Alcoholic liver disease and how to avoid them

Cirrhosis is present in 12.9% of patients with alcohol use disorder (AUD), defined by the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) as a maladaptive pattern of substance use leading to clinically significant impairment or distress, with severity being graded based on the number of diagnostic criteria met. Once decompensated, advanced ALD is easy to identify, while less severe forms often go unrecognised due to the clinically silent period that precedes decompensation.⁸ As such, there is an urgent need to identify at-risk patients early using validated questionnaires, such as CAGE and AUDIT, with proven high sensitivity and specificity in multiple population settings.¹⁰ Shorter versions, such as AUDIT-C, can also be used as screening tools for “risky drinking”.¹¹ After the identification of at-risk patients, further investigations with liver function tests (LFTs) such as serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST), gammaglutamyl transpeptidase (GGT) and blood cell count with mean corpuscular volume should be performed along with tests representing liver synthetic function such as INR and albumin levels, albeit taking into consideration their low sensitivity and specificity.^12,13 Blood tests should always be accompanied by measurement of liver stiffness since advanced liver cirrhosis may present with normal LFTs.7 In fact, it is well demonstrated that patients with ALD are diagnosed at more advanced stages of liver disease in comparison to other aetiologies.^14,15 To address this issue, the National Institute for Health and Care Excellence (NICE) recommends that all patients with harmful drinking habits undergo screening with transient elastography and repeat every 2 years if ALD is present or AUD persists.¹⁶ Early recognition and referral are cost-effective.¹⁷

Case study
In 2016, a 38-year-old male patient was admitted with clinical evidence of severe alcoholic hepatitis (AH). The patient drank about 200 g of alcohol a day, with frequent episodes of binge drinking. Two years before, he experienced a similar episode that resolved with corticosteroid treatment and abstinence. However, abstinence was not sustained, and he relapsed into heavy drinking. Once more, the patient recovered with supportive treatment and abstinence, but this time, he was transferred to an alcohol rehabilitation clinic, where he remained for 4 weeks. This course of action was very effective; the patient managed to become abstinent and remains so at present. He also got a job, got married, gained an additional 10 kilos in weight, and currently has no clinical evidence of cirrhosis.

Alcohol abstinence is a critical goal in the management of ALD patients, as it improves prognosis across all stages, reducing decompensation rates and improving survival, even if advanced liver disease or severe portal hypertension are present.¹⁸ Degré et al. demonstrated that following an episode of AH, 45% of patients relapsed after 3 months and 70% within 5 years of follow-up. In fact, up to 60% of patients who relapsed died during the follow-up.¹⁹ It is very difficult to ascertain abstinence, and there is an unmet need for good biomarkers. The alcohol biomarker phosphatidyl ethanol (PEth) is probably one of the most promising, but it is not available in most centres.²⁰

There has also been evidence supporting the use of brief interventions to reduce alcohol consumption on a case-by-case basis.^21,22 Whenever a patient with AUD is identified, a brief intervention using the 5 A’s model (Ask about use, Advice to quit or reduce, Assess willingness, Assistance, Arrange follow-up), paired with a motivational component, can reduce drinking by an average of 57 g per week.²² The focus should be on establishing goals that the patient agrees on, with an empathetic and collaborative perspective on patient’s expectations. Medical therapy with naltrexone, acamprosate and nalmefene can be seen as useful adjuncts, but the benefits tend to be moderate and should not be the sole treatment.^23,24

On a population level, strong advocacy efforts for policies that focus on pricing, limitation of access and marketing of alcoholic beverages are ongoing and need to be reinforced. Minimum unit pricing (MUP) has been shown to reduce alcohol consumption in British Columbia by up to 32% after one year of implementation, with a decrease of 8.95% and 9.22% in acute and chronic alcohol-attributable hospital admissions, respectively.²⁵ In a recent systematic review, it was shown that doubling alcohol taxes or introducing MUP reduced consumption by 10%.²⁶ Other policies suggested by the WHO that have proven benefits are advertising bans, restrictions on alcohol purchase through minimum legal age, and laws punishing driving under the influence of alcohol.^26,27

In patients with AUD, the most likely aetiology of cirrhosis is ALD, but approximately 20% of AUD patients with abnormal LFTs have other competing aetiologies. Levin et al. demonstrated through consecutive liver biopsies of patients with >80 g of daily alcohol consumption that only 80% of patients presumed to have ALD had histological features to support the diagnosis.²⁸ Due to the low sensitivity of clinical signs and medical history in distinguishing aetiologies of chronic liver disease (CLD), EASL recommends that patients with abnormal LFTs and AUD be routinely screened for other causes of CLD, including testing for hepatitis B virus (HBV) and hepatitis C virus (HCV) serology, autoimmune markers, transferrin, and transferrin saturation, a1-antitrypsin and in some cases also caeruloplasmin.⁷

A frequent concurrent liver noxa is the presence of cardiometabolic risk factors. The new nomenclature of steatotic liver disease (SLD) introduces the concept of metabolic and alcohol-related/associated liver disease (MetALD), where increased use of alcohol (>20g/day or >140g/week in women and >30g/day or >210g/week in men), coexists with cardiometabolic risk factors.²⁹

No epidemiological data is yet available with this new classification in Europe, but in the USA, the National Health and Nutrition Examination Survey, conducted between March 2017 and March 2020, demonstrated that the age-adjusted prevalence of MetALD was 15.2%, 2.5% of which presented with clinically significant fibrosis (liver stiffness > 6.5 kPa).³⁰ A recent Swedish register-based cohort study showed that in presumed metabolic dysfunction-associated steatotic liver disease (MASLD) patients, the concurrent diagnosis of ALD and AUD is common (17.2%) and that these patients have a higher likelihood (adjusted Hazard Ratio of 5.8) of major adverse liver outcomes in comparison to patients with only MASLD.³¹

Other concurrent aetiologies of liver disease may be frequent in patients with excessive alcohol consumption and lead to a worse prognosis, thus demanding systematic exclusion.

Case study
During my first year of residency, we received a patient who flew from Angola for evaluation of hepatic encephalopathy. He was abusing alcohol, with obvious cirrhosis manifested with ascites and grade II hepatic encephalopathy. He also had a transferrin saturation of 55% and a ferritin level of 352 ng/ml, so I hypothesised he would have haemochromatosis alongside alcoholic cirrhosis. Obviously, I was wrong, and the not-so-inexpensive homeostatic iron regulator (HFE) gene test produced a negative result.

The discovery of the genetic basis of haemochromatosis has brought some clarification to the once confusing high prevalence of siderosis in ALD.³² Iron metabolism markers, such as ferritin and transferrin saturation, proved to be elevated in one-third to two-thirds of ALD patients.^32,33 This reflects an iron overload status that may be, at least in part, explained by the negative regulation of hepcidin.³⁴ Actually, ALD patients heterozygous for the C282Y mutation in the HFE gene failed to show increased hepatic iron stores when compared with ALD patients homozygous for the wild-type allele.³² The same is not so clear for patients carrying the H63D mutation.³⁵ Conversely, the finding that haemochromatosis patients are often excessive drinkers is explained by the fact that alcohol acts as a potent co-factor in the development of cirrhosis.³⁶ Thus, although we can state that there is an association between haemochromatosis and alcohol abuse, this seems to represent not a cause-and-effect relationship but rather an aggregation of co-factors in advanced liver disease patients.

It is valuable to remember that the magnitude of ferritin elevation is usually lower in ALD patients (10–500 μg/L) than in homozygous haemochromatosis patients (500–10,000 μg/L).³² The same is true for increased transferrin saturation values in ALD (20–60%) and homozygous haemochromatosis (60–100%).³² In conclusion, cut-off values may need adjustment to a higher threshold in ALD patients, and a clinical suspicion must be supplemented by other clinical clues (e.g. ethnicity, active versus inactive alcohol consumption). A good strategy may be to reevaluate iron parameters only after several weeks of abstinence, and only if they remain elevated should hemochromatosis be considered.

Case study
A few years ago, a 38-year-old female was admitted with jaundice in the setting of excessive alcoholic intake. She had no previous history of liver disease but stated a prolonged 80 g/day alcohol consumption pattern. In the setting of macrocytosis, mildly elevated transaminase levels (AST>ALT) and elevated total bilirubin levels, this drinking pattern supported the diagnosis of AH (Maddrey score of 23). The patient also presented with mild thrombocytopenia and equivocal abdominal ultrasound findings (diffuse hyperechogenicity, hepatomegaly and mild ascites) but no oesophageal varices nor other stigmata of portal hypertension. A diagnosis of cirrhosis was unclear, and elastography revealed a stiffness value of 47 kPa. After discharge and several months of sustained abstinence, liver stiffness values decreased to 8 kPa, and no stigmata of cirrhosis were identifiable. In addition, the platelet count, which was low at admission (102,000/mcL), normalised after 3 weeks of abstinence.

Alcohol intake has been associated with elevated liver stiffness, with studies showing an average decline in liver stiffness values of 10% with abstinence.^37,38 Thus, the presence of fibrosis may be overestimated by liver stiffness values by as much as 27%.³⁹ Significant inflammation and congestion, as is found in AH, may indeed explain the elevated liver stiffness in these patients. Fernandez et al. demonstrated a positive correlation between AST and liver stiffness levels, which may be more pronounced in advanced stages of fibrosis.^40,41 Another probable consequence of this inflammatory milieu present in AH liver is an increase in both sinusoidal resistance and flow, resulting in portal hypertension even in the absence of cirrhosis.⁴²

Another well-known toxic effect of alcohol is direct (but reversible) myelosuppression, leading to ineffective megakaryopoiesis, which, in conjunction with a reduced platelet lifespan, results in worsening thrombocytopenia.⁴³ Summing up, liver stiffness values and thrombocytopenia should always be interpreted carefully in the setting of active alcohol intake, especially if they indicate the possibility of significant fibrosis/cirrhosis. It is thus always best to repeat measurements after abstinence and reduced inflammation, preferably after a reduction of AST to <100U/L.³⁹

Case study
A male patient with advanced chronic liver disease (ACLD), presumably associated with chronic hepatitis C (non-viraemic after treatment one year before), was admitted to the emergency department with haematemesis. Oesophageal variceal bleeding was confirmed and successfully managed by band ligation. During rounds the next morning, the patient presented with jaundice (total bilirubin 5.4 mg/dL) and mildly elevated aminotransferase levels (with an AST:ALT ratio >2), thrombocytopenia (45,000/mcL) and leukocytosis (12.4 x109 cel/L). Over the days that followed, his bilirubin level went up (19 mg/dL), and his international normalised ratio (INR) and creatinine levels also worsened. The patient was then in grade III-IV hepatic encephalopathy. The clinical question followed: why he would be in acute-on-chronic liver failure (ACLF)? Is he infected? What about alcohol? His previous registries stated he misused alcohol. On admission, the patient denied excessive alcohol consumption, but his wife confirmed he had gone back to drinking heavily. So, is this, in fact, AH?

AH is a clinical syndrome, and jaundice is its cardinal symptom. Most of its clinical features were present in this patient (jaundice, AST:ALT elevation 2–5 times the upper limit of normal (ULN) in a ratio of 2:1 and an inflammatory response syndrome) and can appear in active drinkers or patients who have very recently ceased alcohol use.⁷ Severe AH is one of the main precipitating events for the development of ACLF. The clinical picture may worsen with progression to ACLF and is associated with complications such as gastrointestinal bleeding, encephalopathy and hepatorenal syndrome. Histologically proven AH is estimated to comprise 6% of patients with decompensated cirrhosis and up to a quarter of ACLF patients.⁴⁴ Diagnosis based on classic clinical criteria such as those mentioned above were found to misdiagnose AH in up to 25% of cases,^45,46 thus, screening for alcohol consumption is essential, using questionnaires such as CAGE and AUDIT.

As the aforementioned clinical case shows, diagnosis may be difficult to establish due to its similarity with ACLF alone. Therefore, a biopsy should be done in equivocal cases.⁷ Figure 1 shows the classic histopathology findings in AH. A transjugular route should be preferred due to the increased risk of bleeding in these patients. A recent study comprising of 184 patients with AH showed that ductular bilirubinostasis (DB) was the only histological feature more frequently present in patients with ACLF compared to patients without ACLF (50.7% vs 30.6%, p = 0.003).⁴⁷ Interestingly, Altamirano et al. verified that the presence and type of bilirubinostasis have a marked influence on short-term mortality.⁴⁸

Figure 1 | Histopathology images of a liver biopsy sample from a patient with alcoholic hepatitis. Haematoxylin and eosin (H&E) stain. Asterisks denote Mallory bodies and full arrows denote steatosis.

a | A PMN infiltrate (double arrowhead) can be seen surrounding the Mallory bodies and sclerosing hyaline necrosis is also visible (rhombus); steatosis is mainly microvesicular.

b | A hepatocyte with ballooning degeneration (arrowhead); steatosis is mainly macrovesicular. Image courtesy of Dr Adília Costa.

The cornerstone of AH treatment is abstinence. Many therapies have been proposed, such as nutritional treatment, pentoxifylline and N-acetylcysteine, but the best evidence-based therapy, although controversial, is corticosteroid therapy. The evidence regarding steroids in severeAH severeAH has been conflicting, with multiple studies and meta-analyses demonstrating diverging findings.^49–51 The most consistent finding is that prednisolone reduces 28-day mortality in patients with high Maddrey Discrimination Function (MDF) (>32 points) or concomitant hepatic encephalopathy. Simultaneously incorporating other scores, such as MELD and ABIC, further improves severity stratification and prognosis.⁵² The MDF >32 group can be further stratified into responders by a Lille model score <0.45 by day 7 to help the clinician decide if they should maintain prednisolone.^53,54 Foncea et al. showed that Day 4 Lille score predicted responders in 90% of patients when compared to Day 7 Lille and could be used to avoid prolonged futile use of corticoid therapy,⁵⁵ but these findings have not yet been incorporated into current guidelines.

Despite the apparent short-term improvement in mortality with prednisone, the STOPAH trial showed that corticosteroids do not seem to influence medium-term (90 days) and long-term (1 year) mortality, a finding consistent with other previous studies.⁵⁶ The major setback of corticosteroid use is the increased risk of infection and worsening/lack of efficacy in the setting of concomitant complications, such as gastrointestinal haemorrhage and hepatorenal syndrome.⁵⁶ Concurrent infection is not an absolute contraindication to corticoid therapy, although it is a feared complication, as the three-month incidence of infectious complications is reported to be up to 65%,⁵⁷ carrying a 30% increase in mortality.⁵² Thus, when evaluating a candidate for corticosteroid treatment, a complete and thorough sepsis work-up should be undertaken and sequentially repeated during the duration of treatment.⁷ If the sepsis work-up is negative, then therapy should not be postponed based on elevated inflammatory markers since, in AH, they are a sign of systemic inflammatory response syndrome and do not always correspond to occult infection. Louvet et al. demonstrated that patients whose infection is diagnosed on initial evaluation and successfully controlled by appropriate antibiotic therapy may be given corticosteroids without increasing mortality.⁵⁷ Interestingly, when infection develops during corticosteroid therapy, it is associated with higher mortality, but this is only significant in the group of responders.⁵⁶ Given that 90% of infections in AH are bacterial, prophylactic antibiotic therapy has been proposed, but recent clinical trials showed no benefit in adding amoxicillin–clavulanic acid to prednisone.⁵⁸ In conclusion, a thorough sepsis work-up should be performed (prior to and during treatment with corticosteroids), and treatment with prednisone 40mg/day given to those who are eligible (MDF >32 points or concomitant hepatic encephalopathy) and reevaluated with the Lille score on the 7th day of treatment. Expectations regarding long-term outcomes should be tempered, and abstinence should always be reinforced as the most efficacious measure.

The nutritional status of patients with ALD is of paramount importance. Up to 50% of outpatients with advanced ALD and most hospitalised patients with AH have evidence of clinically significant nutritional depletion.^59,60 The pathophysiology is multifactorial, with reduced appetite due to upregulation of inflammation, intestinal dysfunction due to small bacterial overgrowth, epithelial damage and pancreatic insufficiency, and nutritional imbalance due in part to the empty calories from alcohol.⁶¹

Protein-energy malnutrition as defined by altered body composition due to an imbalance of energy, protein and micronutrients is frequent in ALD patients and can occur at any stage.⁶² These patients have higher rates of infectious complications and liver decompensation, and malnutrition is a recognised independent predictor of survival in cirrhosis and after liver transplantation.⁶²

EASL released its first guidelines on nutrition in chronic liver disease in 2019 and recommends screening for malnutrition in patients with cirrhosis, irrespective of body mass index (BMI).⁶²

There is no gold standard for nutritional assessment and no specific tool designed for ALD patients. BMI correlates poorly with nutritional status due to anasarca and sarcopenic obesity.⁶³ The Liver Disease Undernutrition Screening Tool⁶⁴ and the Royal Free Hospital Nutritional Prioritisation Tool⁶⁵ have both shown good sensitivity, are easy to implement, and have been shown to relate to clinical outcomes.⁶¹

Nutritional counselling by a multi-disciplinary team should be advised to all cirrhotic patients with signs of malnutrition, and a nutritional plan should be established.⁶² Some general rules apply when counselling the patients in our practice, such as maintaining a regular eating pattern with meals or snacks every 2-3 hours, avoiding protein restriction, implementing a bedtime snack, and avoiding adding salt to food.

There is also increasing evidence of the benefits of exercise, although previously, there was some controversy related to the theoretical risk of increased portal pressures during exercise and the potential harm of inducing variceal bleeding. In a systematic review it was found that in selected patients with cirrhosis, there was a benefit of exercise in endurance and functional outcome measures without adverse effect. In fact, since sarcopenia is so important in advanced liver disease, this beneficial effect is to be expected.⁶⁶