Portal Hypertension - State of the Art 1.partK.- J. Paquet
Department of Surgery and Vascular Surgery
Heinz - KALK - hospital
D - 97688 Bad Kissingen, Germany
Acute bleeding from varices in the esophagus is a catastrophic complication of portal hypertension and has supposed a presistant and often frustrating challange to gastroenterologists, endoscopists, surgeons and general practitioners alike for more than half a century. Today, the origin and pathophysiology of this often lethal complication of portal hypertension, the main cause of which is alcoholic cirrhosis in Europe, United States and civilized countries and schistosomiasis in developing countries, is wel understood, yet treatment remains a matter of continued debate.
Pathophysiology of portal hypertension: increased portal vascular resistance and hemodynamic circulationPortal hypertension is characterised by alteration in the splanchnic and systematic circulation associated with the development of portosystemic collateral channels, the most important of which are found in lower esophagus and stomach. Bleeding from these gastro - esophageal varices represents the major clinical complication and over the last decade there has been considarable interest in the pharmacological management of this condition.
The factors underlying the development and maintenance of portal hypertension and the pathogenesis of variceal rupture as yet not fully understood. Whilst an increase in portal vascular resistance, as a consequence of liver disease, appears to be the primary event in the majority of cases, increasing attention has focused on the potential importance of enhanced circulation levels of vasoactive compounds coupled with a proposed reduction in vascular sensivity to endogenous vasoconstrictors. Consequently, portal hypertension is now being more widely considered as a multi - organ disorder associated with changes in blood flow within both systemic and splachnic vascular beds.
The presence of a hyperdynamic state (high cardiac output/low vascular resistance) in association with septic shock and multiple - organ failure in the critically ill patient is well recognized (1,2). Observation in animal models and portal - hypertensive patients indicate that a similar hyperdynamic circulation develops as a consequence of significant portosystemic shunting, irrespective if the presence or absence of the underlying cirrhosis (3-8). The systemic circulation is characterized by peripheral vasodilatation resulting in a reduction in peripheral vascular resistance and enhanced cardiac output while portal venous inflow increases as a consequence of splanchnic hyperaemia. Consequently, the increase in portal venous flow may contribute to the maintenance of portal hypertension.
Indeed, a similar consensus is emerging as regards understanding the sequence of events in the pathophysiology of ascites formation in cirrhosis. According to this hypothesis, peripheral arterial vasodilatation, which is a general component of the hyperdynamic circulation, is implicated as the initiating pathogenic event which predisposes to sodium and water retention (9). The current concepts advanced to explain the hyperdynamic state implicate the combination of enhanced circulating levels of endogenous vasodilators and a reduced vascular sensitivity to vasoconstrictors.
Variceal bleeding: pathogenesis and risk factorsThe gastro - esophageal junction represents the clinical most significant area of anastomosis between the portal and systemic circulations. A number of studies have characterized the venous anatomy and highlighted the potential importance of the pallisade and perforating zone in the development and rupture of varices (10-12). A portal pressure of at least 12mm Hg is reguired for esophageal varices to develop (13). Progressive variceal dilatation subsequently depends on portal blood flow and local anatomical factors. The factors which predispose to and precipitate variceal hemorrhage have still not been completely and clearly identified.
The erosive theory whereby it was proposed, that variceal bleeding may be precipitated by peptic erosion of a varix (14) has been largely discounted by more recent studies (15). Important are levels of portal pressure, reflecting in most circumstances intravariceal pressure (16-18), variceal size (19-23), variceal wall structure and tension (22) and liver function (23).
In summary, the pathogenesis of variceal rupture is still not understood. The accumulating evidence indicates that the varix most likely to rupture is large with socalled telangiectasies, mini - varices of cherry red spots (20,24,25) on its top and in a Child-Pugh C cirrhotic patient with a high portal pressure. More detailed morphological investigation of the angioarchitecture of bleeding and non-bleeding varices in conjugation with rupture, provocation tests in animal models may help elucidate the pathogenesis of this important and potentially life threatening clinical problem.
Treatment with vasoactive drugs and other pharmacological agentsThe inability of any kind of treatment to occlude savely and permanently and acutely spurting varix or provide long-term benefit without causing to the patients treatment-related complications has been the main drawback of therapeutic efforts. The hypothesis of variceal rupture discussed before is true, the aim of pharmacological treatment should be to reduce variceal wall tension, portal pressure, splachnical circulation or to prevent any abrupt increase in these parameters. Some vasoconstrictor drugs are currently used in order to achieve these goals and in the attempt to stop the acute bleeding. All these agents decrese either portal pressure and azygos blood flow. Vasopressin although effective, has significant cardiac and gastrointestinal adverse effects that discourage its use. Combination with nitroglycerine reduces it averse effects while maintaining or even enhancing the reduction in portal pressure. Glypressin which acts as a slow-release preparation of vasopressin has a longer duration of action and can be administered as single intravenous injections instead of these drugs on systematic circulation leads to an overlapping spectrum of untowards effects. Somatostatin and the synthetic octapeptide octreotide displays similar pharmacological effects on splachnic hemodynamics but have a better tolerability profile. Thanks to its longer duration of action and ease of administration, octreotide could become the drug of choice for the early, prehospital management of bleeding varices (26). A different approach to the pharmacological treatment of variceal bleeding may be the use of compounds, like metoclopramide and domperidone, that increases the lower esophageal sphincter pressure (LESP), thereby reducing the inflow of blood into the submucous venous plexus of the esophagus and hense into the esophageal varices. However more studies are needed before these compounds be considered a real alternative to the above established drugs (27).
Emergency management of acute esophageal variceal hemorrhage with special reference to endoscopic sclerotherapyAll patients with suspected bleeding from esophageal varices require immediate hospitalisation if possible in the intensive care unit. If expert treatment is not available patients should be transferred to a center with appropriate treatment facilities as soon as the bleeding is controlled because subsequent management is difficult and may require more than one complex therapy. Although variceal bleeding may stop spontaneously in up to 60% of patients (28) it is not yet possible to predict which patients continue to bleed and will require further emergency therapy.
EndoscopyImmediate endoscopy is essential. At least 30% of patients with suspected variceal bleeding do not have varices. Patients shown to have varices those whose varices have stopped bleeding and those who varices but are bleeding from another lesion. Approximately one third of patients fall in each group (29). Once variceal bleeding is confirmed therapy should be initiated immediately.
The therapy options currently available for acute variceal hemorrhage are summarized in table 1. The treatment with vasoactive drugs has been covered before.
Table 1 - Therapeutic options for acute variceal bleeding
|Pharmacologic agents||Vasopressin / Glypressin
Vasopressin / Nitroglycerin comined
Somatostatin and Octreotide
Metoclopramine and Domperidone
|Ballon tamponade||Sengstaken-Blakemore tube
|Endoscopic scleroterapy(transhepatic obliteration of varices)|
Baloon tamponadeMost episodes of bleeding from esophageal and fundal varices can be controled initially by the inflation of a correctly inserted balloon. By our experience continous bleeding indicates either inadequate position of the tube or bleeding from lesions other than esophageal or subcardial varices (30,31). Thus there is enough time for resuscitation and management planning. Rebleeding after deflation of the balloon is frequent - 30 to 60% - and mortality is high in absence of additional therapy (32). Therefore preferably within 6 to 24 hours a definite procedure should be undertaken to achieve more lasting control. The Linton-Nachlas tube appears to be superior to the Sengstaken-Blakemore tube for gastric varices while both types exhibit similar efficacy in the control of the bleeding from esophageal sites.
Emergency endoscopic sclerotherapyEmergency endoscopic sclerotherapy can be performed immediately at the time of the first diagnostic endoscopy (32) as preferred and recommended by our group or it can be delayed until the variceal hemorrhage has been controlled by conservative measures with or without the use of pharmacological agents or balloon tamponade. The use of immediate sclerotherapy as practized by our group has gained popularity (33) and has even been advocated as the treatment of the choice. It requires a high degree of skill. We recommend its use whenever possible since it provides instant control of hemorrhage. In three controlled trials (32,34,35) (table 2) - the first was performed by our group - it could be shown that it is significantly superior to balloon tamponade in reducing the recurrences of hemorrhage and in two studies (32,34) improving survival.
Table 2 - Results of controlled randomised trials of emergency sclerotherapy of bleeding esophageal varices in comparison to balloon tamponade
|References||No. of pats.||Sclerotherapy i.v.,p.v.||Hemostasis (immediate) (%)-sc/c||Survival rate after 1 year - %|
|Paquet, Feussner 1985||21 (22)||Polidocanol 0,5 + 1% p.v.||90 (55)||79 (38)|
|Larson et al. 1986||21 (22)||Tetradecylsulfat 3% i.v.||85 (47)||62 (54)|
|Moreto et al. 1988||23 (20)||Tetradecylsulfat 3% i.v.||83 (37)||47 (41)|
c - control
Emergency endoscopic sclerotherapy should be only performed by experts or very skillful endoscopists. If these conditions are not fulfilled we recommend to manage the acute variceal hemorrhage by pharmacological therapy and/or balloon tamponade and to transfer the patient to a special center. Thus, delaying sclerotherapy until the initial hemorrhage has been brought under control remains the most widely practice option. After sclerotherapy has controlled the hemorrhage we support the view that at least two to four additional sessions are necessary to obliterate the varices by intravariceal or combined injection or to protect them by scar tissue by paravariceal injection.
Emergency injection sclerotherapy is not always successful. Failure can be defined as an acute episode of variceal bleeding that occurs within hours or days of emergency injection sclerotherapy and requires blood transfusion and subsequent emergency management. If immediate therapy failed during the first three days, bleeding should be first controlled by balloon tamponade before the patients again subjected to sclerotherapy. Our group prefers the use of Linton-Nachlas tube. If recurrence of variceal hemorrhage occurs three days or later after the first emergency sclerotherapy this option can be used again. By these treatment modalities the success rate of hemostasis is more than 90% in the approximately 10 to 20% of patients who have further episodes of acute variceal bleeding after two sclerotherapy treatments during a single hospital admission if the patients who are in Child-Pugh A class are excluded (33,36). These patients constitue the group that does not respond to emergency sclerotherapy and our group and other recommend that bleeding in such patients be temporary controlled with balloon tamponade should be followed by one of the major surgical procedures (33,37) or TIPS (38, 39). Unfortunately it is not yet possible to predict during the initial evaluation which patients will not respond to sclerotherapy.
Long-term endoscopic sclerotherapyAlthough repeated injecton sclerotherapy is the most widely practised long-term treatment after variceal bleeding it is not yet proved to be the single most effective form of management. Five major controlled trials (Table 3) have evaluated long-term endoscopic sclerotherapy for the pravention of reebleeding (40 - 44).
Table 3 - The effect of elective endoscopic sclerotherapy on the incidence rebleeding and survival in randomized controlled studies of patients with variceal bleeding and cirrhosis of the liver
|References||No. of pats. (sc/c)||Bleeding rate after 1 year(%)(sc/c)||Survival rate after 2 year(%)(sc/c)c|
|Westaby, McDougall et al. (1983)||50 (60)||49 (79) significantly different||78 (43) significantly different|
|Terblanche et al. (1983)||38 (37)||67 (82) significantly different||45 (45) significantly nodifferent|
|Copenhagen-ES-Trial (1984)||93 (94)||31 (60) significantly diff. from 40 days||78 (65) significantly diff. after 40 days|
|Söderlund (1985)||54 (53)||46 (66) significantly different||82 (58) significantly different|
|Korula et al. (1985)||56 (60)||47 (71) significantly different||51 (35) significantly diff. only if urgent shunts are exluded|
The percentage were calculated by adding the number of individuals who had rebled or survived and dividing this by the total number of patients in the studies.
In all these studies the frequency of rebleeding was reduced in sclerotherapy patients although in three trials (42-44) the difference to the control group was only significant when the total number of bleeding episodes was considered rather than the number of patients who bled. However up to 50% of patients rebled on chronic sclerotherapy although most of the rebleeding occurs before complete eradication or protection by scar tissue of the varices has been achieved and is of minor severity in all these studies. The benefitial effect of long-term sclerotherapy on survival as convincingly demostrated in the trial from the King°s College Hospital, London (40) and confirmed by our group (45,46) is nevertheless debated.
Endoscopic variceal ligationAlthough endoscopic sclerotherapy remains the standard treatment of acute variceal bleeding considerable interest in endoscopic variceal ligation (EVL) was stimulated by the publication of the first randomised controlled trial comparing EVL with sclerotherapy (47). EVL was as effective as sclerotherapy for the control of acute variceal bleeding (86 vs.77%, not significant). Esophageal varices were eradicated by the randomly assigned technique and patients were observed for a mean of about one year. Bleeding reccured in 48% of the sclerotherapy patients as compared with 36% of the EVL-treated patients.
More important, patients treated by sclerotherapy had significantly higher mortality rate (22 vs.2%, p less than 0,001), with complication of a more serious nature.
However, these high figures of rebleeding and complications in sclerotherapy patients do not correspond with other figures in literature being significantly lower (29,32-35,40-44). Additionally, studies in dogs with large gastric varices showed that gastric ulceration was more common and the ulcers larger and deaper with variceal ligation than with sclerotherapy or cyanoacrylate, suggesting that EVL may not be appropriate for gastric varices (48).
Transjugular intrahepatic portosystemic shuntsInterest in the transjugular intrahepatic portosystemic shunt (TIPS) procedures (38,39) continues to mount with the publication by Ring et al.(49) of their results in 13 patients referred for liver transplantation. These patients had either active (n = 6) or recurrent (n = 7) variceal bleeding despite sclerotherapy. In 13, bleeding was controlled by the TIPS procedure, with 1 patient having recurrent bleeding 102 days later because of shunt occlusion. Although the autors report no complication related to the procedure and no encephalopathy and eight patients who did not receive a liver transplant subsequent presentation from several centers report an incidence of post-TIPS encephalopathy (new or worsening) of 15 to 25% (50-52). It tends to be more common in older patients (51,52) and is usually responsive to medical therapy. Randomised controlled trials are needed before TIPS procedure can be accepted as standard therapy.
Prevention of recurrent variceal hemorrhage1 - Role of endoscopic sclerotherapy and beta-blockers Both non-selective beta-blockers and sclerotherapy are effective in decreasing the risk of recurrent variceal hemorrhage but their success in enhancing survival has been less dramatic. In a trial of patients with decompesated liver cirrhosis, Dasarathy et al.(53) found sclerotherapy superior to propranolol in preventing recurrent variceal bleeding and improving survival. Of the six published studies comparing beta-blocker and sclerotherapy for prevention of recurrent hemorrhage related to portal hypertension, on favoured sclerotherapy (53), one favoured propranolol (54) and four reports had not statistical difference (55-58). Only the study by Dasarathy et al.(53) reported a difference in survival rate. Although a metaanalysis showed a trend in favour of sclerotherapy (Fir.4), the heterogenity of the trial suggests that if one therapy is superior, the differences are clinically insignificant.
Vinel et al.(59) reported that the combination of propranolol and sclerotherapy is superior to sclerotherapy alone in preventing recurrent hemorrhage from all sites (18 vs. 39%, p Ă 0,005) but that mortality rate were similar in both groups.
In contrast Ink et al.(60) showed a non-significant benefitial trend in favour of sclerotherapy combined with propranolol when compared with propranolol alone for prevention of rebleeding from all sites. However, the combination was significantly superior in preventing rebleeding from esophageal varices. No survival differences were noted. It appears that when propranolol is continued after varices have been obliterated or protected by scar tissue by sclerotherapy, a therapeutic benefit from combination therapy may be achieved.
2 - Role of surgery Because a main cause of death in patients who bleed from varices is liver failure, it would seem even more logical to treat both, portal hypertension and hepatocellular dysfunction by liver transplantation instead of pharmacological, endoscopic or other surgical treatment. In practice however, a given form of therapy may not be suitable for all patients. In other words treatment should be tailored to individual patients.
The results of seven controlled trials (61-67) comparing endoscopic sclerotherapy (ES) with operative shunting are summarized in table 4.
Table 4 - Results of randomized controlled trials comparing long-term sclerotherapy with shunt surgery in the prevention of recurrent variceal hemorrhage
|Investigator||Reference||Therapy||Number of pats.||Rebleeding %||Salvage operat. %||Survival %||Follow-up months|
|Warren et al.||31||EST, DSRS||36 35||53 x 3||31||75 x 51||26 26|
|Cello et al.||32||EST, PCVS||28 24||50 x 21||11||32 17||17 17|
|Teres et al.||33||EST, DSRS||48 42||38 x 14||6||68 71||27 27|
|Rikkers et al.||34||EST, S||30 27||57 x 19||7||61 56||25 25|
|Henderson et al.||35||EST, DSRS||37 35||59 x 3||35||65 x 43||61 (median)|
|Spina et al.||36||EST, DSRS||20 20||35 x 5||-||95 100||29 24|
|Planas et al.||37||EST, PCVS||35 34||40 x 2,9||-||79 83||19 21|
EST - endoscopic sclerotherapy
DSRS - distal splerorenal shunt
PCVS - porto-caval shunt
S = DSRS 23 patients, nonselective shunt 4
p less than 0,05
In all trials shunt operation were superior to endoscopic sclerotherapy in preventing rebleeding. Three studies (62-64,66,67) analysed the costs of both forms of therapy but did not find a difference (62,64,67). Likewise in five of seven studies there was no significant difference in survival. In the remaining two studies (61,65), both from the Emory group in Atlanta,U.S.A., the endoscopic sclerotherapy group had an improved survival. However, in both of these studies, approximately one third of the ES patient had to undergo salvage shunting for uncontrollable rebleeding. In fact surgery appears to play the most important role un reducing mortality. The author concluded the ES with surgical back-up provides the best results with regards to survival.
Similar results of the treatment of long-term sclerotherapy failures by elective shunt operations have been published by our group (36). From January 1,1983 to January 1,1993 692 consecutive patients were admitted with bleeding esophageal varices. Fourteen patients had to be excluded from the prospective evaluation because of uncontrollable hemorrhage during the first 12 to 24 hours or refusal of treatment. 311 Child-Pugh C patients were excluded from the control trial. In the remaining 367 patients, 182 of them were Child-Pugh A and 185 B, endoscopic sclerotherapy was successful in 194. In 173 patients, with at least two rebleeding despite long-term sclerotherapy, specific selection criteria were used to assess suitability for elective narrow-lumen mesocaval interposition (NLMS) or distal splenorenal shunt (DSRS):
1 - Liver volume, sonographically determined as 1000 - 2500ml 2 - Portal perfusion index over 30% at sequential scintigraphia
3 - No activity or progression of cirrhosis seen at laparoscopy and biopsy
4 - No stenosis of the hepatic arterial circulation, suitable lumen and lengths of splenic vein, angiographic studies.
Modalities of therapy in 367 patients with Child-Pugh A and B classification and types of shunt procedures: 69 patients refused shunt operation and 26 did not fulfil selection criteria. In this group ES was continued. 88 patients were shunted: 54 narrow-lumen mesocaval shunts (NLMS), 32 distal splenorenal shunts (DSRS), 1 portocaval and 1 splenorenal Linton shunt. The continued sclerotherapy and shunt group were comparable concerning number, demographic characteristics, etiology, severity and histology of liver disease (table 5).
Table 5 - Demographic characteristic, etiology, severity and histology of liver disease, early and long-term results in 173 patients, either selected for endoscopic sclerotherapy or shunt operation
41 pats. (48%)
40 pats. (46%)
other types of cirr.
mortality at 30 days
late mortality (up to Jan.1,1993) p = 0,01
number of patients
There was no significant difference of mortality at 30 days (5 vs.7%). 29 from the surviving 31 patients receiving ES (36% died during the late follow-up. Mean follow-up time in both groups was 43 months. Four patients of the ES and three of the shunt group were lost to follow-up after 18 to 39 months. Seventeen shunt patients died during the late follow-up (18%). The cumulative survival curve calculated using the method of Kaplan-Meier demostrates significantly better survival (p Ă 0,01) in favour of the patients selected for shunt after sclerotherapy had failed.
3 - Transsection with or without devascularisation Many surgical procedures intend to a "ablate, decongest or disconnect" the dilated venous collaterals of the esophagus and stomach have been conceived. All of these approaches have the advantage of preserving portal blood flow to the liver. In Europe and the U.S.A. older ablative procedures are no longer used because of prohibitively high morbidity and mortality. However in varices secondary to schistosomiasis these procedures provide acceptable results (68,69). New techniques and procedures have led to a renewed pursuit of this approach in the United Kingdom (70,71), Japan (72,73) and Germany by our group (74-76).
Esophageal transsection and reanastomosis have been performed using a staple gun (70,77) or sutures. In general an ablative operation alone without a procedure to completely disconnect the mucosal, submucosal and extrinsic venous systems of the esophagus does not control recurrent bleeding (78,79). The possible exception to this in variceal bleeding secondary to hepatic schistosomiasis where results of ablative surgery are very good (68,69). The combination of devascularisation and transsection should hypothetically further reduce the changes of recurrent bleeding. The Sugiura operation and its modification has been used in Japan but mostly for elective operation (72,73). Sugiura°s group reports excellent results (6% incidence of rebleeding with up to 15 years of follow-up) although 185 of the 636 patients reported in his series had not bled and were operated on prophylactically. In contrast when used in patients with alcoholic cirrhosis the results have not been nearly as favourable (80,81) as confirmed by our group (74,75).
(Pokračování v čísle 94-4.)
Současný stav antimikrobní profylaxe v biliární chirurgiiVyhnánek,F., *Lochmann,O.
Chirurgická klinika 3.LF UK v Praze
* Ústav lékařské mikrobiologie 2.LF UK v Praze
SouhrnV přehledu je zhodnocen současný stav antimikrobní profylaxe v biliární chirurgii. Pro volbu antimikrobní prevence je nezbytné posoudit některé limitující faktory jako je výskyt bakterií ve žlučníku a žlučových cestách, jejich aktuální citlivost, rizikové faktory vzniku infekce u operovaného nemocného. Autoři podle zkušeností doporučují jako antibiotikum první volby ke krátkodobému profylaktickému podání cefalosporiny II. generace (cefamandol, cefoxitin, cefuroxim).
antimikrobní profylaxe - biliární chirurgie - cefalosporiny
SummaryCurrent status in prophylactic antibiotics in biliary surgery is analyzed. It is necessary to consider some limiting factors such the incidence of microbes in the gallbladder and bile ducts, their current sensitivity, risk factors of infection at the patient operated on for the election of antibacterial prophylaxis. Authors recommended by their experience for short antibacterial prophylaxis cephalosporins the II.generation (Cefamandol, Cefoxitin, Cefuroxim).
Key words: antibacterial prophylaxis - biliary surgery - cephalosporins