Article Text

Download PDFPDF

COVID-19 and liver disease
Free
  1. Jean-François Dufour1,
  2. Thomas Marjot2,3,
  3. Chiara Becchetti4,5,
  4. Herbert Tilg6
  1. 1Hepatology, Department of Biomedical Research, University of Bern, Bern, Switzerland
  2. 2Oxford Liver Unit, Oxford University Hospitals NHS Foundation Trust, John Radcliffe Hospital, Oxford, UK
  3. 3Nuffield Department of Medicine, Translational Gastroenterology Unit, University of Oxford, Oxford, UK
  4. 4Department of Hepatology and Gastroenterology, ASST Grande Ospedale Metropolitano Niguarda, Bern, Italy
  5. 5Department of Visceral Surgery and Medicine, Inselspital, University Hospital of Bern, Bern, Switzerland
  6. 6Department of Internal Medicine I, Gastroenterology, Hepatology, Endocrinology and Metabolism, Medical University Innsbruck, Innsbruck, Austria
  1. Correspondence to Professor Jean-François Dufour, University of Bern, Bern 3012, Switzerland; jf.dufour{at}svmed.ch

Abstract

Knowledge on SARS-CoV-2 infection and its resultant COVID-19 in liver diseases has rapidly increased during the pandemic. Hereby, we review COVID-19 liver manifestations and pathophysiological aspects related to SARS-CoV-2 infection in patients without liver disease as well as the impact of COVID-19 in patients with chronic liver disease (CLD), particularly cirrhosis and liver transplantation (LT). SARS-CoV-2 infection has been associated with overt proinflammatory cytokine profile, which probably contributes substantially to the observed early and late liver abnormalities. CLD, particularly decompensated cirrhosis, should be regarded as a risk factor for severe COVID-19 and death. LT was impacted during the pandemic, mainly due to concerns regarding donation and infection in recipients. However, LT did not represent a risk factor per se of worse outcome. Even though scarce, data regarding COVID-19 specific therapy in special populations such as LT recipients seem promising. COVID-19 vaccine-induced immunity seems impaired in CLD and LT recipients, advocating for a revised schedule of vaccine administration in this population.

  • COVID-19
  • cirrhosis
  • liver transplantation

Statistics from Altmetric.com

COVID-19: liver manifestations and pathophysiological aspects

Hepatic manifestation of COVID-19 infection in patients without liver disease

It has been recognised in the very first clinical reports on COVID-19 that the liver is affected by SARS-CoV-2 infection. Early clinical studies have claimed that SARS-CoV-2 infection was accompanied by acute liver injury (ALI) as reflected by an increase especially in aspartate aminotransferase (AST) and less in alanine aminotransferase (ALT).1–4 Elevated transaminases affected up to 50% of infected subjects and clearly correlated with severity of disease.5 Transaminase elevation during COVID-19 is mostly mild and reversible and elevations are usually <5 times the upper limit of normal in up to 80% of infected patients.6 Severe ALI (>20 upper limit of normal transaminase levels) is uncommon occurring in <0.1% of infected patients.7 ALI reflected by increased transaminases and bilirubin was related to disease severity and poor outcome of COVID-19.8 Increases in gamma glutamyl transferase (γGT) and alkaline phosphatase (AP) are less frequently observed and found more in the later course of disease, whereas moderate increases in AST, ALT and bilirubin are a very common feature in COVID-19 disease.9 Rare cases of acute liver failure caused by SARS-CoV-2 have been reported.10 Other rare clinical case presentations with liver involvement include febrile hepatitis,11 acute cholecystitis with detection of viral RNA in the gallbladder wall,12 hepatic artery thrombosis13 or an entity called post-COVID-19 cholangiopathy, which is characterised by severe cholestasis and structural bile duct alterations after recovering from life threatening disease.14–16 Interestingly, there have been a very few reports suggesting that COVID-19 infection might reflect ‘a final hit’ leading to de novo manifestation of immune-mediated liver disease such as primary biliary cholangitis (PBC) (one case)17 or autoimmune liver disease (two cases).18

Detection of SARS-CoV-2 in the liver: is it really there?

As one of the key host receptors for this virus, that is, ACE2 is expressed in the liver, especially in cholangiocytes, it has been assumed in the early phase of the pandemic that SARS-CoV-2 might be detectable in the liver.19 Liver progenitor cells also express the SARS-CoV-2 priming molecule transmembrane serine protease 2 (TMPRSS2).20 Wang and colleagues detected in postmortem liver biopsies from two subjects by performing ultrastructural examination via electron microscopy typical coronavirus particles in the cytoplasm of hepatocytes.21 Another early study described hepatic steatosis, ductular fibrosis, acute liver cell necrosis, lobular cholestasis, central vein thrombosis and lymphocytic infiltrates as key features in an autopsy study of COVID-19.22 Chornenkyy and colleagues reported detection of SARS-CoV-2 by RT-PCR using formalin-embedded liver tissue and mild focal portal tract inflammation.23 A larger study investigated liver tissues from 60 patients.24 SARS-CoV-2 was detected at RNA and/or protein levels (nucleocapsid–protein) in 22% of patients and 2/5 bile samples were positive for SARS-CoV-2 RNA. Besides macrovesicular and microvesicular steatosis, microthrombotic pathology dominated in the liver accompanied by activation of intrahepatic stem cell compartment and features of aberrant regeneration.24 Not all liver histopathological studies showed detectable virus particles in the liver.25 In this autopsy study, 42% of patients (n=24) exhibited evidence of severe hepatic necrosis, obviously a key histological feature besides macrovesicular/microvesicular steatosis, data which are also supported by a recent systematic review.26 Interestingly, certain viral antigens (nucleocapsid and spike protein) were detectable in the liver up to 6 months after recovering from COVID-19 disease.27 A single-cell atlas study investigating autopsy tissue samples observed that viral RNAs are enriched in mononuclear phagocytes and endothelial lung cells but viral RNA was not detectable in the heart, liver or kidneys.28 Wanner and colleagues detected SARS-CoV-2 viral RNA in 69% of 45 autopsy cases and, furthermore, were able to isolate infectious SARS-CoV-2 from 2 out of 3 autopsy livers and lungs, proposing potential infectivity of postmortem liver tissue.29 Therefore, there exists some evidence that SARS-CoV-2 might infect the liver at least in severe COVID-19 disease.

Pathophysiological concepts: a ‘cytokine storm’ is attacking the liver

It remains unclear whether hepatic involvement during COVID-19 reflects a direct cytopathic effect of the virus, an exaggerated systemic immune response or a combination of insults, including drug-induced liver injury (figure 1). COVID-19 disease is associated with a ‘cytokine storm’ and a massive acute-phase reaction defined by the release of very high levels of the proinflammatory cytokines, tumour necrosis factor (TNF), interleukin 1 (IL-1) and IL-6 paralleled by very high levels of C reactive protein (CRP) and ferritin.30 The nature of this ‘cytokine storm’ is still poorly understood. SARS-CoV-2 is able to activate C-type lectins and Tweety family member 2 by the spike protein on myeloid cells resulting in a strong proinflammatory cytokine response.31 A large-scale single-cell transcriptome atlas revealed that SARS-CoV-2 was found in both epithelial and immune cells, and megakaryocytes and certain monocytic cells in the peripheral blood were identified as major cytokine producers.32 SARS-CoV-2 spike protein interacts with the CD42 receptor to activate platelets and thereby promotes platelet–monocyte interactions via P-selectin/PGSL-1 and CD40L/CD40 to result in a hypercytokine secretion by monocytes.33 Another involved pathway as key driver of systemic inflammation by this virus could reflect complement hyperactivation controlled by the interferon (IFN)–Janus kinase 1/2 (JAK1/2)–STAT1 signalling system and NF-kB.34 Hyperinflammatory severe lung disease in COVID-19 can be due to mutations that affect type I IFN immunity35 or autoantibodies against type I IFNs36 proposing that this disease has features of an autoimmune disease caused by deficient type I IFN immunity.37 The importance of the IFN pathway has also been shown in severe COVID-19 (autopsy cases), where livers exhibited a significant upregulation of type I and II IFN response and IFN-related JAK–STAT signalling.29 Overt inflammation and cytokine activation in this disease may also contribute to vascular damage accompanied by endothelial injury, hypercoagulation and arterial/venous embolism paralleled by activation of various immune cells and platelets finally causing clot formation.38 39

Figure 1

Pathophysiological concepts of SARS-CoV-2 liver infection. SARS-CoV-2 infection is frequently associated with ALI evidenced by increased transaminases in patients without and with prior liver disease. Key host receptors for this virus such as ACE2 receptor and the SARS-CoV-2 co-receptor transmembrane serine protease 2 (TMPRSS2) are expressed on various liver cells such as hepatocytes, cholangiocytes, immune cells or liver progenitor cells. ALI is associated with increased hepatic and systemic cytokine production such as IL-1, TNF or IL-6. Furthermore, severe SARS-CoV-2 infection is characterised by an upregulation of hepatic type I and II IFN response and associated JAK–STAT signalling. Monocytes and macrophages are considered the major cytokine producers in the so called ‘cytokine storm’ frequently observed in SARS-CoV-2 infection. Massive cytokine production contributes to liver cell death. ALI, acute liver injury; ALT, alanine aminotransferase; AST, aspartate aminotransferase; JAK, Janus kinase; IFN, interferon; IL-1, interleukin 1; IL-6, interleukin 6; TNF, tumour necrosis factor.

Several studies have now reported that ALI correlates with degree of systemic inflammation as for example reflected by high-sensitivity CRP (hsCRP) values.40 41 In a large cohort of patients with COVID-19, systemic inflammation as detected by increased levels of IL-6, hsCRP or ferritin was highly correlated with degree of ALI as assessed by AST levels.42 This does, however, not rule out a direct cytopathic effect of SARS-CoV-2 onto the liver but proposes that systemic inflammation might contribute to ALI. Such a concept is also supported by the fact that AST elevations take place very early in the course of infection. Increased cholestatic enzymes (γGT and AP) in the later course of disease might also be mediated by massive systemic inflammation as proinflammatory cytokines cause hepatocellular cholestasis by downregulation of hepatobiliary uptake and excretory systems.43 To summarise, SARS-CoV-2 infection has been associated with a highly unusual and overt proinflammatory cytokine profile, that is, ‘cytokine storm’, suggesting that this tremendous proinflammatory attack contributes substantially to observed early and late hepatologic abnormalities, although it can currently not be ruled out that a direct cytopathic effect of this virus also contributes to early pathologies (figure 1).

Key messages box

COVID-19: liver manifestations and pathophysiological aspects

  • Elevated transaminase affects up to 50% of infected subjects and correlates with severity of disease.

  • Severe ALI is rare and correlates to disease severity and poor outcome of COVID-19.

  • Cholestatic pattern is less frequent in COVID-19 and found more in the later course of disease.

  • Liver injury during SARS-CoV-2 infection is probably multifactorial, including mechanisms as direct cytopathic effect of the virus, exaggerated systemic immune response, vascular damage and coagulopathy and drug-induced liver injury.

COVID-19 in patients with chronic liver disease

Cirrhosis

Outcomes

With the onset of the pandemic, the hepatology community worked quickly to define the risk of SARS-CoV-2 acquisition and adverse COVID-19 outcome conferred by pre-existing chronic liver disease (CLD). Data from large case series and population-level electronic healthcare records during the first global surge suggested that patients with CLD and cirrhosis were not over-represented, implying that these conditions did not increase susceptibility to infection.44 45 Indeed, one large North American study found that patients with cirrhosis had lower risk of SARS-CoV-2 positivity, which likely reflected heightened vigilance, increased testing and greater patient adherence to public health advice.46 However, once infected, it has become clear that patients with cirrhosis are at an increased risk of adverse COVID-19 outcomes, including death. This has been established through multiple strands of evidence, including international registry findings, large observational cohorts and population-level data.

Overall, mortality in patients with cirrhosis following SARS-CoV-2 infection was found to be 32% in a large registry cohort of 729 patients across 29 countries with case fatality incrementally increasing with each Child-Pugh (CP) class (CLD without cirrhosis: 8%, CP-A: 19%, CP-B: 35%, CP-C: 51%).47 Similar stepwise trends were observed in the rates of intensive care unit (ICU) admission, renal replacement therapy and invasive mechanical ventilation. Furthermore, the risk of mortality in those with decompensated cirrhosis (CP-B and CP-C) was significantly elevated compared with contemporaneous patients without cirrhosis testing positive for SARS-CoV-2 after matching for age and comorbidity. Reports of high COVID-19 mortality in cirrhosis have also been replicated both in an exclusively Asian registry48 and in several multicentre cohort studies across different geographical regions.49–51 In a retrospective study in Northern Italy during the early phase of the pandemic, Iavarone et al reported a 30-day mortality of 30%, which was significantly higher than a historical cohort of patients with cirrhosis hospitalised with bacterial infection.49 Outcome data in patients with CLD across 21 North American institutions also found that decompensated (but not compensated) cirrhosis was an independent risk factor for death.50 This study also found a sevenfold increased risk of death from COVID-19 in patients with hepatocellular carcinoma (HCC) compared with the rest of the CLD cohort suggesting that this population may be uniquely susceptible to complications of SARS-CoV-2 infection. Subsequently, a large retrospective French cohort of >259 000 inpatients with COVID-19, including >15 000 with pre-existing CLD, demonstrated that patients with decompensated cirrhosis were at an increased adjusted risk for COVID-19 mortality.52 These findings contrast with one nationwide Swedish CLD cohort which failed to demonstrate significant associations between cirrhosis and COVID-19-related mortality.53 However, this study only included patients with a histologically proven CLD diagnosis prior to 2017, and, therefore, advanced liver disease may be under-represented owing to this group not being subjected to biopsy or dying before the start of the pandemic. Lastly, population data derived from the electronic health records of >6 million UK adults have indicated an elevated adjusted HR for both hospitalisation and death from COVID-19 in patients coded as having cirrhosis.54 Taken together, cirrhosis, and particularly decompensated cirrhosis, should be regarded as a risk factor for severe COVID-19 and death.

Clinical course and longer-term prognosis

The clinical course of COVID-19 in patients with cirrhosis has several hallmarks (figure 2). First, acute hepatic decompensation is a common presenting feature occurring in up to 46% of patients, typically with new or worsening ascites and/or hepatic encephalopathy (HE).47 In 20%–58% of cases, this can occur in the absence of typical respiratory symptoms of COVID-19.47 49 Presentation with GI symptoms is more frequent in patients with CLD than matched controls47 and is associated with a more severe disease trajectory in those with cirrhosis,50 a phenomena which is replicated in the general population55 and thought to be related to greater gut permeability, electrolyte disturbance and systemic inflammatory burden. Acute-on-chronic liver failure (ACLF) is also well recognised, being reported in up to 12%–50%47–49 51 of patients with decompensated cirrhosis and COVID-19. Several well-established prognostic scoring models have been applied to cirrhosis in the context of COVID-19 with CLIF-C ACLF score and CLIF organ failure scores outperforming MELD, NACSELD and Child-Pugh score in International and Latin American cohorts, respectively.47 56 Indeed, the chances of recovery diminish rapidly in parallel with organ support requirements. For example, patients with Child-Pugh C cirrhosis have only a 21% chance of survival if admitted to ICU, dropping to 10% in the event of mechanical ventilation.47

Figure 2

COVID-19 in patients with cirrhosis. This figure summarises the clinical presentation and disease trajectory in patients with COVID-19 and cirrhosis and outlines possible underlying pathogenic mechanisms and available treatment options. ACLF, acute-on-chronic liver failure; HCC, hepatocellular carcinoma; HE, hepatic encephalopathy; mAb; monoclonal antibodies; RAS, rening–angiotensin–aldosterone. Figure made using biorender.com.

Despite SARS-CoV-2 infection triggering acute hepatic decompensation, the ultimate cause of death in patients with cirrhosis is predominantly respiratory failure (71%) followed by liver-related complications (19%).47 There are likely to be multiple overlapping mechanisms linking hepatic dysfunction and lung injury, including cirrhosis-associated immune dysfunction, coagulopathy and altered pulmonary dynamics secondary to ascites and HE.57 Given that gut microbiota composition has been shown to modulate the host immune response to COVID-19,58 it is plausible that dysbiosis and intestinal permeability associated with cirrhosis may also have a deleterious impact, although this remains to be specifically defined. Interestingly, the hallmarks of decompensated cirrhosis, including renin–angiotensin–aldosterone system activation, endothelial dysfunction and systemic inflammation, are all mirrored in the pathophysiology of COVID-19.59 These changes also drive disordered coagulation and in a large nationwide cohort study in France, Mallet et al reported a modest but significant increase in rates of pulmonary emboli in CLD versus patients with non-CLD with COVID-19, which was associated with mortality.52 In addition, this study introduced the concept of limited ‘therapeutic effort’ for patients with cirrhosis and alcohol-related liver disease (ArLD), whereby a these patients had a lower chance of mechanical ventilation and a higher risk of death. This is echoed in international registry data reporting large geographical variability in rates of ICU admission for patients with CLD and cirrhosis despite similar country-by-country rates of mortality (available as conference abstract only).60 Therefore, the elevated COVID-19 mortality in cirrhosis may be accounted for both by biological factors and the nuances of healthcare delivery. Alongside the direct effects of SARS-CoV-2 infection on patients with cirrhosis, the pandemic has had a widespread collateral impact on patients due to disruptions in healthcare provision. This has resulted in delayed presentation and hospitalisation of sicker patients with advanced cirrhosis, increased liver-related mortality and a reduction in overall patient satisfaction.59 61 The strain on healthcare resources and prioritisation of social distancing has also led to interrupted HCC surveillance programmes leading to delayed presentation and increased tumour size at diagnosis compared with the pre-pandemic era.62 63

Despite COVID-19 in patients with cirrhosis being associated with a high immediate risk of death, in those who survive the initial insult, rates of mortality and re-admission at 90 days appear comparable to those with cirrhosis alone.64 Therefore, following the acute infective period, SARS-CoV-2 infection does not seem to precipitate liver disease progression beyond the natural history of cirrhosis. Nonetheless, hepatic MRI changes, including increased T1 signalling, elevated fat fraction and hepatomegaly, have been detected in 10%–28% of otherwise healthy patients up to 4 months after recovery from acute COVID-19.65 66 The longer-term clinical implications of these radiological features following COVID-19 remains to be determined in patients with and without underlying CLD. Furthermore, these hepatic changes may not be specific to COVID-19 and may also be present in patients recovering from other severe systemic insults, although this remains underexplored and is not accounted for in current studies.

It is important to stress that much of our understanding of the disease course in patients with COVID-19 and CLD or cirrhosis is derived from studies conducted in the era preceding COVID-19 vaccination and the emergence of viral variants, including Delta and Omicron. Notably, the effect of the highly prevalent Omicron variant in patients with CLD and cirrhosis, and the modifying impact of vaccination in these individuals, remains to be elucidated.

Non-alcoholic fatty liver disease

The impact of non-alcoholic fatty liver disease (NAFLD) on COVID-19 outcomes has been heavily scrutinised due to its high global prevalence and associations with well-established risk factors for severe COVID-19, including obesity, type 2 diabetes (T2D), cardiovascular disease and hypertension.44 However, deciphering an independent effect of NAFLD on COVID-19 disease course has been challenging due to multiple confounding cofactors, reverse causality from SARS-CoV-2-induced steatosis and heterogeneity in diagnostic criteria and populations investigated. As a result, findings from clinical studies have been inconsistent. Several observational cohorts have demonstrated a significant increase in the risk of severe COVID-19 in patients with NAFLD,67–69 which is corroborated by two meta-analyses of epidemiological studies.70 71 In contrast, NAFLD was found not being significantly associated with severe COVID-19 or death after controlling for comorbidities in a Middle Eastern cohort.72 Several groups have also demonstrated a lack of association between gene variants associated with NAFLD (PNPLA3, TM6SF2, MBOAT7 and GCKR) and severe COVID-19.73 74 Indeed, one study using UK biobank data observed a possible protective immunomodulatory effect of PNPLA3, with the rs738409 G allele being associated with a reduced risk of COVID-19 mortality and hospitalisation.74 Two-step Mendelian randomisation techniques have also proven useful in examining the causal relationships between NAFLD and COVID-19 susceptibility and severity. This approach uses genetic variants as instrument variables to draw causal inferences between risk factors and health outcomes thereby overcoming challenges with confounding. Using large genome wide association datasets of European ancestry, Mendelian randomisation has revealed no evidence to support a causal relationship between NAFLD and development and severity of COVID-19.75 It, therefore, appears unlikely that the presence of NAFLD alone significantly impacts the disease course and outcome from COVID-19, although these patients are likely to be more vulnerable overall due to the presence of comorbidity and the risk of underlying cirrhosis.

Alcohol-related liver disease

Both international registry data and multicentre studies have identified ArLD as having independent associations with COVID-19 mortality after controlling for important cofactors, including liver disease severity.47 50 52 The precise mechanisms through which alcohol consumption and ArLD impact on the pathogenesis of COVID-19 remain unclear, although may be underpinned by functional immunosuppression and poor nutritional status. In addition, as with cirrhosis, patients with ArLD and COVID-19 in a large retrospective French cohort were found to be significantly less likely to receive mechanical ventilation.52 This negative association was far greater than that observed with any other comorbidity or category of Charlson Comorbidity Index suggesting that mortality in hospitalised patients with ArLD and COVID-19 may partly be explained by differential allocation of healthcare resources. These findings are alarming particularly since the incidence of harmful drinking, ArLD and alcohol-related hospital admissions has dramatically increased since the onset of the pandemic.76

Autoimmune liver disease

Understanding the clinical impact of pre-existing immunosuppression in COVID-19 remains complex. Various concerns have been raised in specific disease groups; for example, the use of maintenance corticosteroids and thiopurines has been associated with more severe disease in patients with rheumatoid conditions and inflammatory bowel disease, respectively.77 78 In contrast, the disease course in patients on immunosuppression following solid organ transplantation (SOT), including liver transplantation (LT), appears similar to non-immunosuppressed individuals (discussed below).79 80 Interval meta-analyses also suggested immunosuppressed patients are not at significantly increased risk of SARS-CoV-2 infection or severe COVID-19.81 82

In an international cohort of 70 patients with autoimmune hepatitis (AIH) and COVID-19 (86% immunosuppressed), there were no differences in rates of major adverse outcomes, including hospitalisation, ICU admission and death compared with those with other aetiologies of liver disease.83 When compared with patients without liver disease in propensity score matched analysis, patients with AIH had higher rates of hospitalisation but no increased risk of ICU admission or death. Importantly, independent risk factors for death in patients with AIH with COVID-19 were age and baseline liver disease severity but not the use of immunosuppression. Similar findings were reported in a multicentre cohort of 110 patients with AIH who also had comparable outcomes to other liver disease types despite high rates of immunosuppression.84 However, a larger retrospective study from the same group, including 254 patients with AIH with COVID-19, did show that baseline treatment with systemic glucocorticoids (median dose: 5 mg/day) or azathioprine (median dose: 75 mg/day) was associated with COVID-19 severity compared with no treatment.85 Data for patients with PBC and primary sclerosing cholangitis (PSC) are limited, although one nationwide study in Spain did observe a higher cumulative incidence of hospitalisation and mortality in patients with PBC compared with the general population although interpretations are limited by the lack of adjustment for comorbidities.86

Key messages box

COVID-19 in patients with CLD

  • Cirrhosis, and particularly decompensated cirrhosis, should be regarded as a risk factor for severe COVID-19 and death.

  • Clinical features of COVID-19 in patients with cirrhosis include acute hepatic decompensation mostly with new or worsening ascites and HE, ACLF and GI symptoms.

  • The cause of death in patients with cirrhosis is predominantly respiratory failure followed by liver-related complications.

  • Regarding specific aetiologies of liver disease, NAFLD alone probably does not significantly impact COVID-19 course and outcome, although these patients are likely to be more vulnerable due to presence of at-risk comorbidity.

  • ArLD has been associated with COVID-19 mortality after controlling for relevant cofactors, including baseline liver disease severity.

  • Immunosuppressed patients with autoimmune liver disease do not seem having a worse prognosis compare to other liver disease aetiologies.

COVID-19 and liver transplant

The COVID-19 pandemic deeply affected SOT and among others, LT. On the one hand, clinicians had to manage SARS-CoV-2 infections in both patients on the LT waiting list and recipients, while on the other hand they had to deal with potential SARS-CoV-2 positive donor offers. Both of these issues represented unprecedented challenges for all healthcare practitioners working in this field (table 1).

Table 1

Unmet needs and possible action in the setting of LT

Worldwide, there was an initial decline in LT activity across all transplant centres. In the USA, suspension of live donor transplantation were 68% in March 2020, while restrictions/suspension of deceased donor transplantation reached 73%.87 Despite a low response rate (about 23%), data from a web-based survey, including three global areas (North, Central and South America, Europe and the rest of the world), showed significant differences in the number of candidates on the waiting list between the early pandemic and the pre-pandemic periods, as well as the number of LTs performed.88 These trends from the early phase likely reflected caution about the safety of LT recipients and a reallocation of human and instrumental resources dedicated to COVID-19 healthcare. Later, as the activity of LT centres gradually turned to normality, increasing waitlist registration rates were reported in the USA, where LT rates from deceased donors even exceeded those in 2019.89 These data could be related to the relevant increase in the listing rate for LTs with alcoholic aetiology as an indirect effect of the COVID-19 pandemic itself due to widespread increase in harmful alcohol consumption.90 91 For example, combined public health data from 18 population surveys in the UK demonstrated a 25% increase in alcohol sales and 59% increase in high-risk drinking habits in 2021 compared with 2020.92 Compared with the pre-pandemic period, Cholankeril et al reported a significant increase in both ArLD listing (+7.26%; p<0.001) and LT performed (+10.67%; p<0.001) during the pandemic. Interestingly, the greatest increase occurred in young adults (+33%) and in patients with severe alcohol-associated hepatitis (+50%).91 Expanding the LT indication for acute alcoholic hepatitis in recent years93 94 has certainly contributed to this shift in the epidemiology of LT waiting lists. In the COVID-19 era, aggravating factors such as the psychological burden and delays in healthcare have contributed to an increase in alcohol use disorder, as suggested by some reports.95

Recently, there has been a shift from an initial rejection of offers of SARS-CoV-2 infected organs to a current use of such organs in the clinical practice of most centres, according to local protocols. This was supported by encouraging information on the safety profile and acquisition of SARS-CoV-2.96 97 In a worldwide survey conducted at the end of 2020, only 12%–17% of the centres transplanted organs from previously SARS-CoV-2-infected donors.88 However, more recent data suggest that SOT from such donors may be feasible and safe, at least in the short-term follow-up.98 While active COVID-19 still remains a clinical contraindication to both donation and allocation in standard activity, large-scale vaccination campaigns, more efficient therapies for COVID-19, as well as the different penetrance of SARS-CoV-2 variants are leading to the adoption of standardised protocols, tailored to available resources, for this particular allocation.99

SOT and, more specifically, LT recipients infected with SARS-CoV-2 have drawn the attention of the scientific community since the beginning of the pandemic (figure 3). Although a relatively small population, these patients are in fact frequently exposed to healthcare facilities, burdened by several comorbidities, and exposed to immunosuppression therapy (IS). IS has proven to be a controversial topic from the outset as a possible worsening or conversely therapeutic approach for SARS-CoV-2.100 Therefore, several studies aimed to investigate the clinical outcome and susceptibility to SARS-CoV-2 infection of LT recipients to date. Susceptibility of LT recipients was reported higher in European cohorts from Spain.101 However, this apparent, increased susceptibility could be a consequence of a low viral screening threshold for LT recipients. The clinical presentation of COVID-19 in LT recipients does not seem dissimilar to that of patients with non-LT, with respiratory symptoms being the pivotal ones. However, several studies highlighted a higher rate of GI manifestations in this population than in patients with non-LT (30% vs 12%, respectively; p<0.0001),102 with rates of diarrhoea ranging from 22% to 30% among the former.103

Figure 3

COVID-19 in LT recipients. This figure summarises the clinical presentation and disease trajectory in patients with COVID-19 and LT recipients. LT, liver transplantation.

Outcomes after COVID-19 in patients with SOT have often been reported and described as worse when compared with the general population.104 105 In a matched cohort of 2307 SOT recipients (of whom 240 LT recipients) compared with 231 047 non-transplant patients with COVID-19, the authors reported higher COVID-19-related mortality for the SOT group. However, propensity-matched analyses revealed that this increased risk is secondary to the higher burden of comorbidities, whereas no difference was found in terms of intubation or mechanical ventilation at 30 days (relative risk (RR): 1.04; 95% CI 0.86–1.26) or 60 days (RR: 1.03; 95% CI 0.86–1.24) between the 2 groups after adjustment for such confounders.105 Other reports confirmed that age and other comorbidities had a much greater impact on the outcomes than that conveyed by LT itself.101 102 106 107

The ongoing knowledge on the interplay between LT and IS was further clarified by other reports. Indeed, a deleterious role of mycophenolate mofetil (MMF) was suggested by an initial experience in Spain, where IS containing MMF was found to be an independent predictor of severe COVID-19 (RR: 3.94; 95% CI 1.59–9.74; p=0.003).101 On the other hand, a protective role of tacrolimus (TAC) was advocated in the European experience of Belli et al,108 where the use of TAC in the IS regimen had a positive independent effect on survival (HR 0.55; CI 95% CI 0.31 to 0.99). A meta-analysis on the available literature concluded discouraging a complete withdrawal of IS in LT recipients with COVID-19, and suggested instead to consider MMF discontinuation and replacement with other IS regimens in selected cases, based on the severity of the disease.109 However, data regarding immunosuppression should be confirmed by independent and contemporary cohorts. Limiting aspects are in fact dosage, plasma levels and duration of intake that may have influenced these results.

Subsequently, the scientific community’s attention shifted to the immunological response, both humoral and T cell mediated, of LT recipients to SARS-CoV-2. Early reports had particularly focused on the humoral response mediated by anti-nucleocapsid antibodies, detecting lower levels in LT recipients than in non-LT controls.110 However, this type of antibodies was later shown to be poorly involved in the actual response to infection.111 A more comprehensive report analysing antibodies against the nucleocapsid protein, spike protein of SARS-CoV-2 and their neutralising activity in LT recipients with confirmed SARS-CoV-2 infection (COVID-19 LT) compared with immunocompetent patients (COVID-19 immunocompetent) demonstrated that LT recipients, despite immunosuppression and less symptoms, were able to mount a detectable anti-nucleocapsid antibody titre in 80% of the cases. When the anti-spike antibody response was considered, no difference in positivity rate was found between the COVID-19-LT and COVID-19-immunocompetent cohorts (97.1% vs 100%, p=0.314). Functional antibody testing showed neutralising activity in 82.9% of LT recipients versus 100% in COVID-19-immunocompetent cohort (p=0.024).112 Anti-spike antibodies were also longitudinally monitored in the work of Caballero-Marcos et al. The authors showed no differences between COVID-19-LT recipients and controls regarding the prevalence of anti-spike immunoglobulin (IgG) antibodies at 3 (94.8% vs 96.8%; p=0.12) and 6 months after the infection (90.1% vs 94.4%; p=0.10), while a difference was observed at 1 year (88.2% vs 100%; p=0.02), although the anti-spike titre was similar between the 2 groups at each interval time considered.113 On the other hand, the small reports available on the analysis of the T cell-mediated response in LT recipients showed detectable, although rapidly declining,114 SARS-CoV-2 T cell-mediated immunity after a median of 3 months from COVID-19, with no meaningful differences with immunocompetent patients.115

Key messages box (figure 3)

COVID-19 and liver transplant

  • Following an initial decline in transplant activity, there was then an increase during the pandemic. These trends could be related to the relevant increase in the listing rate for LTs with alcoholic aetiology as an indirect effect of the COVID-19 pandemic.

  • Donors infected with SARS-CoV-2 could represent a lifesaving opportunity to implement the donor pool. However, studies that include long-term follow-up are needed.

  • In the early pandemic, LT recipient outcome was not necessarily worse than that of the general population. In addition to the purely respiratory manifestations, GI symptoms should be investigated in patients with LT.

  • The natural immunologic response to SARS-CoV-2, both serologic and T cell mediated, in LT recipients appears to be only modestly reduced compared with that of healthy controls. However, some delay in mounting this response and a more rapid decline over time were observed.

Treatment of COVID-19 in LT recipients and patients with CLD

Considerable progress has also been made in the field of COVID-19 treatment and the first experiences in SOTs appear promising. Nevertheless, there is minimal data on clinical outcomes among SOT recipients treated with differing therapies for COVID-19. As a consequence, consensus guidelines recommend that clinicians should treat SOT recipients with COVID-19 using a similar approach to non-transplant patients, considering steroids as valuable alternative to manage both COVID-19 and IS adjustment, bearing in mind the importance of drug–drug interaction (table 2).116 Although very small, a positive appraisal was observed with tocilizumab,117 while experience with convalescent plasma appeared less effective.118 In the field of antivirals, small positive experiences with remdesivir119 are reported. However, the quality of these studies is burden by several limitations (eg, small sample size and retrospective studies). Although very few data are available in the context of SOTs, the use of newly approved oral antivirals such as molnupiravir120 appears particularly promising. Paxlovid (co-packaged combination of nirmatrelvir, a second-generation protease inhibitor, and ritonavir) is expected to strong interact with calcineurin and mammalian target of rapamicyn inhibitor. Therefore, there are recommendations against co-administration.121–123 In a retrospective study, liver and kidney transplant recipients with COVID-19 treated with monoclonal antibody (bamlanivimab or casirivimab-imdevimab) experienced lower hospitalisation rate from 32% to 15% (p=0.045) and no mortality (13% vs 0%, p=0.04).124 Another report from the Mayo Clinic group confirmed comparable favourable effect in SOT, additionally reporting 10 adverse events potentially attributable to monoclonal antibody therapy, none rated as severe.125 The Food and Drug Administration has recently issued an Emergency Use Authorization (EUA) for the investigational long-acting monoclonal antibodies, tixagevimab and cilgavimab, for pre-exposure prophylaxis of COVID-19 in subjects that have either a history of severe allergy to tolerate the vaccine or moderate/severe immune compromission, comprising SOT.126 Although there is a lack of data on this specific topic, pre-exposure prophylaxis seems particularly appealing in a context of preventive strategies.

Table 2

Drugs fully authorised or authorised for emergency use in COVID-19 (table updated in mid-February 2022)121 160 161

Safety and efficacy data relating to novel and emerging COVID-19 therapies in patients with CLD and cirrhosis are limited. However, there are some specific considerations, which may influence choice of treatment agent in these patient groups. Despite Paxlovid trials demonstrating no increased risk of liver biochemistry abnormalities compared with placebo,127 both the nirmatrelvir and ritonavir component of this drug undergo extensive metabolism by hepatic cytochrome P450 enzymes leading to concerns about drug accumulation and toxicity in patients with decompensated cirrhosis. This is consistent with well-established concerns regarding the use of similar protease inhibitors in patients with decompensated HCV cirrhosis.128 Conversely, there is currently no reason to suspect that molnupiravir or remdesivir are particularly unsafe in patients with cirrhosis and clinical trials to date have not demonstrated any significant increased risk of hepatic adverse events with these medications compared with placebo.129–131 Passive immunisation using monoclonal antibodies may also be an attractive treatment option for patients with cirrhosis early in the disease course owing the increased likelihood of suboptimal humoral responses to COVID-19 vaccination (discussed below). If available, early testing of SARS-CoV-2 spike IgG antibody titres following a positive COVID-19 diagnosis may help clinical decision-making, with the greatest therapeutic benefit likely to be observed in seronegative individuals. Promising immunomodulatory therapies in the setting of established COVID-19 include JAK1/2 inhibitors (JAKI) (eg, baricitinib) and IL-6 receptor antagonists (eg, tocilizumab). Use of these agents in clinical trials for rheumatoid conditions have revealed an association with liver biochemistry abnormalities and should, therefore, be used with caution in patients with CLD and COVID-19 alongside close monitoring of liver parameters.132 133 Importantly, HBV reactivation has been documented with the use of JAKI and tocilizumab and, therefore, patients should be screened for HBsAg and anti-HBc prior to treatment initiation in order to guide decisions around prophylactic nucleoside analogue therapy.134 135

Key messages box

Treatment of COVID-19 in LT recipients and patients with CLD

  • Safety and efficacy data relating to novel and emerging COVID-19 therapies in patients with CLD, cirrhosis and LT are limited. For LT, concerns are also related to drug interactions with immunosuppressive therapy.

  • Considering antivirals drugs particular attention should be paid to co-packaged combination of nirmatrelvir and ritonavir. In patients with cirrhosis caution should be used, whereas in LT recipients should not co-administer with immunosuppressants due to their strong interactions.

  • Monoclonal antibodies have seemed very promising; however, the spread of newer variants (eg, Omicron) has made them less effective. Pre-exposure prophylaxis seems particularly promising in a context of preventive strategies in LT recipients and patients with cirrhosis, although there is a lack of data.

  • HBV reactivation has been documented with the use of baricitinib and tocilizumab and therefore patients should be screened for HBsAg and anti-HBc prior to treatment initiation.

Vaccination against SARS-CoV-2 in patients with liver disease

The rapid and large-scaled introduction of mRNA vaccines changed the pandemic trajectory protecting the population against severe COVID-19. When it comes to discuss vaccination against COVID-19 in patients with CLD and LT, two aspects have to be considered, namely, safety and efficacy. Regarding safety, vaccines against COVID-19 are not live vaccines and, therefore, can be used in immunosuppressed patients and patients with CLD and LT are not at higher risk of vaccination-related complications. Several case reports documented an elevation of the liver tests after vaccination,136 rarely some of these patients developed a clinically apparent hepatitis with features of AIH and in one case with antimitochondria antibodies.137 138 It is yet unclear whether this side effect is associated with an underlying immune-mediated disease and predictive parameters remain to be identified.

Regarding efficacy, the disease behaviour of COVID-19 is known to be substantially attenuated in vaccinated compared with unvaccinated individuals including those with CLD and LT. In a retrospective analysis of US veterans with cirrhosis, receptiont of even a single mRNA vaccine dose not only reduced rates of SARS-CoV-2 infection and but markedly improved rates of hospitalisation and death in those developing breakthrough COVID-19.139 These benefits should help cirrhotic patients to accept the vaccination. The same authors found that two doses of COVID-19 mRNA vaccine were associated with a lower incidence of COVID-19 infection, symptomatic COVID-19 and COVID-19-related death in LT recipients.140 In patients with CLD and LT recipients, vaccination against SARS-CoV-2 appears to result in favourable outcomes as attested by the absence of mechanical ventilation, ICU or death among fully vaccinated patients.141 Therefore, AASLD and EASL recommend prioritisation for vaccination of patients with CLD, including those with immune-mediated liver diseases, HCC and LT recipients.142 143 Treatment of underlying liver disease does not have to be withheld while receiving COVID-19 vaccination. LT candidates should receive the mRNA COVID-19 prior transplantation, if possible, to improve their immune response. COVID-19 vaccination is recommended in stable LT recipients as early as 6 weeks post-LT and immunosuppression should bot be reduced because of the vaccination.

In CLD and LT recipients, vaccine-induced immunity seems different compared with the general population as several reports confirmed a low humoral144 145 and T cell-mediated146 response. In the general population, the antibody response after two doses of the anti-SARS-CoV-2 mRNA Pfizer-BioNTech BNT162b2 and Moderna−1273 vaccines is excellent147 148; however, it is unsatisfactory in patients with LT and with CLD,146 149 150 with the exception of NAFLD.151 In case of AIH, the response to SARS-CoV-2 vaccine appears to be impaired in comparison to PBC and PSC even in the absence of immunosuppression.152

Immunogenicity after receiving the second Pfizer-BioNTech BNT162b2 SARS-CoV-2 vaccine dose among LT recipients was significantly lower with positive serology in only 47.5% (p<0.001). Antibody titre was also significantly lower in this group (mean: 95.41 AU/mL vs 200.5 AU/mL in controls, p<0.001).144 Thuluvath et al found that 61.3% of LT recipients and 24% of those with CLD (with or without cirrhosis) had poor antibody responses (undetectable or suboptimal).153 Guarino et al reported that 3 months after 2 doses of BNT162b2 mRNA vaccine 75% of 492 LT recipients had detectable antibodies (anti-spike protein); however, the titre was significantly lower than in the control group. Multivariate analysis revealed that factors associated with non-response were older age, shorter time from LT and IS with antimetabolites.154 A booster dose of an mRNA vaccine may increase significantly the neutralisation antibody titres against the new SARSCoV-2 B.1.1.159 Omicron variant as it does in the general population.155 A booster dose of mRNA anti-SARS-CoV-2 vaccines in LT recipients appears to be immunogenic and safe.156 157 The results of these studies on immunological response appear promising when contextualised in the current pandemic setting where hybrid immunity (mediated by infection and vaccine) becomes increasingly frequent. Encouraging data are available on the attenuation of severe forms in vaccine-LT-treated patients.158 However, considering that several reports have confirmed a low humoral-mediated and T cell-mediated response in patients with CLD and LT, a revised schedule of vaccine administration was advocated in these populations.159

Key messages box

Vaccination against SARS-CoV-2 in patients with liver disease

  • Vaccination in patients with liver disease is generally considered safe and effective and should, therefore, be strongly recommended.

  • Vaccination-related complications include elevation of the liver tests. Cases of hepatitis with features of AIH have been reported; however, the causative relationship needs to be elucidated.

  • Regarding efficacy, the disease behaviour of COVID-19 is known to be substantially attenuated in vaccinated compared with unvaccinated individuals including those with CLD and LT, however vaccine-induced immunity seems lower compared with the general population as several reports confirmed a low humoral-mediated and T cell-mediated response.

  • Revised vaccine schedules, including boosting with additional doses, seem promising in increasing humoral response in these populations.

Conclusion

In a matter of few months, the amount of knowledge collected on COVID-19 in the field of hepatology is impressive. It is important to remind that the majority of knowledge available referred to studies conducted in the era preceding COVID-19 vaccination and the emergence of viral variants, including Delta and Omicron. Bearing in mind this limitation, we can conclude that SARS-CoV-2 infection has been associated with an inflammatory state, which probably contributes to the observed liver abnormalities. Liver abnormalities have adverse prognostic implications. Risk factors for severe COVID-19 and death include CLD, especially decompensated cirrhosis. While LT does not represent a risk factor per se of worse outcome, LT patients still remain a delicate population, particularly considering the setting of vaccination. COVID-19 vaccine-induced immunity seems be impaired in CLD and LT recipients, advocating for a revised schedule of vaccine administration in this population and particular caution in both populations. As this infection remains a matter of concern, we need to continue the effort to better inform and take care of our patients with liver disease.

Ethics statements

Patient consent for publication

Ethics approval

Not applicable.

References

Footnotes

  • Twitter @dufour_jf, @tom_marjot, @BecchettiChiara

  • Contributors Each author wrote a part of the manuscript. All authors reviewed/edited the whole paper several times.

  • Funding TM receives funding via a Wellcome Trust Clinical Research Training Fellowship (reference number: 102176/B/13/Z) and has received registry grant funding from the European Association for the Study of the Liver (EASL) (reference number: 2020RG03). HT is supported by the excellence initiative VASCage (Centre for Promoting Vascular Health in the Ageing Community), an R&D K-Centre (COMET programme (Competence Centers for Excellent Technologies)) funded by the Austrian Ministry for Transport, Innovation and Technology, the Austrian Ministry for Digital and Economic Affairs and the federal states Tyrol, Salzburg and Vienna.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.