Portal hypertensive bleeding in cirrhosis: Risk stratification, diagnosis, and management: 2016 practice guidance by the American Association for the study of liver diseases

This guidance provides a data-supported approach to risk stratification, diagnosis, and management of patients with cirrhosis and portal hypertension (PH). A guidance document is different from a guideline. Guidelines are developed by a multidisciplinary panel of experts who rate the quality (level) of the evidence and the strength of each recommendation using the Grading of Recommendations Assessment, Development, and Evaluation system. A guidance document is developed by a panel of experts in the topic, and guidance statements, not recommendations, are put forward to help clinicians understand and implement the most recent evidence. This guidance focuses on PH, varices, and variceal hemorrhage (VH), and statements are based on the following: (1) review of the recent literature using PubMed, giving more weight to large, well-designed, prospective trials and well-performed meta-analyses; (2) several consensus conferences among experts; and (3) the authors' years of experience caring for patients with cirrhosis and varices. Management of ascites and encephalopathy is addressed in other documents. When little or no data exist from well-designed, prospective trials, emphasis is given to results from large series and reports from recognized experts. In this case, clinical studies needed to clarify that management are specified in a section on future research. Practice guidelines for the diagnosis and treatment of gastroesophageal VH were published in 2007, endorsed by the American Association for the Study of Liver Diseases (AASLD), American College of Gastroenterology, American Gastroenterological Association, and American Society of Gastrointestinal Endoscopy (ASGE).1 Since then, a number of randomized, controlled trials (RCTs) have advanced our approach to managing VH. Additionally, four international consensus conferences were held since then, where experts in the field evaluated the changes in pathophysiology, diagnosis, and management of varices and VH. These include two AASLD/European Association for the Study of the Liver single-topic conferences in 2007 (many of the recommendations from this conference were incorporated into the aforementioned guidelines)2 and in 2013, and two Baveno consensus conferences in 20103 and in 2015.4 In this updated practice guidance, recommendations derived from these consensus conferences were also incorporated, particularly those from the latest Baveno conference that took place in Baveno, Italy, in April 2015. Perhaps the most relevant change in these recommendations has been the recognition of the different stages of cirrhosis,5 so that recommendations are now focused on risk stratification and individualizing care for PH. Intended for use by health care providers, this guidance identifies preferred approaches to the diagnostic, therapeutic, and preventive aspects of care of patients with PH. As with other guidance documents, it is not intended to replace clinical judgment, but rather to provide general guidance applicable to the majority of patients. They are intended to be flexible, in contrast to formal treatment recommendations or standards of care, which are inflexible policies designed to be followed in every case. Clinical considerations may justify a course of action that differs from this guidance. Cirrhosis is a chronic condition with a high mortality. It constitutes the fifth-leading cause of adult deaths and ranks eighth in economic cost among the major illnesses.6 Cirrhosis is a heterogeneous disease that cannot be studied or managed as a single entity and is classified in two main prognostic stages: compensated and decompensated cirrhosis.5, 7 This classification depends on the presence or absence of clinically evident decompensating events (specifically ascites, VH, and encephalopathy [HE]), with a median survival in the compensated stage that exceeds 12 years, whereas it is only 1.8 years in patients who develop decompensation.8 The Child-Turcotte-Pugh (CTP) classification has been used to stratify patients with cirrhosis. Patients with cirrhosis belonging to the CTP-A class are compensated, whereas those in the CTP-B/C class are mostly decompensated. PH is the initial and main consequence of cirrhosis and is responsible for the majority of its complications. In fact, it has been shown that portal pressure (PP), determined by the hepatic venous pressure gradient (HVPG), is better than liver biopsy in predicting development of complications of cirrhosis in patients with chronic liver disease (CLD) without cirrhosis on liver biopsy.9 Therefore, a new entity denominated compensated advanced chronic liver disease (cACLD) has been proposed, emphasizing that PH may occur before a formal anatomical diagnosis of cirrhosis is established.4 This entity would encompass patients with cirrhosis and those with advanced liver fibrosis with PH (HVPG > 5 mm Hg). For ease of understanding, in the rest of this guidance, the entity of cACLD will be referred to as compensated cirrhosis (CC), both terms being interchangeable and acceptable by consensus.4 The stage of CC is asymptomatic, and it is the longest stage. Pathophysiological mechanisms are evolving at this stage, and therefore several substages are being recognized. Based on PP, patients with CC can be divided into those with mild PH (HVPG > 5 but < 10 mm Hg) and those with clinically significant portal hypertension (CSPH), defined by an HVPG ≥10 mm Hg. CSPH is associated with an increased risk of developing varices,10 overt clinical decompensation (ascites, VH, and HE),11 postsurgical decompensation,12 and hepatocellular carcinoma (HCC).13 This substaging is not only prognostically important, but, as mentioned below, the mechanisms maintaining PH at these substages are different, and therefore their therapeutic approach will be different. CSPH is present in approximately 50%-60% of patients with CC without gastroesophageal varices (GEV).10 Patients with GEV have, by definition, CSPH, because patients with GEV have an HVPG of at least 10 mm Hg.14, 15 Prognosis is worse in patients with CC with GEV compared to those without GEV.16, 17 Therefore, among patients with CSPH, two substages are recognized based on the absence or presence of GEV. It is important to recognize that although PH and its direct consequences (varices) form the bases of staging in CC, liver insufficiency, even at this stage, plays an important role, given that serum albumin and the Model for End-Stage Liver Disease (MELD) score are also independent predictors of decompensation.11 VH constitutes a decompensating event, but its mortality differs whether it presents as an isolated complication of cirrhosis (20% 5-year mortality) or whether it presents in association with other complications (over 80% 5-year mortality).8 Whereas in the past, emphasis had been placed on managing the direct complications of PH, varices, and VH, it is now clear that these complications cannot be considered in an isolated manner. Rather, they should be considered in the context of advances in the staging of cirrhosis and in the context of other complications of cirrhosis that may occur concomitant or subsequent to development of varices and VH.4 Stages of PH in cirrhosis are depicted in Fig. 1, and goals of therapy at each stage are shown in Table 1. Stages and substages of cirrhosis. The two main stages are the compensated and decompensated stages. The latter is characterized by the presence of clinically overt complications: ascites, VH, or HE. The compensated stage is the longest stage, and it is asymptomatic. There are at least two main substages of compensated cirrhosis with different prognostic and predominant pathophysiological mechanisms: patients with mild PH and those with CSPH. Patients in the latter stage are at risk of developing decompensation, particularly those who have GEV. The decompensated stage is much shorter and can rapidly progress to a stage of further decompensation in which renal failure (HRS) and liver failure (encephalopathy and jaundice) develop, leading to a high mortality. GEV are present in approximately 50% of patients with cirrhosis, but this depends on the clinical stage. In patients with CC, GEV are present in 30%-40%, whereas they can be present in up to 85% of patients with decompensated cirrhosis.18, 19 In patients with CC, varices develop at a rate of 7%-8% per year,10 and progression from small to large varices occurs at a rate of 10%-12% per year, with decompensated cirrhosis being an independent predictor of progression.20 VH occurs at a rate of around 10%-15% per year and depends on the severity of liver disease, size of varices, and presence of red wale marks (areas of thinning of the variceal wall).21, 22 Six-week mortality, which is now recognized as the primary endpoint to assess the impact of therapies for acute VH,4 ranges between 15% and 25%.23-25 Other factors associated with poor outcomes in patients with VH are the presence of bacterial infections and an HVPG >20 mm Hg, which is mostly observed in patients belonging to the CTP-C class.26, 27 If untreated, recurrent VH occurs in 60% of patients, usually within 1-2 years of index hemorrhage.28 Obesity and alcohol use are associated conditions of prognostic relevance in patients with cirrhosis, independent of etiology. Obesity has been shown to predict worsening of liver fibrosis, cirrhosis decompensation, and lack of regression of cirrhosis in patients with viral cirrhosis,29-31 whereas even moderate alcohol intake can lead to worsening PP and has been shown to worsen prognosis of hepatitis C virus (HCV)- and nonalcoholic steatohepatitis (NASH)-related cirrhosis.32, 33 Therefore, although beyond the scope of this guidance, weight loss and alcohol abstinence are important considerations in patients with cirrhosis. PP increases initially as a consequence of an increased intrahepatic resistance to portal flow attributed to structural mechanisms (e.g., fibrous tissue, vascular distortion from regenerative nodules, and microthrombi; Fig. 2). This “structural” component, which explains around 70% of the increased intrahepatic resistance, could be targeted by treating the etiology of cirrhosis, the use of antifibrotic agents, and even anticoagulants.34 However, at least one third of the increased intrahepatic resistance is attributed to an increased intrahepatic vascular tone, which, in turn, is attributed to endothelial dysfunction resulting mostly from reduced nitric oxide (NO) bioavailability.35 This “functional” component is amenable to vasodilators (such as nitrates, alpha-adrenergic antagonists, and angiotensin-2 blockers).36 These drugs should not be used alone, given that they also cause systemic vasodilatation, decrease arterial blood pressure, and may worsen sodium retention. A conceptually more appealing approach to ameliorate the functional component is to use drugs that will reduce PP by improving endothelial dysfunction, such as statins.37 An added advantage of these drugs is that, by causing intrahepatic vasodilatation, they may improve hepatic blood flow and liver function. Statins in particular also have antifibrotic properties.34 Pathogenesis of PH and sites of action of currently recommended therapies to reduce PP or obliterate varices. In cirrhosis, PP increases initially as a consequence of an increased intrahepatic resistance to portal flow attributed to structural mechanisms (e.g., fibrous tissue, regenerative nodules) and an increased intrahepatic vascular tone (functional component). One of the initial consequences of PH is the formation of portosystemic collaterals. Concomitant or even preceding development of collaterals, splanchnic vasodilatation occurs, leading to increased flow into the gut and into the portal venous system. Vasodilation leads to activation of neurohumoral and vasoconstrictive systems, sodium and water retention, increased blood volume, and increased cardiac output; that is, a hyperdynamic circulatory state that further increases portal venous inflow and PP. Additionally, activated vasoconstrictive systems to further contribute to intrahepatic vasoconstriction. Treatment of etiology, by ameliorating fibrosis/inflammation, target the mechanical component of the increased intrahepatic resistance. Vasodilators (like the α-adrenergic blocking effect of carvedilol) target its functional component (this is the site of action of statins). NSBBs (β2-adrenergic blocking effect), SMT, and VP act by causing splanchnic vasoconstriction, thereby reducing portal venous inflow. NSBBs also act by decreasing cardiac output (β1-adrenergic blocking effect). The TIPS connects the hypertensive portal vein with a normotensive hepatic vein, thereby bypassing the site of increased resistance. Varices can be obliterated either endoscopically (EVL or cyanoacrylate injection) or by an endovascular approach (BRTO). One of the initial consequences of PH is the formation of portosystemic collaterals, the most important being those that develop through the coronary and/or short gastric veins and constitute GEV. Although formation of collaterals had been assumed to be the result of dilatation of preexisting vascular channels, research studies have implicated a process of neoangiogenesis.38 Concomitant or even preceding the development of collaterals, splanchnic vasodilatation occurs, leading to increased flow into the gut and into the portal venous system. Therefore, even when portal flow is entirely diverted through collaterals, PH persists.39 Increased splanchnic NO production is the main factor that leads to vasodilatation and increased splanchnic blood flow. Hyperglucagonemia and neoangiogenesis further contribute to the increased splanchnic blood flow that maintains the portal hypertensive state.38 Vasodilation occurs not only in the splanchnic, but also in the systemic circulation (manifested clinically as arterial hypotension), leading to activation of neurohumoral and vasoconstrictive systems, sodium and water retention, increased blood volume, and increased cardiac output, that is, a hyperdynamic circulatory state that further increases portal venous inflow and PP. Additionally, norepinephrine, angiotensin-2, and antidiuretic hormone (activated neurohumoral and vasoconstrictive systems) further contribute to intrahepatic vasoconstriction. Drugs that act by causing splanchnic vasoconstriction, such as non-selective beta-blockers (NSBBs; propranolol, nadolol, and carvedilol), vasopressin (VP), and its analogue, terlipressin, and somatostatin (SMT) and its analogues (octreotide, vapreotide) are known to reduce PP and constitute the current mainstay in the treatment of varices and VH. Given that these drugs act by decreasing flow to the splanchnic circulation and liver, an improvement in liver function would not be expected. β-1 adrenergic blockade decreases portal flow through a decrease in cardiac output, and β-2 blockade decreases portal flow through splanchnic vasoconstriction by unopposed α-adrenergic activity. Therefore, it is essential that beta-blockers used in the treatment of PH be nonselective. Importantly, the effect of NSBBs in decreasing flow is more related to their β-2 blocking effect rather than to their β-1 effect40 and explains the lack of correlation between decreases in PP and decreases in heart rate.41 Carvedilol, an NSBB with anti-α1 adrenergic (vasodilator) activity, acts as an NSBB decreasing portal flow, but also acts as a vasodilator (intrahepatic circulation). HVPG response is greater with carvedilol than with propranolol or nadolol, but, given its vasodilatory properties, carvedilol is associated with a greater decrease in mean arterial pressure (MAP).42 It has been recently shown that patients with mild PH (HVPG > 5 but < 10 mm Hg) have a normal cardiac index (i.e., they have not yet developed the hyperdynamic circulatory state), whereas those with CSPH, especially if varices are present, have already developed a hyperdynamic state. Accordingly, response to NSBB in patients with mild PH is suboptimal compared to that of those with CSPH,43 indicating that there is no role for NSBB in the setting of mild PH. Endoscopic variceal ligation (EVL) is a local therapy that consists of placing rubber bands around esophageal varices (EV) in repeated sessions until they become obliterated. Because it is a local therapy that has no effect on PH, recurrence of varices is the rule, and patients require indefinite endoscopic monitoring. Local therapies for management of gastric (mostly cardiofundal) varices consist of the (1) transendoscopic obturation by injection of cyanoacrylate glue into the varices or (2) transvenous obliteration by instillment of sclerosants and/or liquid embolic agents into a gastro-/splenorenal collateral through the left renal vein aided by balloon occlusion, that is, balloon occluded retrograde transvenous obliteration (BRTO).44 In patients with decompensated cirrhosis, placement of the transjugular intrahepatic portosystemic shunt (TIPS) by interventional radiological techniques that consist of connecting the hypertensive portal vein with a normotensive hepatic vein by a coated stent causes a significant decrease, and even normalization, of PP. Therefore, in patients with functional TIPS stents, there is no need for other therapies for PH (e.g., NSBB, EVL). PH is defined as a portal pressure gradient (the difference in pressure between the portal vein and the hepatic veins) greater than 5 mm Hg. The best method to assess PP is through the catheterization of the hepatic vein with determination, through a balloon catheter, of the HVPG, which is the difference between the wedged (or occluded) hepatic venous pressure and the free hepatic venous pressure.45 Normal HVPG is 3-5 mm Hg. It should be underlined that the wedged (occluded) pressure (and, consequently, the HVPG) is a measure of sinusoidal pressure and does not provide useful data in prehepatic or presinusoidal PH (Table 2). An HVPG over 5 mm Hg identifies patients with cACLD/CC secondary to conditions associated with sinusoidal hypertension (Table 2). As mentioned above, PH is further defined as mild PH (HVPG > 5 but < 10 mm Hg) and as CSPH (HVPG ≥ 10 mm Hg). Above this threshold of 10 mm Hg, all the complications of PH are more likely to appear (varices, clinical decompensation). In patients with GEV (who, by definition, have CSPH), an HVPG > 12 mm Hg identifies bleeding risk, mostly because there is clear evidence that shows that reducing the HVPG to levels of 12 mm Hg or below is associated with protection from variceal hemorrhage (VH).28 An HVPG > 16 mm Hg indicates a higher risk of death.46 As mentioned previously, an HVPG ≥20 mm Hg predicts failure to control bleeding, early rebleeding, and death during acute VH,27, 47 and in patients with cirrhosis awaiting liver transplantation, each 1-mm-Hg increase in HVPG predicts a 3% increase in the risk of death in a median follow-up of 19 months.48 Despite the crucial role of HVPG in the determination of CSPH and other outcomes, HVPG measurements require specific expertise, are invasive, relatively expensive, and not available in all centers. Therefore, HVPG measurements are not considered standard of care for every patient with cirrhosis, particularly because noninvasive or surrogate indicators are increasingly utilized at most centers. In a step-wise diagnostic approach, specific signs of PH should be first looked for on physical examination. They include spider nevi or visible abdominal portosystemic collaterals. The absence of physical signs cannot be used to rule out CSPH. Among laboratory data, a low platelet count is the most common laboratory sign of PH; it correlates slightly with HVPG and with the presence of GEV. However, taken alone, it is not accurate enough to either diagnose or exclude CSPH or GEV. On the other hand, the combination of platelet count with other unrelated noninvasive tests (NITs) improves the noninvasive diagnosis of CSPH.49 Ultrasound provides safe and inexpensive imaging evidence of morphological abnormalities associated with cirrhosis and PH. The presence of portocollateral circulation on ultrasound, computed tomography (CT), or magnetic resonance imaging (recanalized paraumbilical vein, spontaneous splenorenal circulation, and dilated left and short gastric veins) or the finding of a reversal of flow within the portal system is 100% specific for CSPH50 and is sufficient to diagnose CSPH. Several other sonographic signs of PH have been described, such as dilatation of portal vein and the reduction of portal vein velocity (or their combination as congestion index of the portal vein).51, 52 Although splenomegaly taken alone is a sensitive, but nonspecific, sign of PH, the size of the spleen should be routinely reported, because, when combined with platelet count and liver stiffness, it provides accurate data on the presence of CSPH/varices.49, 53 The ability to assess liver stiffness (LS), a physical property of liver tissue influenced by the amount of liver fibrosis content, has represented a major advance in this field. LS by transient elastography (TE; FibroScan) has proved very accurate for discriminating patients with and without CSPH, with a mean area under the receiver operating curve (AUROC) of 0.93 in a recent meta-analysis (based on five studies including 420 patients)54 and can be currently considered the backbone of the noninvasive diagnosis of PH. However, most of the data have been obtained in patients with untreated viral cirrhosis and alcoholic cirrhosis. Data regarding other etiologies and data in patients who have eliminated HCV require further investigation. Most studies have shown that the best LS cutoff to detect CSPH is >20-25 kilopascals (kPa), with a diagnostic accuracy over 90%.55, 56 In a prospective study, HVPG ≥10 mm Hg and LS ≥21 kPa were equally effective in predicting decompensation.57 In a large study, an LSPS (liver stiffness [in kPa] × spleen size [in cm]/platelet count [in number/mm3] score) > 2.06 was 90% specific in ruling in CSPH with a positive predictive value of >90%.49 Importantly, these measures/scores have to be considered in the context of clinical parameters. In this sense, a recent prospective study described a sequential screening-diagnostic strategy based on LS measurements assessed in the context of the presence of any ultrasound abnormality and/or a platelet count <150,000/mm3 and identified the subgroup of patients with CC in whom CSPH would be more likely.56 Spleen stiffness (SS) measurement by TE has been recently proposed as a novel parameter more tightly related to PH, with promising results.58, 59 In fact, SS > 54 kPa was better than LS and similar to HVPG in predicting first clinical decompensation in one study. However, SS cannot be measured by TE without a separate ultrasound exam and cannot be measured if the spleen is not significantly enlarged. Therefore, SS measurements by TE cannot be recommended in clinical practice. Newer sonoelastographic methods allow direct visualization of the liver and spleen, facilitating SS measurement. Evidence is still limited, but point shear wave elastography (SWE; ARFI; Siemens, Germany)60 and two-dimensional real-time SWE (Aixplorer; Supersonic Imagine, France)61, 62 show promising results with higher applicability and similar accuracy in the prediction of CSPH. Magnetic resonance elastography (MRE) is an emerging technique that provides data on LS and SS of much larger areas of the liver and spleen compared to ultrasound-based techniques. Although MRE has been shown to be accurate in the staging of liver fibrosis,63 data regarding its diagnostic performance in the diagnosis of CSPH are still very limited, with one study showing that LS determined by MRE predicted onset of clinical decompensation in patients with CC.64 More studies are needed in this field. Determining the presence and size of varices and presence of red wale marks requires esophagogastroduodenoscopy (EGD), an invasive and expensive procedure that is not free of risks. Many studies have looked for noninvasive ways of determining the presence of high-risk varices (medium/large varices, i.e., those requiring prophylactic therapy) so as to circumvent the need for screening endoscopy. The discriminative accuracy of NITs in predicting the presence of any GEV is limited (AUROC between 0.71 and 0.84),55 and the use of NITs to diagnose GEV is not recommended. However, NITs are accurate to rule out high-risk varices in patients with CC. In particular, LS combined with platelet count correctly identifies patients at very low risk (<5%) of having high-risk varices.56, 65 These data have been obtained mostly from patients with untreated viral cirrhosis. Data in patients with NASH cirrhosis, cholestatic liver disease, and in those with HCV-related cirrhosis achieving sustained virological response (SVR) are needed. By consensus among experts, and after review of the literature, it was proposed that patients with CC with LS <20 kPa (determined by TE) and a platelet count >150,000/mm3 were very unlikely to have high-risk varices (<5%), and endoscopy could be safely avoided in them.4 Unpublished studies have validated these cutoffs and report that 20%-25% of EGDs can be circumvented. In patients with cirrhosis secondary to hepatitis B, an LSPS (liver stiffness [in kPa] × spleen size [in cm]/platelet count [in number/mm3] score) < 3.5 was accurate in ruling out high-risk varices.53 Whether this cutoff can be applied to patients with cirrhosis attributed to other etiologies remains to be established. Because measurements of SS are more feasible with ARFI, irrespective of spleen size, this technology is a promising tool in diagnosing and ruling out high-risk varices and compares favorably to other NITs in Asian studies60; however, data in European and American patients are lacking. Patients without evidence of CSPH should be monitored to identify onset of the syndrome. Even if data on this specific aspect are lacking, data from published abstracts suggest that LS and platelet count monitoring could be useful. The appearance of new portosystemic collaterals during follow-up has been shown to be associated with variceal formation and growth,66 as is progressive spleen enlargement.67 Therefore, when performing screening for HCC, imaging evidence of worsening PH should be specifically sought. Patients without varices on screening endoscopy constitute an area of uncertainty, given that their natural history has not yet been fully elucidated, particularly with the emergence of therapies that eliminate the etiologic agent.68 Experts' opinion suggests that if liver injury is ongoing (e.g., active drinking in alcoholics and lack of SVR in HCV) and/or cofactors of disease are present (e.g., obesity, alcohol), surveillance endoscopy should be repeated at 2-year intervals. Otherwise, in the absence of ongoing injury, 3-year intervals are considered sufficient.4 Although probably reasonable, there are no data to support discontinuing screening endoscopies if several of them are negative for varices. In patients with small varices on screening endoscopy who are not candidates for primary prophylaxis (see below), repeat endoscopy is recommended. It has been suggested that if the liver injury is ongoing (e.g., active drinking in alcoholics and lack of SVR in HCV) and/or cofactors of disease are present (e.g., obesity), surveillance endoscopy should be repeated at yearly intervals. Otherwise, in the absence of ongoing injury, 2-year intervals are considered sufficient.4 Because development of decompensation could indicate worsening of PH and liver dysfunction with a higher incidence of cirrhosis, patients with no or small varices on screening endoscopy should have a repeat endoscopy performed when and if decompensation develops. Changes in HVPG, spontaneous or during pharmacological therapy, have been shown to be predictive of outcomes. In patients with a history of VH, a decrease in HVPG to less than 12 mm Hg or a decrease greater than 20% from baseline significantly reduces the risk of recurrent hemorrhage, ascites, encephalopathy, and death.69, 70 In patients with CC, reductions in HVPG >10% from baseline have been associated with a reduction in development of varices,10 first VH, and death.71 Recent studies show that the need for separate HVPG procedures to assess response to therapy can be obviated by assessing the acute hemodynamic response to intravenous propranolol (0.15 mg/kg) during a single procedure, but this requires further investigation.71, 72 Unfortunately, there have been no NITs (e.g., Doppler, LS) that correlate with changes in HVPG. As mentioned above, therapy of varices and VH should be stratified according to the different clinical stages of cirrhosis and PH that are shown in Table 1. The objective of therapy for patients at an early stage is to prevent the development of later stages. Varices and VH should be managed in the context of the presence (or absence) of other complications of cirrhosis/PH (e.g., ascites, encephalopathy), and therefore the status (compensated or decompensated) of the patient with varices/VH should be always considered in the selection of the different therapies. In the compensated patient, the ultimate objective is to prevent decompensation; that is, the objective is not only to prevent varices or VH, but also to prevent the other complications of cirrhosis. In addition to specific therapies that will be outlined below, in the compensated patient, every effort should be taken to eliminate the etiologic agent and to correct associated aggravating conditions, such as alcohol, obesity, and drug-induced liver injury, given that these measures, in themselves, can decrease portal pressure and reduce the risk of decompensation. This stage is defined by an HVPG >5 but < 10 mm Hg. Patients in this stage do not have varices or ot