1 Portal Hypertension and Gastrointestinal Bleeding Jaime Bosch, M.D., Ph.D., 1,2 Juan G. Abraldes, M.D., 1,2 Annalisa Berzigotti, M.D., 3,4 and Juan Carlos Garcia-Pagan, M.D. 1,2 ABSTRACT Variceal bleeding is one of the most serious complications of portal hypertension. The driving force for the development of varices is an increase in portal pressure. As portal hypertension progresses, varices dilate until they finally rupture and bleed. This sequence of events might be prevented by achieving a sufficient decrease in portal pressure or by acting locally at the varices with endoscopic treatments. This article reviews the rationale for the management of portal hypertension and the current recommendations for the prevention and treatment of variceal bleeding. KEYWORDS: Cirrhosis, varices, endoscopic band ligation, TIPS, b-adrenergic blockers, hepatic venous pressure gradient Portal hypertension is a common clinical syndrome defined by a pathologic increase of portal venous pressure. As a consequence, the gradient between portal pressure (PP) and inferior vena cava pressure (IVC) (portal pressure gradient, PPG) is increased above the upper normal value of 5 mm Hg. Portal hypertension can be due to many different causes, at prehepatic, intrahepatic, and posthepatic sites (Table 1). Cirrhosis of the liver accounts for 90% of cases of portal hypertension in Western countries. The importance of portal hypertensive syndrome is defined by the frequency and severity of its complications, including upper gastrointestinal bleeding from ruptured gastroesophageal varices, ascites, and hepatorenal syndrome, which represent the leading causes of death and of liver transplantation in patients with cirrhosis. This article focuses on esophageal varices and gastrointestinal bleeding in cirrhosis. Portal hypertension is considered to be clinically significant (CSPH) when PPG, or its clinical equivalent hepatic venous pressure gradient (HVPG), exceeds 10 to 12 mm Hg, since this is the threshold for the clinical manifestations of portal hypertensive syndrome to appear. The vast majority of patients with cirrhosis develop CSPH along the course of the disease, and data from a recent multicentric study indicate that CSPH is already present at diagnosis in 60% of histologically proven, well-compensated cirrhosis cases. 1 NATURAL HISTORY AND PROGNOSIS Variceal bleeding is the last step of a chain of events initiated by an increase in PP, followed by the development and progressive dilation of varices until these finally rupture and bleed. According to recent data, 2 the appearance of varices in compensated patients indicates a change from a clinical stage with a very low risk of death at 1 year (stage 1; 1% risk) to an intermediaterisk stage (stage 2; 3.4% risk). The occurrence of variceal bleeding is a dreadful event, marking the progression to 1 Hepatic Hemodynamic Laboratory, Liver Unit, Institut d Investigacions Biomediques August Pi i Sunyer (IDIBAPS), University of Barcelona, Spain; 2 Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBEREHD), University of Barcelona, Spain; 3 Centre Diagnostic per la Imatge, Hospital Clinic, Barcelona, Spain; 4 Department of Internal Medicine, Cardioangiology, Hepatology, Policlinico S. Orsola, University of Bologna, Italy. Address for correspondence and reprint requests: Professor Jaime Bosch, M.D., Ph.D., Hepatic Hemodynamic Laboratory, Liver Unit, Hospital Clínic, Villarroel 170, Barcelona, Spain ( Complications of Cirrhosis; Guest Editor, Pere Ginès, M.D. Semin Liver Dis 2008;28:3 25. Copyright # 2008 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) DOI /s ISSN
2 4 SEMINARS IN LIVER DISEASE/VOLUME 28, NUMBER Table 1 Etiology of Portal Hypertension According to the Site of Increased Resistance to Portal Blood Flow Prehepatic Splenic vein thrombosis Portal vein thrombosis Congenital stenosis of the portal vein Extrinsic compression of the portal vein Arteriovenous fistulae Intrahepatic Cirrhosis (viral, alcoholic, biliary, metabolic) Partial nodular transformation Nodular regenerative hyperplasia Congenital hepatic fibrosis Peliosis hepatic Polycystic disease Idiopathic portal hypertension Hypervitaminosis A Arsenic, copper sulfate, vinyl chloride monomer poisoning Granulomatous diseases (sarcoidosis, tuberculosis, primary biliary cirrhosis, schistosomiasis) Amyloidosis Mastocytosis Rendu-Osler-Weber syndrome Liver infiltration in hematologic diseases Acute fatty liver of pregnancy Severe acute viral and alcoholic hepatitis Chronic active hepatitis Hepatocellular carcinoma Cyanamide toxicity Veno-occlusive disease Posthepatic Hepatic veins thrombosis (Budd-Chiari syndrome) Congenital malformation and thrombosis of the inferior vena cava Constrictive pericarditis Tricuspid valve diseases decompensation of the liver disease and to a very high risk of death stage (stage 4; 57% risk) (Fig. 1). Prevalence and Formation of Esophageal Varices When cirrhosis is first diagnosed, varices are present in 30 to 40% of compensated patients and in 60% of decompensated. 1,3,4 Clinical-hemodynamic correlations have shown that varices may appear when HVPG increases to over 10 mm Hg. 1,5 Since portal hypertension eventually develops in almost every patient with cirrhosis it is thought that, if cirrhotic patients are followed long enough, virtually all of them will develop varices. 6 In those cirrhotic patients without varices at first endoscopy, the annual incidence of new varices is a mean of 7%, ranging from 5 to 10% in published series. 6 8 An HVPG over 10 mm Hg is a strong predictor for the development of varices, 1 confirming that PP plays a pivotal role as the driving force for the development of collaterals. Progression of Esophageal Varices from Small to Large Once developed, varices increase in size from small to large before they eventually rupture and bleed. Studies assessing the progression from small to large varices are controversial, showing rates of progression of varices ranging from 5 to 30% per year The most likely reasons for such variability are differences in patient selection and follow-up endoscopy schedule. 10 The factor that has been most consistently associated with variceal progression is Child-Pugh class. 7 9,11 Other factors include alcoholic etiology of cirrhosis and presence of red wale markings in the esophageal varices. 8,10 Changes in HVPG (either spontaneous or caused by drug therapy or transjugular intrahepatic portal-systemic shunts [TIPS]) are usually accompanied by parallel variations in the size of the esophageal varices, which are significantly reduced when HVPG decreases below 12 mm Hg. 12,13 Thus, an increased HVPG plays a key role both in development and progression of the varices. Incidence and Risk Indicators of First Bleeding from Esophageal Varices First variceal bleeding has an incidence of 4% per year; this risk increases to 15% per year in patients with medium-large varices. 4 The most important predictive factors related to the risk of bleeding are variceal size, severity of liver dysfunction as assessed by the Child-Pugh classification, and presence of red signs. 14 Variceal size and red color signs are associated with increased bleeding risk probably because they reflect direct parameters determining variceal wall tension (radius, wall thickness), which is the decisive factor determining variceal rupture. In addition, many studies have shown that variceal bleeding only occurs if the HVPG reaches a threshold value of 12 mm Hg. 5,12,15 Conversely, if the HVPG is substantially reduced (below 12 mm Hg or by more than 20% of baseline levels), there is a marked reduction in the risk of bleeding. 12,16 This is of outstanding importance, since it demonstrates that the portal hypertension syndrome is reversible by pharmacological treatment effectively decreasing PP (Fig. 2). DIAGNOSIS HVPG Measurement Measurement of HVPG by hepatic vein catheterization is an objective and quantitative equivalent of PP in
3 PORTAL HYPERTENSION AND GI BLEEDING/BOSCH ET AL 5 Figure 1 Clinical stages of cirrhosis according to clinical features Four clinical stages can be identified, each with distinct clinical characteristics and a markedly different prognosis. Each stage is defined by the presence or absence of complications of cirrhosis. As shown, the onset of esophageal varices in compensated patients indicates a change from a clinical stage with a very low risk of death at 1 year to an intermediate risk stage, while the occurrence of variceal bleeding situates the patient in the highest risk stage. (Modified from D Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic indicators of survival in cirrhosis: a systematic review of 118 studies. J Hepatol 2006;44: ). 174 cirrhosis. 17 HVPG has proved to add prognostic information in many settings, including compensated cirrhosis, 18 acute variceal bleeding, 19 and patients awaiting liver transplantation. 20 However, as previously mentioned, not all patients with increased HVPG have varices. Patients with CSPH are at high risk of varices and should undergo endoscopic screening. Endoscopy The current consensus is that every cirrhotic patient should be endoscopically screened for varices at the time of diagnosis. 21 The aim of the screening for esophageal varices is to detect patients requiring prophylactic treatment. In patients without varices on initial endoscopy, further evaluations should be performed to detect the Figure 2 Natural history of esophageal varices as a function of variceal wall tension. The solid line shows the evolution of untreated patients: variceal wall tension increases markedly as a result of an increase of hepatic venous pressure gradient (HVPG) above 10 mm Hg, enlargement of the varix, and decrease of wall thickness. Once wall tension increases to values exceeding the elastic limit of the vessel, the patient may experience his or her first bleeding episode. Afterward, the patient remains at risk unless wall tension is decreased by means of medical therapy (decrease of HVPG to < 12 mm Hg or by > 20% versus baseline value) (discontinuous line). Primary prophylaxis (dotted line) prevents or delays wall tension from reaching the rupture point, so reducing the bleeding risk. (Modified from Escorsell A, Bosch J. Pathophysiology of variceal bleeding. In: Groszmann RJ, Bosch J, eds. Portal Hypertension in the 21st Century. Dordrecht/Boston/London: Kluwer Academic Publishers; 2004: )
4 6 SEMINARS IN LIVER DISEASE/VOLUME 28, NUMBER development of varices. The current consensus is that endoscopy should be repeated after 2 to 3 years in patients without varices at the first endoscopy. The expected incidence of moderate/large varices and/or variceal bleeding in these patients (and, thus, the risk of leaving patients without prophylaxis) is less than 10% at 3 years. 8,10 In those centers in which hepatic hemodynamic studies are available it is advisable to measure HVPG. An HVPG over 10 mm Hg indicates a more rapid progression to complications of cirrhosis, and calls for shorter surveillance intervals. 22 In patients with small varices on initial endoscopy, the aim of subsequent evaluations is to detect the progression of small to moderate/large varices because of its important prognostic and therapeutic implications. Based on an expected 10 to 15% per year rate of progression of variceal size, endoscopy should be repeated every 2 years in patients with small varices. In patients with advanced cirrhosis or red wale marks, a 1-year interval might be recommended. 8,10 Endoscopic Videocapsule Patients are frequently intolerant to repeated conventional endoscopies, and often require sedation. Recently, endoscopic videocapsule has been suggested, as this may improve patients tolerance. Once swallowed, the videocapsule records images at predetermined intervals. In the two published studies, capsule endoscopy allowed a correct identification of varices in 80% of cases 23,24 ;inoneof these studies, capsule endoscopy proved a correct identification of red wale marks. 23 However, it may not be as good in assessing variceal size and it may have poor accuracy in identifying the presence of hypertensive gastropathy and gastric varices. 25 Therefore, endoscopic videocapsule cannot be currently recommended as the routine screening method for the evaluation of gastroesophageal varices or portal hypertensive gastropathy in cirrhosis. Noninvasive Tests The ideal method to diagnose and follow up portal hypertension should be reproducible, inexpensive, and noninvasive (Fig. 3). No noninvasive procedure has proved, so far, to be accurate enough to avoid endoscopy in patients with negative indicators. 26,27 However, some clinical, laboratory, and imaging variables may help to select patients with high risk for varices. 28,29 CLINICAL SIGNS AND LABORATORY FINDINGS The presence of splenomegaly, or enlarged spleen size, is the clinical sign most often reported in studies on Figure 3 Existing diagnostic tools in the clinical assessment of portal hypertension. The ideal method to diagnose portal hypertension should be noninvasive, accurate, objective, and reproducible. HVPG, hepatic venous pressure gradient; CT, computed tomography; MRI, magnetic resonance imaging; CDUS, color-doppler ultrasonography.
5 PORTAL HYPERTENSION AND GI BLEEDING/BOSCH ET AL 7 noninvasive prediction of portal hypertension and esophageal varices. Platelet count, and platelet to spleen ratio, 29 have been suggested as useful to select patients with high probability of having large esophageal varices. However, the cut-offs vary a lot across studies (platelet count from 68,000 to 160,000/mm 3 ), indicating low specificity. In addition, in viral cirrhosis, platelet count can decrease independently from the development of portal hypertension. Other laboratory data independently correlated with the presence of varices or large varices are indicators of liver failure, such as hypoalbuminemia, low prothrombin time, and hyperbilirubinemia, 30 or their combination in the Child-Pugh score. 31 Some indicators of liver fibrosis, such as hyaluronic acid level 32 or the Fibrotest, have also been suggested to correlate with the presence of esophageal varices. IMAGING TECHNIQUES Ultrasonography (US)/color Doppler-US (CDUS) is the preferred initial examination in patients with suspected portal hypertension. US can detect signs of portal hypertension such as splenomegaly, presence of portocollateral vessels, and ascites, and complement this information with data on liver size, echotexture, and margins, which can suggest underlying cirrhosis. Among the US parameters, spleen length is the strongest independent predictive marker of esophageal varices. Portal vein diameter (above 13 mm), portal blood flow velocity (mean velocity below 12 cm/s), 33 lack or reduced respiratory variations of splenic and superior mesenteric vein diameter, 34 congestion index of portal vein, altered hepatic venous Doppler pattern, increased intraparenchymal hepatic and splenic artery impedance, 35 increased intraparenchymal renal artery impedance, 36 and reduced mesenteric artery impedance 37 have all been found to correlate with the presence of CSPH and esophageal varices. However, these variables do not show a satisfactory predictive accuracy in independent sets of patients, probably due to the high interobserver and intraobserver variability. Computed tomography (CT) scan and MRI (magnetic resonance imaging) allow a good visualization of the portal venous system, and both techniques permit the diagnosis of esophageal varices. 38,39 The measurement of azygos blood flow by angio-mr has shown a correlation with the presence of esophageal varices and the risk of variceal bleeding. However, the technical complexity and cost of these techniques makes their use unlikely for the screening of gastroesophageal varices in cirrhosis. LIVER STIFFNESS (FibroScan) Transient elastography (FibroScan 1 ) is a noninvasive method recently developed for assessing liver fibrosis. 40 Because fibrosis and scarring of the liver tissue are major factors contributing to the development of portal hypertension, liver stiffness measurement has theoretical potential for the evaluation of portal hypertension. Recent studies demonstrate that liver stiffness shows a good correlation with HVPG: in a study in liver transplant recipients with hepatitis C virus recurrence, a liver stiffness value 8.74 kpa had a sensitivity and specificity of 90% and 81% for the diagnosis of increased PP (HVPG 6 mm Hg). 41 In three different studies, two published as abstracts 42,43 and one as a full paper, 44 the authors found that liver stiffness can detect accurately an HVPG over 12 mm Hg in patients with cirrhosis or long-lasting chronic liver disease. However, different cut-off values were found, namely 17 kpa, 23 kpa, and 13.6 kpa. Interestingly, in the study by Vizzutti and colleagues, 44 it was shown that above the threshold value of 13.6 kpa there was not a good correlation between liver stiffness and HVPG, probably since structural abnormalities within the liver (e.g., fibrosis) are not the only mechanisms promoting portal hypertension, which is also determined by the magnitude of portocollateral blood flow and changes in hepatic vascular tone (which are not reflected by the Fibroscan). As for the prediction of esophageal varices, the study by Kazemi and associates suggested that liver stiffness might be used to select those patients likely to have large varices and requiring screening endoscopy. 45 In this study, however, endoscopic assessment was retrospective, and transient elastography performed no better than other simple noninvasive variables that are routinely obtained in these patients, such as blood test and ultrasonography. Additionally, transient elastography was not discriminative to detect small varices, since the best cut-off for detecting all types of varices overlapped with that for detecting cirrhosis. 46 Further studies are therefore needed before recommendation can be given on the routine use of Fibroscan for the detection of CSPH and varices. RATIONAL BASIS OF THERAPY In cirrhosis, portal hypertension is due to increased resistance to portal blood flow through the cirrhotic liver, and increased blood flow in the portal and collateral circulation. Portal hypertension can therefore be attenuated by decreasing intrahepatic resistance, reducing portocollateral blood flow, or both 47 (Fig. 4). These targets can be addressed by means of medical treatment, local treatments, or derivative procedures. The initial factor in the pathophysiology of portal hypertension is the increase in vascular resistance to portal blood flow. In cirrhosis this increase in resistance occurs at the hepatic microcirculation (sinusoidal portal hypertension). Increased hepatic vascular resistance in cirrhosis is not only a mechanical consequence of the hepatic architectural disorder caused by the liver disease, as there is also a dynamic component due to the active
6 8 SEMINARS IN LIVER DISEASE/VOLUME 28, NUMBER Figure 4 Rational basis of therapy. A mechanical increase in hepatic resistance due to structural alterations of liver parenchyma is the initial factor that leads to portal hypertension. This might be overcome by TIPS placement. The recognition of a functional and, thus, reversible component in the increased hepatic resistance sets the rationale for the use of vasodilators in portal hypertension. Furthermore, portal hypertension leads to a cascade of disturbances in the splanchnic and systemic circulation characterized by vasodilation, sodium and water retention, and plasma volume expansion that lead to an increase in portal blood inflow that contributes to maintain and worsen portal hypertension. This sets the rationale for the use of vasoconstrictors and diuretics in the treatment of portal hypertension. Varices are the result of the development of portosystemic collaterals. Endoscopic treatments are directed at locally eradicating the varices, but do not decrease portal pressure. TIPS, transjugular intrahepatic portal-systemic shunts; PSS, portal-systemic shunt; CO, carbon monoxide; NE, norepinephrine; VP, vasopressin; A-II, angiotensin-ii; Na, sodium. contraction of portal/septal myofibroblasts, activated hepatic stellate cells, and vascular smooth muscle cells in portal venules. 48 This increased hepatic vascular tone is due to an imbalance between endogenous vasodilatory and vasoconstrictor stimuli, the former being insufficient to counteract the influence of the latter. 49 Indeed, while vasoconstrictors are increased, intrahepatic nitric oxide (NO) production is clearly decreased. 49,50 This deficient intrahepatic NO production is due to endothelial dysfunction in the liver microvascultature 50,51 and may also favor local thrombosis and fibrogenesis. 49 This finding provides a rational basis for drug therapies aimed at increasing NO within the liver either by the use of NO donors or by enhancing e-nos activity with statins. Oxidative stress is in part responsible for reduced intrahepatic NO bioavailability, which provides a rational basis for testing antioxidants, such as ascorbic acid. 52 The increased resistance through the cirrhotic liver can also be targeted by mechanical means, bypassing the liver. This can be achieved by portal-systemic shunt surgery and by TIPS. These procedures are highly effective in decreasing PP, but have the detrimental effect that, by further decreasing portal blood flow through the liver and by increasing portal-systemic shunting, they may enhance liver failure and facilitate hepatic encephalopathy. The second factor contributing to portal hypertension is an increased blood flow through the portal venous system. This is due to splanchnic arteriolar vasodilation, probably due to excessive release of endogenous vasodilators (endothelial and neurohumoral) Splanchnic hyperemia contributes to the increase in PP and explains why portal hypertension persists, despite the establishment of an extensive network of portosystemic collaterals that may divert over 80% of the portal blood flow. The increased portal venous inflow can be corrected pharmacologically by means of splanchnic vasoconstrictors, such as vasopressin and its derivatives, somatostatin and its analogues, and nonselective b- adrenergic blockers. These are the drugs that have more widely been used in the treatment of portal hypertension. Splanchnic vasodilation is in part due to an increased release of NO in splanchnic arteries, which is amenable to pharmacological manipulation. However, this presents the difficulty of inhibiting NO synthesis selectively in the splanchnic arterial bed, which is not feasible at present. Splanchnic vasodilation is accompanied by increased cardiac index and hypervolemia,
7 PORTAL HYPERTENSION AND GI BLEEDING/BOSCH ET AL 9 representing the hyperkinetic circulatory syndrome associated with portal hypertension. An expanded blood volume is necessary to maintain the hyperdynamic circulation, which provides a rationale for the use of low-sodium diet and spironolactone to attenuate the hyperkinetic syndrome and the PP elevation in patients with cirrhosis. 56 Combined pharmacological therapy consists in associating vasoconstrictive drugs, which will decrease portal blood inflow, and vasodilators, which may reduce the intrahepatic vascular resistance. The best-known example is the association of propranolol or nadolol and isosorbide-5-mononitrate (ISMN). Therapeutic strategies aimed at decreasing PP have the advantage of potentially preventing other portal hypertension-related complications, such as portal hypertensive gastropathy, ascites, spontaneous bacterial peritonitis, and hepatorenal syndrome. Local treatments at the varices complete the spectrum of treatments for complications of portal hypertension. These treatments do not decrease PP and therefore have no potential for preventing other complications of portal hypertension. Endoscopic therapy is directed at eradicating the varices either by injecting a variety of irritating substances into or around the varices to promote thrombosis and fibrosis, or by ligating the varices using elastic bands. It is possible that the efficacy of endoscopic therapy can be enhanced if combined with an agent that effectively lowers PP. Finally, hemostatic agents, such as recombinant activated factor VII, are being explored as adjuvants to conventional therapy to arrest variceal bleeding in patients with poor liver function. 57,58 Clinical Scenarios The treatment of portal hypertension includes the prevention of variceal hemorrhage in patients who have never bled, the treatment of the acute bleeding episode, and the prevention of rebleeding in patients who have survived a bleeding episode from esophageal or gastric varices. The main difference among these scenarios is that natural history and prognosis are quite different from one to another. THE ACUTE BLEEDING EPISODE Natural History and Prognosis Ruptured esophageal varices cause 70% of all upper gastrointestinal bleeding episodes in patients with portal hypertension. 59 Thus, in any cirrhotic patient with acute upper gastrointestinal bleeding, a variceal origin should be suspected. Diagnosis is established at emergency endoscopy based on observing one of the following: (1) active bleeding from a varix (observation of blood spurting or oozing from the varix) (nearly 20% of patients); (2) white nipple or clot adherent to a varix; (3) presence of varices without other potential sources of bleeding. INITIAL CONTROL OF BLEEDING Because variceal bleeding is frequently intermittent, it is difficult to assess when the bleeding stops and when a new hematemesis or melena should be considered an episode of rebleeding. Several consensus conferences have addressed this issue and set definitions for events and timing of events related to episodes of variceal bleeding. 21 Using these definitions, data from placebo controlled clinical trials show that variceal bleeding is spontaneously controlled in 40 to 50% of patients. 60 Currently available treatments increase control of bleeding to 80% of the patients. 59 EARLY REBLEEDING The incidence of early rebleeding ranges between 30% and 40% in the first 6 weeks. The risk peaks in the first 5 days, with 40% of all rebleeding episodes occurring in this very early period, remains high during the first 2 weeks, and then declines slowly in the next 4 weeks. After 6 weeks the risk of further bleeding becomes virtually equal to that before bleeding. 61 Currently available treatments have reduced the 6-week rebleeding rate to 20%. 59 Early rebleeding is a strong predictor of death from variceal bleeding. Prognostic indicators for early rebleeding were assessed in most studies together with initial failure to control bleeding and 5-day risk for death, forming a composite end point referred to as 5-day failure. Bacterial infection, 62,63 active bleeding at emergency endoscopy, 59,63 Child-Pugh class or score, 59,63,64 AST levels, 59 presence of portal vein thrombosis 59 and an HVPG > 20 mm Hg measured shortly after admission 19,64,65 have been reported as significant predictors of risk for 5-day failure. MORTALITY Mortality from variceal bleeding has greatly decreased in the last two decades, from 42% mortality in the Graham and Smith study in to the current 15 to 20%. 59,66 68 This is probably due to implementation of effective treatments (endoscopic and pharmacological therapies and TIPS), as well as improved general medical care (i.e., antibiotic prophylaxis). Since it may be difficult to assess the true cause of death (i.e., bleeding versus liver failure or other adverse events), the general consensus is that any death occurring within 6 weeks from hospital admission for variceal bleeding should be considered as a bleeding-related death. 21 Immediate mortality from uncontrolled bleeding is in the range of 4 to 8%. 4,59 Prehospital mortality from variceal bleeding might be around 3%. 69 Like the risk for rebleeding, the risk for mortality peaks the first days after bleeding, slowly declines thereafter, and after 6 weeks becomes constant
8 10 SEMINARS IN LIVER DISEASE/VOLUME 28, NUMBER and virtually equal to that before bleeding. 60,61 At present, only 40% of the deaths are directly related to bleeding, while most are caused by liver failure, infections, and hepatorenal syndrome. 59 Thus, although there is still room for improving hemostatic treatments, to substantially decrease mortality from variceal bleeding, therapies should be able to prevent deterioration of liver and renal function. The most consistently reported death risk indicators are Child-Pugh classification or its components, blood urea nitrogen or serum creatinine, active bleeding on endoscopy, hypovolemic shock, and hepatocellular carcinoma. 59,61,62,70 Prognostic indicators gathered in the early follow-up include early rebleeding, bacterial infection, and renal failure. 70 From these data it is clear that management of bleeding cirrhotic patients should be aimed not only at controlling the bleeding, but also at preventing early rebleeding, infection, and renal failure. Treatment of Acute Variceal Bleeding Acute variceal bleeding should be managed in an intensive care setting by a team of experienced medical staff, including well-trained nurses, clinical hepatologists, endoscopists, interventional radiologists, and surgeons. Lack of an appropriate facility or staff demands immediate referral. Decision-making shall follow the guidelines set up in a written protocol developed to optimize the resources of each center. GENERAL MANAGEMENT The general management of the bleeding patient is aimed at correcting hypovolemic shock (with judicious volume replacement and transfusion) and at preventing complications associated with gastrointestinal bleeding (bacterial infections, hepatic decompensation, renal failure), which are independent of the cause of the hemorrhage and demand immediate management. Initial resuscitation should follow the classic Airway, Breathing, Circulation scheme, and is aimed at restoring an appropriate delivery of oxygen to the tissues (which depends on SaO 2, cardiac output, and hemoglobin concentration). Airway should be immediately secured, especially in encephalopathic patients, since the patient is at risk of bronchial aspiration of gastric content and blood. This risk is further exacerbated by endoscopic procedures. Endotracheal intubation is mandatory if there is any concern about the safety of the airway. Blood volume replacement should be initiated as soon as possible with plasma expanders, aiming at maintaining systolic blood pressure around 100 mm Hg. Avoiding prolonged hypotension is particularly important to prevent infection and renal failure, which are associated with increased risk of rebleeding and death. 70 Transfusion policy should be conservative, since it may induce rebound increases in PP and rebleeding. 71,72 The use of vasopressin analogues or somatostatin blunt the increase in PP induced by volume expansion. 73,74 Blood transfusion should aim at maintaining the hematocrit at 0.21 to 0.24 (Hb 7 to 8 g/l), 21 except in patients with rapid ongoing bleeding or with ischemic heart disease. The role of platelet transfusion or fresh frozen plasma administration has not been assessed appropriately. The use of recombinant activated factor VII (rviia, Novoseven 1 ), which corrects prothrombin time in cirrhotics, 75 has been assessed in two randomized controlled trials. The first trial showed, in a post hoc analysis, that rfviia administration may significantly improve the results of conventional therapy in patients with moderate and advanced liver failure (stages B and C of the Child-Pugh classification) without increasing the incidence of adverse events. 57 A more recent trial tested rviia in patients with active bleeding at endoscopy and with a Child-Pugh score 8 points. This trial failed to show a benefit of rviia in terms of decreasing the risk of 5-day failure but improved 6-week mortality. 58 Infection is a strong prognostic indicator in acute variceal bleeding. 63 The most frequent infections are spontaneous bacterial peritonitis (50%), urinary tract infections (25%), and pneumonia (25%). The use of prophylactic antibiotics in patients with acute variceal bleeding has been shown to reduce both the risk of rebleeding 76 and mortality. 77 Therefore, antibiotics should be given to all patients from admission. Quinolones are frequently used due to their easy administration and low cost. 78 In high-risk patients (hypovolemic shock, ascites, jaundice, or malnutrition) intravenous ceftriaxone has recently been shown to be superior to oral norfloxacin. 79 Variceal bleeding can trigger hepatic encephalopathy. However, there are no data to support the prophylactic use of lactulose or lactitol. 21 SPECIFIC THERAPY FOR CONTROL OF BLEEDING Initial therapy for acute variceal bleeding is based on the combination of vasoactive drugs with endoscopic therapy. Rescue therapies for failures include balloon tamponade and portal-systemic shunts, either surgical or TIPS. Pharmacological Therapy Vasoactive drugs act to reduce variceal pressure by decreasing variceal blood flow. The selection of the drug depends on the local resources. Terlipressin should be the first choice if available, since it is the only drug that has been shown to improve survival. 60,80 Somatostatin, octreotide, or vapreotide are the second choice. 60,81 If these drugs are not available, vasopressin plus transdermal nitroglycerin is an acceptable option. 60 Terlipressin is a long-acting triglycyl lysine derivative of vasopressin. Clinical studies have consistently
9 PORTAL HYPERTENSION AND GI BLEEDING/BOSCH ET AL 11 shown less frequent and severe side effects with terlipressin than with vasopressin, even when vasopressin is associated with nitroglycerin. The most common side effect of this drug is abdominal pain. Serious side effects such as peripheral or myocardial ischemia occur in less than 3% of the patients. 82 Terlipressin may be initiated as early as variceal bleeding is suspected at a dose of 2 mg/4 hours for the first 48 hours, and it may be maintained for up to 5 days at a dose of 1 mg/4 hours to prevent rebleeding. 82 Compared with placebo or nonactive treatment, terlipressin significantly improves the rate of control of bleeding and survival. 83 This is the only treatment that has been shown to improve prognosis of variceal bleeding in placebo-controlled randomized clinical trials (RCTs) and meta-analysis. 60,83 Terlipressin is as effective as any other effective therapy, including endoscopic injection sclerotherapy (EIS), and is safer than vasopressin þ nitroglycerin and EIS. 60,82,83 The overall efficacy of terlipressin in controlling acute variceal bleeding at 48 hours is 75 to 80% across trials, 83 and 67% at 5 days. 82 Terlipressin is also useful in hepatorenal syndrome. 84 Thus the use of terlipressin for variceal bleeding may prevent renal failure, which is frequently precipitated by variceal bleeding. 70 Somatostatin is commonly used as an initial bolus of 250 mg followed by a 250-mg/h infusion that is maintained until the achievement of a 24-hour bleedfree period. The bolus injection can be repeated up to three times in the first hour if bleeding is uncontrolled. Therapy may be further maintained for up to 5 days to prevent early rebleeding. 85 The use of higher doses (500 mg/h) causes a greater fall in HVPG and translates into increased clinical efficacy in the subset of patients with more difficult bleedings (those with active bleeding at emergency endoscopy). 86 Major side effects with somatostatin are rare. Minor side effects, such as nausea, vomiting, and hyperglycemia occur in up 30% of patients Several randomized controlled trials showed that somatostatin significantly improves the rate of control of bleeding compared with placebo or nonactive treatment. 60,81 However, despite the beneficial effect on control of bleeding, somatostatin did not reduce mortality. 60 Somatostatin has been compared with terlipressin and no differences were found for failure to control bleeding, rebleeding, mortality, or in the incidence of adverse events in both treatment groups. 60 Octreotide is a somatostatin analogue with longer half-life. This, however, is not associated with longer hemodynamic effects than somatostatin. 88 The optimal doses are not well determined. It is usually given as an initial bolus of 50 mg, followed by an infusion of 25 or 50 mg/h. 81 As with somatostatin, therapy can be maintained for 5 days to prevent early rebleeding. The safety profile of octreotide is close to that of somatostatin. The efficacy of octreotide as a single therapy for variceal bleeding is controversial. No benefit from octreotide was found in the only trial using octreotide or placebo as initial treatment, 89 which may be due to rapid development of tachyphylaxis. 88 However, RCTs using octreotide on top of sclerotherapy have shown a significant benefit in terms of reducing early rebleeding. 90 It has been speculated that this may be related to its sustained ability to prevent postprandial increase in PP. 81 Mortality, however, was not affected. 60,90 These results suggest that octreotide may improve the results of endoscopic therapy but has uncertain effects if used alone. When compared with other vasoactive drugs, octreotide was better than vasopressin and equivalent to terlipressin, again suggesting a clinical value from the use of octreotide, although all these studies were underpowered and none was double blind. 60 Endoscopic Therapy Both sclerotherapy and endoscopic band ligation (EBL) have been shown to be effective in the control of acute variceal bleeding. Two randomized trials specifically compared band ligation and sclerotherapy in acute variceal bleeding. 91,92 In one of them all patients also received pharmacological therapy (somatostatin). 92 In eight additional trials these two modalities were compared both in acute bleeding and in the prevention of rebleeding. Meta-analysis shows that EBL is better than sclerotherapy in the initial control of bleeding, and is associated with fewer adverse events and improved mortality. Additionally, sclerotherapy, but not EBL, may induce a sustained increase in PP. 93 Therefore EBL should be the endoscopic therapy of choice in acute variceal bleeding, though injection sclerotherapy is acceptable if band ligation is not available or is technically difficult. Endoscopic therapy can be performed at the time of diagnostic endoscopy, early after admission, provided that a skilled endoscopist is available. This is important since there has been an increased frequency of aspiration pneumonia since emergency endoscopic therapy has become a universal practice. Current Recommendation for Initial Treatment The current recommendation is to combine these two approaches, starting vasoactive drug therapy early (ideally during the transfer to the hospital, even if active bleeding is only suspected) and performing EBL (or injection sclerotherapy if band ligation is technically difficult) after initial resuscitation when the patient is stable and bleeding has ceased or slowed (Table 2). The rationale for that comes from several RCTs demonstrating that early administration of a vasoactive drug facilitates endoscopy and improves control of bleeding and 5-day rebleeding. 80,87,94,95 Drug therapy also improves the results of endoscopic treatment if started just after sclerotherapy or band ligation. 60,90 Conversely, the association of endoscopic therapy also improves the efficacy of vasoactive treatment. 87 However, this combined approach failed to significantly improve 6-week mortality with
10 12 SEMINARS IN LIVER DISEASE/VOLUME 28, NUMBER Table 2 Treatment of Acute Variceal Bleeding: Recommendations The best approach is the combined use of a pharmacological agent, started from admission (or even during transfer to hospital), and an endoscopic procedure. Terlipressin, somatostatin, octreotide, and vasopressin þ nitroglycerin (in this order of preference) may be used; drug therapy should be maintained for at least 2 to 5 days. Endoscopic band ligation or injection sclerotherapy (in this order of preference) are the endoscopic treatments of choice in bleeding esophageal varices; in bleeding gastric varices the best endoscopic choice is obturation with cyanoacrylate. Prophylaxis of infection with broad spectrum antibiotics should be given to all patients. TIPS should be used as a rescue procedure when medical and endoscopic therapies fail; patients bleeding from gastric varices may require an earlier decision for TIPS. TIPS, transjugular intrahepatic portal-systemic shunts. respect to endoscopic therapy 96 or a vasoactive drug 87 alone. The optimal duration of drug therapy is not well established and requires evaluation. Current recommendation is to maintain the drug for 2 to 5 days, to cover the period of maximum risk of rebleeding. 21 RESCUE THERAPIES: TAMPONADE, SURGERY, AND TIPS In 10 to 20% of patients, variceal bleeding is unresponsive to initial endoscopic and/or pharmacologic treatment. If bleeding is mild and the patient is stable, a second endoscopic therapy (if technically possible) might be attempted. If this fails, or bleeding is severe, the patient should be offered a derivative treatment, before the clinical status of the patient further deteriorates. Balloon tamponade achieves hemostasis in 60 to 90% of variceal bleedings 3 but should only be used in the case of a massive bleeding, for a short period of time (less than 24 hours) as a temporal bridge until definite treatment is instituted. Bleeding recurs after deflation in over half of the cases and severe complications are common. A recent report suggests that the use of esophageal covered stents might achieve hemostasis in most patients with refractory bleeding, 97 with the advantage over tamponade of less severe complications despite much longer periods of treatment. Adequately designed trials should confirm these findings. Both TIPS and surgical shunts are extremely effective for controlling variceal bleeding (control rate approaches 95%), but due to worsening of liver function and encephalopathy mortality remains high. 3,98 TIPS is the first choice, since most patients requiring rescue treatment have advanced liver disease. However, rarely, if ever, will a patient with a Child-Pugh score over 13 survive TIPS. This clearly indicates that some patients do not benefit from TIPS in this setting, and sometimes it is difficult to make a clinicallybased decision. Prognostic scores 99 may provide objective parameters to ease the decision of not offering invasive treatments in difficult cases. SPECIAL PROBLEMS IN MANAGEMENT The Patient with Poor Prognostic Indicators The current consensus is to apply to all patients the same treatment. However, it is plausible that patients with poor prognostic indicators might benefit from a more aggressive therapeutic approach ab initio. This was recently explored in a randomized trial in which patients with high PP (> 20 mm Hg) were randomized to receive standard therapy or TIPS. Those who underwent early TIPS had significantly less treatment failure and lower mortality than patients undergoing standard therapy. 65 However, the standard therapy used in the control arm of this trial was only endoscopic therapy, which is not the current standard of combination of vasoactive drugs and endoscopic treatment. 21 An ongoing multicenter study will answer whether early TIPS (performed with covered stents) is superior to combination therapy in high-risk patients (ISRCTN ). Gastric Varices Gastric varices develop in 20% of patients with portal hypertension. 100 They are the source of 5 to 10% of all upper digestive bleeding episodes in patients with cirrhosis. The risk of gastric variceal bleeding is lower than that of esophageal variceal bleeding, but gastric variceal bleeding, in particular that from fundal varices, tends to be more severe, to require more transfusions, and to have a higher mortality rate. 100 Fundal varices account for 1 to 3% of variceal bleeds. The optimal treatment of gastric fundal varices has not been determined, since there are only a few RCTs and most data come from retrospective series and case reports. The initial treatment is similar to that of esophageal variceal bleeding, including the administration of a vasoactive drug (either terlipressin, somatostatin, or a somatostatin analogue) and general management for hemodynamic stabilization and the prevention of complications. However, the efficacy of vasoactive drugs has not been assessed specifically. Balloon tamponade, with the Linton-Nachlas tube, has been used with limited success, 101,102 but may serve as a bridge to derivative treatments in massive bleedings. Some endoscopic therapies are promising, but quality information is scarce, and most studies include both fundal varices and gastroesophageal varices. Sclerotherapy, variceal obturation with tissue adhesives ( glue injection ), thrombin, elastic band ligation, and ligation with large detachable snares have been reported. 103 In most uncontrolled series, cyanoacrylate is highly effective, on the order of 90%. 104 Two recent randomized trials compared banding ligation with
11 PORTAL HYPERTENSION AND GI BLEEDING/BOSCH ET AL 13 cyanoacrylate injection in patients with acute variceal bleeding. In one trial cyanoacrylate injection was more effective and safer than band ligation in the control of acute bleeding, and was associated with less rebleeding. 105 In another trial both treatments were equally effective in the control acute of bleeding, but rebleeding was more frequent in the ligation group. 106 In another RCT by Sarin and coworkers, cyanoacrylate was better than alcohol injection in achieving initial hemostasis and in achieving faster variceal obliteration. 107 These trials suggest that cyanoacrylate injection is the endoscopic therapy of choice in acute bleeding from gastric varices. This technique requires expertise, and is usually not feasible during active bleeding. TIPS is very effective in the treatment of bleeding gastric varices, with more than a 90% success rate for initial hemostasis and a low rebleeding rate. 108,109 TIPS is usually associated with the embolization of the collateral vessels that feed the varices. Derivative and devascularization surgery are also effective, but with limited applicability in advanced cirrhosis. The authors recommendation is to start with a vasoactive drug. If bleeding is not controlled and if an expert endoscopist is available, variceal obturation with tissue adhesives should be attempted. In cases of massive bleeding or after failure of previous therapies, TIPS (or surgical shunt in Child-Pugh class A patients) is mandatory. In some patients, especially in those with massive initial bleeding or in centers with reduced endoscopic experience in the treatment of gastric varices, TIPS may be used even before attempting endoscopic therapy. Portal Hypertensive Gastropathy Portal hypertensive gastropathy (PHG) is a macroscopic finding of a characteristic mosaic-like pattern of the gastric mucosa ( mild PHG), red-point lesions, cherry red spots, and/ or black-brown spots ( severe PHG). 110 These lesions, however, are not entirely specific, that is, they can occur in the absence of portal hypertension. In PHG there is marked dilation of the vasculature of the gastric mucosa and submucosa, together with an increased blood flow and tendency to decreased acid secretion. PHG is unrelated to Helicobacter pylori infection. The overall prevalence of PHG in patients with cirrhosis strongly correlates with the severity of the disease and ranges between 11% and 80%. 110 The incidence of acute bleeding is low (less than 3% at 3 years) with a mortality of 12.5%; for chronic bleeding the incidence is around 10 to 15% at 3 years. In acute bleeding, PHG b-adrenergic blockers, somatostatin, octreotide, vassopressin, terlipressin, and estrogens have been proposed based on their ability to decrease gastric perfusion in this condition However, only one uncontrolled study so far has evaluated one of these drugs (somatostatin) in acute bleeding from PHG. 115 Hemostasis was achieved in all patients. PREVENTION OF FIRST BLEEDING Prophylactic Therapy: Who Benefits? PATIENTS WITHOUT VARICES The observation that nonselective b-blockers attenuated or delayed the development of collaterals in experimental models of portal hypertension 116 prompted studies to investigate whether these agents could prevent the development of esophageal varices in patients with cirrhosis. A recent study randomized 213 patients with cirrhosis and portal hypertension but without varices to receive timolol (a nonselective b-adrenergic blocker) or placebo for a median of 55 months. 1 The primary end point was development of esophageal varices or variceal hemorrhage. The rate of development of the primary end point did not differ between the two treatment groups (intention to treat). The incidence of secondary end points (i.e, ascites, encephalopathy, liver transplantation, or death) was also not significantly different. Adverse events were more frequent in the timolol group. Therefore, b- adrenergic blockers cannot be recommended for the prevention of the development of esophageal varices. Recent studies have shown that blockade of the vascular endothelial growth factor signaling cascade is highly effective in reducing the formation of collaterals in experimental models, 117,118 but no study has explored this clinically. A different approach, which is easily available, is to prevent the progression of cirrhosis (i.e., abstinence in alcoholics, antivirals in viral cirrhosis, lifestyle change in nonalcoholic steatohepatitis, corticosteroids in autoimmune hepatitis, phlebotomies in hemochromatosis, copper chelators in Wilson s disease). PATIENTS WITH VARICES In the past, only patients with medium to large varices were considered for prophylactic treatment of variceal bleeding. This was because most studies with b-adrenergic blockers were performed in patients with medium to large varices, while the beneficial effects of b- blockers are less clear in patients with small varices. 60 However, the classification of varices according to their size is very subjective. In fact, in the recent Baveno IV consensus conference it was not possible to agree on a definition of small and big varices. 21 On the other hand, it is well established that small varices with red signs or in Child-Pugh class C patients have a bleeding risk similar to that of big varices. 14 Additionally, b-adrenergic blockers may reduce the rate of progression from small to large varices, and decrease the incidence of variceal bleeding in patients with small varices. 7 Therefore, the current recommendation is to extend prophylactic treatment to patients with small varices with red signs or Child-Pugh C. 21
12 14 SEMINARS IN LIVER DISEASE/VOLUME 28, NUMBER Screening for Varices: When and How As previously discussed, the current consensus is that every cirrhotic patient should be endoscopically screened for varices at the time of diagnosis. 7 In patients without varices on initial endoscopy, a second (follow-up) evaluation should be performed after 2 to 3 years. 21 Since endoscopy is unpleasant for the patient, and screening in all cirrhotic patients is a substantial burden, empirical b- blocker therapy for all patients has been proposed. Two studies suggest that this strategy is cost-effective, 119,120 but a third suggested that empiric b-blockers are costeffective only in patients with decompensated cirrhosis. 121 The lack of effectiveness of b-adrenergic blockers to prevent the development of varices, and the high rate of side effects observed in well-compensated patients, 1 call into question the use of b-adrenergic blockers without screening. Treatments for the Prevention of First Bleeding: b-adrenergic Blockers versus EBL Nonselective b-adrenergic blockers (propranolol or nadolol) have been shown to reduce the risk of first variceal bleeding (from 24 to 15% after a median follow-up of 2 years). 60 Mortality was also reduced (from 27 to 23%) though this did not achieve statistical significance. 60 It is important to note that b-blockers are among the safest and least expensive drugs in Europe. However, about 25% of cirrhotic patients with medium/large esophageal varices may have either contraindications for the administration of nonselective b-blockers or cannot tolerate these drugs, and the degree of protection afforded (40% relative risk reduction) is far from ideal, which has stimulated the search for alternative treatments. EBL is effective in preventing the first variceal bleeding in patients with medium to large varices. 122 Sixteen trials have compared EBL with b-adrenergic blockers as a first-line option for primary prophylaxis of variceal bleeding The meta-analysis of these trials shows an advantage of EBL over b-adrenergic blockers in terms of prevention of first bleeding, without differences in mortality (Fig. 5). Six of these trials are only available in abstract, which makes it difficult to evaluate their quality, and most trials were underpowered or lack any sample size calculation (11 out of 16 included fewer than 100 patients). Four of the trials were prematurely stopped after an interim analysis, due to futility in three cases 125,131,138 or to an apparently significant benefit of banding in a small study. 130 When meta-analysis is restricted to published studies with a quality score above 5 (out of 10), there is no significant benefit from EBL. These data have led to considerable controversy over whether ß-blockers should remain the first treatment option to prevent first variceal bleeding, since in addition to the higher efficacy of banding ligation, adverse events requiring treatment discontinuation were significantly more frequent in patients treated with nonselective b-blockers. 139 However, the type and severity of side effects was different between the two therapies. 140 Indeed, most side effects related to b-blockers (hypotension, tiredness, breathlessness, poor memory, insomnia) were subjective and were easily managed by adjusting the dose or discontinuing b- blockers, did not require hospital admission, and resulted in no fatalities. In contrast, side effects related to EBL included 12 bleeding episodes and 1 esophageal perforation. In most cases, these complications required hospitalization and blood transfusion and resulted in 3 deaths. 125,138 Further, because of the short duration of follow-up in most of these studies, the long-term safety and benefits of prophylactic EBL are still uncertain. On the contrary, long-term safety and efficacy of nonselective b-adrenergic blockers are well established. 141,142 Bleeding episodes after discontinuing b-blockers because of side effects have also been considered as an argument against their use as first choice. However, most of these bleedings occurred months or years after b-blockers were withdrawn, suggesting that if an alternative therapy, such as EBL, had been offered to these patients, some of these bleeding episodes could have been prevented. The cost-effectiveness of EBL versus ß-blockers for primary prophylaxis has been compared in three decision-analysis studies. Different assumptions on the incidence of variceal bleeding, quality of life with each treatment, mortality, or other portal hypertensive complications led to conflicting conclusions. 119,121,143 The recommendation made at the 2005 Baveno consensus conference is that nonselective b-blockers should be considered as first-choice treatment to prevent first variceal bleeding, while EBL should be offered to patients with medium/large varices and contraindications or intolerance to b-blockers (Table 3). 21 Even in this case, a recent trial shed some doubts on the use of EBL. In that trial EBL was compared with no treatment in patients with contraindications to b-blockers. The trial was prematurely stopped due to a high rate (12%) of iatrogenic bleeding in the band ligation group. 144 However, the early termination of the trial frustrated the possibility of obtaining clear-cut evidence to support a recommendation. The combination of EBL plus b-adrenergic blockers appears to offer no benefit in terms of prevention of first bleeding when compared with EBL alone. 145 There is a lower rate of recurrence of varices in patients treated with EBL plus propranolol but at the expense of more side effects. Probably more studies would be required, although these are unlikely to be performed due to the very large number of patients that would be needed.
13 PORTAL HYPERTENSION AND GI BLEEDING/BOSCH ET AL 15 Figure 5 Prevention of first variceal bleeding. Meta-analysis (random effects model) of randomized controlled trials comparing endoscopic band ligation with b-adrenergic blockers in the prevention of first variceal bleeding. (A) Bleeding. (B) Mortality. EBL, endoscopic band ligation; RR, relative risk; CI, confidence interval. Unanswered Issues HVPG MONITORING Three longitudinal studies have demonstrated that a sufficient HVPG decrease (to 12 mm Hg or below or of more than 20% from baseline) is associated with an effective protection from first variceal bleeding. 12,18,142 This leads to the question of whether HVPG measurements should be used to monitor PP response to drug treatment in clinical practice. Two simulation analyses have yielded conflicting results, one suggesting that HVPG monitoring might be cost-effective in primary prophylaxis, 146 and the other arriving at the opposite conclusion. 147 The main problem is that the assumptions of these analyses (i.e., how to manage nonresponders to medical treatment) have never been tested in randomized controlled trials. A recent study of HVPG-guided therapy suggested that the shift of nonresponders from b-blockers to EBL does not improve the outcome. 148 The issue will remain hypothetical until HVPG-guided therapy is proven to be better than empirical approaches in randomized controlled trials.
14 16 SEMINARS IN LIVER DISEASE/VOLUME 28, NUMBER Table 3 Prophylaxis of First Variceal Bleeding: Recommendations Patients without varices should be screened endoscopically for the appearance of varices every 2 to 3 years. Patients with moderate/large varices should be treated with a nonselective b-blocker if there are no contraindications. Patients with small varices with red signs or with advanced liver failure (Child-Pugh C) are at similar risk of bleeding as those with moderate/large varices and should be considered for preventive therapy. Patients with moderate/large varices with contraindications to or who cannot tolerate b-blockers should undergo endoscopic band ligation; band ligation might be used as the first choice in patients with large varices depending on patient s preferences and local resources. If no bleeding occurs treatment should be maintained for life (unless the liver disease improves and significant portal hypertension disappears). THE COMBINATION OF b-adrenergic BLOCKERS WITH NITRATES The addition of ISMN has been shown to significantly increase the long-term HVPG response to b-adrenergic blockers. 149 However, it is less clear whether the greater effect of this combination on PP translates into a greater clinical efficacy in primary prophylaxis. An open trial comparing nadolol to nadolol þ isosorbide mononitrate demonstrated a significantly lower rate of first bleeding in the combination group, without survival advantage. 150,151 However, a subsequent double-blind, placebo-controlled study including a large number of patients failed to confirm these results. 152 ENDOSCOPIC TREATMENT: HOW FREQUENT, HOW TO MONITOR THE TREATMENT There is no agreement on how frequently the varices should be ligated in the initial course of eradication. Some authors wait a minimum of 1 month between banding procedures, 130 while others perform EBL on a weekly basis. 125 Reported frequency of complications with the former strategy was lower than with the latter. A recent trial evaluated the effectiveness and complications of EBL performed every 2 weeks compared withevery 2 months. This trial included patients with and without previous bleeding, though most patients were treated for primary prophylaxis. 153 The 2-month interval scheme obtained a higher total eradication rate and a lower recurrence rate. No patient in either group developed variceal bleeding. Thus, although admittedly weak, current evidence favors monthly intervals. Follow-up endoscopies should be performed every 6 months and varices should be re-eradicated upon recurrence. This is in marked contrast to prophylaxis with b-blockers, in which no follow-up endoscopies are needed. PREVENTION OF RECURRENT BLEEDING FROM ESOPHAGEAL VARICES Patients surviving a first episode of variceal bleeding have a high risk of recurrent bleeding, of over 60% within 1 year from the first bleeding. Because of this, all patients surviving a variceal bleeding should receive active treatments for the prevention of rebleeding 21 (Fig. 6, Table 4). Figure 6 Reported variceal rebleeding rate with different treatment options. Constructed from data of randomized clinical trials for secondary prevention of variceal rebleeding. EIS, endoscopic injection sclerotherapy; BB, beta-blockers; EBL, endoscopic band ligation; ISMN, isosorbide-5-mononitrate; TIPS, transjugular intrahepatic portal-systemic shunts; DSRS, distal splenorenal shunt; HVPG, hepatic venous pressure gradient. (Based on data from Bosch J, Garcia-Pagan J. Prevention of variceal rebleeding. Lancet 2003;361: )
15 PORTAL HYPERTENSION AND GI BLEEDING/BOSCH ET AL 17 Table 4 Prevention of Variceal Rebleeding: Recommendations All patients surviving a bleeding episode should be treated to prevent recurrent bleeding; in patients who bled and were not receiving b-blockers, either b-blockers ISMN, band ligation, or both can be used. Patients who bled while on b-blocker or under prophylactic EBL should be treated with the combination of EBL plus b-blockers. Patients who are intolerant to or who have contraindications to b-blockers should be treated with EBL. The efficacy of TIPS in the prevention of recurrent bleeding from esophageal varices is not different from that of shunt surgery (distal splenorenal shunt or 8-mm Hg graft shunt), especially since the introduction of PTFE-covered stents. Failures of medical therapy are preferably managed with PTFE-TIPS; this is the only option in poor surgical candidates. Patients who survive a bleeding episode and have advanced liver failure (Child-Pugh score > 8 or MELD > 16) should be considered for liver transplantation. ISMN, isosorbide-5-mononitrate; EBL, endoscopic band ligation; TIPS, transjugular intrahepatic portal-systemic shunt; PTFE, polytetrafluoroethylene; MELD, model for end-stage liver disease. Pharmacological Treatment The efficacy of nonselective b-blockers in the prevention of variceal rebleeding has been proven in many RCTs, and these drugs are now widely accepted as the first-line pharmacological therapy in this setting 21. Several metaanalyses have consistently found a marked benefit from b-blockers showing a reduction in rebleeding rate from 63% in controls to 42% in treated patients. 60 Mortality was significantly reduced, from 27 to 20% 60 and mortality from bleeding was also significantly reduced. 154 Betablockers have been compared with endoscopic variceal sclerotherapy in the prevention of rebleeding. No significant differences were found either for rebleeding or for mortality but side effects were significantly less frequent and severe with b-blockers. 60 The combined administration of propranolol or nadolol plus ISMN was introduced after demonstrating that ISMN enhanced the PP-reducing effect of nonselective b-blockers. 149 There is insufficient information on whether this translates into a clinical advantage, since there are only two studies of ISMN associated with propranolol 155 or nadolol 156 versus the corresponding b-blocker alone, in the prevention of rebleeding. One of these studies showed significant benefit, but not the second (which is still available only in abstract form). 156 However, the association of ISMN with propranolol or nadolol has been found to be superior to endoscopic sclerotherapy in one study. 157 In addition, the association of propranolol/nadolol and ISMN has been compared with EBL in four studies Meta-analysis of these four studies has shown no significant differences between the treatments in preventing rebleeding or in mortality. Compared with TIPS, the combination of ISMN and propranolol is less effective for the prevention of rebleeding but it is associated with significantly less encephalopathy, similar mortality, and much lower cost. 162 Treatment dosage and schedules for propranolol, nadolol, and ISMN are the same as for the prevention of first bleeding. Endoscopic Treatment Endoscopic injection sclerotherapy of esophageal varices significantly reduces both the rebleeding and death risk. It takes four to six endoscopic sessions to eradicate varices, but recurrence of varices occurs in nearly 40% of patients within 1 year from eradication. This requires further endoscopic sessions to maintain eradication. The most serious side effects of therapy are dysphagia, esophageal stenosis, and bleeding from esophageal ulcers, which may account for as much as 14% of all the rebleeding episodes. As noted above, sclerotherapy has no advantage over drug therapy and causes more frequent and severe side effects. Endoscopic banding ligation has been proven superior to sclerotherapy. 140 Complications are significantly less frequent and severe with banding ligation. Therefore, banding ligation should be preferred to sclerotherapy. 21 Surprisingly, despite decreasing rebleeding rates, endoscopic ligation does not significantly improve survival compared with sclerotherapy. Although variceal eradication is achieved with a lower number of EBL sessions than with sclerotherapy, there is evidence that it is associated with higher recurrence of varices. 140 Combined Endoscopic Treatment In the approach using banding ligation combined with sclerotherapy, sclerotherapy has been added (either simultaneously or after the reduction of variceal size to small) to EBL and compared with band ligation alone, yielding contrasting results. The meta-analysis of these studies does not show any benefit either for rebleeding or for mortality, and importantly, it shows a trend toward an increasing complication rate with combination endoscopic therapy. Combined Endoscopic and Pharmacological Treatment The association of injection sclerotherapy and b-blockers has been compared with either sclerotherapy or b-blockers alone. The meta-analysis of the RCTs comparing combination therapy with sclerotherapy alone showed a significant reduction of the rebleeding
16 18 SEMINARS IN LIVER DISEASE/VOLUME 28, NUMBER risk with combination therapy, but no differences in mortality. 3 Also, when compared with b-blockers alone, combination therapy significantly reduced the rebleeding risk without advantage for survival. 60 Two RCTs have shown that the combination of banding ligation plus b-blockers is superior to banding ligation alone in terms of recurrence of varices and recurrence of bleeding. 163,164 In addition, in one RCT, adding band ligation to nadolol plus ISMN was shown to be superior to nadolol þ ISMN alone in preventing variceal rebleeding. However, no significant differences were observed when considering rebleeding episodes of any cause. This was due to a greater number of ulcerrelated bleeding episodes in the treatment arm including banding. 165 All together, these results emphasize that the possible greater clinical efficacy of the combination of endoscopic plus drug treatment in the prevention of rebleeding should be further evaluated. However, until then the recommendation is to use the combination in patients who bled under either treatment alone. In these patients, the substitution of the failing treatment with the other should not be preferred to their combination. TIPS TIPS has been compared with sclerotherapy and with banding ligation showing almost consistently that TIPS is superior to either endoscopic therapy for the prevention of rebleeding. Similarly, TIPS has also been shown to be superior to the combination of ISMN and propranolol. 162 Not surprisingly, this impressive efficacy in preventing recurrent bleeding is associated with a marked increase in the risk of encephalopathy without reducing either overall mortality or mortality from bleeding. TIPS has been compared with surgical shunts in two RCTs. 166 In the first study, TIPS was compared with 8-mm portocaval H-graft shunt. 166,167 A significantly lower rebleeding rate was found with the surgical shunt. Significantly more patients required liver transplant in the TIPS group than in the surgical shunt group. There was no difference in mortality. The composite end point of failures which included rebleeding, shunt thrombosis, deaths, and need for transplant, was significantly higher for the TIPS. In the second trial TIPS was compared with the distal splenorenal shunt (DSRS) in Child-Pugh class A and B patients. 167 No significant differences in rebleeding rate (5.5% in the DSRS group, and 9% in the TIPS group), incidence of hepatic encephalopathy, liver transplantation, or mortality was found. However, to obtain these results the reintervention rate was significantly higher in the TIPS group (82%) than in the DSRS group (11%). However, new polytetrafluoroethylene (PTFE)-covered stents have solved one of the major drawbacks of TIPS use of bare stents, that is, its high rate of occlusion or dysfunction. Indeed, a multicenter RCT reported much lower obstruction and reintervention rates with the use of PTFE stents, which was associated with lower rates of recurrent bleeding or ascites without an increase in the incidence of encephalopathy, than with the use of bare stents. 168 These results suggest that the small disadvantage of TIPS versus surgical shunt would be overcome by the use of PTFE-covered stents. HVPG-Guided Therapy in the Prevention of Rebleeding Pharmacological (or spontaneous) reduction of HVPG to < 12 mm Hg or by 20% of the baseline value (HVPG responders) markedly decreases the risk of rebleeding and also reduces mortality. 169 Indeed, achieving such a hemodynamic response is associated with a rebleeding risk that is even lower than that achieved using surgical shunts or TIPS 170 (Fig. 7). As a consequence, to add further treatment (i.e., band ligation) to these patients is likely to not enhance efficacy but to increase the side effects. On the other hand, it is still unclear whether patients with an insufficient hemodynamic response to pharmacological therapy would benefit from adding other drugs or alternative treatments. As previously discussed, a decision analysis does not clearly suggest that HVPG-guided therapy is costeffective in the primary prevention of variceal bleeding. However, probably because of the higher risk of variceal hemorrhage in secondary prophylaxis, a recent study concluded that the use of nonselective b-blockers þ - long-acting nitrates offering EBL to HVPG nonresponders was cost-effective and only marginally more expensive than EBL or pharmacological treatment without HVPG monitoring. 171 However, decision analysis models are limited by the many assumptions on which these are based. Because of these, the value of HVPGguided therapy must be demonstrated in randomized clinical studies. A first attempt was done by Bureau and colleagues 148 In this study, HVPG nonresponders to b-blockers were given ISMN and if their condition remained unchanged, they were shifted to EBL. Although the study included low numbers, it suggested that band ligation is not a good alternative in HVPG nonresponders. Preliminary data from a Spanish multicenter RCT comparing nadolol þ ISMN with nadolol þ ISMN þ EBL 165 found no significant differences in rebleeding rates in HVPG nonresponders treated with drugs alone or with drugs þ EBL, suggesting that adding EBL may not be the best alternative to reduce rebleeding in HVPG nonresponders. It is therefore possible that more effective and aggressive therapies are required to reduce the high rebleeding tendency of HVPG nonresponders (46 to 65% in a recent survey 170 ). The recent study by Gonzalez et al found a 19%
17 PORTAL HYPERTENSION AND GI BLEEDING/BOSCH ET AL 19 Figure 7 Meta-analysis of the randomized clinical trial (RCTs) comparing b-blockers þ isosorbide-5-mononitrate (ISMN) with endoscopic band ligation (EBL) of esophageal varices to prevent recurrent variceal bleeding and mortality. Each RCT is identified by the name of the first author and the year of publication. Squares indicate the odds ratio (OR) with the two treatments for each RCT. Horizontal bars denote the 95% confidence interval (CI) of the OR. The vertical line represents the line of identity of effect of the two treatments. The diamonds represent 95% CIs of the pooled OR. rebleeding rate in the HVPG nonresponders that were shifted to receive TIPS, suggesting that this might be an option. 172 Unfortunately, the study included a small number of patients and had no control group. Therefore, investigators should continue to study HVPG-guided therapy in order to be able to make clear recommendations. Treatment of Gastric Varices The optimal treatment to prevent gastric (fundal) variceal bleeding is not well defined, due to the relatively low incidence of this condition. In clinical practice, nonselective b-blockers or the tissue adhesive isobutyl-2- cyanoacrylate (bucrylate) are used as first line therapy. PTFE-TIPS or shunt surgery is recommended in failures. Treatment of PHG Prevention of recurrent bleeding from PHG should be based on nonselective b-blockers, at the same dosage as for treating esophageal varices. 173 Adequate iron supplementation may be useful to prevent or correct chronic iron-deficient anemia in patients with severe PHG. 173 Rare patients who have repeated severe bleeding from PHG despite pharmacological therapy may benefit from endoscopic ablation, either by argon plasma coagulation, neodymium:yttrium-aluminum-garnet (Nd:YAG) laser, or heater probe. If this is not feasible or fails, TIPS may be considered as an alternative therapy. PHG reverses after liver transplantation. ACKNOWLEDGMENTS We thank Ms. Clara Esteva for expert secretarial support. This study was supported in part by grants from the Ministerio de Educación y Ciencia (SAF-04/04783), and from Instituto de Salud Carlos III (FIS 06/0623 and 05/0519). ABBREVIATIONS CDUS color Doppler ultrasound CSPH clinically significant portal hypertension CT computed tomography DSRS distal splenorenal shunt EBL endoscopic band ligation HVPG hepatic venous pressure gradient ISMN isosorbide-5-mononitrate IVC inferior vena cava MRI magnetic resonance imaging
18 20 SEMINARS IN LIVER DISEASE/VOLUME 28, NUMBER NO PHG PP PPG PTFE RCT TIPS US REFERENCES nitric oxide portal hypertensive gastropathy portal pressure portal pressure gradient polytetrafluoroethylene randomized clinical trial transjugular intrahepatic portal-systemic shunts ultrasound 1. Groszmann RJ, Garcia-Tsao G, Bosch J, et al. Betablockers to prevent gastroesophageal varices in patients with cirrhosis. N Engl J Med 2005;353: D Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic indicators of survival in cirrhosis: a systematic review of 118 studies. J Hepatol 2006;44: D Amico G, Pagliaro L, Bosch J. The treatment of portal hypertension: a meta-analytic review. Hepatology 1995;22: D Amico G, Luca A. Natural history clinical-haemodynamic correlations: prediction of the risk of bleeding. Baillieres Clin Gastroenterol 1997;11: Garcia-Tsao G, Groszmann RJ, Fisher RL, Conn HO, Atterbury CE, Glickman M. Portal pressure, presence of gastroesophageal varices and variceal bleeding. Hepatology 1985;5: Christensen E, Fauerholdt L, Schlichting P, Juhl E, Poulsen H, Tygstrup N. Aspects of the natural history of gastrointestinal bleeding in cirrhosis and the effect of prednisone. Gastroenterology 1981;81: Merkel C, Marin R, Angeli P, et al. A placebo-controlled clinical trial of nadolol in the prophylaxis of growth of small esophageal varices in cirrhosis. Gastroenterology 2004;127: Merli M, Nicolini G, Angeloni S, et al. Incidence and natural history of small esophageal varices in cirrhotic patients. J Hepatol 2003;38: Zoli M, Merkel C, Magalotti D, et al. Natural history of cirrhotic patients with small esophageal varices: a prospective study. Am J Gastroenterol 2000;95: de Franchis R. Evaluation and follow-up of patients with cirrhosis and oesophageal varices. J Hepatol 2003;38: Cales P, Oberti F, Payen JL, et al. Lack of effect of propranolol in the prevention of large oesophageal varices in patients with cirrhosis: a randomized trial. French-Speaking Club for the Study of Portal Hypertension. Eur J Gastroenterol Hepatol 1999;11: Groszmann RJ, Bosch J, Grace ND, et al. Hemodynamic events in a prospective randomized trial of propranolol versus placebo in the prevention of a first variceal hemorrhage [see comments]. Gastroenterology 1990;99: Vorobioff J, Groszmann RJ, Picabea E, et al. Prognostic value of hepatic venous pressure gradient measurements in alcoholic cirrhosis: a 10-year prospective study. Gastroenterology 1996;111: North Italian Endoscopic Club. Prediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices: a prospective multicenter study. The North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices [see comments]. N Engl J Med 1988;319: Casado M, Bosch J, Garcia-Pagan JC, et al. Clinical events after transjugular intrahepatic portosystemic shunt: correlation with hemodynamic findings. Gastroenterology 1998;114: Feu F, Garcia-Pagan JC, Bosch J, et al. Relation between portal pressure response to pharmacotherapy and risk of recurrent variceal haemorrhage in patients with cirrhosis. Lancet 1995;346: Bosch J, Garcia-Pagan JC, Berzigotti A, Abraldes JG. Measurement of portal pressure and its role in the management of chronic liver disease. Semin Liver Dis 2006;26: Merkel C, Bolognesi M, Sacerdoti D, et al. The hemodynamic response to medical treatment of portal hypertension as a predictor of clinical effectiveness in the primary prophylaxis of variceal bleeding in cirrhosis. Hepatology 2000;32: Moitinho E, Escorsell A, Bandi JC, et al. Prognostic value of early measurements of portal pressure in acute variceal bleeding. Gastroenterology 1999;117: Ripoll C, Banares R, Rincon D, et al. Influence of hepatic venous pressure gradient on the prediction of survival of patients with cirrhosis in the MELD era. Hepatology 2005; 42: de Franchis R. Evolving consensus in portal hypertension: report of the Baveno IV consensus workshop on methodology of diagnosis and therapy in portal hypertension. J Hepatol 2005;43: Groszmann RJ, Merkel C, Iwakiri Y, et al. Prevention of the formation of varices (pre-primary prophylaxis). In: de Franchis R, ed. Proceedings of the IV Baveno International Consensus Workshop on Definitions, Methodology and Therapeutic Strategies. Oxford: Blackwell-Science; 2005: Eisen GM, Eliakim R, Zaman A, et al. The accuracy of PillCam ESO capsule endoscopy versus conventional upper endoscopy for the diagnosis of esophageal varices: a prospective three-center pilot study. Endoscopy 2006;38: Lapalus MG, Dumortier J, Fumex F, et al. Esophageal capsule endoscopy versus esophagogastroduodenoscopy for evaluating portal hypertension: a prospective comparative study of performance and tolerance. Endoscopy 2006;38: Frenette C, Kuldau JG, Hillebrand D, Lane J, Pockros PJ. Comparison of esophageal pill endoscopy and conventional endoscopy for variceal screening. Hepatology 2006; 44(suppl 1):445A 26. D Amico G, Morabito A. Noninvasive markers of esophageal varices: another round, not the last. Hepatology 2004; 39: de Franchis R. Noninvasive diagnosis of esophageal varices: is it feasible? Am J Gastroenterol 2006;101: Chalasani N, Imperiale TF, Ismail A, et al. Predictors of large esophageal varices in patients with cirrhosis [see comments]. Am J Gastroenterol 1999;94: Giannini E, Botta F, Borro P, et al. Platelet count/spleen diameter ratio: proposal and validation of a non-invasive parameter to predict the presence of oesophageal varices in patients with liver cirrhosis. Gut 2003;52: Sanyal AJ, Fontana RJ, Di Bisceglie AM, et al. The prevalence and risk factors associated with esophageal
19 PORTAL HYPERTENSION AND GI BLEEDING/BOSCH ET AL 21 varices in subjects with hepatitis C and advanced fibrosis. Gastrointest Endosc 2006;64: Schepis F, Camma C, Niceforo D, et al. Which patients with cirrhosis should undergo endoscopic screening for esophageal varices detection? Hepatology 2001;33: Vanbiervliet G, Barjoan-Marine E, Anty R, et al. Serum fibrosis markers can detect large oesophageal varices with a high accuracy. Eur J Gastroenterol Hepatol 2005;17: Zoli M, Iervese T, Merkel C, et al. Prognostic significance of portal hemodynamics in patients with compensated cirrhosis. J Hepatol 1993;17: Bolondi L, Gandolfi L, Arienti V, et al. Ultrasonography in the diagnosis of portal hypertension: diminished response of portal vessels to respiration. Radiology 1982;142: Sacerdoti D, Merkel C, Bolognesi M, Amodio P, Angeli P, Gatta A. Hepatic arterial resistance in cirrhosis with and without portal vein thrombosis: relationships with portal hemodynamics. Gastroenterology 1995;108: Berzigotti A, Casadei A, Magalotti D, et al. Renovascular impedance correlates with portal pressure in patients with liver cirrhosis. Radiology 2006;240: Taourel P, Blanc P, Dauzat M, et al. Doppler study of mesenteric, hepatic, and portal circulation in alcoholic cirrhosis: relationship between quantitative Doppler measurements and the severity of portal hypertension and hepatic failure. Hepatology 1998;28: Cho KC, Patel YD, Wachsberg RH, Seeff J. Varices in portal hypertension: evaluation with CT. Radiographics 1995;15: Johnson CD, Ehman RL, Rakela J, Ilstrup DM. MR angiography in portal hypertension: detection of varices and imaging techniques. J Comput Assist Tomogr 1991;15: Ziol M, Handra-Luca A, Kettaneh A, et al. Noninvasive assessment of liver fibrosis by measurement of stiffness in patients with chronic hepatitis C. Hepatology 2005;41: Carrion JA, Navasa M, Bosch J, Bruguera M, Gilabert R, Forns X. Transient elastography for diagnosis of advanced fibrosis and portal hypertension in patients with hepatitis C recurrence after liver transplantation. Liver Transpl 2006;12: Lemoine M, Katsahian S, Nahon P, et al. Liver stiffness measurement is correlated with hepatic venous pressure gradient in patients with uncomplicated alcoholic and/or HVC related cirrhosis. Hepatology 2006;44(suppl 1):204A 43. Bureau C, Metivier S, Peron JM, et al. Prospective assessment of liver stiffness for the non-invasive prediction of portal hypertension. J Hepatol 2007;46(suppl 1):S Vizzutti F, Arena U, Romanelli RG, et al. Liver stiffness measurement predicts severe portal hypertension in patients with HCV-related cirrhosis. Hepatology 2007;45: Kazemi F, Kettaneh A, N kontchou G, et al. Liver stiffness measurement selects patients with cirrhosis at risk of bearing large oesophageal varices. J Hepatol 2006;45: Bosch J. Predictions from a hard liver. J Hepatol 2006; 45: Bosch J, Abraldes JG, Groszmann RJ. Current management of portal hypertension. J Hepatol 2003;38:S54 S Pinzani M, Gentilini P. Biology of hepatic stellate cells and their possible relevance in the pathogenesis of portal hypertension in cirrhosis. Semin Liver Dis 1999;19: Wiest R, Groszmann RJ. The paradox of nitric oxide in cirrhosis and portal hypertension: too much, not enough. Hepatology 2002;35: Gupta TK, Toruner M, Chung MK, Groszmann RJ. Endothelial dysfunction and decreased production of nitric oxide in the intrahepatic microcirculation of cirrhotic rats. Hepatology 1998;28: Rockey DC, Chung JJ. Reduced nitric oxide production by endothelial cells in cirrhotic rat liver: endothelial dysfunction in portal hypertension. Gastroenterology 1998;114: Hernandez-Guerra M, Garcia-Pagan JC, Turnes J, et al. Ascorbic acid improves the intrahepatic endothelial dysfunction of patients with cirrhosis and portal hypertension. Hepatology 2006;43: Vorobioff J, Bredfeldt J, Groszmann RJ. Hyperdynamic circulation in a portal hypertensive rat model: a primary factor for maintenance of chronic portal hypertension. Am J Physiol 1983;244:G52 G Pizcueta MP, Piqué JM, Bosch J, Whittle BJ, Moncada S. Effects of inhibiting nitric oxide biosynthesis on the systemic and splanchnic circulation of rats with portal hypertension. Br J Pharmacol 1992;105: Pizcueta P, Piqué JM, Fernández M, et al. Modulation of the hyperdynamic circulation of cirrhotic rats by nitric oxide inhibition. Gastroenterology 1992;103: Garcia-Pagan JC, Salmeron JM, Feu F, et al. Effects of low-sodium diet and spironolactone on portal pressure in patients with compensated cirrhosis. Hepatology 1994;19: Bosch J, Thabut D, Bendtsen F, et al. Recombinant factor VIIa for upper gastrointestinal bleeding in patients with cirrhosis: a randomized, double-blind trial. Gastroenterology 2004;127: Bosch J, Thabut D, Albillos A, et al. Recombinant Factor VIIA (RFVIIA) for active variceal bleeding in patients with advanced cirrhosis: a multi-centre randomized double-blind placebo-controlled trial. J Hepatol 2007;46(suppl 1):S D Amico G, de Franchis R. Upper digestive bleeding in cirrhosis: post-therapeutic outcome and prognostic indicators. Hepatology 2003;38: D Amico G, Pagliaro L, Bosch J. Pharmacological treatment of portal hypertension: an evidence-based approach. Semin Liver Dis 1999;19: Graham DY, Smith JL. The course of patients after variceal hemorrhage. Gastroenterology 1981;80: Bernard B, Cadranel JF, Valla D, Escolano S, Jarlier V, Opolon P. Prognostic significance of bacterial infection in bleeding cirrhotic patients: a prospective study. Gastroenterology 1995;108: Goulis J, Armonis A, Patch D, Sabin C, Greenslade L, Burroughs AK. Bacterial infection is independently associated with failure to control bleeding in cirrhotic patients with gastrointestinal hemorrhage. Hepatology 1998;27: Abraldes JG, Aracil C, Catalina MV, et al. Value of HVPG predicting 5-day treatment failure in acute variceal bleeding: comparison with clinical variables. J Hepatol 2006;44(suppl 2):12A 65. Monescillo A, Martinez-Lagares F, Ruiz-del-Arbol L, et al. Influence of portal hypertension and its early decompression
20 22 SEMINARS IN LIVER DISEASE/VOLUME 28, NUMBER by TIPS placement on the outcome of variceal bleeding. Hepatology 2004;40: Carbonell N, Pauwels A, Serfaty L, Fourdan O, Levy VG, Poupon R. Improved survival after variceal bleeding in patients with cirrhosis over the past two decades. Hepatology 2004;40: Chalasani N, Kahi C, Francois F, et al. Improved patient survival after acute variceal bleeding: a multicenter, cohort study. Am J Gastroenterol 2003;98: Stokkeland K, Brandt L, Ekbom A, Hultcrantz R. Improved prognosis for patients hospitalized with esophageal varices in Sweden Hepatology 2006;43: Nidegger D, Ragot S, Berthelemy P, et al. Cirrhosis and bleeding: the need for very early management. J Hepatol 2003;39: Cardenas A, Ginès P, Uriz J, et al. Renal failure after upper gastrointestinal bleeding in cirrhosis: incidence, clinical course, predictive factors, and short-term prognosis. Hepatology 2001;34(4 Pt 1): McCormick PA, Jenkins SA, McIntyre N, Burroughs AK. Why portal hypertensive varices bleed and bleed: a hypothesis. Gut 1995;36: Castaneda B, Debernardi-Venon W, Bandi JC, et al. The role of portal pressure in the severity of bleeding in portal hypertensive rats. Hepatology 2000;31: Villanueva C, Ortiz J, Miñana J, et al. Somatostatin treatment and risk stratification by continuous portal pressure monitoring during acute variceal bleeding. Gastroenterology 2001;121: Morales J, Moitinho E, Abraldes JG, Fernandez M, Bosch J. Effects of the V1a vasopressin agonist F-180 on portal hypertension-related bleeding in portal hypertensive rats. Hepatology 2003;38: Ejlersen E, Melsen T, Ingerslev J, Andreasen RB, Vilstrup H. Recombinant activated factor VII (rfviia) acutely normalizes prothrombin time in patients with cirrhosis during bleeding from oesophageal varices. Scand J Gastroenterol 2001;36: Hou MC, Lin HC, Liu TT, et al. Antibiotic prophylaxis after endoscopic therapy prevents rebleeding in acute variceal hemorrhage: a randomized trial. Hepatology 2004; 39: Bernard B, Grange JD, Khac EN, Amiot X, Opolon P, Poynard T. Antibiotic prophylaxis for the prevention of bacterial infections in cirrhotic patients with gastrointestinal bleeding: a meta-analysis. Hepatology 1999;29: Rimola A, Garcia-Tsao G, Navasa M, et al. Diagnosis, treatment and prophylaxis of spontaneous bacterial peritonitis: a consensus document. International Ascites Club. J Hepatol 2000;32: Fernandez J, Ruiz-del-Arbol L, Gomez C, Durandez R, et al. Norfloxacin vs ceftriaxone in the prophylaxis of infections in patients with advanced cirrhosis and hemorrhage. Gastroenterology 2006;131: Levacher S, Letoumelin P, Pateron D, Blaise M, Lapandry C, Pourriat JL. Early administration of terlipressin plus glyceryl trinitrate to control active upper gastrointestinal bleeding in cirrhotic patients. Lancet 1995;346: Abraldes JG, Bosch J. Somatostatin and analogues in portal hypertension. Hepatology 2002;35: Escorsell A, Ruiz-del-Arbol L, Planas R, Albillos A, et al. Multicenter randomized controlled trial of terlipressin versus sclerotherapy in the treatment of acute variceal bleeding: the TEST study. Hepatology 2000;32: Ioannou GN, Doust J, Rockey DC. Systematic review: terlipressin in acute oesophageal variceal haemorrhage. Aliment Pharmacol Ther 2003;17: Uriz J, Ginès P, Cardenas A, et al. Terlipressin plus albumin infusion: an effective and safe therapy of hepatorenal syndrome. J Hepatol 2000;33: Escorsell A, Bordas JM, del Arbol LR, et al. Randomized controlled trial of sclerotherapy versus somatostatin infusion in the prevention of early rebleeding following acute variceal hemorrhage in patients with cirrhosis. Variceal Bleeding Study Group. J Hepatol 1998;29: Moitinho E, Planas R, Bañares R, et al. Multicenter randomized controlled trial comparing different schedules of somatostatin in the treatment of acute variceal bleeding. J Hepatol 2001;35: Villanueva C, Ortiz J, Sabat M, et al. Somatostatin alone or combined with emergency sclerotherapy in the treatment of acute esophageal variceal bleeding: a prospective randomized trial. Hepatology 1999;30: Escorsell A, Bandi JC, Andreu V, et al. Desensitization to the effects of intravenous octreotide in cirrhotic patients with portal hypertension. Gastroenterology 2001;120: International Octreotide Varices Study Group, Burroughs AK. Double blind RCT of 5 day octreotide versus placebo, associated with sclerotherapy for trial failures. Hepatology 1996;24:352A 90. Corley DA, Cello JP, Adkisson W, Ko WF, Kerlikowske K. Octreotide for acute esophageal variceal bleeding: a metaanalysis. Gastroenterology 2001;120: Lo GH, Lai KH, Cheng JS, et al. Emergency banding ligation versus sclerotherapy for the control of active bleeding from esophageal varices. Hepatology 1997;25: Villanueva C, Piqueras M, Aracil C, et al. A randomized controlled trial comparing ligation and sclerotherapy as emergency endoscopic treatment added to somatostatin in acute variceal bleeding. J Hepatol 2006;45: Avgerinos A, Armonis A, Stefanidis G, et al. Sustained rise of portal pressure after sclerotherapy, but not band ligation, in acute variceal bleeding in cirrhosis. Hepatology 2004;39: Avgerinos A, Nevens F, Raptis S, Fevery J. Early administration of somatostatin and efficacy of sclerotherapy in acute oesophageal variceal bleeds: the European Acute Bleeding Oesophageal Variceal Episodes (ABOVE) randomised trial. Lancet 1997;350: Cales P, Masliah C, Bernard B, et al. Early administration of vapreotide for variceal bleeding in patients with cirrhosis. French Club for the Study of Portal Hypertension. N Engl J Med 2001;344: Bañares R, Albillos A, Rincón D, et al. Endoscopic treatment versus endoscopic plus pharmacologic treatment for acute variceal bleeding: a meta-analysis. Hepatology 2002;35: Hubmann R, Bodlaj G, Czompo M, et al. The use of selfexpanding metal stents to treat acute esophageal variceal bleeding. Endoscopy 2006;38: