Advances in the understanding and management of alcohol-related liver disease
BMJ 2023; 383 doi: https://doi.org/10.1136/bmj-2023-077090 (Published 20 November 2023) Cite this as: BMJ 2023;383:e077090
- 1Department of Metabolism, Digestion and Reproduction, Imperial College London, London, UK
- 2Department of Brain Sciences, Imperial College London, London, UK
- Correspondence to: M Thursz m.thursz{at}imperial.ac.uk
Abstract
Alcohol-related liver disease (ALD) is a major cause of liver-related morbidity and mortality. Epidemiological trends indicate recent and predicted increases in the burden of disease. Disease progression is driven by continued alcohol exposure on a background of genetic predisposition together with environmental cofactors. Most individuals present with advanced disease despite a long history of excessive alcohol consumption and multiple missed opportunities to intervene. Increasing evidence supports the use of non-invasive tests to screen for and identify disease at earlier stages. There is a definite role for public health measures to reduce the overall burden of disease. At an individual level, however, the ability to influence subsequent disease course by modifying alcohol consumption or the underlying pathogenic mechanisms remains limited due to a comparative lack of effective, disease-modifying medical interventions. Abstinence from alcohol is the key determinant of outcome in established ALD and the cornerstone of clinical management. In those with decompensated ALD, liver transplant has a clear role. There is consensus that abstinence from alcohol for an arbitrary period should not be the sole determinant in a decision to transplant. An increasing understanding of the mechanisms by which alcohol causes liver disease in susceptible individuals offers the prospect of new therapeutic targets for disease-modifying drugs. Successful translation will require significant public and private investment in a disease area which has traditionally been underfunded when compared to its overall prevalence.
Introduction
Alcohol-related liver disease (ALD) encompasses a spectrum of clinical presentations and histopathological lesions attributable to the effects of excess alcohol consumption on the liver. The development of steatosis with excessive alcohol consumption occurs in up to 90% of individuals.12 In a subgroup of patients the development of steatosis is accompanied by the development of ballooned cells, hepatocyte death, and a predominantly neutrophilic lobular inflammation—defining the histological lesion of steatohepatitis due to ALD.3 This may progress to fibrosis and ultimately cirrhosis with the concomitant risk of developing decompensated liver disease and hepatocellular carcinoma. A proportion of patients may present with the distinct clinical syndrome of alcohol-related hepatitis (AH) characterized by the rapid onset of jaundice and other features of liver failure in the context of chronic, ongoing, heavy alcohol use and carries a high risk of short term mortality.34
The term “alcoholic” is pejorative and associated with significant stigma. This has been recognized by specialists and specialist bodies, which have sought to address it through a change in terminology. The European Association for Study of the Liver (EASL) has recommended use of the term “alcohol-related,”3 while the American Associated for the Study of Liver Disease (AASLD) advocates the term “alcohol-associated.”5 The terminology used in this article is defined in table 1.
Recommended terms and their abbreviations
In this review we have tried to summarize recent developments in the understanding of ALD from a public health perspective through to our biological understanding of the disease pathogenesis. We have summarized areas of clinical management, highlighting the prioritization of achieving alcohol abstinence, which may be of interest to a wide range of primary and secondary care physicians.
Sources and selection criteria
The MESH search terms “Fatty Liver, Alcoholic,” “Hepatitis, Alcoholic,” “Liver Cirrhosis, Alcoholic,” and “Liver Diseases, Alcoholic” were used as a backbone in PubMed searches to identify relevant articles published between 1960 and May 2023. Where relevant, these were combined using negative logic with the MESH term “Fatty Liver, Non-alcoholic” to avoid retrieving articles related to non-alcohol related fatty liver disease. This backbone query was supplemented with specific terms pertaining to the key topic of each subsection. Results were filtered to include primary research articles, clinical trials, systematic reviews, meta-analyses, and guidelines published in the English language. We prioritized publications within the past decade but also included older articles that were considered landmark trials that changed the treatment paradigm of ALD and alcohol use disorder (AUD). We excluded articles published in non-peer reviewed journals, case reports, and case series.
Epidemiology
Alcohol consumption is a risk factor for myriad acute and chronic health issues. Analyses for the Global Burden of Disease Study indicate that in 2016 alcohol use was the seventh leading risk factor for death and disability-adjusted life-years,6 with alcohol use disorders estimated to have an age standardised prevalence of 1321 per 100 000 people.7 In 2017 cirrhosis caused by ALD had an estimated global age-standardised prevalence of 320 per 100 000 and was responsible for 27.3% of the annual 1.32 million cirrhosis-related deaths.8 ALD is the primary etiology in a substantial proportion of liver transplants—36.7% in the US in 20169 and around 40% in Europe in 2020.10 In the UK recent National Health Service Blood and Transplant figures indicate that alcohol was the primary cause of liver disease in 27% of elective transplants.11
There is a strong correlation between population levels of alcohol consumption and the incidence and prevalence of ALD.12 Changes in population-level drinking behavior preface disease epidemiology. At a global level, per capita alcohol consumption has increased significantly over the past 30 years with further increases predicted.13 These predictions show regional variation, and, while drinking behavior in the European region may remain stable, marked increases have been forecast in the Americas without policy intervention.13 Modeling studies indicate that in the US this could translate into an 84% increase in the age-standardised mortality due to ALD.14 Emerging data suggest that the covid-19 pandemic may have served to entrench drinking behaviors with an increase in alcohol consumption seen in those with high risk drinking before the pandemic in the US15 and Europe.1617
The role of public health in tackling alcohol-related harm
Given the strong links between alcohol consumption and ALD at a population level, there is a strong rationale for evidence based and effective public health measures to reduce alcohol-related harms. Policies are in place to restrict access, such as where and who can purchase alcohol, pricing and taxation, and advertising bans.1819 Notably, although education about alcohol and its harms is popular, delivered alone without other interventions it shows limited evidence of effectiveness.20
There is a wealth of evidence showing that selling alcohol for a minimum unit price (MUP) results in reductions in alcohol consumption and harms.21 In the UK, an MUP of 50 pence/unit of alcohol (8 g) was enacted in Scotland in May 2018 (after a lengthy legal challenge by the alcohol industry) followed by Wales in 2020. Consultations are ongoing about an MUP in Northern Ireland, while England has no plans to introduce the policy. Evaluations of the impact in Scotland and Wales have reported a decrease in amount of alcohol purchased, use, and heavy drinking.22 However, reduction in alcohol harms is not universal, with studies variously indicating no reduction in consumption in younger men or men living in more deprived areas,23 no reduction in emergency department attendance,24 and unaltered consumption or severity of alcohol dependence among people drinking at harmful levels.25 Further, not all changes were beneficial: reduced spending on food and utilities and increased borrowing were seen in those drinking at harmful levels who struggled to afford their alcohol.25 Improvements in other alcohol related harms after introduction of an MUP took a few years to accrue in Canada, but reductions in numbers of patients discharged in Glasgow with alcohol-related liver disease have already been reported.2627
Pathophysiology of alcohol-related liver disease
Intestinal microbiome and gut barrier function (fig 1)
Microbial dysbiosis in the intestine. (1) Alcohol is converted to acetaldehyde in intestinal epithelial cells. Acetaldehyde induces cellular damage, including proteins that maintain tight junctions between epithelial cells, leading to loss of mucosal integrity. (2) Multiple changes in the composition of the intestinal microbiome are observed in patients with alcohol-related liver disease and alcohol-related hepatitis. (3) Dysbiosis of the microbiome leads to further damage to the epithelial barrier and loss of mucosal integrity. (4) Bacterial products, including lipopolysaccharides and toxins such as the Enterococcus faecalis cytolysin toxin cross the damaged epithelium into the mesenchymal blood vessels and subsequently reach the liver via the portal venous system. (5) Intact bacteria may breach the epithelial barrier and be phagocytosed by monocytes or neutrophils. Failure to mount an effective oxidative burst in phagocytic cells then allows dissemination of bacteria through the circulation.
Heavy alcohol consumption is associated with numerous changes in the composition of the intestinal microbiome.28 Although the biological significance of these changes is not fully understood, there are several lines of evidence which implicate this dysbiosis in the pathogenesis of ALD. If the fecal microbiome of patients with AH is transferred into a mouse model of ALD then the severity of liver steatosis and inflammation are significantly worse than the fecal microbiome from healthy individuals.29 This seems to indicate that the microbiome itself is pathogenic. A study by Duan and colleagues demonstrated that the prevalence of Enterococcus faecalis in fecal samples from patients with ALD, and particularly those with AH, is much higher than in healthy populations.30 Furthermore, a proportion of E faecalis bacteria encode a secreted cytolysin protein which is toxic to hepatocytes; carriage of the cytolysin-producing E faecalis was associated with higher mortality compared with patients with AH who did not carry this strain.30
Alcohol is directly toxic to the intestinal epithelium and causes a loss of mucosal integrity resulting in increased gut permeability.31 Epithelial cells in the stomach and upper gastrointestinal tract metabolise alcohol to acetaldehyde, which may damage the cell, but lack the ability to convert this to relatively inert acetate. In addition to this biochemical toxicity, the abnormal microbiome is also thought to contribute to the breakdown in mucosal integrity, possibly through the production of proteolytic enzymes which damage intercellular tight junctions. Increased permeability allows the translocation of microbial products or possibly intact microbes into the portal venous system, from where they can reach the hepatic sinusoids. Interestingly, in the presence of an intact intestinal mucosa the E faecalis cytolysin does not seem to cause hepatotoxicity.
Kupffer cells, a specialized tissue resident macrophage which lives in the hepatic sinusoids, respond to microbial products via pathogen-associated molecular pattern receptors. Primarily Kupffer cells respond to lipopolysaccharide from the bacterial cell wall through Toll-like receptor (TLR) 4. TLR-4 signaling stimulates the production of pro-inflammatory cytokines and chemokines including IL-1b, IL-6, IL-8, and TNFa.32 This results in tissue inflammation, including the recruitment of monocytes and polymorphonuclear cells to the liver.
Oxidative stress and lipid metabolism (fig 2)
Effects of alcohol on the liver. (1) Alcohol metabolism results in the accumulation of lipids within hepatocytes due to excess calorie consumption, increased lipogenesis, and reduced lipid oxidation. (2) Generation of excess reactive oxygen species through the metabolism of alcohol in the cytochrome P450 system leads to widespread damage of intracellular organelles. The resulting ballooned hepatocytes are a histological hallmark of steatohepatitis. Increased senescent cells in the liver produce pro-inflammatory cytokines which activate stellate cells. (3) Lipopolysaccharide and other bacterial products in the portal venous blood interact with toll-like receptors on (4) macrophages (including Kupffer cells) in the hepatic sinusoids, resulting in the secretion of inflammatory cytokines and chemokines. (5) Chemokine secretion leads to infiltration of the hepatic parenchyma by inflammatory cells including lymphocytes and neutrophils. (6) Stellate cells, resident in the space of Disse between endothelial cells and parenchymal cells are activated by inflammatory signals transforming them into a myofibroblast phenotype. Activated stellate cells secrete collagen, resulting in fibrosis and eventually cirrhosis. (7) Inflammation and epigenetic changes in the hepatocytes leads to loss of cell identity (de-differentiation). Reduction in expression of multiple metabolic pathways, including bile acid transporters, is responsible for jaundice and liver failure.
Alcohol consumption results in several metabolic changes in the liver cells. At low concentrations alcohol is metabolized by alcohol dehydrogenase to acetaldehyde and by acetaldehyde dehydrogenase to acetate. Acetate is an energy source which can be used in the Krebs cycle but may also be used in lipogenesis. Heavy alcohol consumption is commonly a cause of obesity and liver steatosis due to the excess calorie consumption. Acetaldehyde may form adducts with proteins and other macromolecules within the cells, causing cell damage through loss of function and possibly stimulating autoimmune responses due to the creation of neoantigens. In excess, the acetaldehyde dehydrogenase pathway is fully saturated and alcohol is metabolised in the cytochrome P450 system via the enzyme CYP2E1.33 This metabolic pathway generates an excess of reactive oxygen species resulting in damage to cell membranes, organelles, and macromolecules. This oxidative damage can be countered through oxidative defense systems, particularly through provision of glutathione. In patients who drink heavily, dietary habits may result in lower levels of glutathione and its precursors. Furthermore, in patients with cirrhosis the key glutathione generating enzyme, methionine adenosyl transferase 1, is expressed at markedly reduced levels, resulting in reduced capacity for oxidative defense.34
Lipid homeostasis is more widely deranged in patients with ALD, with overall increased lipogenesis and reduced lipid oxidation.35 Ethanol metabolism increases the ratio of NADH:NAD within the cell, which inhibits β-oxidation of fatty acids. Alcohol exposure also increases the expression of the transcription factors SREBP-1c and ChREBP which regulate lipogenesis.3637 Excess lipid within the cell causes a stress response in the endoplasmic reticulum and mitochondrial damage. In addition to the intrinsic damage modulated by deranged lipid metabolism, steatotic hepatocytes seem to be more sensitive to external toxic signals, leading to apoptosis and other forms of cell death.35 The importance of deranged lipid homeostasis in the pathogenesis of ALD is underlined by genetic associations in loci related to lipid metabolism, namely the genes PNPLA3, TM6SF2, HSD17B13, and MBOAT7.38 The PNPLA3 variant has reduced lipase activity, resulting in reduced secretion of very low density lipoprotein and excess accumulation of lipid within the cell.
Loss of hepatocyte differentiation and failure of regeneration
AH, the most severe form of ALD, presents with liver failure, characterized by low levels of clotting factors, albumin, and other serum proteins as well as high levels of bilirubin.39 Not all patients with AH have cirrhosis, although advanced fibrosis is invariably present, suggesting that the features of liver failure cannot be fully explained by paucity of hepatocytes. Epigenetic modification of the hepatocyte nuclear factor 4 a (HNF4a) gene locus may provide a link between inflammation and loss of hepatocyte function.40 HNF4a is a key transcription factor which up-regulates hepatocyte functions such as albumin production, gluconeogenesis, and bile acid transport. In AH it has been shown that a fetal isoform (P2) of HNF4a is produced in response to the cytokine transforming growth factor b (TGFb), which has the opposite effect to the normal P1 isoform, therefore suppressing hepatocyte differentiation and function. In addition, patients with ALD have a high proportion of hepatocytes expressing markers of cellular senescence indicating that they would not be able to respond to increasing hepatocyte demand through proliferation.41 An alternative pathway of liver regeneration through the generation of hepatocytes from the pluripotential stem cell niche adjacent to the portal tracts is observed in patients with ALD and specifically AH. Histologically this process, characterized by bile ductular proliferation, is associated with a poor prognosis as the stem cells seem to be unable to differentiate into mature hepatocytes at a sufficient rate.42
Natural course of alcohol-related liver disease
The disease progression of ALD is well defined but displays considerable variability. A recent systematic review and meta-analysis of histological studies of ALD incorporated data from 37 studies representing around 7500 patients.2 The study reported abnormal liver histopathology in approximately 85% of individuals reporting excessive alcohol consumption (≥60 g/day of ethanol for included studies). A quarter of biopsies demonstrated steatohepatitis, and over half showed fibrosis or cirrhosis. In the subgroup of studies reporting outcome data, the annualized rate of progression to cirrhosis was 8-10% in those with steatohepatitis or pre-cirrhotic fibrosis. Liver-related mortality is related to the presence of progressive liver disease on biopsy and correlates with fibrosis stage, at least in patients with compensated liver disease.24344 The importance of drinking behavior in determining disease progression cannot be overstated. Abstinence from alcohol is a fundamental determinant of liver-related mortality at almost all stages of disease.2434546 A recent Swedish study reported that a histological diagnosis of ALD is associated with a dramatic increase in the risk of mortality compared with population controls (five year cumulative mortality 40.9% v 5.8%), particularly from liver-related events and within in the first year after diagnosis.47 This finding may be attributed in part to the comparatively late stage of disease at diagnosis (52% prevalence of cirrhosis at baseline), but the increased risk of mortality was not confined to this group. In comparison with non-alcohol related fatty liver disease, ALD typically presents at a later stage, is more rapidly progressive, and is associated with greater mortality from liver-specific causes.4849 Estimates of the incidence of hepatocellular carcinoma in patients with ALD cirrhosis vary, though many studies report an annualised risk in the range 1.5-3%.50515253
Alcohol-related hepatitis is a less common presentation of ALD, with population level estimates of the annual incidence at 25-45 per 100 000 person years.5455 However, the severity of disease means AH may account for up to 20% of hospital admissions for ALD.56 Affected individuals were typically men in their fifth sixth decade. However, several epidemiological studies in the US and Europe indicate a concerning rise in the incidence of AH in younger age groups and among women.545657 Mortality from AH is high and tightly linked to the degree of liver dysfunction at presentation. Severe AH (defined by Maddrey’s discriminant function (MDF) >32, model for end-stage liver disease (MELD) >18, or Glasgow alcoholic hepatitis score (GAHS) >8) confers a high risk of short term mortality which nears 30% at 90 days5859 and has changed little in the 50 years since the condition was first described.60 Non-severe AH (MDF <32) is not, however, benign and confers a substantial mortality risk of around 20% at one year.46
Variability in disease development and progression is, in part, genetically determined. Large scale genome-wide association studies have identified several loci associated with ALD disease risk, many of which are associated with lipid metabolism. These include PNPLA3, TM6SF2, HSD17B13, MBOAT7, MARC1, and SERPINA1, all of which show similar associations with non-alcohol related fatty liver disease.616263 Genetic studies also indicate certain loci also influence disease progression and the development of complications. The variant rs738409 in PNPLA3 has been associated with age of onset, disease progression, and survival even in patients with established cirrhosis.6465666768 Variants in PNPLA3, TM6SF2, HSD17B13, TERT, and WNT3A-WNT9A are associated with the risk of hepatocellular carcinoma.6970
Screening for alcohol-related liver disease
The large proportion of patients with ALD who present with advanced disease coupled with high mortality in the period immediately after diagnosis47 underscore the need to identify patients at earlier disease stages. If abstinence can be achieved then disease progression will likely be halted or even reversed, with positive impact on clinical outcomes.24344 ALD cirrhosis develops over years of heavy drinking during which affected individuals may have multiple healthcare contacts, each representing a missed opportunity to intervene.7172 Of patients with healthcare attendances due to alcohol related issues, 7-16% will develop cirrhosis in the next 8-12 years.7374 Multiple hospital contacts with alcohol-related issues defines a population at further increased risk.74 Screening need not be confined to the healthcare setting, workplace-based studies have demonstrated a substantial case acquisition rate.75
Non-invasive tests for diagnosis and staging of alcohol-related liver disease
The gold standard for the diagnosis and staging of ALD is liver biopsy. However, as an invasive test requiring specialist training and conferring a risk of morbidity, it is neither practical nor appropriate to perform in all patients with suspected ALD and is rarely undertaken in clinical practice.447677 In most clinical scenarios the pertinent questions are the extent of fibrosis and the presence or absence of cirrhosis as these features identify those patients with progressive disease and at greatest risk of adverse liver-related outcomes.77
In patients with established or decompensated chronic liver disease, clinical stigmata, potentially in conjunction with biochemical and radiological features, may be sufficient to establish a diagnosis of cirrhosis. Establishing the degree of fibrosis at earlier stages of disease is more challenging. Liver ultrasound demonstrates acceptable specificity in identifying steatosis, with sensitivity proportional to the severity of steatosis.78 Although many clinicians factor the presence of imaging features suggestive of cirrhosis into their diagnostic algorithm, there are limited data regarding the diagnostic performance of ultrasound for ALD cirrhosis.79 Other tests to detect and estimate the degree of underlying fibrosis in ALD may be derived as composite scores from commonly measured biochemical tests (namely, the aspartate aminotransferase to platelet ratio index (APRI) and the fibrosis-4 (FIB4) index), proprietary tests of panels of fibrosis biomarkers (enhanced liver fibrosis (ELF) test, FibroTest), or based on measurement of liver stiffness using elastography (Fibroscan, shear wave elastography, magnetic resonance elastography, acoustic radiation force impulse imaging). Serum-based tests lend themselves to larger scale deployment as the need for specialist equipment and training presents challenges to widespread deployment of elastography-based techniques. The diagnostic performance of ELF and FibroTest for ALD cirrhosis is good (area under the receiver operated characteristic (AUROC) >0.8) and superior to that of indirect indices of liver fibrosis derived from routine biochemical tests.8081 A cut-off of 10.5 for ELF has been reported as having a high negative predictive value in a population of patients where the anticipated prevalence of advanced fibrosis is low, indicating its utility as a screening test.80
Several studies have evaluated the utility of liver stiffness measurement (LSM) by transient elastography for diagnosis of ALD fibrosis and cirrhosis. Individual studies and meta-analyses indicate good diagnostic performance, particularly when applied to rule out cirrhosis or advanced fibrosis.80828384 The optimal LSM cut-off to diagnose significant ALD is unclear, but values in the range 12-15 kPa appear to predict advanced fibrosis or cirrhosis (≥F3) with good sensitivity and reasonable specificity.8384 There is etiology-specific variation in the LSM values that predict significant liver disease, and those for ALD are generally higher than those for other liver diseases.85 LSM may be affected by factors independent of the degree of underlying fibrosis: these include hepatic congestion, cholestasis, alcohol, recency of food intake, and hepatic inflammation. Accordingly, steatohepatitis on biopsy and higher serum levels of aspartate aminotransferase (AST) and bilirubin have all been associated with higher LSM values and the need to use higher cut-off values for any given degree of fibrosis.84 LSM falls with detoxification from alcohol, with the degree of fall linked to the serum AST. Repetition of LSM after at least two weeks of abstinence may allow patients to be more accurately classified.8286
Modeling suggests that screening for ALD fibrosis using non-invasive tests is cost-effective in populations with excess alcohol consumption under a variety of different test combinations and assumptions regarding ability to affect drinking behavior.87 A recent study demonstrated the feasibility of screening for ALD and non-alcoholic fatty liver disease using non-invasive tests in the general population in Denmark.88 However, it is likely that consistent and widespread implementation of this form of screening will be challenging. A recently reported artificial intelligence-based model demonstrated significantly better performance in predicting significant liver disease in at-risk individuals in a primary care population than models based on blood-based indices.89 Combined with the digitisation of healthcare records, this raises the exciting prospect of integrating screening models into clinical systems.
Risk stratification in alcohol-related liver disease using non-invasive tests
Increasing literature exists about the prognostic information offered by many of these non-invasive tests in ALD. FIB-4, ELF and Fibroscan have all been associated with adverse liver-related events (viz. mortality, development of variceal hemorrhage, ascites or hepatocellular carcinoma) though typically with moderate prognostic performance.8190 Nonetheless it appears that LSM >15k Pa or ELF >10.5 define populations of patient with ALD at dramatically increased risk of disease complications.91 The development of clinically significant portal hypertension is associated with an increased risk of adverse liver related events including variceal hemorrhage and portends a poor prognosis. The Baveno VI consensus guidelines indicated that patients with LSM <20 kPa and a platelet count >150×109/L had a low probability of having varices requiring therapy.92 Studies examining the performance of these criteria have reported a high negative predictive value in patient populations including those with ALD indicating their prognostic value in this setting.9394 These thresholds were restated in Baveno VII and expanded with recommendations that LSM >25 kPa was sufficient to rule in clinically significant portal hypertension and identify a population of patients with ALD cirrhosis likely to have endoscopic evidence of portal hypertension and an increased risk of decompensation.95 A recent study established a series of serum protein panels which were able to predict liver biopsy features and subsequent clinical events in patients with ALD with a greater accuracy than existing non-invasive tools.96 Subject to validation these panels may form the basis of non-invasive panels offering greater diagnostic and prognostic information and the opportunity to study earlier stages of ALD before the development of cirrhosis.
Definition and classification of alcohol use disorder
In ICD-11, harmful use of alcohol is diagnosed when alcohol consumption has caused damage to a person’s physical or mental health or has resulted in behavior leading to harm to the health of others.97 Alcohol dependence is a complex syndrome that includes “impaired control over their alcohol use” as a key feature which may not necessarily be accompanied by craving. Other features are alcohol use taking precedence over other activities and physiological neuroadaptations—tolerance and withdrawal. “Binge” is a term widely used, but definitions differ across the world. In the UK, binge drinking is drinking six or more units of alcohol for a woman and eight or more units for a man in a single session. If conveying an “on-off” pattern, bingeing may mean the individual is alcohol dependent as they cannot control their consumption when they start drinking. While previously ICD and DSM diagnostic categories were similar, DSM-5 is now different to ICD-11 with alcohol use disorder seen as a continuum from mild (harmful) to severe (dependent) rather than distinct entities as in ICD-11.9798 As many people who drink excessively have comorbid psychiatric and physical disorders, including other substance misuse, these must also be assessed and managed alongside their alcohol problems. Any safeguarding concerns such as domestic violence or risk to minors should also be explored.
Neurobiology in alcohol use disorders
Characterizing the neurobiology of alcohol use disorder informs optimization of current methodology and development of new approaches to prevent and treat addiction. The pleasurable effects of alcohol are mediated by increasing endorphins and dopamine levels in the brain in the so called pleasure-reward dopaminergic mesolimbic system.99100 As alcohol consumption becomes more chronic, neuroadaptations occur. Blunted function in μ-opioid receptor and in dopaminergic systems are present in alcohol dependence and may also be implicated in vulnerability to addiction.101102103 Chronic alcohol consumption also results in increased activity of the κ-opioid receptor and dynorphin system, which is associated with dysphoria to counter the pleasure from endorphin-stimulated μ-opioid receptor system.100
In alcohol dependence, blunted activity is seen in the striatum of the mesolimbic system in response to non-salient rewards such as money, but, notably, heightened activity is seen in response to salient (ie, alcohol) cues.104105106107 Treatment, whether pharmacological (such as naltrexone) or psychological, has been shown to attenuate this hyper-reactivity to alcohol cues.107108 Decreased function in prefrontal cortical areas contribute to difficulties in, for example, decision making, response inhibition, error monitoring, and salience attribution that are commonly seen in addiction.100 Many neuroimaging studies have also shown dysregulated function in prefrontal cortical regions in response to cues and craving.104107
Many people drink alcohol for its anxiolytic and sedative effects. These effects are primarily mediated by the brain’s inhibitory GABA-benzodiazepine system.99 Alcohol acutely boosts activity of the GABA-benzodiazepine receptor and acutely antagonises the brain’s main excitatory glutamatergic system, particularly the NMDA receptor, resulting in reducing brain activity. Regular consumption of alcohol results in tolerance because of reduced sensitivity of the GABA-benzodiazepine receptor and an up-regulation of NMDA glutamate receptors. This greater number of NMDA glutamate receptors is thought to underpin amnesia during drinking episodes. These neuroadaptations associated with tolerance also drive withdrawal symptomatology in the absence of alcohol. Such symptoms include tremor, sweating, and anxiety as well as potentially life-threatening complications such as seizures and delirium tremens. The amygdala and associated neurotransmitters (such as dynorphin, noradrenaline, and substance P/NK1) play a key role in mediating many of aspects of withdrawal such as negative mood and heightened stress responses.109
Brain atrophy, commonly seen with chronic alcohol consumption, likely results from hyperglutamatergic activity and neuroinflammation.110111 It is important to recognize and inform your patient that functional and structural recovery occurs with abstinence, particularly in the frontal cortex.112113
Treatment of alcohol use disorder
If an individual has ALD, abstinence from alcohol will facilitate recovery, and they should be given information about local addiction services. It is important to determine if someone is alcohol dependent and if they have had previous complications during withdrawal, such as a seizure or delirium tremens. If they have, they will need medical support (see below) to safely stop drinking and may need to be admitted as an inpatient if they previously experienced complications, are unwell, or their home environment is not supportive. Although withdrawal from alcohol may be routinely managed by hepatology teams, any patient should also be referred to an addiction service to ensure that appropriate levels of support and treatment are offered to achieve and maintain abstinence.
Psychosocial interventions, self help, and peer support
A range of approaches have been shown to be effective in reducing unhealthy consumption of alcohol and in those with chronic liver disease.114 In someone who is drinking at a harmful level, a brief intervention based on the acronym FRAMES is appropriate (Feedback (on the risk from their damaged liver etc), Responsibility, Advice, Menu, Empathy, Self efficacy).114115116 Completing a drink diary (when, how much, why) is also effective in allowing the patient and practitioner to understand drinking behaviors and how to potentially modify them.
Motivational interviewing can be used to support changes in behavior through bringing to the fore and resolving discrepancies between the patient’s goals and actions, as well as any ambivalence. A cognitive behavioral approach aims to identify core beliefs or underlying assumptions that can be seen as irrational and resulting in a dysfunctional relationship with alcohol. The individual is given homework between sessions, which are of limited number. Cognitive behavior therapy is also widely used to treat depression and anxiety, both of which are highly comorbid with alcohol use disorder. Relapse prevention involves training the patient to identify high risk situations and develop coping skills and alternative strategies to drinking. Contingency management is based on using positive reinforcement, such as vouchers for maintaining abstinence, to change behavior.
Other approaches include social behavior and network therapy, which aims to develop supportive non-alcohol social networks or couples or family therapy to facilitate change within close relationships. Many patients have experienced trauma in their life, and increasingly any planned treatment is now conceptualized as “trauma-informed” to take account of this. In addition to individual work, complementary group work is generally offered. Because of the covid-19 pandemic, services now may deliver treatment and support online rather than in person. A mix of these different psychosocial approaches are used clinically rather than rigidly sticking to one model. A systematic review of 13 observational studies and RCTs of 1945 patients with chronic liver disease explored the impact of integrated combination psychotherapy with cognitive behavior therapy, motivational enhancement therapy (MET), and comprehensive medical care. Regarding induction of abstinence, more patients reported abstinence in four RCTs from the intervention (45.4%, range 25.4-74%) than control groups (36.7%, range 19.7-50%), with observational studies also indicating benefit. Regarding maintenance of abstinence, the only RCT reported MET was non-significantly superior (84.8% v 93.4%) with one observational study showing benefit of intervention on relapse compared with control (16.4% v 35.1%).117
For those who are dependent on alcohol, peer support organisations such as 12-step Alcoholics Anonymous (https://www.alcoholics-anonymous.org.uk/) and SMART recovery (https://smartrecovery.org.uk/) play a valuable role in recovery, particularly in providing advice and a non-alcohol network of support and friendship to those who may be isolated.118 Meetings are widely available, some of which cater for particular groups (women only, LGBT) and can be accessed in person or online. In addition, organisations such as Al-Anon (https://www.al-anonuk.org.uk/), Al-Ateen (https://www.al-anonuk.org.uk/alateen/), or NACOA (https://nacoa.org.uk/) provide support for friends and relatives, including children, of someone with alcohol dependence.
Pharmacotherapy
Medication plays a key role in managing withdrawal and relapse prevention in moderate to severe alcohol dependence alongside psychosocial interventions.114119 There is very limited evidence to support the use of medication in reducing harmful non-dependent drinking, and it is not generally recommended.114
Detoxification
Alcohol withdrawal is potentially life threatening, and people who are alcohol dependent commonly require medication such as a reducing regimen of a benzodiazepine to prevent these complications.114120 In those with hepatic impairment, consideration can be given to using lorazepam or oxazepam due to their pharmacokinetics. Although anticonvulsants such as carbamazepine have been used in some countries for alcohol detoxification, their use is not widespread in the UK.119
Preventing complications: thiamine
Thiamine is an essential cofactor for enzymes involved in brain metabolism of glucose, and in its absence anaerobic metabolism ensues, resulting in brain damage. Thiamine deficiency can be manifest as a range of signs and symptoms, but at its most serious, Wernicke’s encephalopathy, it is a medical emergency. The triad of ophthalmoplegia (nystagmus on lateral gaze), ataxia, and confusion are present in only about 10% of cases so a high index of suspicion is needed. It has been proposed that presence of any two of the criteria are sufficient for diagnosis.121 Treatment of thiamine deficiency is with the parenteral thiamine preparation Pabrinex until no improvement is seen.119 Everyone should be assessed for risk of thiamine deficiency—for example, poor nutrition or signs suggestive of deficiency (such as peripheral neuropathy)—so that they can receive Pabrinex.119 It must be remembered that thiamine deficiency is not only seen during alcohol withdrawal and may occur with another condition increasing metabolic demands, such as infection. Therefore, any change in an individual’s condition such as increase in confusion or ataxia should prompt assessment of potential thiamine deficiency. Parenteral thiamine must be used if someone is thiamine deficient or at risk of deficiency, absorption of oral thiamine through the gut and transport into the brain is insufficient to replenish stores.122 For all other patients with alcohol use disorder undergoing detoxification, oral thiamine can be used.119
Relapse prevention
As an adjunct to psychosocial interventions, medication for relapse prevention should be considered for everyone with moderate to severe alcohol dependence. Based on extensive evidence from RCTs, the National Institute for Health and Care Excellence (NICE) recommends the NMDA receptor modulator (glutamate modulator) acamprosate or the non-selective opiate antagonist naltrexone as first line to support abstinence and to start when abstinent.114 Meta-analyses of clinical trials suggest that acamprosate compared with placebo is efficacious in supporting abstinence (RR = 0.83 (95% CI 0.77 to 0.88)), while naltrexone prevents a lapse becoming a full-blown relapse (RR = 0.83 (0.75 to 0.91)).119 Disulfiram, an aldehyde dehydrogenase inhibitor, was suggested by NICE as a second line treatment to support abstinence because of its potential adverse effects if combined with alcohol and the range of conditions in which it is absolutely or relatively contraindicated. In the UK, baclofen, a GABA-B agonist, is used off-label based on some evidence showing efficacy in supporting abstinence, particularly among those who are more severely dependent and anxious and have not responded licensed medications, although not all meta-analyses have found baclofen superior to placebo.123124125126127 Other medications such as topiramate or gabapentin have been recommended as second line in the US but are not widely used or recommended in the UK.128
All the above relapse prevention medications are started once the person is abstinent and are to be taken every day. In contrast, nalmefene is licensed to start while a person is still drinking and “as needed.” Nalmefene is another opiate antagonist with slightly different pharmacology to naltrexone as it is a partial agonist at κ-opioid receptors rather than an antagonist. Nalmefene is licensed for use in alcohol dependence to support reduction in drinking in those who do not need immediate detoxification and whose drinking has not changed with psychosocial interventions. Combinations of relapse prevention medications can be used, but there is little evidence to guide which work best.
All of these relapse prevention medications have been used in patients with abnormal liver function tests, and improvement is generally reported in clinical trials, but careful consideration is required if they are to be used for patients with severe hepatic impairment, and SmPCs (summaries of product characteristics), reviews, and guidelines should be consulted.114119123129 Because of the risk of hepatotoxicity, disulfiram and naltrexone should not be used in severe or acute hepatic impairment, acute hepatitis, or hepatic failure. On the other hand, trials of baclofen have been conducted safely in those with cirrhosis.130
It is notable, however, that increasingly fewer people are offered relapse prevention medication or continue beyond a few weeks of taking the medication despite their effectiveness in improving outcomes in those with liver disease.131132133 Barriers to overcome include lack of knowledge of and confidence in the evidence base, concerns about cost, and philosophical views about “medicalizing” addiction and not using medication.134
Integration of hepatology and addiction services in secondary or primary care
To ensure optimal likelihood of recovery, people with ALD need to be supported to achieve abstinence from alcohol and to maintain it. This is best achieved via integrated cooperation of hepatology, addiction services, and general practice.135136 Despite this, such integration is rarely evident, with both patient-related and clinician-related barriers including stigma in patients and clinicians.137 In parts of the UK, addiction services are now led by the third sector and outside the NHS, whereas hepatology services are within the NHS. This has had an immense impact on working together. For instance, staff in NHS and third sector involved in treatment of an individual may be able to communicate only with explicit patient consent, which might not be given.
Many general hospitals have “alcohol care teams” which provide valuable expertise, assistance, and links with primary care and other community-based teams during any inpatient stays in a general hospital.138 A joint outpatient clinic with hepatology and addiction services is a useful model to deliver and coordinate care, particularly for those patients reluctant to attend a community-based addiction service. Similarly, an alcohol specialist embedded in primary care, working alongside other professionals, can result in a more coordinated delivery of care.139 While many hepatologists and GPs are comfortable prescribing medication for detoxification and relapse prevention, not all feel adequately skilled, so most of the prescribing is initiated in an addiction service. The patient’s GP is generally requested by community addiction services to organize any blood tests and physical investigations as required.
In the UK, most dedicated inpatient units have closed, resulting in waiting lists of several weeks; this means that treatment for alcohol dependence, even complex cases, is largely community based. Because of changes in how addiction services are delivered in the UK, it is notable that most inpatient alcohol detoxifications now take place in non-specialist settings.140 Similarly, relapse prevention treatment, such as psychosocial treatment, occurs primarily in the community. Assessment for residential rehabilitation can take many weeks, and, even when it is approved, waiting lists are long. Many community-based services offer a similar range of one-to-one and group sessions that someone may receive in a residential setting.
As most addiction services rely on self referral, some patients may not go through with presenting to a local service despite expressing a willingness to seek treatment to their hepatologist or GP. Having an alcohol specialist in the hospital or primary care can provide greater support to increase the likelihood that the patient engages with addiction treatment. While addiction services will include input from psychiatry, psychology, and mental health nurses, assessment and treatment of any substantial psychiatric disorder will need referral to a mental health service. For some patients, a regular multi-professional meeting may be required to ensure all mental and physical health and social care needs are being optimally met along with management of any risks.
Treatment of alcohol-related liver disease
At all stages of ALD, the key therapeutic intervention is to achieve alcohol abstinence. Even at the stage of decompensated cirrhosis, the withdrawal of alcohol makes a substantial impact on patient survival (see above). Unlike those at other stages of disease, patients with AH may continue to deteriorate and experience high mortality for up to 90 days after presentation despite cessation of alcohol. Urgent treatment is therefore required to treat AH specifically while not neglecting the need to support patients becoming abstinent after discharge from hospital.
Despite the high prevalence of ALD cirrhosis, there have been remarkably few clinical trials of therapeutic interventions. The nature of the condition, variation in natural course in response to abstinence, and the complicated nature of patients’ lives when suffering with addiction undoubtedly make the conduct of clinical trials challenging. However, lack of investment by public and commercial funders also plays an important role. One randomized controlled trial, conducted over 20 years ago, randomized 123 patients with Childs-Pugh class A or B cirrhosis to treatment with the antioxidant S-adenosyl methionine (S-AME) or placebo. Overall there was a numerical mortality advantage (16% v 30%, P=0.077), but in patients in the Child-Pugh C class there was a significant mortality advantage (12% v 29%, P=0.025) for S-AME treatment.141 Unfortunately, this study has never been repeated, although S-AME is available in many countries as an over-the-counter medication.
A single double-blinded placebo-controlled trial evaluated the treatment of 68 patients (v 68 controls) with non-cirrhotic ALD using rifaximin to improve gut barrier function and reduce hepatic inflammation. The primary endpoint was improvement in histologically assessed liver fibrosis. While the primary endpoint was not achieved, the rate of fibrosis progression was reduced in the rifaximin treated group (0.42 (95% CI 0.18 to 0.98), P=0.044).142
AH is classified as severe when the Maddrey’s discriminant function, based on measurement of bilirubin and prothrombin time, is ≥32. The mortality for AH is around 20% at one month and 30% at three months.143 The only treatment currently recommended is corticosteroids, such as prednisolone, with the aim of reducing inflammation in the liver.3 In large trials and meta-analyses of 2111 patients, prednisolone treatment results in a small, short term survival advantage at one month (hazard ratio 0.64 (95% CI 0.48 to 0.86)) but no benefit at three months.59144 This loss of benefit is probably due to the increased rates of severe infections seen in patients treated with steroids.145 However, subgroup analyses in patients who were enrolled in the STOPAH trial suggest that those with a neutrophil:lymphocyte ratio between 5 and 8 or those with a serum level of the cytokeratin 18 M30 fragment (CK18-M30) >5000 may derive a durable benefit from prednisolone treatment.146147
Transplantation
An orthotopic liver transplant is an option for patients with decompensated ALD cirrhosis and severe AH. Outcomes from transplantation for ALD cirrhosis are good, with one-year and five-year graft survival of 84% and 73% respectively.148 Most patients with decompensated cirrhosis due to ALD will recompensate once they achieve alcohol abstinence, but approximately 10% will continue to deteriorate. Nevertheless, in the UK ALD is the most common indication for liver transplantation, representing 27% of all adult transplants in 2021-22 in the UK.11 In many transplant centers a six-month abstinence rule is applied; this allows time for the effects of abstinence to be fully assessed, but, more controversially, this period of abstinence is thought to predict alcohol behavior after transplantation.149 Selection for transplantation and priority for organ allocation is based on measures of residual liver function, commonly the Model for End Stage Liver Disease (MELD) based on creatinine, bilirubin, INR, and sodium. A MELD score of 20-29 indicates a mortality risk of 19.6% over three months and is a level at which transplantation may be considered.150 In contrast to the US and much of the rest of Europe, the UK system employs a variant of the MELD score termed UKELD. A score greater than 49 is associated with a 9% one-year risk of mortality and represents the minimum criterion for listing.151
Disease recurrence after transplantation is a common and a widely accepted risk in most types of liver disease. However, in ALD the resumption of alcohol use after transplantation is often viewed as a calamity in a way that is, perhaps, disproportionate to its risk. A recent meta-analysis indicated that only around a fifth of individuals resumed alcohol consumption within five years of transplantation for decompensated ALD cirrhosis; recurrence of heavy drinking was less prevalent at 14%.152 Psychiatric comorbidities, pre-transplant abstinence of less than six months, lack of social support including unmarried status, and smoking were associated with an increased risk of relapse. Harmful levels of drinking after transplantation may lead to an accelerated course of liver disease. Patients transplanted for ALD are at higher risk of cardiovascular disease and malignancy than patients transplanted for other indications.
For many years, the six-month abstinence “rule” excluded transplantation in patients with severe AH despite the high short term mortality. A European consortium conducted landmark studies demonstrating that early transplantation improved survival compared with matched controls (77 ± 8% v 23 ± 8%, P<0.001 at six months) with only three of the 26 patients resuming any level of alcohol consumption.153 Subsequent studies have replicated these results.154 Despite this apparent effectiveness, controversy persists. The prognostic scoring systems used in AH are not sufficiently precise to identify those patients who will die without transplantation.155 In a recent study, only four of 49 patients listed for early transplant recovered without transplant. However, 34 of the 95 patients who were not listed experienced a spontaneous recovery, though only a minority (7/34, 20%) achieved recompensated liver function.156 The selection criteria for transplant candidates are not well defined, leading to heterogeneity between centers. Additionally, the recent surge in incidence of severe AH, secondary to mental health disorders arising during the covid-19 pandemic, have placed enormous additional demands on the limited supply of donor organs, leading to prolonged waiting times for patients with other liver disorders on transplant waiting lists.157 Finally, a combination of legislation and professional attitudes make access to transplantation unavailable in some countries or centers. The UK, for example, currently has no program for transplantation for patients with AH.
Emerging treatments
As our understanding of the pathogenesis of AH improves, new treatment options are emerging. Therapeutic manipulation of the gut microbiome is a rational approach to treatment, although the sophisticated tools required for precision treatment are not yet available. However, early studies with fecal microbial transplantation (FMT) have been promising. A trial in India of FMT derived from healthy relatives in patients with AH and contraindications to steroids demonstrated remarkable improvement in survival relative to historical controls.158 Clearly these results need to be replicated in prospective randomized trials, and it will be important to identify the crucial mechanisms conferring benefit to develop more specific interventions. The identification of the E faecalis cytolysin toxin provides one opportunity for more precise treatment targeting. Potentially this strain of E faecalis may be targeted using highly selective phages which could eliminate the bacterium from the host microbiome.30
Alternative targets for treatment are the regenerative failure and epigenetic dysregulation. Larsucosterol is an epigenetic modifier which has been evaluated in a phase 2 trial in patients with severe AH.159 Early responses, judged by the Lille score, were superior to historical controls, providing support for an ongoing phase 3 study. Interleukin 22 (IL-22) is a pleotropic cytokine which stimulates both intestinal and hepatic epithelial regeneration. The IL-22 agonist F-652 has been evaluated in a phase 2 trial in AH demonstrating superior Lille responses and MELD score improvements relative to propensity-matched historical controls.160
Guidelines
Several specialist societies and healthcare bodies produce and maintain clinical management guidelines for ALD. The latest guidelines from the European Association for the Study of the Liver (EASL),3 American Association for the Study of Liver Disease (AASLD),5 American College of Gastroenterology (ACG)161 and Latin American Association for the Study of the Liver (ALEH)162 are summarized and compared in Table 2.
A comparative assessment of recommendations from different specialist society guidelines on the management of alcohol-related liver disease (ALD)
Conclusions
The link between alcohol excess and the development of liver disease is long established. However, population-level changes in drinking behaviors and a lack of concerted global and national public health interventions mean that ALD disease burden is rising. Significant recent efforts and technological developments offer the opportunity to use non-invasive tools to screen for, diagnose, and stage ALD before the development of cirrhosis, giving a window for intervention before the development of established disease. Alcohol abstinence is and always will be the mainstay of disease therapy. Brief interventions help reduce harmful alcohol use, but a more holistic and integrated approach to the management of alcohol use disorders, incorporating primary care, mental health, addiction and liver specialists with consideration of pharmacotherapy, is likely necessary to improve outcomes in those with severe dependence. Beyond this, there is a total absence of disease-modifying treatments in ALD. As the body of knowledge regarding the pathogenetic mechanisms increases, the opportunity to identify new targets for therapies does too. The shared genetic liability with non-alcoholic fatty liver disease, another liver disease with significant unmet clinical need, hints at the possibility that certain treatments may show common benefit in both conditions. Translating these opportunities into successful treatment strategies will require an enabling environment, which, in large part, requires a commitment from policymakers to give ALD the attention and resource commensurate to its prevalence and associated disease burden.
Questions for future research
Are we able to develop and implement early diagnosis strategies to prevent progression to identify harmful drinking behavior and prevent end-stage liver disease?
Can we develop a strategy to reduce stigmatization in alcohol use disorder among medical professionals and public?
Can the intestinal microbiome be manipulated to treat or prevent alcohol-related liver disease without resorting to fecal microbial transplantation?
Is it possible to develop pharmacotherapy to preserve or restore liver function while awaiting resolution of alcohol use?
Can we improve the pharmacotherapy for alcohol use disorder in patients with advanced liver disease?
Glossary of abbreviations
AASLD: American Association for the Study of Liver Disease
AH: Alcohol-related hepatitis
ALD: Alcohol-related liver disease
APRI: AST to platelet ratio index
ASH: Steatohepatitis due to alcohol-related liver disease
AUD: Alcohol use disorder
CIWA: Clinical Institute Withdrawal Assessment for Alcohol protocol
DSM: Diagnostic and Statistical Manual of Mental Disorders
EASL: European Association for the Study of the Liver
FIB-4: Fibrosis-4
HCC: Hepatocellular carcinoma
ICD: International classification of disease
LSM: Liver stiffness measurement
MOR: Mu opiate receptor
MUP: Minimum unit price
TLR: Toll-like receptor
VTA: Ventral tegmental area
Acknowledgments
We acknowledge support from the NIHR-Imperial Biomedical Research Centre. We are grateful for critical input from Dr Stephen Atkinson.
Footnotes
State of the Art Reviews are commissioned on the basis of their relevance to academics and specialists in the US and internationally. For this reason they are written predominantly by US authors
Contributors: All authors contributed to the planning, writing and review of the article. The corresponding author acts as the guarantor for the article and attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.
Competing interests: ALH declares honorarium from Lundbeck for lecture/training and royalties from editorship of Edwards' Treatment of Drinking Problems: A Guide for the Helping Professions. MT declares consultancy fees from GSK, Durect, Intercept, and Surrozen.
Patient involvement: No patients were asked for input in the creation of this article.
Provenance and peer review: Commissioned; externally peer reviewed.