Top level research in hepatology: COVID-19 and beyond

The world has now been living with COVID-19 for more than a year; it has had an enormous impact not only on our daily lives, but also on healthcare and research. When the first wave hit the healthcare system, oncological care was, however, one of the areas of medicine that was felt to be “essential” (whatever that may imply). In the current issue of JHEP Reports, S. Muñoz-Martinez et al.¹ report on a survey among 76 high volume centres worldwide investigating the impact of the COVID-19 pandemic on the care for patients with hepatocellular carcinoma (HCC). The results are worrying: more than 80% of centres made adjustments to their screening programmes, diagnostic procedures were substantially delayed and although between 60 and 80% of centres maintained several types of treatment, most of the procedures were rescheduled, e.g. over 60% of surgical procedures. Transplant programmes were put on hold in 44% of the centres. Many centres could continue to include patients in clinical trials, although that activity was also substantially reduced. This data is worrisome, as it is not to be excluded that these changes might have an impact on the long-term outcomes of patients with HCC. The CERO-19 (Liver Cancer Outcome in the COVID-19 pandemic) project continues to collect data that will inform us on these latter issues, but the data presented here should already alarm us and be taken into account when deciding on health policy measures to deal with future waves of COVID-19.

In the same issue, Ganne-Carié et al.² tackle another important issue in HCC, namely aetiology-related differences. In this study, the authors focused on HCCs that were diagnosed during 6-montly surveillance in patients with cirrhosis related to viral hepatitis, alcohol use or a combination of both as an underlying aetiology. The first interesting finding was the lower 5-year incidence in those with alcohol-related liver disease, compared to viral or mixed aetiology. At diagnosis, tumour burden was comparable between aetiologies, but early BCLC stages were nevertheless less frequent in the alcohol-related liver disease group. The latter is explained by the poorer physical condition with more impaired ECOG performance status. Median survival in these cases was lower, but when adjusting for tumour size and ECOG score, as well as Child-Pugh score, there was no aetiology-specific difference, suggesting that the poorer condition of the patients was of more impact than the aetiology. Of course, alcohol abuse comes with a psychosocial and nutritional context that needs to be considered, besides the specific mechanisms of alcohol toxicity.

The COVID-19 pandemic has also had a huge impact on daily life, with a risk of weight gain and sedentary behaviour that can further aggravate the problems associated with obesity, including non-alcoholic fatty liver disease (NAFLD). The relation between weight and NAFLD is complex. Overweight and obesity are clearly risk factors for NAFLD, and weight loss can induce histological improvement, even in terms of fibrosis regression.^3,4 But it is not all about weight. Oh et al.⁵ compared 54 obese Japanese men who participated in a closely supervised aerobic exercise programme but did not reduce their caloric intake (they even slightly increased their energy intake) to 24 men who followed a dietary restriction regimen without changing their physical activity level. Both groups improved on non-invasive tests that reflect NAFLD severity, but in the exercise group, weight loss was minimal, and the authors could hence assess the weight-independent effects of exercise. The latter point is crucial, as it pinpoints the added value of this study: several studies have already shown that exercise is beneficial in NAFLD,⁶ but the methodological rigour of the current study addresses specific questions that to date were difficult to reliably answer, and in particular to separate the effect of exercise on weight from weight loss-independent effects that might be introduced by exercise. The authors hereby demonstrate substantial weight loss-independent effects on liver steatosis and liver stiffness, as well as on other parameters like the FibroScan-AST score. Furthermore, the authors provide mechanistic insights that help explain the observations, by showing that the improvements in hepatic parameters were closely associated with anthropometric changes, e.g. reduction in adipose tissue and preservation of muscle mass, with increases in muscle strength, reductions in inflammation and oxidative stress and changes in organokine concentrations (such as adiponectin and myostatin) during the exercise regimen. This all stresses the importance of the adipose tissue-muscle-liver interaction in NAFLD pathogenesis that goes far beyond a simplistic view on body weight as a driver of NAFLD.^3,4

The search for a pharmacological treatment for NAFLD, focusing on the more severe subtype of non-alcoholic steatohepatitis (NASH) with some degree of fibrosis, faces many challenges. The complexity of disease pathophysiology is one of the main reasons that no drug has specifically been approved for the treatment of NASH, and several trials (both phase II and III) recently failed to reach their primary endpoints. One of the molecules that is currently being studied in phase III, is the liver-targeted fatty acid synthetic enzyme stearoyl CoA desaturase 1 (SCD1) inhibitor aramchol. In phase II, positive effects on liver fat content but also on steatohepatitis were observed.⁷ There was also a trend, albeit not significant, suggesting a beneficial impact on fibrosis. SCD1 is mainly expressed in hepatocytes, but Bhattacharya et al.⁸ report on the effects of aramchol in hepatic stellate cells (HSCs) in vitro. In primary human HSCs, aramchol significantly reduced SCD1 mRNA whilst upregulating peroxisome proliferator-activated receptor (PPAR) gamma. The latter effect was also observed in primary human hepatocytes, highlighting the potential importance of the PPAR pathways.⁹ This was accompanied by a decrease in expression of several genes implicated in fibrogenesis, both at the mRNA and protein levels. Experiments with SCD1 knockdown or overexpression in another human HSC cell line further support the role of SCD1 and the effect of aramchol on HSC and fibrosis. This illustrates that aramchol not only targets hepatocytes but also HSCs, hence potentially adding a direct antifibrotic effect to the overall action on NASH. Although this proof of efficacy needs to be confirmed in an ongoing phase III trial, these data provide at least further support for the study of aramchol as a treatment for NASH and fibrosis.

NAFLD, by definition, excludes the consumption of a relevant amount of alcohol, making the diagnosis exclusionary, but in real life, several causes of fatty liver disease can co-exist. Whether this then translates into progressive disease depends, to some degree, on the presence of genetic factors that can be relevant to several causes of chronic liver disease. Bianco et al.¹⁰ review genetic polymorphisms that play a role in both alcohol-related and NAFLD, or their combination, focusing on PNPLA3 and MBOAT7, with an emphasis on how these genetic variants impact on disease pathophysiology. They also look in depth at the role of interleukin-32, which has emerged as a key regulator of obesity-driven inflammation, and further highlight the importance of the adipose tissue-liver axis, particularly in NAFLD.⁴

As also mentioned in the paper by Bianco et al.¹⁰ there is an ongoing debate about changing the name of NAFLD to a name (and corresponding set of criteria) that better reflects our current understanding of NAFLD as fatty liver disease developing in the context of metabolic derangements. Interestingly, and that is why it is important not to forget the old literature, clinicians already linked fatty liver to problems of obesity, unhealthy macronutrient dietary composition and sedentarism, even in children, a long time ago. Such reports already date from the first half of the 19^th century, as elegantly reviewed by Ayonrinde.¹¹ This puts the current discussion, with claims about novel insights, into a different historical perspective – these historical data may help us find consensus about a new name for the most prevalent chronic liver disease of modern times.

Severe alcoholic hepatitis remains a challenge for treating clinicians and is associated with a high mortality rate. Despite intensive research efforts, not so much progress has been made over the last decade. Corticosteroids remain the cornerstone of therapy, but given the potential side effects, a better understanding of disease mechanisms and treatment response could help guide improved disease management. Sharma et al.¹² examined (by RNA-seq) the peripheral blood mononuclear cells of 37 patients with severe alcoholic hepatitis, treated with corticosteroids, at baseline and at 7 days, where treatment response was assessed based on the Lille score.¹³ Interestingly, the most upregulated gene panels related to the lymphoid lineage and their activity and proliferation (T and B cells and NK cells), rather than myeloid cells, with striking differences at baseline between future responders and non-responders. Furthermore, the presence or absence of changes over time in the activity of functional gene modules related to innate and adaptive immunity clearly differentiated responders vs. non-responders. Hence, this study gives further insight into the immunological mechanisms at play in severe alcoholic hepatitis, but also, if validated in larger cohorts, offers the potential to better select patients with the best benefit/risk ratio for corticosteroid therapy.

It is well known that hepatocytes within the liver are functionally different according to their zonal distribution. This is important, as for many studies on e.g., gene expression, we rely on tissue homogenates in which zonal differences within the liver are not captured. Also, when we analyse the different cell compartments separately, still we mostly rely on a mixture of cells resulting from a homogenate without zonation. Single cell techniques allow for a more granular approach. Payen et al.¹⁴ used a tissue dissociation protocol on freshly explanted human livers. First, they confirmed functional clustering of the hepatocytes, in line with the zonation landmark genes that were previously described. But most notably, HSCs also clustered in 2 distinct groups, which they called HSC1 and HSC2. This clustering appeared not to be driven by activation status genes. Based on further analysis and immunostaining on biopsy samples, HSC1 appeared to be localized in the portal and centrolobular areas, whereas the second type was more often located around the sinusoids. Further functional characterization showed that both types are the most important contributors to extracellular matrix formation, but furthermore that HSC1 are more involved in glycosaminoglycan metabolism whereas HSC2 have antigen presenting cell properties. This study is hence an important contribution to our understanding of the heterogeneity in liver cell function, especially in HSCs, which are key drivers of fibrosis formation and hence disease progression in many chronic liver diseases. Also, in light of the ongoing search for antifibrotic therapies, these findings could significantly contribute to drug development.

Cellular senescence has been shown to be a potentially relevant mechanism in many (not only age-related) diseases. Cholangiocyte senescence has emerged as an important feature in the pathogenesis of primary sclerosing cholangitis (PSC), for which pharmacotherapy is still an unmet need. The secretome of senescent cells (senescence-associated secretory phenotype) contributes to an altered, pro-inflammatory, microenvironment and is supposed to play a role in inflammation and fibrogenesis. In the current issue of JHEP Reports, Alsuraih et al.¹⁵ shed further light on this mechanism and the potential benefit of its pharmacological modulation. They used 2 novel genetic models to manipulate p16^Ink4a expression in the well-established PSC model of multidrug-resistance 2 (Mdr2)^-/-mouse. p16INK4a is highly expressed in senescent cells and plays a role in cell cycle regulation. Genetic removal of p16-positive senescent cholangiocytes (by targeted apoptosis) markedly reduced peribiliary inflammation and fibrosis. These and other results from the study by Alsuraih et al.¹⁵ clearly support the concept that senescent cholangiocytes play an important role in PSC pathogenesis, and that inhibiting senescence or targeted removal of these senescent cells (senolytic therapy) improves inflammation and fibrosis, suggesting this has potential implications in the clinic. In this same issue, Cazzagon et al.¹⁶ provide clinical data that are nicely in line with the exhaustive preclinical work of Alsuraih et al.¹⁵ On liver samples of a cohort of 35 patients with PSC, they demonstrated that p16 expression in native bile ducts and in ductular reaction correlated with fibrosis, hepatitis activity and bile duct loss. Also, p21, another marker of senescent cholangiocytes, correlated with some of these features, as well as with clinical outcome parameters, further stressing the importance of cholangiocyte senescence and the clinical relevance of the aforementioned preclinical data. Several senolytic compounds have recently been developed. Targeting apoptosis is probably not feasible in patients due to the toxicity of compounds developed so far, but fisetin, a senolytic compound with a flavonoid structure, was shown to be beneficial.¹⁵ As flavonols have a good safety and tolerability profile, further studies are warranted to determine its therapeutic potential.

The editorial team aims to share the highest quality innovation in clinical and experimental hepatology with our readership – the selection of papers included in the current issue of JHEP Reports certainly meet this goal. We hope you enjoy it!