Potential conflict of interest: Dr. Bosch consults for and received grants from Chiasma and Microtech. He also consults for Gilead and Norgine.
Cirrhosis is a major health problem, being the 5th cause of death in the U.K. and 12th in the U.S., but 4th in the 45 to 54 age group. Until recently cirrhosis was considered a single and terminal disease stage, with an inevitably poor prognosis. However, it is now clear that 1-year mortality can range from 1% in early cirrhosis to 57% in decompensated disease. As the only treatment for advanced cirrhosis is liver transplantation, what is urgently needed is strategies to prevent transition to decompensated stages. The evidence we present in this review clearly demonstrates that management of patients with cirrhosis should change from an expectant algorithm that treats complications as they occur, to preventing the advent of all complications while in the compensated phase. This requires maintaining patients in an asymptomatic phase and not significantly affecting their quality of life with minimal impairment due to the therapies themselves. This could be achieved with lifestyle changes and combinations of already licensed and low-cost drugs, similar to the paradigm of treating risk factors for cardiovascular disease. The drugs are propranolol, simvastatin, norfloxacin, and warfarin, which in combination would cost £128/patient annually—equivalent to U.S. $196/year. This treatment strategy requires randomized controlled trials to establish improvements in outcomes. In the 21st century, cirrhosis should be regarded as a potentially treatable disease with currently available and inexpensive therapies. (HEPATOLOGY 2012;56:1983–1992)
Cirrhosis is the severest form of liver disease and is the common outcome of different mechanisms and etiologies of chronic liver injury. It is the 5th cause of death in the U.K., which is still increasing,1 and 12th in the U.S., but 4th in the 45 to 54 age group,2 such that cirrhosis is a major public health problem. The only definitive treatment is liver transplantation, but the need for organs greatly exceeds the supply, so that deaths on the waiting list continuously grow.
Until a decade ago cirrhosis was considered a single and terminal disease stage carrying a uniformly bad prognosis, progressing from compensated to decompensated cirrhosis, heralded by the development of ascites, variceal bleeding, jaundice, or encephalopathy.3 More recently, a new classification has been proposed in order to improve prediction of prognosis, based on clinical criteria with distinctly different survival likelihood's between four stages4: stage 1 (compensated with no esophageal varices) with an estimated 1% mortality rate per year, and stages 2 (varices), 3 (ascites), and 4 (gastrointestinal bleeding) with an annual mortality rate that increases from 3.4% to 20% to 57%, respectively.4 A fifth stage due to infections5 and renal failure6 with a 67% 1-year mortality rate further refines this classification.
The high morbidity and mortality of cirrhosis led us to highlight the need for preventive therapies in such patients, in order to delay the advent of complications, using the paradigm of prevention of cardiovascular diseases,7 according to which interventions are offered to individuals with a 10% 10-year risk of adverse events.8 In cirrhosis, combination therapies with already licensed, inexpensive, and relatively safe drugs could be given in a highly cost-effective way.7 Preventive strategies can be considered in three groups, starting from the least invasive: lifestyle changes, specific treatments according to disease etiology, and nonspecific therapies. In this review we elaborate on the likely effectiveness and evaluation of this approach using new data.
The features of metabolic syndrome are pathophysiologically linked with nonalcoholic liver disease, but also impact the progression of chronic liver disease in general. Insulin resistance is associated with more severe fibrosis in patients with chronic hepatitis C (CHC), is linked with obesity in reducing the likelihood of antiviral response,9 and predicts the occurrence of hepatocellular carcinoma (HCC) and liver-related mortality.10 Diabetes confers an increased hazard ratio of 3.28 for the development of HCC in CHC.11 Indeed, diabetes, insulin resistance, and obesity were associated with a worse overall and also liver-related mortality in 2,061 patients with chronic liver disease of different etiologies.12 Metabolic syndrome is associated with more severe fibrosis in nonalcoholic fatty liver disease and chronic viral hepatitis,13 and is an independent risk factor for cirrhosis in chronic hepatitis B (CHB).14 Metabolic syndrome was also significantly associated with the development of both HCC and intrahepatic cholangiocarcinoma in a study of 4,392 patients with such tumors.15 This strengthens the reported independent association between diabetes and HCC16; indeed, patients with metabolic syndrome may develop HCC without significant fibrosis.17 More important, in a prospectively followed well-defined cohort, obesity had a deleterious effect on the evolution of compensated cirrhosis of all etiologies, independent of portal pressure and liver function, with clinical decompensation occurring in 43% and 31% of obese and overweight patients, respectively, compared with 15% of patients with normal body mass index (BMI) after a median follow-up of 59 months.18 Despite the well-established association of obesity with fibrosis progression and development of HCC, there are few trials and little hard evidence of the effects of weight loss in patients with chronic liver disease. Even in nonalcoholic steatohepatitis (NASH), existing trials were poorly conducted with too few patients and insufficient surrogate markers of efficacy as endpoints.19 Indeed, the effect of treating obesity and/or insulin resistance in patients with cirrhosis has not been addressed in randomized controlled trials (RCTs) and could prove to be a valid therapeutic target. Such trials should only include obese patients with compensated cirrhosis, as some experts do not advise weight loss in decompensated disease.20
Alcohol abstinence is of cardinal importance, especially in patients with alcoholic liver disease. Even moderate alcohol consumption worsens portal hypertension in patients with alcohol-induced cirrhosis,21 whereas continued heavy drinking is associated with increased mortality.22
Cigarette smoking has specific deleterious effects, and is associated with more severe fibrosis in patients with CHC,23 primary biliary cirrhosis,24 and NASH.25 Daily cannabis use is associated with more severe steatosis and fibrosis in CHC.26 There may also be an increased risk of HCC among heavy smokers with CHB.27 Therefore, smoking cessation strategies may have additional benefits in patients with chronic liver disease, which should be evaluated in cirrhosis.
Coffee has antioxidant effects in the liver, in part by up-regulating glucuronidation.28 In a large population study, among 5,944 individuals with high risk of liver injury coffee consumption was associated with a lower risk of elevated alanine aminotransferase (ALT).29 Coffee consumption was associated with a significant reduction in risk of fibrosis in patients with NASH.30 In CHC, consumption of more than three cups of coffee per day was an independent predictor of sustained virological response (SVR) to antiviral treatment31 among 885 treated patients and reduced histological activity.32 In patients with various etiologies of chronic liver disease, consumption of more than two cups per day was associated with a reduced risk of HCC33 and of advanced fibrosis.34 In patients with cirrhosis, dark chocolate blunted the postprandial increase in hepatic vein pressure gradient (HVPG) without changing postprandial hyperemia, suggesting an improvement of intrahepatic endothelial dysfunction.35 A similar blunting of postprandial HVPG was noted after intravenous administration of vitamin C.36 These nutritional components warrant further evaluation in long-term epidemiological follow-up, but it is unlikely that RCTs will be performed, as no pharmaceutical product is involved. In the absence of robust evidence, patients can be given choices that they may wish to follow despite the lack of high-quality evidence.
Maintenance of nutrition is important in cirrhosis; a late carbohydrate meal at night reduces gluconeogenesis from protein breakdown, improves nutrition as shown in an RCT, and is an easy lifestyle change.37
Vaccination against hepatitis A and B, flu, and pneumococcus should be offered as early as possible, ideally in stage I compensated cirrhosis, because the antigenic response becomes weaker as cirrhosis progresses.38 However, vaccination policies are poorly adhered to: a recent population survey in the U.S. revealed that only 20% and 32% of patients with chronic liver disease received vaccination for hepatitis A and B, respectively.39
Advice to patients with cirrhosis should not be limited to avoiding alcoholic beverages. All patients with chronic liver disease should be vaccinated and counseled about weight reduction if they are obese and have compensated cirrhosis. Information on the potential benefit from smoking cessation and drinking at least three cups of coffee per day should be given, as these might be of value as evidenced by epidemiological studies in large populations. The therapeutic potential of intensive lifestyle counseling and behavioral therapy ideally should be evaluated in future trials, although we acknowledge they may be difficult to carry out.19
Antiviral treatment has changed the natural history of CHB, even in patients with established cirrhosis. Lamivudine significantly reduced the incidence of decompensation and HCC in patients with CHB-related cirrhosis40; moreover, HVPG decreased by at least 20% in 76% of patients with CHB-related cirrhosis after a year of lamivudine therapy.41 New antiviral drugs, namely entecavir and tenofovir, have reduced the rate of viral resistance and therefore all patients with CHB cirrhosis should be on long-term antiviral treatment.42
In patients with CHC-related cirrhosis, antiviral treatment is more difficult and is currently offered in compensated disease. Although response rates are relatively low, the use of protease inhibitors, namely telapravir and boceprevir, greatly increases the rate of SVR.43, 44 Elimination of interferon in future regimens should allow treatment in decompensated cirrhosis. Only the attainment of SVR improves outcomes and reduces the likelihood of developing varices,45 reduces clinical decompensation, and improves survival.46 Statin therapy and good diabetes control both improve the chances of SVR.47 Low vitamin D levels have also been associated with low response to antiviral therapy.48 Therefore, in this difficult to treat group weight loss counseling, increasing coffee consumption, supplementation with vitamin D, achieving good diabetic control, and use of statins could optimize outcomes with antiviral treatment.
The diagnosis of cirrhosis is a watershed in patients with chronic liver disease, as it heralds the initiation of screening for esophageal varices and HCC. Until recently, the only established preventative treatment was nonselective beta-blockers (NSBBs) for primary prevention of bleeding in patients with varices.49 However, evidence is emerging that established drugs, such as NSBBs, statins, antibiotics, and anticoagulants, may have expanded indications in patients with cirrhosis irrespective of etiology and prevent or delay the advent of complications.7
Portal hypertension is the pathophysiological consequence of cirrhosis and is due to increased intrahepatic resistance and increased portal flow. The increased resistance is due to the structural distortion of cirrhosis but also to increased intrahepatic vascular tone and accompanying endothelial dysfunction.50 The increased intrahepatic vascular tone is responsible for 30% of the magnitude of portal hypertension and is principally due to changes in endothelial vascular factors resulting in vasoconstriction, among which the most important is a decrease in nitric oxide production in the intrahepatic circulation.51 However, there is an overproduction of nitric oxide in the splanchnic circulation that constitutes a paradox in cirrhosis, and is responsible for the systemic vasodilatation and hyperdynamic circulation that are pathogenetic in the development of ascites and renal dysfunction.51 Varices form when HVPG ≥10 mmHg and other complications of cirrhosis also occur once this threshold is reached.52
NSBBs are highly effective in both primary and secondary prevention of variceal bleeding by reducing HVPG.49, 50 A single double-blind RCT of timolol versus placebo showed that the development of varices does not occur if HVPG remains <10 mmHg, although a therapeutic benefit of timolol in preprimary prophylaxis could not be demonstrated, possibly due to sample size problems.53 Carvedilol is an NSBB with weak vasodilating properties owing to alpha-1 blockade, which may act on the intrahepatic circulation. It is more effective than propranolol in reducing HVPG54 and hepatic vascular tone55; in one randomized study it was better than band ligation for primary prevention of variceal bleeding.56
Besides its effects in lowering HVPG, NSBB markedly decreased blood flow through gastroesophageal collaterals and varices in patients with cirrhosis, as assessed from measurements of azygous blood flow,57 an effect more pronounced that the decrease in portal blood flow, which explains why in some patients NSBB decrease variceal pressure more than HVPG, and in others solely variceal pressure.58 A substantial long-term decrease in variceal pressure (>20% of baseline), analogous to HVPG, is associated with a marked decreased risk of bleeding.59 The combination of HVPG and variceal pressure measurements under NSBB increases the sensitivity and specificity of the pressure changes in identifying patients who will not bleed over a 3-year period (5% versus 75% in nonresponders using both measurements), thus markedly decreasing the gray indeterminate zone for prognostication.60
However, NSBBs may have additional beneficial effects such as reducing the incidence of bacterial translocation and spontaneous bacterial peritonitis (SBP), which were not anticipated when first used. Bacterial translocation is a critical event in the natural history of cirrhosis and consists of the translocation of gut bacteria and bacterial products to the mesenteric lymph nodes. This translocation is the net result of increased intestinal permeability, bowel dysmotility, and bacterial overgrowth that occur in cirrhosis.61 A meta-analysis demonstrated a significant effect of NSBBs in reducing the incidence of SBP in patients with ascites.62 This “additive” effect of NSBBs over and above their effect in reducing portal pressure is probably due to increased bowel motility and thus reduced bacterial translocation. Indeed, in a rat model of cirrhosis splanchnic sympathectomy resulted in reduced bacterial translocation.63 Long-term follow up of patients randomized to endoscopic band ligation or NSBBs for secondary prophylaxis of variceal bleeding showed increased survival in the NSBB group despite a higher rate of rebleeding, demonstrating an additive survival benefit of NSBBs.64
Recently, evidence of a potential antiangiogenic effect of NSBBs has also emerged, which could slow down the development of collaterals. Propranolol significantly reduces the size of severe hemangiomas of infancy; proposed mechanisms include decreased expression of vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) genes, and the triggering of apoptosis of capillary endothelial cells.65 All of the above suggest a pleotropic action of NSBBs over and above HVPG reduction, similar to that of statins in preventing cardiovascular events.
Indeed, in patients with compensated cirrhosis and large varices treated with NSBBs, an HVPG decrease greater than 10% significantly reduced the risk of developing ascites during a mean 53 months of follow-up.66 There is a single report of reduced survival in patients with refractory ascites receiving NSBB,67 but this is unsubstantiated by others and may be related to dose of NSBBs and more advanced disease stage.68 The same group found that NSBBs caused paracentesis-induced circulatory dysfunction in patients with refractory ascites,69 but this had minimal clinical impact. Of note, these observations pertain solely to patients with advanced cirrhosis, for whom a preventive strategy is not warranted, as liver transplantation is indicated.
In patients with cirrhosis there is increased susceptibility to infection, related to the degree of liver dysfunction.70 Once an infection occurs, mortality is increased 4-fold, and of those infected 30% will die within a month and in those surviving, 30% within a year.5 In advanced cirrhosis with ascites, translocation of bacteria and bacterial products occurs, especially endogenous endotoxin from the gut, which is not adequately cleared by the sinusoidal Kupffer cells.71 Kupffer cells undergo inflammatory activation in parallel with the development of portal hypertension.72 Endotoxin also binds to Toll-like receptor 4 (TLR4) and activates quiescent hepatic stellate cells, thus promoting fibrogenesis.73 Furthermore, liver endothelial cells express TLR4, which is involved with regulating angiogenesis.74 Neutrophil dysfunction is invariably present, in both stable and decompensated patients75; infection causes an excessive proinflammatory response of mononuclear cells.76 Bacterial translocation in itself, without overt infection, is associated with aggravation of peripheral vasodilatation and worsening of intrahepatic endothelial dysfunction.77 Therefore, endotoxin and probably other bacterial products may cause a vicious cycle of worsening hemodynamics, enhanced angiogenesis and fibrogenesis, and endothelial dysfunction, accelerating deterioration of the clinical course.
Oral quinolones are the standard of care for secondary prevention of spontaneous bacterial peritonitis78 and are used for primary prophylaxis in selected cases.79 Norfloxacin compared to placebo ameliorated the hyperdynamic circulation of cirrhosis without detriment to the splanchnic and renal hemodynamics, and slightly reduces HVPG.80 Genetic factors may predispose to SBP, so new assays may be used in the future to better define candidates for primary antibiotic prophylaxis.81
Although quinolones are established therapies, there is increasing interest in rifaximin, an antibiotic with minimal systemic absorption and reduced resistance. In RCTs rifaximin was shown to be an effective preventive therapy for both overt82 and minimal hepatic encephalopathy83 and to improve health-related quality of life,84 probably through gut decontamination and subsequent reduction in bacterial products and ammonia concentrations. This effect is not limited to preventing encephalopathy but may extend to preventing infection. Rifaximin significantly decreased HVPG after a 1-month administration in 30 patients with decompensated alcoholic cirrhosis.85 Thus, there is a rationale for RCTs of rifaximin for primary or secondary prevention of SBP and hepatic encephalopathy, and antibiotics combined with NSBBs for the prevention of variceal bleeding. Emergence of resistance to rifaximin or norfloxacin is a concern and can only be evaluated in an RCT with long follow-up. Another potential antibiotic agent is colistin, which disrupts lipopolysaccharide (LPS) in the gut, and was found to reduce the blood levels of nitrites, nitrates, and endotoxin in 15 patients.86
Another approach to modify bacterial translocation is to change the gut flora using probiotics. In one small study, probiotic administration for 4 weeks restored neutrophil phagocytic capacity in eight patients with cirrhosis, possibly by changing interleukin (IL)-10 secretion and TLR4 expression.87 Probiotic yogurt administration in another RCT significantly reversed minimal hepatic encephalopathy.88 However, all probiotic studies to date have included small numbers of patients with different etiologies and stages of liver disease and variable probiotic regimens. They require well-conducted adequately powered RCTs, as there could be unexpected effects.89
Statins reduce cardiovascular events over and above the lowering of LDL cholesterol, and are used for primary prevention of these events. In cirrhosis, statins improve portal hypertension by reducing the intrahepatic vascular resistance without affecting systemic hemodynamics.50 In a rat model of cirrhosis, simvastatin selectively increased nitric oxide (NO) availability in the liver circulation.90 In 13 patients with cirrhosis, simvastatin increased the hepatosplanchnic output of NO products and decreased hepatic resistance.91 The above observations were confirmed in an RCT of 59 patients, in which simvastatin (initially 20 mg/day subsequently increased to 40 mg/day) given for 1 month significantly reduced HVPG and, most important, this effect was additive to that of NSBBs and was associated with a marked improvement in liver clearances, indicating a potential for improved liver function.92 Ongoing long-term follow-up RCTs will confirm whether statins decrease the incidence of bleeding and other complications.
The effects of statins are not confined only to portal hypertension. In a large population of 6,515 patients with diabetes, statin use was associated with a significant reduction in the risk of HCC, having an adjusted odds ratio of 0.74.93
Importantly, statins have been shown to be safe in patients with liver disease with a similar complication rate to patients without liver disease.94, 95
Recently, the understanding of coagulopathy in cirrhosis has changed. Cirrhosis is no longer considered a hypocoagulable state, as both pro- and anticoagulant proteins are reduced, and the hemostatic balance is maintained in many cases, but set to a lower point, with a lower threshold for tipping toward thrombosis or bleeding.96 Thus, thrombin generation in vitro in stable cirrhosis is normal97 or increased,98 provided there are a sufficient number of platelets.99 Indeed, population studies show that cirrhosis does not protect patients from deep vein thrombosis,100 and the risk for venous thromboembolism is greater in patients under 45 years of age compared to noncirrhotic controls.101
Moreover, recent evidence suggests an active role of coagulation in the pathogenesis of liver fibrosis.102 Thrombin, a key protease in the coagulation cascade, promotes the profibrogenic role of activated hepatic stellate cells and in animal models is implicated in fibrosis progression.102 Thrombotic risk factors were independently associated with more severe fibrosis in cohorts of patients with viral hepatitis103 and nonalcoholic fatty liver disease.104 In addition, the “parenchymal extinction” theory suggests that intrahepatic microvascular occlusion is part of the natural history of cirrhosis and contributes to the aggravation of fibrosis and worsening of liver function.105 An RCT evaluating warfarin in precirrhotic stages is currently under way (Table 1).
Table 1. Preventative Trials in Cirrhosis as Listed in clinicaltrials.gov
Diet and exercise
Portal and systemic hemodynamics
Pre-primary prophylaxis of varices
Recurrence of variceal bleeding
Simvastatin + NSBB
Improvement of hemodynamic response
Change in HVPG
Change in HVPG
Complications of portal hypertension
Carvedilol vs. propranolol
Change in HVPG
Currently, the only indication for anticoagulation in cirrhosis, used in some centers, is in patients with portal vein thrombosis on the transplant waiting list in order to prevent extension of the thrombus106: warfarin significantly increased portal vein recanalization and was not associated with increased bleeding risk in such patients.107 In an RCT, preemptive enoxaparin administration prevented the development of portal vein thrombosis in patients awaiting liver transplantation with no associated side effects; furthermore, fewer decompensating events were documented in the treatment group, thus implying an additional benefit.108 The choice of anticoagulant raises further interest, given the effectiveness of the new direct thrombin and anti-Xa inhibitors.109 These drugs are administered orally, have fixed dosing, and do not require measurement of blood coagulation assays, thus addressing the drawbacks of warfarin and low molecular weight heparin. Given the above, there is a rationale for setting up an RCT of warfarin or low-molecular weight heparin in early stage cirrhosis, to prevent further fibrosis progression and advent of complications after eradicating esophageal varices. The presence of portal hypertensive gastropathy is still a concern in such patients, as use of anticoagulants might aggravate occult blood loss and anemia, although neither were reported in the randomized trial in patients with late-stage cirrhosis.108
Future Therapies and Drugs to Avoid in Cirrhosis
Apart from the above-mentioned drugs, which are relatively safe and already licensed, other attractive candidates as future therapies have also emerged. Pentoxyfylline, a tumor necrosis factor alpha (TNF-α) inhibitor with an excellent safety profile, was evaluated in an RCT including patients with Child-Pugh C cirrhosis.110 Short-term mortality was not improved but the risk of renal dysfunction and other complications was reduced. Although there is a theoretical risk of bacterial infections and sepsis from TNF inhibition, this was not observed. Further trials should include patients with earlier stages of cirrhosis and longer follow-up, with close monitoring for infections.
Angiotensin receptor blockers (ARB) and angiotensin converting enzyme (ACE) inhibitors are effective in reducing HVPG in patients with Child-Pugh A cirrhosis and are safe, but not in more severe stages, due to reduction in glomerular filtration rate (GFR), so they should be evaluated as an adjunct to NSBBs or as an alternative if there is intolerance.111
Metformin for controlling diabetes and treating insulin resistance could potentially decrease the risk of HCC. In a prospective cohort study of 100 patients with type 2 diabetes and CHC, treating diabetes with metformin was independently associated with a reduced incidence of HCC and liver-related death or transplantation.112 A similar effect in lowering the incidence of HCC with antidiabetic therapy was observed in a population study of nearly 20.000 patients with diabetes.113 There is a risk of lactic acidosis, but this is clinically relevant only when GFR <30 mL/min,114 which is not seen in compensated cirrhosis.
Proton-pump inhibitors should be used cautiously and only if indicated due to the higher risk of SBP and Clostridium difficile infection.115, 116 Nonsteroidal antiinflammatory drugs in all stages of cirrhosis and ACE inhibitors in patients with ascites should also be avoided as they increase the incidence of renal failure.117
Cirrhosis can no longer be considered a single disease stage; progression from one stage to another reflects worsening prognosis with significant differences in 1-year survival.4-6 The evidence presented clearly demonstrates that management of patients with cirrhosis should change from an expectant algorithm that treats complications as they occur, to preventing the advent of all complications while in the compensated phase. This requires maintaining patients in an asymptomatic phase and not significantly affecting their quality of life with minimal impairment due to the therapies themselves. Relatively safe, effective, and already licensed drugs, which even in combination are inexpensive, are readily available to evaluate in this setting.7 The new treatment paradigm has implications for early diagnosis of cirrhosis in primary care or in screening. The potential use of noninvasive markers is promising, once they are sufficiently validated and threshold cutoffs are established.118, 119
This treatment paradigm requires RCTs to prove efficacy, evaluating combined therapies as is already the case for preventing cardiovascular disease, but also to evaluate safety, particularly when drugs are used in combination and for the long term. We have calculated the annual cost of combination therapy with propranolol, simvastatin, norfloxacin, and warfarin in the U.K. as £128/annum/patient—equivalent to U.S. $196/year or €154 /year.7 The design of these randomized trials should be based on the existing clinical stages of cirrhosis, namely, compensated with no varices, compensated with varices, with ascites, and with a history of bleeding.4 This makes enrolment simple and criteria well defined. Table 1 summarizes current trials in patients with cirrhosis, as listed in clinicaltrials.gov, whereas Figs. 1 and 2; summarize this new treatment paradigm.
NSBBs should be tested in phase III trials in patients of all stages of cirrhosis, with the possible exception of refractory ascites, to prevent bacterial translocation and infection together with norfloxacin or rifaximin with an assessment of a reduction of HVPG and bleeding. Statins should be tested in phase III trials to prevent HCC, as an antifibrotic, and for preprimary, primary, and secondary prevention of bleeding. Warfarin should be evaluated for prevention of worsening fibrosis in patients with stage I cirrhosis. Rifaximin should be evaluated for preventing bacterial translocation and as primary or secondary prophylaxis of SBP. All obese or overweight patients with compensated cirrhosis should be advised to lose weight and all patients should be encouraged to stop smoking and drink at least 3 cups of coffee per day. All these drugs and lifestyle changes should be tried in various combinations according to the clinical stage of cirrhosis to test for additive effects. As all the above drugs are licensed, these trials would have immediate applicability as well as feasibility. International collaboration would guarantee fast enrolment and more rapid confirmation of therapeutic benefit. In the 21st century, liver cirrhosis should be regarded as a treatable disease with currently available therapy and not an irreversible disease that leads inevitably to liver transplantation.