should ever be made before it has been shown by venography that the vein at the proposed site of anastomosis is patent, and that it is in such a position that it will effectively drain the portal circulation, particularly the region of the cardiac end of the stomach In trans splenic venography in a patient with a normal portal circulation the splenic vein appears as a cured sessel about 1 cm in diameter

·52
MANAGEMENT OF PORTAL IIYPERTENSION
Fig 35
Trans splenic senogram in a patrent with a
normal portal circulation
Fig 36 Trans splenic senogram in a patient with
cirrhosis of the filer without demonstrable Sancts the
portal veins in both lohes of the enlarged liner are sho in

·MANAGEMENT OF PORTAL HYPERTENSION 53 running across the epigastrium (Fig 35) It is joined by the superior mesenteric vein slightly to the left of the mid-line, the blood entering from this rein causing a filung defect which is carried a little way up the portal ven and gradually fades away The portal vein runs upwards and to the right and divides into its tuo branches It has been said that there is normally streamlınıng in the portal vein blood from the splenic sein usually passing into the left branch of the portal vein while that from the superior mesenteric vein passes mainly to the right branch Venography does not bear this out, in normal cases muting of the opacified and clear bloods seems to be complete in the portal veins and both lobes fill equally (Fig 36), sometimes thrombosis in one branch may prevent filling of an entire lobe (Fig 9) By the time the injection is completed the smaller branches of the portal vein in the Iner are filled and the appearance is that of vessels dividing repeatedly, each dn ision galing rise to tuo of small calibre Normally the opaque medium flows rapidly through the laer and later plates shou a confused pattern of filling of portal and hepatic seins and smusoids In cases of portal hy pertension some dilatation of the splenic and portal veins may be seen, while the splenic vein is not only lengthened but rendered more tortuous by the shifting of the hilum of the spleen to the right, which occurs as that organ enlarges The flow of blood is retarded and much of the opaque medsum enters collateral channels, and in fact in some cases very little may reach the portal rein and the liver, resulting in faint shadows in these sites in spite of the fact that the veins are fully patent In an extreme case there may be a reversal of blood flow in part of the portal circulation and this possibility must be borne in mind AL, a boy age 7 years had an attack of melaena in June 1956 which was severe cnough for him to require a blood transfusion Hus hser and spleen were found to be enlarged " During subsequent months he had repeated haematemeses and blood transfusion Ils treatment was complicated by a persistent salmonella infection A year after his first melaena he was transferred with a view to surgical treatment The luer function tests were 2!l normal Intra splenic venography (Fig 37a) showed side splenic and left gastric veins, but no opaque medium entered the portal semn or the lier When the abdomen was explored a venogram by a jejunal sein (ig 37b) showed a healthy portal vein, filling of the inver and of a tortuous umbilical sein, though some of the medium injected by thus route also entered the left gastric vein and the oesophageal His portal pressure was 25 mm Ilg and a satisfactory end-to-side portacas al anastomosis was carried out He was a case of congenital hbrosis of the inver Varices When there is extra hepatic obstruction the site of the commencement of the blocked vein is usually clearly indicated, a mass of collateral vessels surrounds the obstructed segment When the block 1$ in the portal vein only, come filling of the vessels of the Inter is usually seen If the splenic

·54
MANAGEMENT OF PORTAL HYPERTENSION
(6)
Fig 37 (a) Trans splenic venogram in a boy of
seven with intra hepatic obstruction No opaque
medium reaches the liver as it is all diverted into
collaterals especially the left gastric vein
(0) Venogram using a jejunal vein in the same
patient some of the injection goes by the portal vein
to the her but the remainder is dnected to the left
gastric vein There is also filling
a vein in the
falciform ligament The portal vein 29 fully patent
and a portacaval anastomosis was made

·MANAGEMENT OF PORTAL HYPERTENSION 55 and portal seins are obstructed it is unusual for any of the medium to reach the liver but masses of collaterals may be observed around the pancreas and the fundus of the stomach (Fig 21) Sometimes the collateral circulation in the retro-peritoneal tissues is so well developed that some filling of the left renal vein and the inferior vena cava is seen--a natural but inadequate portacaval anastomosis (rig 25) Occasionally the umbilical vein in the falciform ligament with its origin from one of the two main branches of the portal vein (usually the left) is visible (Fig 28) Irregularity in the wall of the splenic or portal vein or narrowing of the lumen may indicate partial thrombosis, such a mural thrombosis may not prevent the vein being used for an anastomosis of the lumen is adequate, and it may es en be possible to remove some of the clot at the operation In one of our patients the portal vein was found to have a double lumen, old thrombus forming a longitudinal septum in the sein With the information which has been made available by these investigations the clinician will be in a position to advise on the line of treatment to be carried out on any individual patient Surgical treatment may be required as a life-saving matter when haemorrhage has been serious but there may be strong contra indications to a major operation, or particularly to a senous shunt, and the indications have to be carefully weighed in the balance The succeeding chapters will describe the alternative operative measures which are available and the indications for each, for the main factors in the success of surgery in the treatment of portal hypertension are the selection of the patient and of the appropriate operation REFERT VCES STEINER, RE, SHERLOCK, S and TURNER, V D (1957) • Percutaneous Splenic Portal Venography" Fac Radiol , 8 158-177 T'ANNER, N C (1958) Personal Commtmcation TERRY, R (1952) Risks of Needle Biopsy of the Liver" Bnt med , 1 1102-1105 WALKER, R N (1943) "Peritoneoscopy" Proc R Soc Ved , 36 445-448 (1957) "A Review of Trans-splenic Portal Venography in the Investiga tion of Portal Hypertension, Surgical Aspects" ) Fac Radiol,8, 178-180 WALKER RM, MIDDLEWISS, J and NAMSON, E \1 (1953) Tonography by Intrasplenic Injection" Britſ Surg, 40, 392-395 Portal

·56
MANAGEMENT OF PORTAL HYPERTENSION
(6)
(c)
(d)
Fig 38 Shots from a cine angiogram taken with an image intensifier
Trans splenic venography in a case of intra hepatic obstruction
The time interval between each shot was approx i second In (a)

·MANAGEVENT OF PORTAL PEYPERTENSION
5
the portal sen is beginning to fill and it is almost clear in (h)
From (c) { } filling of tortugus collaterals along the lesser curvature
of the stomach can be seen in (g) the portal ein patrern in the lucr
as becom ng hlurred by sinusardal filling
SP

·58
MANAGEMENT OF PORTAL IYI ERTENSION
(a)
(6)
U)

·MANAGEMENT OF PORTAI HYPERTFNSION
59
Fig 39
Shots from a cine-angiogram of another patient, there is a
very large and tortuous left gastric cin, which when filled with
opaque medium hides the shadows of the sertebrae The portal
Sein has emptied at (!) but the opaque medium tema:ns in the
collaterals much longer

·CHAPTER 11 TREATMENT OF PORTAL HYPERTENSION With very few exceptions, the only indication for surgical measures in the present state of our knowledge is the prevention or treatment of haemorrhage from the alimentary canal However, as ascites is so often associated with portal hypertension, a note on this is not out of place Treatment of ascites Some surgeons have advocated operative treatment for the relief of ascites due to cirrhosis of the Iner, but it has been pointed out that such ascites is not due to the venous congestion in the portal system though it may be aggravated by it The presence of ascites due to Iner disease is an indication that there is some defect in the function of the Iner Attempts to treat such ascites by surgical means have been made by many surgeons ever since Talma of Utrecht and Morison of Newcastle inde pendently introduced the operation of omentopexy Drainage of ascitic fluid into the urinary tract or back into the blood stream has been given a trial but these all falled as the openings became occluded by adhesions Others have tried to drain the fluid into the tissues of the abdominal wall so that it may pass back into the circulation by the lymphatics, and I hase attempted to achieve the same result by denuding much of the anterior abdominal wall of its parietal peritoneum Talc has been applied to the surface of the liver to promote vascular adhesions between the iner and the diaphragm, in view of the evidence that the ascites is associated with obstruction to the outflow from the liver, but the results have not met with striking success Ascites is a variable symptom which may disappear spontaneously, so it is difficult to assess the results of any measures for its relief, fortunately in many cases it responds well to treatment by a lox salt diet Very few patients can tolerate a diet which is completely salt free, and it is difficult to devise such a diet which is palatable, but it is possible to reduce the daily intake of sodium 10 og gm and thus should be adopted as a first measure when ascites is troublesome Sometimes it has a dramatic effect, in other cases little benefit is obtained Elimina tion of fluid may be assisted by the use of diuretics such as mersalyl or chlorothiazide Non-operative treatment of portal hypertension It has been traditional to treat patients with cirrhosis of the ler on a lugh protein diet and to gue glucose by mouth Apart from the calories ·TREATMENT OF PORTAL HYPERTENSION 61 that these foods provide there is no evidence that they have any specific effect on the liver Sometimes if the patient has established a good collateral circulation, the high protein may precipitate the onset of neurological symptoms, which may even proceed to coma Apart from the value of a low salt diet and the use of druretics in the treatment of ascites and abstention from alcohol, I am therefore drawn to the conclusion that there are no medical measures which influence the course of the disease in Laennec's cirrhosis of the liver, and there is certainly no medicinal treatment which relieves patients with extra-hepatic obstruction Iron should, of course, be given for anaemia which is the result of haemorrhage When there is a specific cause of fibrosis of the liver it is a different matter Many patients in whom there is an alcoholic factor in the etiology will benefit when they give it up In fact the prognosis of patients with alcoholic cirrhosis seems much worse than those who have cirrhosis following virus hepatitis, and patients with the former condition who have an operation for portal hypertension should be advised of the utmost importance of their gring up their alcohol consumption Schistosomiasis should be treated by the specific drugs advocated for this condition Ilepatolenticular degeneration may be slowed in its course by stimulating the elimination of copper TL - - AL hypersplenism has in the condition whichever the cause, however, blood transfusjon plays an important part and is frequently needed as an emergency for acute haemorrhage, and as a preliminary to operative treatment Operative treatment of portal hypertension Surgical treatment can take three forms The first aims at reducing the portal senous pressure to a safe level, which can only be done by off the flow of portal blood to these vessels The third reduces the blood flow through the liver 1 Means of lonering the portal pressure The nearer the by pass is to the site of obstruction the more efficiently wil it function, thus if the obstruction is in the liver a direct anastomosis between the portal vein and the inferior vena cava will gne the best results It, however, the portal vein is obstructed, the only other ven in the portal system which 19 large enough is the splenic vein, the nearest available sein in the systemic system being the left renal sein A splenorenal anastomosis has the ·62 TREATIENT OF PORTAL HIPERTENSIO disadvantage that the flow in the splenic vein will have to be resorsed, and the operation entails removal of the spleen and consequently some of the already formed collaterals in the peritoneal folds of the spleen 'The tail of the pancreas may distort the splenic vein when the latter is brought down to the renal vein In our experience, splenorenal anastomoses have not had the success which is so striking from the point of vient of preventing haemorrhage after a portacaval anastomosis Other seins of the portal system have been employed and autogenous Tein grafts have been inserted (Rousselot, 1952), but such procedures are very likely to fail as a result of thrombosis and it is doubtful whether they are worth attempting splent cartery Portal vein Inferior --- vena card Lelt renal ven Splenc ven (6) r (2) Fig 40 Portal systemic anastomoses which at present have a place in the treatment of portal hypertension (a) End to-side portacaval anastomosis (6) End-to side splenorenal anastomosis associated with splenectomy 2 Pallatite operations The palliative measures used which do nothing to bring down the portal pressure, are as follows (A) Methods which obliterate the varices (a) Injection through an oesophagoscope of sclerosing solutions (b) Oesophagotomy and ligature of the seins (B) Methods intended to interrupt the flon of blood through the varices (a) Gastric transection (b) Oesophageal transection (c) Devascularisation of the region of the cardia (C) Excision of the sarıx-bearing area Oesophago-gastrectomy ·TREATNET OF PORTAL IIYPERTENSION 63 (D) Methods which promote adhesions with a view to the formation of alternative vascular channels (a) Intraperitoneal, ornentopery or the use of talo (b) Intrathoracic, mediastinal packing It must be remembered that any steps which will divert the flow ot portal blood from the varices into other channels may have the effect of opening up these alternative channels to a greater extent, and as it is only the varices in the region of the cardia which are dangerous, any steps which reduce the danger have a good prospect of prolonging the life of the patient, though it must be realised that by closing one outlet for the portal blood ve may in fact be raising the portal blood pressure still further There are no records in the literature of long-term results of large series of patients treated by any of the above methods, and therefore it must be assumed that up to date no outstanding success can be claimed for any of them wever, some valuable as emergency measure to control acute haemorrhage from varices and the evidence suggests that most of them carry only slight risk and may do something to prolong life them may TU JL occurred It has mainly been employed as a treatment for cirrhosis of the liver rather than for portal hypertension, 1 e as a means of relieving ascites That it does gne some relief cannot be doubted from the large number of cases reported in the literature but the mortality is fugh owing to the risk of hepatic necrosis To me it has always appeared to be wrong to deprive a diseased organ of its only arterial blood supply, and although ascites may be temporarily relieved there is no evidence yet that the operation does anything to prolong life or to improve the function of the liver The results are difficult to assess because many patients with ascites arc temporarily benefited by a laparotomy only When the operation has been done in many cases the ligature has been placed proximal to the gastroduodenal artery and thus has probably little effect in reducing the arterial blood flow through the liver The enthusiasm for this operation in some quarters has been short-lived and the operation cannot be recommended as a means of treating portal hypertension The emergency treatment of haemorrhage It has already been noted that haemorrhage from oesophageal varices • Lo 11 • • In -- -- {{ therapy Local treatment consists in the use of an inflated balloon to cause pressure on the bleeding point or to obstruct the blood flow T ·64 TREATVIENT OF PORTAL HYPERTENSION 11.11 L 1 f.at 1 The Seng veins of the cardia and thus cut off the flow to the varices staken tube combines both these methods, having one balloon for the oesophagus and another for the stomach A balloon in the oesophagus must be used with caution and at least one example of rupture of the oesophagus by its use has been recorded It seems safer, therefore, if the bleeding does not cease by milder methods, to use a gastric balloon with traction and for this purpose a Miller-Abbott tube will suffice It is not easy to maintain this traction comfortably, if the tube is fixed to the check or forehead by adhesive strapping it is likely to cause pain where it presses on the lip, or even cause ulceration in an unconscious patient, while restless patients with neurophathy have been known to pull the tube out esen with the balloon inflated Most patients will not tolerate these tubes for more than two or three days, but as a rule this is long enough for the bleeding to stop, if it has not done so by this time more drastic measures should be employed The tendency has been, as with haematemesis from peptic ulcers, to institute operative treatment earlier and the view 18 now expressed that if the patient is bleeding so severely that the passage of a tube is necessary an open operation should be undertaken Tubes with inflated bags should only be used as a temporary measure to control bleeding while lost blood 18 replaced by transfusion and the patient is being prepared for operation If as a result of previous investigations it is known that the portal sein is patent and the liver function is satisfactory, an emergency portacas al anastomosis may be done if the surgeon is conversant with this operation Alternatively the venogram may be done on the operating table and this is followed immediately by the appropriate operation, if there is no clinical evidence of liver failure and the portal vein is patent, a portacatal anastomosis may be undertaken In such circumstances precautions against the danger of hepatic coma are advisable, blood should be washed out of the stomach and out of the colon, glucose gnen intravenously, and neomycın given by mouth as soon after the operation as the patient is able to swallon MC female, aged 46 had been in hospital a year previously with anaemia and ankle swelling, and was found to have a palpable liver and spleen She had a sudden severe haematemesis, followed by a repetition a month later, for which she was transfused with 4 pints of blood and a Sengstaken tube in serted for three days Only six days later she bled again severely and the tube was reinserted and further transfusion given, in the meantime a veno gram by intrasplenic injection had shown a patent portal rette and large With the tube in place and a blood transfusion running, an ambu, lance journey of more than a hundred miles 33 made, and a portacayal Varices ·time TREATMENT OF PORTAL HYPERTENSION 65 anastomosis was performed next day through a right thoraco-abdominal approach Her gall-bladder which contained stones was removed at the same The portal pressure had been 42 mm Hg before the anastomosis Convalescence was delayed by a left-sided empyema, but after this had been drained she made a good recovery In other cases, if the bleeding persists in spite of other measures, a direct operation on the oesophagus must be considered; this is best done through the left pleural cavity In fact, if no surgeon who has experience of portacaval anastomosis is available, this is the best plan, and if a venous shunt is possible this can be done later after there has been time for a fuller investigation of the patient The question not infrequently arises as to what a surgeon should do who has opened the abdomen at an emergency operation for haematemesis thought to be due to a peptic ulcer, when it is found that there is no ulcer but evidence of portal hypertension In such a case, the steps to be taken must be left to the judgement of the surgeon at the time, they will depend on the patient's general condition, age, and the state of his liver However, if it seems that his haemorrhage is otherwise likely to be fatal, and if there are not too many vascular adhesions, the best plan is to carry out a high gastric transection, with division of the vessels in the lesser omentum and the vasa brevia above the line of section The splenic artery should be divided between ligatures at the same time in order to reduce the blood flowing into the portal circulation At a later date the question of more definitive treatment can be considered When a decision has been reached that major operation is advisable, it is necessary that the patient should be as fit as possible, preliminary blood transfusion may have been necessary and adequate blood should be available for transfusion at the time of operation The patient should have a diet with an adequate protein content, breathing exercises, with the especial object of obtaining maximum expansion of the bases of the lungs should also be instituted REFERENCES BOEREMA, I (1949) Bleeding Varices of the Oesophagus in Cirrhosis of the Liver and Banti's Syndrome" Arch Chir Neerl, 1, 253 260 Cirgical Treatment of Varı ', 27. 422-429 Varices by Injection of a JlIerusing DULULLONI Amer med 45 , 135, 754-757 RIENHOFF, W F (1951) "Ligation of the Hepatic and Splenic Arteries in the Treatment of Portal Hypertension Bull John's Hopk Hosp, 88, 368-375 ROUSSELOT, L M (1952) 'Autogenous Vein Grafts in Splenorenal Anastoa description of techniques and its clinical application in seven patients" Surgery, 31, 403-410 SENGSTAXEN, RW and BLAKEVORE, A H (1950) "Balloon Tamponage for the Control of Hemorrhage from Esophageal Varices” Ann Surg, 131, 781-789. Coinna and Carina D Inn EIN 2 mosis ·64 TREATMENT OF PORTAL HYPERTENSION the hallonn muv he placed in the nesaphame and for then NAPA *lan. The Seng fundus of the stomach, traction on the tube will lead to pressure on the veins of the cardia and thus cut off the flow to the varices staken tube combines both these methods, having one balloon for the oesophagus and another for the stomach A balloon in the oesophagus must be used with caution and at least one example of rupture of the oesophagus by its use has been recorded seems safer, therefore, if the bleeding does not cease by milder methods, to use a gastric balloon with traction and for this purpose a Miller-Abbott tube will suffice It is not easy to maintain this traction comfortably, if the tube is fixed to the check or forehead by adhesive strapping it is likely to cause pain where it presses on the lip, or even cause ulceration in an unconscious patient, while restless patients with neurophathy have been known to pull the tube out, even with the balloon inflated Most patients will not tolerate these tubes for more than two or three days, but as a rule this is long enough for the bleeding to stop, if it has not done so by this time more drastic measures should be employed The tendency has been, as $ith haematemesis from peptic ulcers, to institute operative treatment earlier and the view is now expressed that if the patient is bleeding so severely that the passage of a tube is necessary an open operation should be undertaken Tubes with inflated bags should only be used as a temporary measure to control bleeding while lost blood 15 replaced by transfusion and the patient is being prepared for operation If as a result of previous investigations it is known that the portal vein is patent and the liver function is satisfactory, an emergency portacaval anastomosis may be done if the surgeon is conversant with this operation Alternatively, the venogram may be done on the operating table and this is followed immediately by the appropriate operation, if there is no clinical evidence of liver failure and the portal vein is patent, a portacaval anastomosis may be undertaken in such circumstances precautions against the danger of hepatic coma are adussable, blood should be washed out of the stomach and out of the colon, glucose given intravenously, and neomycin gren by mouth as soon after the operation as the patient 18 able to swallon MC female aged 46, had been in hospital a year previously with aruema and ankle swelling and was found to have a palpable Iner and spleen She had a sudden severe haematemesis, followed by a repetition a month later, for which she was transfused Bith 4 pints of blood and a Sengstalen tube in serted for three days Only six days later she bled again seierely and the tube was reinserted and further transfusion gren, in the meantime a veno gram by intrasplenic injection had shown a patent portal sein and large varices. With the tube in place and a blood transfusion running an ambu, lance journey of more than a hundred miles was made, and a portacaval ·PORTACAYAL ANASTOMOSIS 67 exercised in the selection of the patients Poor selection probably accounts for some of the disappointing results which have been reported but if the indications laid down here are adhered to gratifying results should be obtained The patient should preferably not be above sixty years of age there should be no evidence of recent or rapid deteriora tron in the function of the liver the patient should not be jaundiced and the liver function tests must be adequate Any recent evidence of portal systemic neuropathy should make one very cautious about advising the operation and then it should only be undertaken if the patient is having dangerous haemorrhages As regards the liver function tests the greatest emphasis has been placed on the level of the serum albumin and we have not operated on any patient who had a level below 30 gm per cent Re versal of the albumin globulin ratio need not contra indicate the operation if the albumin level is satisfactory We have only used the bromsulpha lein retention test in cases where there has been considerable doubt con cerning the decision but it is generally accepted that retention of more than 30 per cent after 30 minutes indicates that the venous shunt will carry a high risk In two of our patients who had been jaundiced a few months before the operation suggesting at that time that there was active hepatitis the jaundice recurred a few weeks after the operation and per sisted for many months The flocculation tests are not of great value from the point of view of selection for operation but when they are abnormal the risk of post operative neuropathy is greater However many of our patients who have had gro ly disturbed flocculation tests eg a thymol turbidity of 10 units have done well after the operation A raised seruin alkaline phosphatase level should not deter the surgeon but if it is high a careful search should be made for any obstruction affecting the extra henat lente TA hand oli 1 L and the level has remained high after the operation Technique Anaesthesia The anaesthesia should be light with the use of a short acting relaxant drug as the right pleura will be opened positive pressure will be needed to maintain expansion of the lung Position In our experience the only approach which gives an ade quate exposure of the portal vein is by a right thoraco abdominal incision In a thin patient it may be possible to perform the operation from the front but the increased vascularity of the tissues around the common bile duct renders this approach much more d fficult The patient should be tilted on the left side with pillows under the right shoulder and hop so that the anterior axıllary line is uppermost A table ·CHAPTER VI PORTACAVAL ANASTOMOSIS This operation was first used as an experimental procedure in animals by Eck in Russia in 1877 (for translation Child, 1953), and first applied to man for the relief of ascites by Vidal (1910) in France Poor results followed these early operations and the method fell into disrepute In vessel suture, the operation as practised today came to be established The approach through a sight thoraco-abdominal incision, as first advo cated by Satinsky (1948), as a landmark ishich has made all the difference to the ease and safety of the operation Indications The indications for a portal systemic anastomosis have already been discussed, and it has been concluded that a portacaval anastomosis gnes much better long-term results than a splenoxeral anastomosis from the point of view of pretentmg haemorrhage, and carries no greater operative risk Probably there is a greater incidence of neuropathy after this operation, because it is more efficient as a shunt, and as the purpose of the operation is to prevent haemorrhage, the risk of this complication should be taken into consideration when considering whether a shunt should be created Having decided that this is the right line of treatment, the most efficient anastomosis should be performed Before the operation the patency of the portal vein must be demon strated This will probably have been done by means of trans splenic enography If the spleen has been removed the best course is to com mence the operation by making the anterior part of the incision first, and performing a venogram by injection into a mesenteric vein of one of the eins of the great omentum If the venogram has shown any obstruction of the splenic vein, it is possible that the anastomosis will not dras the splenic part of the portal circulation, and it would be sise to remove ite spleen at a preliminary or a second operation Selection of patients If good results are to be obtained from the operation, both as regards the immediate mortality and the subsequent course, great care must be ·PORTACAVAL ANASTOMOSIS which permits a lateral tilt is an advantage, so that the angle can be adjusted during the operation (Fig 41) Incision This follows the line of the eighth or ninth rib, and is projected forwards beyond the costal margin for about 5 cm The level of the lower of these ribs should be chosen when the liver is much enlarged A few of the anterior fibres of the latissimus dorsi are divided, the external oblique is split and the rectus sheath opened About 12 cm of the rib are removed subperiosteally, and the corresponding costal cartilage split lengthwise (Fig 42) Further forward the internal oblique and transversalis muscles are incised and the peritoneum opened The pleura is opened through the rib bed and the diaphragm incised radially as far as the bare area, vessels on its surface being secured by stitches, one on each side being left long to hold the diaphragm aside This incision of the diaphragm is important as it allows the liver to be displaced upwards, giving access to its under surface Dissection of the portal zein The next step is usually the most difficult part of the operation, the exposure and dissection of the portal vein The foramen of Winslow defined (Fig 43) but sometimes it is found to have been obliterated by adhesions An incision in the peritoneum is carried backwards from the foramen Fig 42 Method of splitting along the lower border of the liver in the costal cartilage lengthwise after removal of the rub This the pouch of Rutherford Morison, allows a more satisfactory repair this incision should be about i cm of the costal margin from the liver so that any vessels, and there may be many collaterals in this fold of peritoneum, can be coagulated or ligatured without damage to the liver The peritoneal incision 15 carried down the posterior margin of the foramen of Winslow, across its base and up the right free margin of the lesser omentum to the hilum of the liver, a branch incision downwards over the inferior vena cava helps to mobilise the duodenum In some cases there are enlarged lymph nodes and many thin-walled collateral vessels in the connective tissue along the portal vein, but it is unusual to find any nodes behind the vein, so that this structure is best 13 L 1 1 but also the gall bladder, the duodenum and adjacent retroperitoneal tissues The common bile duct is retracted forwards and in many cases does not come into view at all Large lymph nodes lying in the line

·68
PORTACAVAL ANASTOMOSIS
(a)
es
(6)
Fig 41
The position of the patient for a portacayal anastomosis
The dotted line marks the line of the incision

·71 PORTACAVAL AVASTOVOSIS of approach should be removed During this dissection oozing from tiny Vessels may be troublesome, but it is best controlled by patient pressure and the cautious use of deatherm On more than one occasion we have inserted a pack and adjourned for refreshment at this stage of the operation, but the knowledge that a patent portal ein has been shown by venography is an encouragement to continue with patient dissection Ilhen once the portal sein is exposed it is found to lie in a tunnel of connective tissue, to which it is only loosely adherent, and it can be dis sected free using small gauze pledglets (Fig 44) Careful watch must be kept for any tributaries as troublesome bleeding will occur if they are torn, they should be divided between ligatures of the finest silk These tributaries are inconstant, but one is not infrequently found entering the left laterai wall of the vein, ie the side away from the operator When there has been some thrombosis in the vein adhesions are found outside the vein to the adjacent connective tissue which make the dissection much more difficult, previous venography will have demonstrated that there is a fairly wide lumen, so in the cases suitable for a portacaval anastomosis any thrombus present is likely to be attached to the wall on one side of the vein only Sometimes mural thrombt can be successfully removed from the vein after it has been divided during the next stage of the operation but if the lumen is adequate and the thrombs are firmly adherent, they should not be disturbed As soon as a ligature can be passed round the vein the latter can be drawn aside which facilitates further dissection upwards and downwards The rein should be freed down to the upper border of the duodenum and upwards to the hilum of the Inver, preferably so that its division into the right and left branches can be defined Ligatures are then placed round these branches but not tied (Fig 45) In some cases the division of the ein is very high and, provided an adequate length of the main sein is freed it is not necessary to expose the branches Exposure of the evena cara Only the anterior surface of this structure need be exposed, and the adventitia should be removed over an area of about 5 cm by 3 cm This means that it is cleared from the level of the entrance of the left renal vein up to its disappearance behind the liver Occasionally one or two small tributaries require to be ligatured and divided, but it is important to make the anastomosis as high as possible in the vena cara This is a convenient time in the operation to measure the portal venous pressure, erther by a canula in a tributary or a needle in the portal sein itself (Fig 6) Making the anastomosis The next step is the division of the portal vein The higature round the upper end of the portal rein, or those on its tuo branches, are tied and a Blalock clamp placed on the main vein as low down as possible, there should be at least I 5 cm of the vein betu een the clamp and the point of division (Fig 46) Blood is aspirated from the

·Lesser
Omantum
woll Bladder
Fig 3
Portacas al anas
tomosis Exposure of the
right free border of the lesser
omentum and the foramen
of 1 inslow
Llver
W Foramen of
Winslow
Liver
4,
Duodenum
Fig 14
Portacaial anas
toTOSIS The portal sein
exposed in the margin of the
lesser omentum

·PORTACAVAL ANASTOMOSIS 73 vein to ensure that the clamp is sufficiently tight before the vein is divided, but if it is suspected that there is some clot in the vein it is better to cut the vein across before the clamp is tightened so that any loose clot may be washed out The vein is now cleanly cut across just below the ligatures and washed clear of blood If there is any recent clot present it is wise to release the clamp to ensure that there is an adequate flow of blood Before doing this two stay stitches are placed in the cut end of the vein, so that its retraction behind the duodenum is prevented, the end of the vein is then compressed between the left thumb and index finger while the clamp is removed On releasing the grasp it will be seen if there is a good flow of blood, and if so, the Blalock clamp is reapplied while the end of the vein 15 again compressed between the thumb and index finger If there is not a good flow, clot may be removed by a strong sucker, or by gall-stone forceps passed along the vein (Fig 47) It is important to ensure that there is a wide lumen before the clamp is reapplied If clot has been present i ml (5000 units) of heparin should be injected into the vein behind the clamp A suitable clamp (such as the Satinsky clamp) is now placed so that it picks up part of the anterior wall of the vena cava, and this is also aspirated to ensure that the clamp has efficient Fig 47 Portacaval anastomosis control (Fig 48) A vertical incision Method of removing blood clot is made in the vena cava, its length from the portal sein being half the circumference of the portal vein (Fig 49) The anastomosis is made with arterial sutures using a continuous everting stitch, this being less likely to initiate thrombosis than a simple over and over stitch It 18 helpful to put a stay stitch in the outer edge of the incision in the inferior vena cava to hold it aside while the posterior suture line is being inserted The medial wall is sutured first using a continuous everting mattress stitch and starting at the top and when the suture has been pulled tight it is secured at the lower end by a lock stitch (Fig 50) A fresh stitch for the lateral wall is tied to the upper end of the medial wall suture, and when the suturing is com pleted it is secured to the lock stitch at the lower end As soon as the suturing is completed the clamps are removed when as a rule there is hardly any leakage, light pressure on the suture line may be necessary for a few minutes, but it is unusual for any additional stitches to be required In many cases it is possible to see through the thin wall of the vena cava the turbulence of blood caused by the portal inflow (Fig 51) 6-PH ·Ligature round left brunch of Portal Volm Ligature round Portal Veln Fig 45 Portacavat anas tomosis The portal vein has been freed, and ligatures have been placed round it, and round its two m3in divisions Ligature round Right branch of Portal Voin hymns Blalock Clamp Fig 46 Portaczial anastomosis A Blalock clamp has been placed on the portal Tein and the ligaturcs tied The vein has been aspirated to ensure that the clamp and Ligatures are tight Caudate Lobe of Liver Lkzatures on branches of Portal Vajn ·PORTACAVAL ANASTOVOSIS 75 If desired a second reading of the portal pressure may be taken at this stage and before the wound is closed a small portion of liver should be taken for histological examination This should be from the surface of the night lobe well anay from the edge or from the gall bladder, in order to obtain a sample which is representatie of the liver as a whole Closure Drainage of the pleura is unnecessary, but a tube should be inserted into the pouch of Rutherford Vorison and brought out through the anterior end of the wound and led to an underwater seal It can usually be removed on the following day Fig 50 Portacaval anastomo The diaphragm is repaired with in The posterior layer of the terrupted stitches and the pleura suture using a continuous mattress stitch closed by suture of the periosteum of the rib bed The remainder of the incision is closed in layers A blood transfusion will have been required during the operation in SIS i COD PV livc Tig 51 Portacaval anasto mosis The completed anas tomosis the portal ein is seen emerging from thind the common bile duct GO Anastomos ·Fig 48 Portacaval anastomosis The portal vein has been divided and a lateral clamp picks up a portion of the anterior wall of the vena cava TVC py Fig 49 Portacasal anastomosis Incision in the wall of the sena cava ·CHAPTER VITI SPLENORENAL ANASTOMOSIS This was the first type of venous shunt to be employed during the period of the revival of interest in vascular surgery which has taken place during the last fifteen years or so, for Blakemore and Lord's first cases were treated by removal of the spleen and the left kidney (1945), and the end-to-end anastomosis of their respective veins It was not long before it was realised that there was no need to sacrifice a healthy kidney and that an anastomosis to the side of the renal vein nas as effective In our experience, however, it has not been such a satisfactory procedure as a portacaval anastomosis, we have only employed it in those cases, mostly children with extra-hepatıc obstruction, in which the portal vein has not been suitable The operation itself involves the removal of the spleen, no small item when there are vascular adhesions, the vein is not so satisfactory as the portal vein for holding sutures and, especially in children, it may not be large enough for a satisfactory opening On the whole, therefore, it is a more formidable operation than a portacaval anastomosis, and should be reserved for those patients in whom the latter is not possible, in whom the liver function is adequate (the same indications apply in this respect as for a portacaval anastomosis, see Chapter VII), and in whom the splenic vein is large It will be seen then that the numbers requiring this operation are few, and in fact during the last seven years we have only performed it ten times compared with eighty-five portacaval anastomoses The operation is best carried out through a left abdomino-thoracic incision along the bed of the ninth or tenth rib The patient is supported with the anterior axıllary line uppermost It is a wise precaution to open the lesser sac through the gastro-splenic ligament early in the operation (Fig 52), and dissect out and ligature the splenic artery before the spleen is handled This reduces the vascularity of the peritoneal folds which carry collateral vessels There are two special points in the removal of the spleen, the first is that much oozing may be avoided if the vessels in the lieno-renal fold of the peritoneum and those passing up from the spleen towards the diaphragm are coagulated by deathermy before they are divided (Fig 53), the second is that the splenic vein must be preserved and handled very gently. The vein is kept intact until all the other attachments of the spleen have been divided, it is then secured by a bull-dog clamp and divided, and any blood distal to the clamp is washed out Between 3 and 5 cm of the vein should be freed, and this may entail ·6 PORTACAIAL ANASTO VOSIS most cases, in order to replace any blood loss During the post-operate period a watch must be kept for a pleural effusion which may require aspiration The most serious complication is the neuropathy which may be precipitated by the operation, the signs and treatment of this are discussed in a later chapter It has not been the custom to administer antı coagulants during the post-operatne period and their use is regarded as unnecessary D 1 TURA & REFERENCES Land inon T W (1945). "The Technique of Vising Viralcaval Shunts for Portal Hypertension Surg Ginec Obstet, 96 375-376 FCI,NI 11677) Loliennung Lotation of the lena Porta" ' l'orn med St Petersburg, 130, 1-2 SATINSKY, V P (1948) "Thoraco abdominal Approach for Portacaval Anasto mosis" Ann Surg, 128, 938-947 VIDAL, E (1910) "Discussion of the Eck Operation" Rer Chir, l'aris, 42 1181-11 82 ·CHAPTER VITI SPLENORENAL ANASTOMOSIS This was the first type of venous shunt to be employed during the period of the revival of interest in vascular Surgery which has taken place during the last fifteen years or so, for Blakemore and Lord's first cases were treated by removal of the spleen and the left kidney (1915), and the end to-end anastomosis of their respectise veins It was not long before it was realised that there was no need to sacrifice a healthy kidney and that an anastomosis to the side of the renal vein was as effective In our experience, however, it has not been such a satisfactory procedure as a portacaval anastomosis, we have only employed it in those cases, mostly children with extra-hepatic obstruction, in which the portal sein has not been suitable The operation itself involves the removal of the spleen, no small item when there are vascular adhesions, the vein is not so satisfactory as the portal vein for holding sutures and, especially in children, it may not be large enough for a satisfactory opening On the whole, therefore, it is a more formidable operation than a portacaval anastomosis, and should be reserved for those patients in whom the latter is not possible, in whom the laser function is adequate (the same indications apply in this respect as for a portacaval anastomosis see Chapter VII), and in whom the splenic iein is large It will be seen then that the numbers requiring this operation are fer, and in fact during the last seven years we have only performed it ten times compared with eighty-five portacar al anastomoses The operation is best carried out through a left abdomino-thoracic incision along the bed of the ninth or tenth rib The patient is supported with the anterior axillary line uppermost It is a vise precaution to open the lesser sac through the gastro-splenic ligament early in the operation (Fig 52), and dissect out and ligature the splenic artery before the spleen is handled This reduces the vascularity of the peritoneal folds which carry collateral vessels There are two special points in the removal of the Ir the vessels in the from the spleen before they are divided (fig 53), the second is that the splenic vein must be preserved and handled very gently The rein is kept intact until all the other attachments of the spleen have been divided, it is then secured by a bull-dog clamp and daysded, and any blood distal to the clamp is washed out Betsscen 3 and 5 cm of the vein should be freed, and this may entail 1 ·SPLEVORE AL ANASTOVOSIS Fig 52 Splenorenal anastomosis Exposure of large collateral vessels in the gastro splenic ligament Fig 53 Splenorenal anastomosis Fxposure of the lieno renal fold of peritoneum by drawing the spleen forwards ·its vein SPLENORENAL ANASTOMOSIS 79 ligature and division of a few small tributaries coming from the tail of the pancreas The next step is the mobilisation of the kidney and the preparation of An incision is made in the peri-renal fascia and the kidney de- • livered from its bed; the vessels of the hilum are next dissected out, and to do this it is helpful to place a tape under the hilum and over the upper and lower poles so that traction on the tape draws the hilum of the kidney up into the wound (Tig 54) Both the artery and the vein should be dissected out as, if the artery is not occluded while the anastomosis is being Fig 54 Splenorenal anastomosis Drawing the kidney forwards by means of a tape passed behind the hilum made, the kidney becomes very engorged At least 3 cm of the vein require to be isolated, which may entail ligature of one or two small tributaries Some surgeons use a small clamp which only occludes part of the lumen, but owing to the size of the vein we have found it simpler to occlude the vein completely The actual anastomosis does not as a rule take longer than twenty minutes, and we have seen no evidence of damage to renal function as shown by post-operative intravenous pyelography, when the renal function has been controlled for as long as this However, if more time is required it is advisable after twenty minutes to release the clamp on the renal artery for a few seconds and thus allow some fresh arterial blood to flow into the kidney When all is ready for the actual anastomosis, an encircling ligature is placed round the renal vein close to the kidney or preferably one round

·80
SPLENORENAL ANASTOMOSIS
Fig 55 Splenorenal anastomosis
Suturing the end of the splenic
Vein to the side of the renal vein
by a continuous everting stitch
Fig 56 Splenorenal anastomosis
The completed anastomosis.

·SPLENORENAL ANASTOMOSIS 81 each of its two main tributaries, and traction of this ligature occludes the Vein A Blalock clamp is very suitable for the proximal end of the vein, as it allows the assistant to hold the sein steady Before these are tightened the bull dog clamp is placed on the renal artery A transverse incision is made in the upper wall of the vein, the splenic vein brought down to it and the anastomosis is made with a continuous everting stitch (Fig 55) If the splenic vein is small it should be cut obliquely to increase the size of the lumen of the anastomosis When the anastomosis is com plete the clamps are removed and the kidney returned to its bed, making sure that there is no kinking of or pressure on the splenic vein (Fig 56) The wound is closed with a drain down to the region of the anastomosis for the first twenty-four hours The post operative management of this operation is the same as that after a portacaval anastomosis REFERENCE BLAKEMORE A H and LORD J W (1945) The Technique of Using Vital Inom Tubes in Establishing Portacaval Shunts for Portal Hypertension' Ann Surg , 122, 476-489 ·CHAPTER IV DIRECT OPERATIONS ON THE VARICES I Injection therapy This was introduced by Crafoord and Frenckner (1939) who used a sclerosing solution injected through a needle passed down an oesophagoscope When the size of these vessels is noted at an operation it must be realised that the prospects of getting adequate thrombosis by this method are remote, the method has been used very little during recent years Macbeth (1955) reported his results in 30 patients but when the difficulties of assessing the prognosis are considered it is not easy to judge the value of this method Repeated passage of the oesophagoscope is necessary, his patients having injections on the average on five separate occasions He aduocates removal of the spleen as an adjunct to injection therapy 2 Portal-azygos disconnection or gastric transection Tanner (1950) described an operation which was intended to dusert the portal blood from the dangerous area of the oesophageal varices, this conststed of dividing all the venous channels in the wall of the body of the stomach, the lesser omentum and the gastro-splenic ligament above the level of the ime of section of the stomach, which he dryded across completely and resutured (Fig 57) Later Tanner (1958) added the freeing of the louer 5 cm of the oesophagus and upper 5 cm of the stomach of all their vascular connections, for injection specimens had shown that blood was still reaching the varices by veins around the cardia, this revised operation has been done by him in il cases which were bleeding at the time with early control in every case In theory this is a good operation, but it is difficult to perform, particularly if there are vascular adhesions to the stomach and spleen, and the senous channels tend to reform across the line of suture In order to present this Meredino and Dillard (1955) have suggested the interposition of a segment of jejunum betueen the stomach and oesophagus, but as the pressure in the jejunal veins is high it hardly seems that the prospects of fresh anastomoses would be lessened by this procedure 3 oesophageal transectron In view of the technical difficulties sometimes encountered in making a portal-azygos disconnection by the abdominal route, I have carried out a number of similar operations through the left pleural cavity, transecting esther the lower end of the oesophagus or the fundus of the stomach after incision of the diaphragm When transecting the stomach by this route care is taken at the same time to donde the upper ends of the lesser omentum and the gastro-splenic liga- ·DIRECT OPERATIONS ON THE VARICES 83 ment as in the operation devised by Tanner In the majority of cases, however it is the oesophagus which has been transected This is a simple and safe operation except when operating as an emergency for haemorrhage It may be said that the veins are di ided too high up above the point where rupture occurs This is a valid argument, but in fact the results of oeso phageal transection have been no more unsatisfactory, in fact a little better than operations on the stomach Though fatal haemorrhage does occasionally occur from varices in the fundus of the stomach, this is rare when compared to the risk of bleeding from the oesophagus It is probable that ज Fig 57 Gastric transection with remoral of the spleen The transection is made about 5 m below the cardia and all vessels to the upper third of the stomach are divided some of the varices below the line of section become thrombosed as a result of stasis of their contained blood after the section In the earlier cases in the series the oesophagus was divided completely across just above the diaphragm and resutured, but after a time it was appreciated that nearly all the varicose channels lie in the mucosa and sub-mucosa, and it was very unusual to find any in the muscular coats There was therefore no point in dividing the muscle, so that in many of the cases the muscular coat has been split lengthwise, and only the mucosa and sub-mucosa divided horizontally and resutured (Fig 58) At the same time any large Sessels on the outer surface of the loner oesophagus are ligatured The technique is as follows The left pleura is opened through the ·84 DIRECT OPERATIOTS ON THE VARICES rib-bed of the eighth or ninth rib, the pleura over the oesophagus inçised, and the lower 5 cm of the oesophagus mobilised and lifted from its bed by tapes A light clamp is then placed across the lower end immediately above the diaphragm, the muscle incised longitudinally for about 5 em and the mucosa, which separates easily from the muscular coat, is lifted out, cut across transversely and resutured by a continuous catgut stitch The clamp is then removed and any bleeding points at the lower end are underrrun by catgut stitches A few sutures bring the muscular fibres together and any large veins on the surface below the lesel of the section are also underrun by sutures The oesophagus is then replaced in the mediastinum, with a few stitches in the mediastinal pleura, and the chest (a) (6) (d) Fig 58 Oesophageal transection (a) The muscle is split for about 5 cm (6) The mucosa is exposed (s) The mucosa and submucosa are cut across (d) The mucosa and submucosa are sesutured with a continuous catgut stitch. closed, with or without a temporary drain, depending on whether all oozing has been completely stopped When a gastric transection is carried out through the thorax, a radial incision is made in the diaphragm leaving the fibres around the oesophageal hiatus The fundus of the stomach is then drawn up and dided between clamps a little belon the cardia After division of the upper short gastric vessels and the upper end of the lesser omentum, the stomach is resutured in two layers and then the diaphragm and the chest wall closed During the last few years this operation has, with one exception, been performed only in children in whom it was thought that oesophageal transection might lead to same structure formation Desophageal or gastric transection has heen aduocated under three circumstances in patients who have had serious blecding first when there ·DIRECT OPERATIONS ON THE VARICES 85 was no vein suitable for an anastomosis (Fig 59) second if the liver function was considered too poor for an anastomosis or third in patients who had had evidence of portal systemic neuropathy It may also be advised in patients who are above the age level at which a venous shunt can be recommended 4 Ligature of the earices Boerema (1949) described as an emergency operation incision of the oesophagus lengthways by a left transpleural Fig 59 Trans splenic venogram in a patent with extra hepatic obstruction although the splenic vein was patent it was considered too small for a splenorenal anastomosis and an oeso phageal transection was done Note the collateral circulation fills the left subcostal vein Ann Sentrus anastomosis after only a short interval 5 Mediastinal packing This operation was suggested by Garlock and Som (1990) as a means of promoting vascular adhesions on the outside of the oesophagus which might drain the blood away from the varices inside ·86 DIRECT OPERATIONS ON THE VARICES and thus reduce the pressure inside them The upper mediastinum was approached through an incision in the neck and the louer oesophagus by a right pleural approach, gauze strips were placed alongside the oesophagus and their ends left protruding through the sound or through a separate stab incision, they were removed after about fourteen days 6 Oesophago-gastrectomy Excision of the varicose bearing area was first recommended by Phemister and Humphreys (1947) and has since been employed by Macpherson and others (1956) It would be thought that new renous channels are just as likely to develop across the new suture line after this operation as after a gastric or oesophageal transection There are very few patients who have had a resection of the fundus of the stomach and cardiac sphincter for other lesions, especially for malignant growths, who are able to enjoy good digestion afterwards The risk of such indigestion is quite a high price to pay for the benefits of the operation which are problematical, and in a condition in which the prognosis is sa unpredictable So far only a few patients treated by this rather drastic method have been recorded and no results are available which would indicate that it offers a greater freedom from recurrence of haemorrhage than other operations which carry less operative risk 7 Promotion of intra peritoneal adhesions Rutherford Morison of Neu castle and Talma of Utrecht independently introduced the operation of omentopexy and scarification of the Iner with a view to the promotion of adhesions which would proude portal systemic anastomoses 1 heur operations were done as a means of treating ascites, particularly associated with alcoholic cirrhosis, and met with some success, especially if the patients abstained from the taking of alcohol Grey Tumer reported some patients who had sun ned for many years after this operation, but the fickle prognosis of the disease makes the assessment difficult As a means of preventing haemorrhage there does not seem to be any place for this type of operation, but where there is evidence that the main cause of the ascites is due to obstruction to the outflow tract from the Iner, and other methods fail to gne relief there may be a justifiable reason for attempting to promote vascular adhesions between the liver and the parietal peritoneum Talc has been employed for this purpose in 2 of my own cases and its use has also been reported by Madden and others (1954), but there are not sufficient cases so far described to judge its long term effect on the course of the disease REFERENCES BOEREXIA, I (1949) 'Bleeding Varices of the Desophagus in Cirrhosis of the Liver and Banti's Syndrome' Arch Chu Neerland, 1, 253-260 CRAFOORD, C and FREACANER P (1939) "Neu Surgical Treatment of Varicose Veins of Oesophagus Acta Otolaryng, 27, 422-429 ·8 DIRECT OPERATIONS ON THE VARICES CRILE G (1957) Transesophageal Ligat on of Bleeding Varices Surgery 42 583 584 GARLOCK H and Sow I L (1950) Further Observations on Packing of Vediastinum for Esophageal Varices I TI orac Surg 19 572 586 LISTOV R R and Ellis D S (1956) Emergency and Definitie Treatment of Bleeding Esophageal Varices Í Amer med Ass 100 1017 1023 VACBETH R (1955) Treatment of esophageal Varices in Portal Hyper tension by means of Sclerosing Injections Brit med ) 2 877 880 MACPHERSOY AIS OWEN J A and IVNES ) (1956) Surgical Treatment of Portal Hypertension Lancet I 353 357 VIADDEN J LLORE J V GEROLD F P and Ravid J M (1994) The Pathogenesis of Ascites and a Consideration of its Treatment Surg Gynec Obstet 99 385 391 MEREDINO E A and DILLARD D I (1955) Sphincter Substitution by an Interposed Jejunal Segment at the Oesophago gastric Junction Surg 142 486 509 PEMISTER D B and HUNIPHREYS EM (1947) Gastro esophageal Re section and Total Gastrectomy in the Treatment of Bleeding Varicose lens in Bantas Syndrome Ann Surg 126 397 410 TANNER C (1950) Discuss on Gastroduodenal Haemorrhage as a Surgical Emergency Proc R Soc Ved 43 147 152 -(1958) Operative Vlanagement of Haematemes s and Mclaena Ann i Colt Srg 22 30-42 ·CHAPTERY PORTAI SYSTEMIC NEUROPATHY It is only in the last few years that it has been shows that some of the neurological symptoms which are found in cases of liver disease are duc to the shunting of portal blood direct into the systemic circulation, a hence it reaches the brain before it has passed through the inser (Sherlock et al, 195+) As the treatment of portal hypertension wluch 18 adı ocated includes methods thich increase the volume of blood which by-passes the Inver it is only to be expected that these neurological effects may follos such treatment (IcDermott et al 1954) Although these manifestations do in fact occur in many patients who have had no operation, and are duc to the natural collaterals which have developed, it has been found that they are the most sersous complication of the shunt operations It folloss therefore that the surgeon must consider this risk in each parucular case sf operation is adussed, and must be able to treat the complication of it arises during the post-operatae period It thus Occurs that the mose efficient the ienous shunt in diverting blood from the dangerous varices, the greater is the risk of neurological complications There can be no doubt that an end-to-side portacar al anastomosis, which drerts all the portal blood from the liver, is much the most efficient as a means of preventing haemorrhage from the alimentary tract in cases of portal hypertension, but it carries a greater risk of neuropathy than the less efficient splenorenal anastomosis The difference is such that some writers baie ndiocated a splenorenal anastomosis of the splenic vein is large, in preference to the portal tein-lena cara type of shunt With this argument I cannot agree, for if the patient's lefe is endangered by haemorrhage, the most efficient operation which will present that haemorrhage should be employed, if the patient has had evidence of neuropathy before operation, then it is probably unwise to add in any way to the venous shunt and an alternative operation on the varices would be indicated Patients whose Iner function is poor have a greater tendency to this type of neuropathy, and this is an argument in fasour of limiting shunt operations to patients whose liver function has not deteriorated too far, as laid down in the indtcations for the operations in Chapter VII The precise nature of the toxic substance or substances sshich cause these effects on the brain has not yet been determined In some respects the symptoms resemble those of ammonia intoxication and many of the patients have raised blood ammoniz levels, but this has not been a con ·89 PORTAL SYSTEMIC NEUROPATHY stant finding; there is no doubt that some protein breakdown products containing nitrogen are responsible, but further work is required to identify them and to find a specific neutralising agent It has been shown that there may be a rise in the level of certain keto acids in the blood (Dauson et al, 1957) but there is no proof that they are the responsible substances Modern therapy, which has met with considerable success, is aimed at preventing such substances gaining access to the circulation rom the alimentary canal Though this concept of portal systemic neuropathy has been generally cepted, patients with Iner failure may have serious disturbances of nsciousness, even when there is little or no collateral circulation and this inay in fact be due to lack of substances in the blood stream which are normally supphed only by the liver, and which are essential for central nerve function It has been suggested that 5-hydroxytryptophane may be such a substance (Bessman, Merlis and Borges, 1957) The symptoms The neuropathy manifests itself in a variety of ways but in the same patient each episode tends to hold to a constant pattern The most usual symptom is increasing drowsiness passing to coma, but some patients may pass very suddenly from an apparently normal state to complete unconsciousness, and equally suddenly revert to their fully alert condition In the mild cases the patients have expressionless features are slow in answering questions and may not always give rational ansuers It is at this time that the characteristic flapping tremor of the hands may be elicited if the patient is esked to hold them out in front of him Other patients may be fuite conscious but irrational and say and do peculiar things, as a result hey may be considered psychotic cases and be sent to a mental hospital An unusual manifestation is excessive salivation, one of our patients suffered severely from this for two years after a portacaval anastomosis had been performed, though in every other way he behased quite normally In a few instances there may be personality changes, and though the patients may seem quite normal to those who do not know them well, their relatnes say that there has been some change in their behaviour, which the patient does not recognise himself The syndrome is much more likely to occur in elderly patients, thus in one series recorded (Summerskill et al, 1956) among 17 patients there were only 2 under the age of 40 years, and the average age 138 49 In 4 of these 17 cases a portacaval anastomosis had been performed, in the remainder there was a natural collateral circulation In the writer's series of portacaval anastomoses the average age of 80 patients was 37 years, but among the 16 who have shown signs of neuropathy after the operation the average age was 42 5 years 7-р Н ·go PORTAL SYSTEMIC NEUROPATHY According to Summerskill et al (1956) the blood ammonia level is usually raised and in the cerebro-spinal Ruid increase in both the ammonia and glutamine levels are constant findings The electroencephalogram shows slowing of the dominant rhythms and srregular wase patterns Factors which precipitate an attack are either a large protein meal or a gastro oesophageal haemorrhage which has the same effect in Introducing protein into the alimentary canal Some patients can tolerate a low protein dset but shox symptoms when this is raised abose a certain level Their tolerance of protein may not remain constant and probably depends on variations in their ner function In the first 80 patients of our series of portacaval anastomoses neuron logical complications have been noted in 8 cases during the early postoperatn e period, all 4 patients ho have died while still in hospital hase succumbed from this cause, it is possible that reath our present knowledge of the ætiology and the treatment atalable, some of them might have been saved Two patients, aged 38 and 31, became comatose on the day following operation, the older patient never fully regained conscsousness and died on the thirty-first post-operative day, no definite cerebral changes being found at auropsy The younger patient who also had bronchiectasis, so St-Lumo in the fifth day at 11s making food progress until died four days later The third fatal case, a man of 43 years lo mau extensue thrombosis in lus portal system had a small haemorrhage on the fifth post operative day and relapsed into coma, dying two days lates The fourth a noması of 57 years showed no abnormal signs for four days but then became drousy and remained in varying states of consciousness unul she died in coma six weeks after her operation In one case there was a bout of coma which lasted from the third to the fifth post operate day and the patient has shown no sign of any recurrence, another case had intermittent coma for three months, but then fully recovered and died of an unrelated cause (ruptured aortic aneursm) to years later As example of the danger of blood in the intestine is given by a man who norrlt fn retra after his portacaval anastomosis underwent a gastrectomy operation Eser since, he has had to restrict his prules that LA liable to get' giddy and fainting" turns Light patients after partscaval anastomosis hase had episodes of neuropathy after they returned home, in tuo, severe attacks leading to coma have happened the 2 feu days of leasing hospital, and patients should be warned against celebrating the occasion by a large protein meal, in nether case isas there been any recurrence, though rhey hase both been followed for two years ·PORTAL SYSTEMIC NEUROPATHY 91 Treatment As some of the patients who have had portal systemic neuropathy have had normal liver function before operation and have been able to take a high protein diet caution should be exercised in the early post operatie course of all patients who have had a shunt operation They should start with a very low protein diet but if all is well this can be increased by about 5 gm protein a day until they are on a normal diet If any signs of drowsiness or unusual mental behaviour appear a protein free diet should be at once reverted to and if the symptoms disappear the increase may be tried again If the state of consciousness deteriorates more active measures must be instituted The colon should be washed out and an enema given in order to remove the protein in the alimentary canal as quickly as possible if there is a possibility that haemorrhage into the stomach has occurred a stomach tube should be passed and the gastric contents aspirated The patient should also be given an antibiotic in order to abol sh the bacterial activity in the intestine at first aurcomycin was given 1 1 - may help to prevent pulmonary complications and in the unconscious patient it may be necessary to aspirate secretions from the trachea Intra venous fluids may be required in order to maintain fluid balance Walshe (1956) has recommended the use of intravenous sodium glutamate 20 gm in a litre of salıne on the basis that this substance removes ammonia by converting it into glutamine we have used it on a number of occasions but are unable to convince ourselves that it has any specific action There are some patients who find themselves unable to take a diet con taining a normal amount of protein and as a result of restriction of protein intake feel some lassitude and are unable to do their work such patients may be able to tolerate a fuller det if they are kept on a small maintenance dose of aureomycin or neomycin and it is worth giving this a trial When advocating a shunt operation the risk of portal systemic neuro pathy should always be kept in mind in some patients who have had severe haemorrhage and for whom a shunt is likely to be the only way of preventing a recurrence this risk will have to be taken in others the risk may seem too great At the moment this complication provides the greatest anxiety concerning thts method of treatment of portal hyper tension ·92 PORTAL SYSTEMIC NEUROPATHY REFERENCES FOTr TL ano Ronate fluo Drac sp "Effect of 5-Hydroxy- Prou Soc Coma” DAU DAX McDERNIOTT, V, ADAYS, RD and RIDDELL, A G (1954) "Ammonia Aletabolism in Man" Ann Surg, 140, 539-554 SHERLOCK, S. SUWNERSKILL, WHY, WHITE, L P and PIEAR, E A (1954) **Portal-Systematic Encephalography Neurological Complications of Liver Disease" Lancet, 2, 453-457 SUMMERSKILL, WH J, DAVIDSOV, L A, SHERLOCK, S and STEINER, R E (1956) "The Neuropsychratric Syndrome associated with Hepatic Cirrhosis and an Extensie Portal Collateral Circulation Quart y lled, 25, 245-266 VALSTIE, J M (1956) "Hepatıc Coma" Post Grad med 9, 32, 467-472 ·CHAPTER XI PORTAL HYPERTENSION IN CHII DREN Portal hypertension presents a number of special problems when it occurs in children so that the subject deserves a chapter to itself It has already been mentioned that it is the most common cause of haematemesis in this age group and any child who has this symptom must be fully investigated with this in mind It is found in practice that in the great majority of all cases of portal hypertension the lesion starts as an intra hepatic obstruction but to this children are an exception owing partly to the relative rarity of Ext a hepat c obst uct on Intra hepat c obst uct on Intra and ext a hepat c obst uction Doubtful 30 cases 121 cases 8 cases 1 case 0-4 5 9 10 14 18 19 20 24 25 29 30 34 35 39 40 44 45 49 50 54 55 $9 60 64 65 69 Years Fig 60 The age at which haematemesis first occurred in 160 patients There 1$ a preponderance of cases of extra hepatic obstruction below the age of 15 years cirrhosis of the liver at this age The other most important factor is the frequency of thrombosis of the portal vein in children giving rise to extra hepatic obstruction There are a number of cases on record in which there has been a definite history of umbilical infection and it is presumed that this has led to a septic phlebitis of the umbilical and portal veins, but in many cases no such history can be obtained yet in my series it is not until the fourth quinquennium of life is reached that the number of cases of haemorrhage due to intra hepatic obstruction exceeds those due to obstruction of the portal vein outside the liver As a result of this dis tinction treatment is a much greater problem in children in spite of the fact that they have no disease of the liver Figure 60 shows the age at which the first haemorrhage occurred in 160 ·92 PORTAL SYSTEMJC NEUROPATHY REFERENCES BESSMAN, S P, MERLIS, J K and BORGES, F (1957) "Effect of 5-Hydroxytryptophane upon Electroencephalogram in Hepatic Coma" Proc Soc erp Biol, 95, 502-504 DAWSON, A M, de GROOTE, J, ROSENTILAL, W. S and SHERLOCK, S (1957) "Blood Pyruvic Acid and Alpha-ketoglutaric Acid Levels in Liver Disease and Hepatic Coma" Lancet, 1, 392-396 DAWSON, A M, MCLAREN, J and SHERLOCK, S (1957) "Neomycin in the Treatment of Hepatic Coma" Lancet, 2, 1263-1268 MCDERMOTT, W V ADAMS, R D and RIDDELL, A G (1954) "Ammonia Metabolism in Man" Ann Surg , 140, 539-554 SHERLOCK, S, SUMMERSKILL, 1 H), 11 HITE, L P and PEAR, E A (1954) "Portal-Systematic Encephalography Neurological Complications of Lover Disease" Lancet, 2, 453-457 SUMINIERSKILL, WH J, DAVIDSON, E A, SHERLOCK, S and STEINER, RE (1956) "The Neuropsychiatric Syndrome associated with Hepatic Cirrhosis and an Extensive Portal Collateral Circulation" Quart y Aled, 25, 245-206 WALSHE, J M (1956) Hepatic Coma" Post Grad med , 32, 467-472 ·CHAPTER VI PORTAL HYPERTENSION IN CHII DREN Portal hypertension presents a number of special problems when it occurs in children, so that the subject deserves a chapter to itself It has already been mentioned that it is the most common cause of haematemesis in this age group and any child who has this symptom must be fully investigated with this in mind It is found in practice that in the great majority of all cases of portal hypertension the lesion starts as an intra hepatic obstruction, but to this children are an exception, owing partly to the relative rarity of 28 Extra hepatic obstruct on Intra hepat c obstruct on B Intra and extra hepat e obstruct on Doubtful 30 cases 121 cases 8 cases 1 case hi 04 59 10-14 15 19 20 24 25 29 30 34 35 39 40 44 45 49 50 54 55 59 60-64 65-69 Years Fig 60 The age at which haematemesis first occurred in 160 patients There 18 2 preponderance of cases of extra hepatic obstruction below the age of 15 years cirrhosis of the liver at this age The other most important factor is the frequency of thrombosis of the portal vein in children giving rise to extrahepatic obstruction There are a number of cases on record in which there has been a definite history of umbilical infection and it is presumed that this has led to a septic phlebitis of the umbilical and portal veins, but in many cases no such history can be obtained, yet in my fies it is not until the fourth quinquennum of life is reached that the number of cases of haemorrhage due to intra-hepatic obstruction exceeds those due to obstruction of the portal vein outside the liver As a result of this distinction, treatment is a much greater problem in children in spite of the fact that they have no disease of the liver Figure 60 shows the age at which the first haemorrhage occurred in 160 ·94 PORTAL HYPERTENSION IN CHILDREN patients who bled from oesophageal or gastric varices, this group is not représentative of bleeding in cases of portal hypertension as Wholc for many of these patients had been selected as suitable for surgical treatment before they uere referred to me, probably many in the older age groups were excluded, but the purpose is to show the relative distribution of the intra- and the extra-hepatic groups It will be noted that all except four of the extra-hepatic cases had their first haemorrhage before the age of 25 and that there was no case of intra-hepatic obstruction who had his first bleeding before the age of 5, though more recently a case of congenital fibrosis of the liver has been seen who bled for the first time at the age of 3 From what has been said it is clear that most cases of portal hypertension in children are due to thrombosis of the portal zein, that they have normally functioning livers and that haemorrhage is the only symptom If untreated most of these children will eventually die of haemorrhage, but if this can be prevented they will lead perfectly normal lies I his is the tragedy, for very often these little patients have no suitable vein in the portal system with which a wide anastomosis can be made to a systemic sein, and thus there is no possibility of bringing the pressure in the portal system down to normal The portal sein is cavernous and replaced by many collateral channels of varying size, but none of them suitable for an anastomosis esther because they are individually too small or because their walls are thin and tear easily in some of these cases the splenic vein is also thrombosed and in the smaller children it is not large enough for an anastomosis which has a good chance of remaining patent Onc therefore has to turn to the less effective operative procedures on the oesophagus or stomach to prevent these recurrent haemorrhages. Jordan and his colleagues (1956) have reported the sesults of g children treated by various methods, 3 had cirrhosis of the Irver and in 1 the portal vein was also cavernous In 2 patients in whom a dilated vasıt around the obliterated portal vein was used for an anastomosis to the sens caya, the result was unsatisfactory When the splenic vein was patent, a splenorenal anastomosis done in three patients gare good results, in a 17-year-old patient in whom bleeding persisted after splenectomy, resort was had to a total gastrectomy, but the follow-up was only over a period of nine months These writers condemn splenectomy alone and consider that, even if bleeding has not occurred, a splenorenal anastomosis should be performed, but they regard + years as the lowest age for this owing the small size of the splenic vein in infants They point out, however, that in some cases of portal sein stricture, this vein may be suitable for a portacaval anastomosis Ekman (1957) reports a very similar experience In 30 patients with extra-hepatic abstruction in whom anastomoses 11 ere made, 22 were under the age of 20 when they had their first haemorrhages Eighteen of these One of the had splenorenal shunts and 4 had makeshift shunt operations ·PORTAL HYPERTENSION IN CHILDREN 95 last group died after the operation and the other 3 have all bled again Amongst the 18 patients with splenorenal anastomoses, 9 out of the 16 Sur lors have had recurrent haemorrhage, and 4 have succumbed Shumaker and King (1952) in a study of 21 patients who started having bleeding from varices before the age of 15 found that in only I was there evidence to suggest intra-hiepatic obstruction Five of these children had po operative treatment and 4 of them died of haemorrhage within a few months of their first attack The only surynor did not have her first bleeding until the age of 14 and further attacks continued at frequent intervals for the next 7 years In 10 patients splenectomy alone was the first operation performed and only 1 in whom the operation was not carried out until she was 32 years old, had not had further bleeding at the time of the report, this again points to the inadequacy of splenectomy as a means of treatment Oesophago-gastrectomy was done in 5 patients with 2 operative or post-operative deaths, all the cases had had previous splenectomy, and 2 of the sun ors had no further bleeding when followed for three and a half and two years respectively Bleeding recurred within one and a half years in 2 cases who had devascularisation of the oesophagus and stomach combined with vagotomy and gastrojejunostomy Eight anastomosis operations were carried out on 7 patients, 4 with splenorenal shunts had done well but in 2 others bleeding recurred, the other 2, in which the portal vein and a mesenteric vein were used for anastomosis, had only been followed for a short time The disappointing results of direct operations on the stomach or oeso phagus have led Koop and Roddy (1958) to try inserting a piece of transplanted colon, retaining its blood supply between the oesophagus and stomach after partial oesophago-gastrectomy, they have done this operation in 5 children, but in none was the follow-up longer than a year at the time of their report In my own series there are 45 patients who had their first haemorrhage from oesophageal varices before they reached the age of 20 In 19 the >bstruction was intra-hepatic and the age of onset is noted in the following able Cause of arrhosis schere known Ao of 4ge of onset of haemorrhage lo cause found 11, 12, 15, 16, 17, 18, 19, 19 19, 19 Congenital fibrosis of the liver 3 5 6,17 Congenital absence of bile ducts Schistosomiasis 13 1 tson's Disease Associated with fibro cystic pancreas 13 Associated with endocrine disturbance 14 Of the cases in which no cause was found, 8 had portacasal anastomoses, and all have done sell except the boy of 12 who died of Inser failure five cases TO 5 1 1 17 ·96 PORTAL HYPERTENSION IN CHILDREN years later The 4 patients with congenital fibrosis associated with proliferation of the bile ducts as described on p 21 did sell after a porticava! anastomosis, in children with haematemesis from varices and a hard lives this condition should be suspected, and as it seems that the function of the Iner does not deteriorate, the operation of portacaval anastomosis 15 strongly indicated, and should be done before secondary thrombosis occurs in the portal sein c. anastomosis was performed After six months the spleen was no longer palpable and during the subsequent four years she has led the normal life of a schoolgirl The histology of the liver is depicted in Fig 61 In the cases of schistosomiasis and hepato lentscular degeneration, a portacaval anastomosts has presented further haemorrhage for periods of six and one year respectively since their operations The other 5 cases were considered unfit for a senous anastomosis in view of the poor iner function in 3 cases, and because there was thrombosis in the portal and splenic ieins in the other 2 Four of them are known to have died As it is not always realised that patients #th congenital absence of the bile ducts can survive for a number of years, the following case is quoted RL was admitted to hospital at the age of 6 weeks, jaundice having been noted at 3 weeks, but hus stools had always been pale At a laparotomy the srall bladder was shrunhen and no common bile duct was found Orez the - Ihla ant . A. the US meru bile ducts in the Ich a case the user e hypertension The 26 patients trith Citra-hepatic obstruction hare on the whole shoun disappointing results of treatment One patient who had portal pyraemia at the age of 15 started bleeding two years later, she has had a splenorenal anastomosis and an oesophageal transection and has remained well for one and a half years Another patient, a boy who has a stricture of the portal sein (Fig 8), had a splenorenal anastomosis five years ago and remains sell A third had pneumococcal peritonitis at the age of 7 months, and probably at that time had portal sein thrombosis Tuo years later his spleen was much enlarged and he had his first haematemesis, bleeding continued at frequent inten als for the next tuo months and he died at the age of 2} following an exploraton operation, his liver was

·PORTAL HYPERTENSION IN CHILDREN
97
Fig 6: Histology of the later in a case of congenital
Fibrosis
(a) Low power, showing many large bile ducts in the
portal tracts and scarcity of portal scins x 30
(6) Higher power showing bile thromb in the smaller
bile ducts
x 70

·98 PORTAL HYPERTENSIOV IN CHILDREN normal but the portal sein kas represented by a spongy fibrous ligament although its branches in the liver were normal The ages of the remaining 23 at the time of onset of haemorrhage are shown in the following table 3 AGE OF OVSET OF ILAEVIORRHAGE IN 26 PATIENTS UNDER THE AGE OF 20 WITH EXTRA HEPATIC OBSTRUCTION Age of onset Mo of cases 2 4 I 4 I 2 6 3 7 1 10 4 IT I 12 2 14 1 IS 4 18 2 In none of these could any cause be traced but one child had had an acute suppurative arthritis of the hip during the first week of life and this may have been due to umbılıcal sepsis another child had acute osteo myelitis of the humerus when only a few months old and it is possible that his portal thrombosis occurred during that illness "The portal vein in each case was replaced by a cavernous mass and in many the splenic vern was also thrombosed In 14 the spleen had been removed and in every case but one the bleed ng has recurred Fig 62 shows the fate of 18 of these patients who have died or who have been followed for more than five years from the time of their first haemorrhage It will be noted that 7 of them are dead and the younger the age at first haemorrhage the more serious the outlook Sıx patients who had splenorenal anasto moses when under the age of ro have all had recurrence of bleeding In this group nine venous anastomoses have been performed but the only ones which have so far proved satisfactory are four done in patients who had their first haemorrhages at 14 17 18 and 18 years respectively None of the splenorenal anastomoses in smaller children have remained patent Sixteen patients with extra hepatic obstruction who fall in this age group have had transection operations either of the oesophagus or stomach all these patients except one had had a previous splenectomy or splenorenal anastomosis and the bleeding had recurred From all these accounts it appears therefore that the outlook for a child with portal hypertension with extra hepatic obstruction is poor unless there is a suitable wide vein in the portal system to which an anastomosis can be ·PORTAL HYPERTENSION IN CHILDREN 99 AGE 24 6.10 12 14 16 18 20 22 24 26 28 30 32 لللللللللللللليللا • Hormorrhage S-Splenectomy A-Venous anastomosis L-Splenic artery ligature T-Gastric or oesophageal transaction Fig 62 The fate of eighteen patients with extrahepatic obstruction who had their first haemorrhage before the age of eventy, and who have been followed for five years or more, or until death made, this is most likely to be a splenorenal anastomosis If the splenic sein is small, the operation is not worth doing lakeshift anastomoses using other vessels have nearly always ended in failure Direct operations on the oesophagus or stomach offer the best hope in such patients In children with intra hepatic obstruction the outlook, as in adults, depends on the function of the liver, but if there are large varices, or there has been bleeding and the liver function is reasonably good, an early portacaval anastomosis is strongly indicated before secondary thrombosis occurs in the portal vein Delay may mean that the only possible vern for an anastomosis becomes occluded, and the only opportunity of reducing the portal hypertension is lost REFERENCES JORDAV, P, PATTON, T B and BENSOY, C D (1956) "Portal Hypertension in Infants and Children " Arch Surg, 72, 879-888 CKMAN, C-A (1957) "Portal Hypertension Acta Chur Scand, Suprl, 222, 143 Stockholm Koop, CE and RoDDY, S R (1958). "Colonic Replacement of Distal Oesophagus and Proximal Stomach in the Management of Biecding Varices in Children" Ann Surg , 247, 17.25 SHUWAKER, H B and KING, 11 (1952)"Portal Hypertension in Children associated with Gastro-oesophageal Haemorrhage Arch Surg. 65, 499-510 ·CHAPTER XII RESULTS OF TREATMENT It has been pointed out hon difficult it is to judge the results of any form of treatment for portal hypertension, there are so many factors to be talen into consideration The question can be simplified to some extent of haemorrhage is regarded as the indication for operatie treatment, for in that case freedom from further bleeding can be taken as the main indica. tion of successful therapy Against this, however, must be set any other consequence of the operation che immediate mortalty, any adverse effects on liver function, the incidence of neuropathy and the long-term survival This can be considered against the background that, without surgical treatment something over so per cent of patients with intra-hepatie obstruction sll die within a year of their first haemorrhage, and Shumaker and King (1952) found that four out of five children with extra hepatic obstruction who had no operative treatment were dead within a few months In Chapters VII to IX many of the methods of treatment which have been aduocated were described, it is unfortunate that in only one of these methods have reports of an adequate follow-up been reported, so this will be considered first Portal-systemic anastomosis The immediate mortality of this operation depends very largely on the selection of the cases, when the operation has been perfarmed in the presence of poor iner function the risk is considerable Thus Blakemore (1952) found a post-operatııe mortality of 37 5 per cent in 16 patsents whose scrum albumin vas below 3 gm per cent and 468 per cent in 32 patients who had ascites which had not responded to medical treatment This contrasts with an early mortality of 177 per cent in his first 135 shunt operations of all types when the serum albumin was above 3 gm per cent Il'hen the operation was performed for extra-hepatic obstruction the overall operatie mortality was only 69 per cent compared with 21 9 per cent in the intra hepatic obstruction cases the most common cause of death was luser failure, which seems to include portal-systemit neuropathy, the other major cause 11.35 operaidye haemorrhage which was more serious in the splenorenal anastomoses than in the portacaval operations Less common causes of post operatuse fatalıtıes have been mesenteric thrombosis, and recurrent gastro-intestinal haemorrhage in his group ·101 mosis RESULTS OF TREATMENT In the writer's series of 111 shunt operations the most serious immediate complication has been portal systemic neuropathy, and the 4 cases of portacaval anastomosis who died from this cause have been described (Chapter 10) During the first three years that these operations were being undertaken (1947 to 1950) 2 patients died from operative haemor thage following splenorenal anastomosis, but there have been no deaths from this cause amongst the last 100 operations Miesenteric thrombosis caused one death a month after a splenorenal anastomosis There have been no deaths directly attributable to the operation amongst the 85 cases of end-to side portacayal anastomosis apart from the 4 cases of neuropathy who died on the 7th, 16th 31st and 39th post-operative days respectively Operative mortalities recorded by other writers are 93 per cent in 75 cases (Ekman) 12 per cent in 141 cases (Linton and Ells) and 15 per cent in 106 cases (Yłunt) In the published series the operative mortality has been lower in the patients with extra-hepatic obstſuction as these patients rarely have any defect of Inver function, in consequence the overall mortahty for splenorenal anastomosis is lower than that for portacaval anasto- However, if these two operations are compared when employed in cases of intra hepatic obstruction it is found that portacaval anastomosis has the lower risk Ekman has clearly demonstrated the vastly superior results of portacaval anastomosis when compared with splenorenal anastomosis when this aspect is considered In his follow-up from one to seven years there was a recurrence in 14 out of 32 patients with splenorenal shunts, com- Pared with none following 20 operations of portacaval shunts Linton and Ellis, with a follow-up extending to nine years had 13 recurrences amongst 66 splenorenal shunts (19 5 per cent) and 3 amongst 26 portacaval shunts (11 5 per cent) Blakemore s figures were 6 after 52 spleno renal shunts (11 5 per cent ) and 3 after 46 portacaval shunts (6 5 per cent) followed between one month and six years Our own experience 19 similar though the figures for splenorenal shunts are small Including the earliest cases where technique was imperfect and in some of which vitallium tubes were used, there are 19 splenorenal anastomoses who surmed operation followed for periods up to ten years, 4 early cases were operated on for ascites and should be excluded, of the others 10 of these have had further haemorrhage but it should be borne in mind that 5 of these patients were children with small splenic veins and the view has already been expressed as a result of this experience that the operation is not worth doing in children As a contrast, amongst 81 patients with portacaval anastomoses who survived operation and have been followed from a few weeks up to seven and a half years only 4 instances of further haemorrhage have occurred One of these was fatal, and another patient died of cerebral thrombosis a few weeks after a haemorrhage, but the other 2 have remained well for five years following oesophageal transections & ·102 RESULTS OF TREATMENT Three patients have died of unrelated disease and 4 others have died of liver failure In a recent survey of all the patients who had had end to-side portacaval Proc Some of the women have gone through successful pregnancies About 10 per cent have some ankle oedema and menstrual irregularities, these probably being a feature of their disturbed liver function In a few cases the liver disease is steadily progressive, particularly if the cause is not removed, those patients with haemochromatosis or Wilson's disease DEATHS Time since No of 1st 2nd 3rd 4th 5th 6th Operation Cases 7th Survivors year year year year (year year year Under l year 20 5 19 1 year 16 15 2 years 7 7 3 years 10 1 2 6 4 years 14 1 2 5 years 8 2 6 6 years 8 6 7 years 2 TOTAL 85 72 ld no V Fig 63 The survival of patients who have had end to side portacaval anastomoses tend to run a slow downhill course, and the same happens to those alcoholics who do not mend their ways Figure 63 shows the survival rate of 85 patients who had end to-side portacaval anastomoses, and Figure 64 gives the results of those patients who were operated on more than five years ago The appearance of the anastomosis in a patient who died three years after her operation is shown in Fig 65, and the senographie changes in another patient in Fig 66 Studies of liver function in most cases show no deterioration after the operation, the leucopenia and thrombocytopenia tend to persist but often in a less severe degree in all cases the spleen has become reduced in size and in about half it becomes impalpable two or three years after the operation ·PORTACAVAL ANASTOMOSI 2 3 4 PATIENTS OPERATED ON MORE THAN FIVE YEARS AGO 3128 10 9 8 7 6 5 3 > 6 7 O 10 MH 20 MS 26 AL 29 WC 27 GP 12 SL 15 JB 18 BB 48 JP JC 29 WJ 37 AS 13 TJ 25 ка 38 BL AC 24 WP 30 OMM 54 12 13 14 IS 16 17 18 CIT RESULTS OF TREATMENT Fig 64 The results of the first eighteen patients treated by end-to-side portacavaf anastomosis Five-year follow-up The vertical line represents the time of operation and the figures along the top the years before and after operation Each vertical stroke represents a haemorrhage The figures on the left gre the age at the time of operation Case 5 died of laver failure S=Splenectomy Case 9 died of haemorrhage SR=Splenorenal anastomosis Cases 14 and 15 died of neuropathy T=Oesophageal transections garding case 3, the last two haemorrhages has e been slight and not required transfusion Eor ·104 RESULTS OF TREATMENT Fig 65 The appearance of the portacaval anasto mosis as seen from the caval side in a patient who died of a ruptured aortic aneurysm three years after her portacaval anastomosis Transection operations and direct operations on the varices Though the immediate mortality of these operations low, whether carried out through the abdomen or thorax they are not nearly so efficient as a patent shunt in preventing further bleeding, the operation, however, carries considerable risk when performed as an emergency procedure in the bleeding patient Thus all 4 patients on whom the writer has carried out oesophageal transections as emergencies have died, but there has been no operative mortality amongst the 36 patients on whom it was an elective procedure Linton and Ellis (1956) however, saved 15 out of 20 patients on whom they carried out emergency suture of the varices The long term follow-up of these cases is also not encouraging though better than the expected survival if no operative procedure was undertahen In my own experience half the patients have had further bleeding but usually it is not so frequent or serious as before the operation, though ·RESULTS OF TREATMENT ros 6 of the 36 surs 11ors have died of haemorrhage between one and seven years after the transection operation Without quoting any figures, Tanner (1958) says that he has had excellent long term effects when gastric transection with freeing the vascular connection of the lower 5 cm of the oesophagus and upper 5 cm of both curvatures of the stomach has been carried out It looks as if this operation, which is more extensive than the oesophageal or gastric transections formerly carried out, will probably give better results, my own view is that the spleen should be removed at the same time if splenectomy has not already been performed Hepatic artery ligation This operation has been performed mainly in patients with rather poor liver function, where ascites has been the main symptom It might therefore be expected that nesther the initial results nor the long term suryasal would be very encouraging, thought in a few cases the relief of ascites has been dramatic Altmeser and his colleagues (1955) in a series of 8 cases, most of whom had both haemorrhages and ascites, had an early mortality of 27 1 per cent, and 50 per cent of their patients were dead skithin three and a half years Results were considered good in only a third of the patients Ruggiert (1955) had a post-operatne mortality of 29 per cent in 24 patients, and half of his cases were dead within seven months, but the other half were regarded as clinical cures and returned to work At the present time there is no means of determining beforehand which patients are likely to do sell, and the results quoted may not be considered good enough to justify the high operatuse risk Desophago-gastric resection Very few reports of the results of this operation are available, so it is difficult to draw any conclusion the immediate mortality has been rather high, and no five-year suryn al figures have been published in cases reported by Macpherson, Owen and Innes (1956) there was no recurrence of bleeding, but the period they were followed was in no case longer than two years It has yet to be shown that the risks of the operation are justified by the long-term results and the disadvantages of the loss of the cardiac sphincter Asterialisation of the portal vein Attempts have been made to improve the blood supply to the liver by making a portacaval anastomosis and joining an artery, either the right renal or a graft from the aorta to the hepatic end of the portal vein has been tried experimentally and in a few patients but the results hase been uniformly disastrous and the operation has no place in the surgical treatment of portal hypertension This

·106
RESULTS OF TREATMENT
(a)
Fig 66 Trans splenic portal venograms (a) before
and (b) after a portacaval anastomosis the time inter
val betrieen the two plates 13 in each case 3 seconds

·RESULTS OF TREATMEYT
107
After the anastomosis the splenic vein clears more
quickly and no opaque medium enters the collaterals

·to8
RESULTS OF TREATMENT
(0)
(d)
ma
Frg 67 Shots from a cine angiogram taken with an image intensifier
Trans splenic venography in a case of intra hepatic obstruction ten

·RESULTS OF TREATMENT
109
(g)
(6)
days after an end to side portacaval anastomos $ (Same patient as
Fig 39 ) In all shots from (d) onwards the vena cava can be seen
but a little opaque medium still enters the tortuous left gastric sein

·IIO RESULTS OF TREATMENT Injection of the varices Most surgeons who have given this method of treatment a trial haie abandoned it owing to the poor results and very fex long-term successes have been reported in the literature To sum up, we may conclude that the best treatment has been shown clearly by the results to be a wide portal systemic anastamosis using either the portal or the splenic veins These operations, unless carried out with meticulous care, are likely to fail owing to thrombosis in the neighbourhood of the anastomosis, and they should only be performed by surgeons with experience of the technique of vascular sutures The risk of occlusion is much greater than in the anastomosis of arteries This line of treatment however, does carry the risk of neuropathy which is occasionally quite disabling or may lead to lesser personality changes, and this risk has to be compared with the prospect of reasonable freedom from further haemorrhage If neither the portal nor the splenic sein is available because they are occluded, direct operations on the varices in the oesophagus or stomach or a gastric transection are worthu hile, but there is at present no one procedure which can be said to gne results definitely better than the others A careful study and assessment of the long term results of these operations is needed during the coming years before the question can be solved In the meantime there is much scope for further study of the pathology of the liver, for in so many patients the prognosis depends on the functional capacity of this organ It is possible that steps to prevent umbilical sepsis may reduce to a limited extent the thrombosis of the portal vein which occurs in infancy and which leads to such tragic consequences as the children grow up REFERENCES 2.- ALTA RC Innect "The BLAX Surg Gynec Ubstet, 94, 4+-+5+ LINTON, R. R and Ellis, DS (1956) 'Emergency and Definitive Trealment of Bleeding oesophageal Varices" Amer med As, 160, 1017- 1023 Tror (16) 'Surgical Treatment atment of Portal Hypertension in Liver Cirrhosis by LALIOK VA! Hepatic Anter" Arck klan Cher, 282, 1003-1013 TANNER NC (1958) Operatase Management of Haematemesis and Melaena Ann R Coll Surg. 22, 30-42 ·INDEX Diagnosis 49 Eck operat on 66 Enteritis Staphylococcal post ope tive 90 Extra hepatic obstruction 1: Alcoholic cirrhosis obstruction due to 19 Anaemia 44 47 Anaesthesia far portacar al anastomosis 67 Anal canal collateral circulation in 39 Anastomosis portacaval 66-76 indications 61 86 neurological sequelae 88-91 operative results roo operative technique 67-70 post-operative complications 76 selection of patients for 61 splenorenal 77 81 Anatomy 1 4 Ascytes 48 mechanism of formation 30 treatment 60 Falcsform gament collateral cifcu son in 3 39 venous circulation 3 Gastric transection in treatment Varices 82 operative results 104 ten leſt distribution/ B le ducts concen tal absence intra hepatic obstruction due to 21 Bulharzia obstruction due to 20 Bliary crecho es obstruction due to 21 Blood clot method of removal 73 count, 50 pressure hepatic vein 8 intra splenic 7 portalienous measurement of 7 9 Budd Chucı disease 26 27 Haemateme$18 49 age of first occurrence 93 OS emergency treatment 63 Hacmochromatosis obstruction due 20 Haemorrhage 40 age of onset 93 98 emergency treatment 63 Haemorrhoids 30 Heparıc artery hgation 63 operatiye results 105 ven obstruction 26 pressure 8 Hepatitis virus obstructron due to Hepato lenncular degeneration struction due to 10 Hypersplenism 4+ 47 Hypertension portal Ste Po hypertension splenic 12 Caput Vledusse 39 Children cirrhosis in 95 hacmorrhage in age of onset 93 98 portal hypertension in 93-9 prognosis 99 porta) thrombosis in 93 94 Currhosts alcoholic obstruction due to 19 bihary obstruction due to 21 in children causes 95 Collateral circulation portal hypes tension 11 32 Colon transplantation 95 Congen tal extra hepatic abstructors 12 fibrosis of laser 95 Cruellhier Baumgarten syndromac 39 10-43 Intra hepatic obstruction 18 Intra peritoneal adhesions promoti for treatment of varices Intra splenuc pressure 7 Laennec s cinthosis 61 Leucopenia 44 Liser bloud low So blood supply 3 capsule pathological changes 26 cirrhosis obstruction due to 19 ·12 INDEX Liter-continued Portal hypertension-continued clinical picture in portal hypertension, treatment, 60-87 47 non-operative, 60 fibrosis, congenital, 95 operative, 61-5, 66-81 obstruction due to, 21 palliative operations, 62 pattern of, 24 results, 100-10 function tests, 50 phlebitis, 16 enous supply, 1 Portal systemic neuropathy, 88 aetiology, 88 Management of portal hypertension, symptoms, 89 49-59 treatment, 91 Niediastinal packing in treatment of thrombosis, 17 varices, 85 vein, arterialisation, operative results, Mesenteric vein, inferior, distribution, 1 IOS blood pressure, measurement, 7 Neuropathy, portal systemic, 88 normal range, 9 aetiology, 88 dissection in portacaval anastosymptoms, 89 mosis, 69 treatment, 91 obstruction, congenital, 12 Nutritional deficiency, hepatic effects, extra-hepatic, 12 19 intra-hepatic, 18 traumatic, 13 Obstruction, extra-hepatic, 12 thrombosis in children, 93, 94 intra-hepatic, 18 venograms, pre- and post operative Oesophageal transection in treatment of appearances, 106-9 varices, 82 venous system, anatomy, 1-4 operative results, 104 physiology, 4-9 Oesophago-gastrectomy in treatment of Portal-azygos disconnection of varices, varices, 86 82 operative results, 105 anastomosis, operatre results, TOO Oesophagus, collateral circulation, 2, 32 veins, anastomosis, 1 varices, 32, 33, 36, 37 Prognosis, post-operative, 100 clinical aspects, 46 venous anastomoses, 2 Radiological eramination, 49 Omentopexy in treatment of varices, 86 Retroperitoneal tissues, collateral circulation, 33 Pathology of portal hypertension, 11-31 Peritoneoscopy, 50 Phlebitis, portal, causing extra hepatic Schistosomiasis, obstruction due to, 20 obstruction, 16 Spleen, blood pressure, measurement, 7 Physiology, 4-9 enlargement, 43, 47 Portacaval anastomosis, 66-76 in portal hypertension, 43, 47 indications, 61, 66 renography, 50-59 Splenorenal anastomosis, 77-81 neurological sequelae, 88-91 Stomach, collateral circulation in, 2, 32 operative results, 100 varices, 32, 34, 35 operative technique, 67-76 clinical aspects, 46 post-operative complications, 76 venous anastomoses, 2 selection of patients for, 61 Portal hypertension, collateral circulation, II, 32 Thrombocytopenia, 44 effects, 32-45 Thrombosis, portal, 17 in children, 93, 94 experimental, 6 in children, 93-9 Transection operations, 82 prognosis, 99 results, 104 management, 49-59 Treatment, 60-87 neurological symptoms, 88-91 non-operative, 60 pathology, 11-31 recurrence, post-operative, 101 palliative operations, 62 operatose, 61-5, 66-81

·Varices, demonstration, 49
gastrze, 32, 34. 35
oesophageal, 32, 33, 36, 37
injection, 82
results, ito
ligature, 85
treatment, operative, 82-6
results, 104, 110
Veins, inferior mesenteric, I
left gastric,
on abdominal wall, 47
INDEX
113
Veinscontinued
portal system, anatomy, 1-4
portal-systemic anastomosis, I
Vena casa, exposure in portacaval
anastomosis, 71
Venograms, pre- and post-operative
appearances, 106-9
Venography, 50-59
Venous system, portal, anatomy, 1-4
physiology, 4-9
Virus hepatitis, obstruction due to, 19

·No text on this page.

·No text on this page.