Defining recompensation of alcohol‐related liver disease: A step beyond decompensated cirrhosis

Alcohol is the main cause of advanced liver disease in the Western world, and it is associated with the highest economic and social impact.1 Characteristically, the disease is underdiagnosed at early stages, and still nowadays, most patients are diagnosed when presenting with the first decompensation of cirrhosis.2 Classical studies on the natural history of alcohol-related cirrhosis were focused on investigating factors associated with liver disease progression or death, considering the disease irreversible in nature. However, some studies performed decades ago in small cohorts already described the clinical phenomenon of improvement of liver function and resolution of decompensations in patients with alcohol-related cirrhosis and alcohol-associated hepatitis.3, 4 More recent studies shed some light on this phenomenon reporting the observation of liver function improvement and resolution of decompensations in patients with alcohol-related cirrhosis awaiting liver transplantation, leading to the removal from the waiting list.5-7 In these studies, albumin levels, indirect biomarkers of portal hypertension such as platelet count and time of alcohol abstinence prior to inclusion in the waiting list were found to be associated with clinical improvement. These studies were useful to describe the clinical relevance of the phenomenon of recompensation, but they also left unanswered questions derived from the fact that they were focused on a very specific population of candidates for liver transplantation. Until a few years ago, the phenomenon of cirrhosis recompensation has been mostly investigated in patients with hepatitis C who experienced a resolution of cirrhosis decompensations after the cure of the virus.8 As a result of the evidence derived from these studies, the Baveno consensus recently provided a definition of recompensation in patients with decompensated cirrhosis of different aetiologies, including clinical criteria, such as controlled ascites without diuretic treatment and absence of hepatic encephalopathy or gastrointestinal bleeding in the prior year in a patient with sustained alcohol abstinence.9 These criteria are very valuable as they provide a homogeneous definition of the phenomenon of recompensation to be used in further studies; however, as these criteria were defined by expert consensus and not specifically created for patients with alcohol-related cirrhosis, there is room for further refinements when more evidence on the natural history of the disease is available. In that regard, the criteria established in the Baveno VII consensus include total alcohol abstinence for establishing the diagnosis of recompensation; however, there is growing evidence that in decompensated patients, liver function may improve after reducing the amount of alcohol consumed without achieving total abstinence, the so-called 'harm reduction' hypothesis.10 Future revisions of Baveno criteria will probably need to consider these specific characteristics of the natural history of alcohol-related cirrhosis. In this context, the study by Hofer and colleagues in Liver International is of great interest,11 as it defines the phenomenon of recompensation in a large cohort of patients with alcohol-related decompensated cirrhosis. The investigators included a well-characterized cohort of patients with baseline portal pressure gradient measurement. The authors report a prevalence of recompensation of 18% after a median follow-up of 2 years, which is significantly higher than that previously reported in studies in liver transplant candidates. However, it should be noted that patients included in this analysis had a less severe liver function impairment, with a median MELD score of 14 points, and most patients being Child-Pugh A or B at inclusion. The authors also show that the prognosis of patients that experience recompensation is markedly better than that of patients who remain decompensated, both in the mid and long term. Also, the study nicely points out that Child-Pugh score and albumin levels, body mass index, mean arterial pressure and portal pressure gradient may help predicting the occurrence of recompensation in the setting of alcohol-related cirrhosis. In this regard, one of the main strengths of the study is that it includes a baseline measurement of portal pressure gradient by hepatic hemodynamics. The association between the degree of portal pressure and the risk of liver disease progression, decompensation and death has been widely investigated in cirrhosis.12 However, the association between portal pressure and the probability of resolution of complications or regression of liver disease has been scarcely investigated, only in studies including a small number of patients. To our knowledge, this is the first time that portal pressure gradient has been described as a predictor of recompensation in a large cohort of patients with alcohol-related decompensated cirrhosis, which is a notable contribution to the field. Interestingly, parameters of liver function such as Child-Pugh score and albumin levels, both previously associated with recompensation in patients listed for liver transplantation, were found as predictive factors of recompensation in this study. These findings support the idea that there is probably a window of opportunity for recompensation in patients with cirrhosis and that it is highly unlikely of happening when progressing beyond a certain threshold of liver function impairment. Finally, an interesting finding is the association found between obesity and the probability of recompensation. No robust information is available on the effect that metabolic risk factors may have on the prognosis of alcohol-related liver disease. Two decades ago, Naveau et al. described obesity as a risk factor for cirrhosis and alcohol-associated hepatitis13; however, no further studies have been published in recent years exploring the effect of metabolic risk factors in advanced stages of alcohol-related cirrhosis. The current study provides new evidence that metabolic risk factors, particularly obesity, have deleterious effects in patients with alcohol-related decompensated cirrhosis, preventing recompensation in these patients. More studies are needed to further investigate the interaction between alcohol and metabolic syndrome in patients with liver disease. This study has some drawbacks, derived from the retrospective nature of the design: the lack of detailed information on alcohological follow-up, and the exclusion of patients with alcohol-associated hepatitis and acute-on-chronic liver failure. Although the probability of recompensation may be relatively low in such advanced stages of the disease, identifying the factors associated with the resolution of complications of cirrhosis may be of high interest in this population, as it may be crucial for the management of these patients, especially considering the widely extended indication for early liver transplantation in patients with alcohol-associated hepatitis. Another potential shortcoming of the study could be derived from the use of the Baveno VII definition of recompensation in the context of alcohol-related liver disease, especially regarding the criteria of sustained alcohol abstinence that may be too restrictive and underestimate the actual prevalence of recompensation in alcohol-related liver disease. In summary, this study represents one more step in the knowledge of the natural history of alcohol-related cirrhosis once clinical decompensations occur. New studies investigating biomarkers of recompensation will be needed in order to improve the risk assessment in patients with alcohol-related liver disease and to gain knowledge on possible therapeutic targets to guide new experimental therapies in this population. While achieving alcohol abstinence will still be one of the main goals, a treatment paradigm shift from preventing progression to favouring regression might be a better strategy moving forward. The authors do not have any disclosures to report. Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.