Advertisement

Overlooked subclinical portal hypertension in non-cirrhotic NAFLD: Is it real and how to measure it?

  • Gyorgy Baffy
    Correspondence
    Corresponding author. Address: Section of Gastroenterology, VA Boston Healthcare System, 150 S Huntington Avenue, Room A6-46, Boston, MA 02130, USA.
    Affiliations
    Department of Medicine, VA Boston Healthcare System and Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA
    Search for articles by this author
  • Jaume Bosch
    Affiliations
    Department of Biomedical Research, University of Bern, Bern, Switzerland

    Institut d'Investigacions Biomediques August Pi i Sunyer and CIBERehd, University of Barcelona, Spain
    Search for articles by this author
Published:October 01, 2021DOI:https://doi.org/10.1016/j.jhep.2021.09.029

      Summary

      Clinical and experimental advances related to the detection, magnitude and pathobiology of subclinical portal hypertension in non-alcoholic fatty liver disease (NAFLD), primarily observed in the presence of non-alcoholic steatohepatitis (NASH), prompt us to revisit current disease paradigms. Hepatic venous pressure gradient (HVPG) has been reported to underestimate portal pressure in NASH-related cirrhosis, while inaccuracy is more likely in non-cirrhotic livers, indicating a potential need for new and preferably non-invasive methods of measurement. Although clinically significant portal hypertension (HVPG ≥10 mmHg) retains its prognostic significance in NASH, subclinical portal hypertension (HVPG 6.0–9.5 mmHg) has been repeatedly detected in patients with NAFLD in the absence of cirrhosis or even significant fibrosis whereas the impact of these findings on disease outcomes remains unclear. Mechanocrine signalling pathways in various types of liver cell reveal a molecular basis for the adverse effects of subclinical portal hypertension and suggest a bidirectional relationship between portal pressure and fibrosis. These findings may guide efforts to improve risk assessment and identify novel therapeutic targets in NAFLD.

      Keywords

      Introduction

      Non-alcoholic fatty liver disease (NAFLD) affects more than a quarter of the world’s population.
      • Younossi Z.M.
      Non-alcoholic fatty liver disease - a global public health perspective.
      The spectrum of NAFLD has been traditionally separated into isolated steatosis and steatohepatitis (NASH), the latter defined by steatosis, inflammation, hepatocellular injury, variable degrees of fibrosis and an increased risk of progression to cirrhosis, portal hypertension and hepatocellular carcinoma.
      • Powell E.E.
      • Wong V.W.
      • Rinella M.
      Non-alcoholic fatty liver disease.
      The individual risks that predispose to these highly heterogeneous outcomes remain incompletely understood. Intriguingly, portal hypertension has been observed in NAFLD despite the absence of cirrhosis or significant fibrosis.
      • Vonghia L.
      • Magrone T.
      • Verrijken A.
      • Michielsen P.
      • Van Gaal L.
      • Jirillo E.
      • et al.
      Peripheral and hepatic vein cytokine levels in correlation with non-alcoholic fatty liver disease (NAFLD)-related metabolic, histological, and haemodynamic features.
      • Francque S.
      • Verrijken A.
      • Mertens I.
      • Hubens G.
      • Van Marck E.
      • Pelckmans P.
      • et al.
      Visceral adiposity and insulin resistance are independent predictors of the presence of non-cirrhotic NAFLD-related portal hypertension.
      • Rodrigues S.G.
      • Montani M.
      • Guixe-Muntet S.
      • De Gottardi A.
      • Berzigotti A.
      • Bosch J.
      Patients with signs of advanced liver disease and clinically significant portal hypertension do not necessarily have cirrhosis.
      • Mendes F.D.
      • Suzuki A.
      • Sanderson S.O.
      • Lindor K.D.
      • Angulo P.
      Prevalence and indicators of portal hypertension in patients with nonalcoholic fatty liver disease.
      Furthermore, hepatic venous pressure gradient (HVPG), as commonly measured, may underestimate portal pressure in certain conditions including NASH.
      • Sourianarayanane A.
      • Talluri J.
      • Humar A.
      • McCullough A.J.
      Stage of fibrosis and portal pressure correlation in nonalcoholic steatohepatitis.
      ,
      • Ferrusquia-Acosta J.
      • Bassegoda O.
      • Turco L.
      • Reverter E.
      • Pellone M.
      • Bianchini M.
      • et al.
      Agreement between wedged hepatic venous pressure and portal pressure in non-alcoholic steatohepatitis-related cirrhosis.
      There is also recent evidence that increased portal pressure promotes fibrosis via mechanosensing pathways in liver sinusoids.
      • Bravo M.
      • Raurell I.
      • Hide D.
      • Fernandez-Iglesias A.
      • Gil M.
      • Barbera A.
      • et al.
      Restoration of liver sinusoidal cell phenotypes by statins improves portal hypertension and histology in rats with NASH.
      ,
      • Hilscher M.B.
      • Sehrawat T.
      • Arab J.P.
      • Zeng Z.
      • Gao J.
      • Liu M.
      • et al.
      Mechanical stretch increases expression of CXCL1 in liver sinusoidal endothelial cells to recruit neutrophils, generate sinusoidal microthombi, and promote portal hypertension.
      These findings suggest that portal hypertension is not simply an endpoint in NAFLD-cirrhosis. Instead, portal pressure begins to rise earlier in the course of disease, it could enhance fibrogenesis through reverse causation, and it may not be fully detected by standard hepatic vein catheterisation. Should we be concerned about the presence of increased portal pressure in non-cirrhotic NAFLD? What would be a safe, easy and accurate way to detect subclinical portal hypertension and utilise it in risk assessment? How could an early rise in portal pressure contribute to NAFLD pathophysiology and could these mechanisms identify novel therapeutic targets? The following is an overview of recent advances on these controversial issues.
      HVPG measurement may underestimate portal pressure in NAFLD-cirrhosis.

      What are the challenges of detecting portal hypertension in NAFLD?

      Portal pressure gradient (PPG) represents the pressure difference between the portal vein and the inferior vena cava and it has a normal range of 1 to 5 mmHg.
      • Bosch J.
      • Iwakiri Y.
      The portal hypertension syndrome: etiology, classification, relevance, and animal models.
      Portal hypertension is defined as a PPG >5 mmHg and clinically significant portal hypertension (CSPH) as a PPG of ≥10 mmHg.
      • Bosch J.
      • Iwakiri Y.
      The portal hypertension syndrome: etiology, classification, relevance, and animal models.
      CSPH is a reliable predictor of oesophageal variceal bleeding and other complications defining clinical decompensation,
      • Silva-Junior G.
      • Baiges A.
      • Turon F.
      • Torres F.
      • Hernandez-Gea V.
      • Bosch J.
      • et al.
      The prognostic value of hepatic venous pressure gradient in patients with cirrhosis is highly dependent on the accuracy of the technique.
      while the significance of subclinical portal hypertension (PPG from 6 to 9.5 mmHg) has not been fully established. HVPG is a surrogate measure of PPG, which is calculated from the difference of wedged and free hepatic venous pressures (WHVP and FHVP, respectively); HVPG has been used to establish the aforementioned clinically relevant thresholds, as well as being widely used as a reference standard for risk prediction, and the assessment of portal hypertension and therapeutic responses.
      • Leung J.C.
      • Loong T.C.
      • Pang J.
      • Wei J.L.
      • Wong V.W.
      Invasive and non-invasive assessment of portal hypertension.
      Theoretical considerations predict that HVPG is less accurate in the healthy liver due to persisting intersinusoidal connections (making HVPG about 1 mmHg less than PPG) while equilibration becomes disrupted and the difference between HVPG and PPG tends to disappear in cirrhosis.
      • Bosch J.
      • Abraldes J.G.
      • Berzigotti A.
      • Garcia-Pagan J.C.
      The clinical use of HVPG measurements in chronic liver disease.
      Intersinusoidal buffering remains significant when presinusoidal factors contribute to portal hypertension as happens in schistosomiasis, sarcoidosis, tuberculosis, nodular regenerative hyperplasia and early stages of primary biliary cholangitis. The accuracy of HVPG measurements has also been questioned in the context of cirrhosis from different causes. In a study of 156 patients with cirrhosis dating back to 1985, WHVP closely correlated with portal pressure measured by direct transhepatic access in alcohol-related cases, but the average difference exceeded 4 mmHg in a subgroup labelled ‘nonalcoholic’ with two-thirds of the cases being of unknown aetiology.
      • Pomier-Layrargues G.
      • Kusielewicz D.
      • Willems B.
      • Villeneuve J.P.
      • Marleau D.
      • Cote J.
      • et al.
      Presinusoidal portal hypertension in non-alcoholic cirrhosis.
      Accuracy of HVPG in NAFLD may further decrease in the absence of cirrhosis.
      Recently, the accuracy of HVPG has garnered attention as the method is being applied to a large number of NAFLD-related cases. In a retrospective review, patients with steatohepatitis had lower HVPG for each stage of fibrosis than patients with chronic hepatitis C (3.4 ± 2.4 vs. 7.5 ± 11 mm Hg/stage; p = 0.01).
      • Sourianarayanane A.
      • Talluri J.
      • Humar A.
      • McCullough A.J.
      Stage of fibrosis and portal pressure correlation in nonalcoholic steatohepatitis.
      While the pattern of fibrosis in chronic hepatitis C and NAFLD is different, these data indicate that pressure is underestimated relative to fibrosis stage in NAFLD or, put differently, higher levels of fibrosis in NAFLD correspond to the same values of portal pressure. Indeed, NAFLD has been characterised by the marked presence of perisinusoidal fibrosis that may contribute to disease progression, but is not accounted for in the original METAVIR fibrosis staging system.
      • Charlotte F.
      • Le Naour G.
      • Bernhardt C.
      • Poynard T.
      • Ratziu V.
      • Group L.S.
      A comparison of the fibrotic potential of nonalcoholic fatty liver disease and chronic hepatitis C.
      One could also speculate that vasoreactive mechanisms in NAFLD diminish the impact of fibrosis on portal pressure. In the context of cirrhosis and decompensation, a recent prospective study of 40 patients with NAFLD-cirrhosis matched to patients with cirrhosis related to alcohol or chronic hepatitis C (both n = 40) again disclosed that discordance between WHVP and portal pressure was more frequent in the NASH group (37.5% vs. 14%; p = 0.003), owing to underestimation by WHVP, with individual differences exceeding 5 mmHg in 15% of patients.
      • Ferrusquia-Acosta J.
      • Bassegoda O.
      • Turco L.
      • Reverter E.
      • Pellone M.
      • Bianchini M.
      • et al.
      Agreement between wedged hepatic venous pressure and portal pressure in non-alcoholic steatohepatitis-related cirrhosis.
      Again, there could be alternative explanations, such as increased variability of HVPG measurements in patients with obesity or increased intra-abdominal pressure due to marked respiratory changes,
      • Reverter E.
      • Blasi A.
      • Abraldes J.G.
      • Martinez-Palli G.
      • Seijo S.
      • Turon F.
      • et al.
      Impact of deep sedation on the accuracy of hepatic and portal venous pressure measurements in patients with cirrhosis.
      but this is unlikely to explain all cases with discrepant results. Post hoc analysis of 475 individuals with advanced NAFLD (F3 and F4 fibrosis) in 2 clinical trials with simtuzumab found that 7 cases of decompensation (14% of all decompensating events) occurred in those with subclinical portal hypertension (median HVPG, 7.5 mmHg; range, 4.0–9.5 mmHg) with a median interval of 4.7 months lapsed between detection of increased HVPG and the clinical event.
      • Sanyal A.J.
      • Harrison S.A.
      • Ratziu V.
      • Abdelmalek M.F.
      • Diehl A.M.
      • Caldwell S.
      • et al.
      The natural history of advanced fibrosis due to nonalcoholic steatohepatitis: data from the simtuzumab trials.
      Although it could be that HVPG increased during the interval in some of these cases, the observation is not unexpected, since CSPH is always preceded by a period of subclinical portal pressure elevation.
      • Ripoll C.
      • Groszmann R.
      • Garcia-Tsao G.
      • Grace N.
      • Burroughs A.
      • Planas R.
      • et al.
      Hepatic venous pressure gradient predicts clinical decompensation in patients with compensated cirrhosis.
      Subclinical portal hypertension (HVPG 6.0–9.5 mmHg) is often detected in non-cirrhotic NAFLD but the significance of this finding remains unclear.
      These observations suggest that portal hypertension in NAFLD may not be fully captured by HVPG. It remains to be seen whether portal hypertension in NAFLD-cirrhosis has a presinusoidal component or if it is a heterogeneous condition in which intersinusoidal channels are unevenly affected (Fig. 1). Greater discrepancies may be seen between HVPG and PPG in non-cirrhotic NAFLD when communication between sinusoids is preserved. Some of these cases may involve non-cirrhotic idiopathic portal hypertension (recently termed porto-sinusoidal vascular disease), which is associated with a pre-sinusoidal pathology and near-normal liver stiffness.
      • De Gottardi A.
      • Rautou P.E.
      • Schouten J.
      • Rubbia-Brandt L.
      • Leebeek F.
      • Trebicka J.
      • et al.
      Porto-sinusoidal vascular disease: proposal and description of a novel entity.
      Novel approaches for safe, easy and reliable portal pressure measurements are therefore eagerly awaited. One such approach is endoscopic ultrasound-guided pressure measurement, which allows for direct access to the portal and hepatic vein
      • Bazarbashi A.N.
      • Ryou M.
      Portal pressure measurement: have we come full circle?.
      ; however, this approach is costly and requires further clinical validation. Like current efforts to optimise non-invasive detection of intracranial pressure, imaging-based techniques that accurately measure PPG would certainly have broader applicability and may eventually enable monitoring of portal hypertension in real-world practice.
      Figure thumbnail gr1
      Fig. 1Direct and indirect measurement of portal pressure.
      Schematic diagram of PP recording sites and hepatic sinusoidal haemodynamics in the presence and absence of cirrhosis. PP is primarily defined by structural and functional components of presinusoidal and sinusoidal vascular resistance (‘gate valves’). In non-cirrhotic liver, sinusoids are connected by intersinusoidal channels (‘relief valves’, top panel), which are interrupted along with sinusoidal narrowing in cirrhosis due to extensive fibrosis and nodule formation (bottom panel). Hepatic vein catheterization enables the calculation of HVPG from readings of FHVP and WHVP, while PPG is derived from readings of directly accessed PV and HV pressure. In cirrhosis, the loss of intersinusoidal communication (X) enables sinusoidal pressure to equilibrate with PP when a wedged catheter blocks sinusoidal flow and creates a static column reaching the PV (dashed lines). Therefore, readings at the sites of PV and WHVP in sinusoidal portal hypertension almost perfectly overlap (bottom panel). HVPG underestimates PPG if sinusoidal resistance is buffered by open intersinusoidal channels, which is more likely to occur in the absence of cirrhosis. Presinusoidal component(s) not detected by the wedged catheter would have a similar effect. In these cases, concordance between readings at the PV and WHVP may diminish to a variable degree (top panel). FHVP, free hepatic venous pressure; HV, hepatic vein; HVPG, hepatic venous pressure gradient; ICV, inferior vena cava; PP, portal pressure; PPG, portal pressure gradient; PV, portal vein; RA, right atrium; WHVP, wedged hepatic venous pressure.

      What is the evidence for portal hypertension in non-cirrhotic NAFLD?

      While portal hypertension predictably develops from extensive fibrosis and distorted liver architecture in cirrhosis, elevated portal pressure of lesser magnitude has been repeatedly demonstrated in NASH in the absence of these dramatic histological changes. In a prospective cohort, 8 out of 40 obese patients diagnosed with NAFLD were found to have subclinical portal hypertension, while 1 patient had CSPH in the absence of cirrhosis.
      • Vonghia L.
      • Magrone T.
      • Verrijken A.
      • Michielsen P.
      • Van Gaal L.
      • Jirillo E.
      • et al.
      Peripheral and hepatic vein cytokine levels in correlation with non-alcoholic fatty liver disease (NAFLD)-related metabolic, histological, and haemodynamic features.
      Another study of 50 patients with biopsy-proven NAFLD reported elevated portal pressure in one-third of patients, which – with respect to histological parameters - only correlated with steatosis.
      • Francque S.
      • Verrijken A.
      • Mertens I.
      • Hubens G.
      • Van Marck E.
      • Pelckmans P.
      • et al.
      Visceral adiposity and insulin resistance are independent predictors of the presence of non-cirrhotic NAFLD-related portal hypertension.
      In a retrospective study of 89 patients with CSPH, histological features of cirrhosis were absent in 16% of patients diagnosed with either NAFLD or nodular regenerative hyperplasia.
      • Rodrigues S.G.
      • Montani M.
      • Guixe-Muntet S.
      • De Gottardi A.
      • Berzigotti A.
      • Bosch J.
      Patients with signs of advanced liver disease and clinically significant portal hypertension do not necessarily have cirrhosis.
      In a similar analysis, fibrosis was mild or absent in 12% of 100 patients diagnosed with NAFLD and clinical or imaging signs of portal hypertension, providing further evidence that increased portal pressure may complicate non-fibrotic NAFLD if steatosis is severe.
      • Mendes F.D.
      • Suzuki A.
      • Sanderson S.O.
      • Lindor K.D.
      • Angulo P.
      Prevalence and indicators of portal hypertension in patients with nonalcoholic fatty liver disease.
      A common finding in these works is that subclinical portal hypertension can develop in patients with NAFLD when steatosis is either the sole histological feature or if it is accompanied by mild or moderate fibrosis. Admittedly, however, the differentiation between mild, moderate or severe fibrosis is subjective and is associated with marked interobserver variability.
      • Dinani A.M.
      • Kowdley K.V.
      • Noureddin M.
      Application of artificial intelligence for diagnosis and risk stratification in NAFLD and NASH- the state of the art.
      Experimental studies support the notion that portal hypertension may begin to develop in the absence of fibrosis. When mice are fed a methionine/choline-deficient (MCD) diet, sluggish sinusoidal flow blocking the passage of red blood cells can be seen several weeks before perisinusoidal fibrosis is evident.
      • McCuskey R.S.
      • Ito Y.
      • Robertson G.R.
      • McCuskey M.K.
      • Perry M.
      • Farrell G.C.
      Hepatic microvascular dysfunction during evolution of dietary steatohepatitis in mice.
      PPG is markedly increased without histological evidence of inflammation or fibrosis in rats kept on an MCD diet for 4 weeks compared to animals on control diet (8.2 ± 1.3 vs. 2.2 ± 1.1 mmHg, p <0.001), in association with disorganised sinusoids and the presence of multiple ‘blebs’, which may represent leakage caused by disruption of the normal sinusoidal wall.
      • Francque S.
      • Laleman W.
      • Verbeke L.
      • Van Steenkiste C.
      • Casteleyn C.
      • Kwanten W.
      • et al.
      Increased intrahepatic resistance in severe steatosis: endothelial dysfunction, vasoconstrictor overproduction and altered microvascular architecture.
      Human NAFLD is better reproduced by the rat model of cafeteria diet (with 65% of the energy derived from fat that is predominantly saturated) in which pericentral liver steatosis co-occurs with increased (non-significantly) portal pressure in the absence of inflammation or fibrosis.
      • Pasarin M.
      • La Mura V.
      • Gracia-Sancho J.
      • Garcia-Caldero H.
      • Rodriguez-Vilarrupla A.
      • Garcia-Pagan J.C.
      • et al.
      Sinusoidal endothelial dysfunction precedes inflammation and fibrosis in a model of NAFLD.
      In a more recent preclinical model of steatohepatitis induced by high-fat glucose-fructose diet, mice were lacking significant fibrosis after 8 weeks when diffuse steatosis, ballooning, lobular inflammation and increased portal pressure were already present.
      • Garcia-Lezana T.
      • Raurell I.
      • Bravo M.
      • Torres-Arauz M.
      • Salcedo M.T.
      • Santiago A.
      • et al.
      Restoration of a healthy intestinal microbiota normalizes portal hypertension in a rat model of nonalcoholic steatohepatitis.
      Thus, multiple lines of evidence suggest that the cause-effect relationship between liver fibrosis and portal pressure in NAFLD is not necessarily unidirectional. However, this is certainly not specific to NAFLD and increased portal pressure can be observed before established cirrhosis in other animal models or human liver diseases of other origin
      Increased portal pressure may elicit mechanocrine signalling pathways and contribute to NAFLD progression.
      .

      What cellular and molecular mechanisms are associated with subclinical portal hypertension in NAFLD?

      While evidence of early portal hypertension in NAFLD continues to accumulate, one may ask whether steatosis or steatohepatitis is sufficient to cause elevated portal pressure in the absence of at least some degree of fibrosis. This is an intriguing dilemma related to the chronology of fibrosis and subclinical portal hypertension (Fig. 2). Key elements of this concept have been applied to cirrhosis in the sinusoidal pressure hypothesis.
      • Mueller S.
      Does pressure cause liver cirrhosis? The sinusoidal pressure hypothesis.
      The question is more than academic since fibrosis is a strong predictor of poor clinical outcomes in advanced NAFLD, while we need reliable biomarkers for risk assessment in earlier stages that represent the vast majority of the NAFLD population. Moreover, the effect of subclinical portal hypertension on the natural course of NAFLD is incompletely understood. While CSPH has been defined by a critical threshold, it is clear that it represents the advanced stage of a continuum that starts with even mildly or moderately increased portal pressure, which worsens and promotes disease progression and increases the risk of adverse clinical outcomes.
      • Ripoll C.
      Noninvasive predictors of fibrosis in NASH with and without cirrhosis, just as good as histology (and hepatic venous pressure gradient?).
      Experimental data suggest a bidirectional relationship between increased sinusoidal pressure and liver fibrosis.
      Figure thumbnail gr2
      Fig. 2Mechanotransduction in the pathophysiology of portal hypertension.
      Multiple lines of evidence indicate that portal pressure in NAFLD and other chronic liver diseases may begin to rise prior to the development of significant fibrosis or cirrhosis. Portal hypertension in these cases is usually mild and may only become clinically significant once extensive fibrosis with tissue remodelling occurs in advanced liver disease. Experimental observations suggest that subclinical portal hypertension in early stages of non-alcoholic fatty liver disease may contribute to disease progression, indicating a role for mechanocrine signals in a bidirectional relationship between sinusoidal pressure and fibrosis (top panel). While mechanosignalling affects all liver cell types, LSECs are primarily subjected to various forms of mechanical stress due to blood flow that may reach the sinusoids at supraphysiological hydrostatic pressure or become turbulent due to intraluminal pathology, such as microthrombosis or neutrophil extracellular traps (middle panel). Key components of mechanocrine signalling include complex cell-cell interactions and molecular pathways that include amplification and feed-forward mechanisms (bottom panel). CSPH, clinically significant portal hypertension; ECM, extracellular matrix; LSEC, liver sinusoidal endothelial cell.
      Liver sinusoidal endothelial cells (LSECs) play a key role in sensing and regulating hepatic blood flow, which has a characteristically steep pressure gradient between terminal hepatic arterioles and portal venules.
      • Poisson J.
      • Lemoinne S.
      • Boulanger C.
      • Durand F.
      • Moreau R.
      • Valla D.
      • et al.
      Liver sinusoidal endothelial cells: physiology and role in liver diseases.
      ,
      • Hammoutene A.
      • Rautou P.E.
      Role of liver sinusoidal endothelial cells in non-alcoholic fatty liver disease.
      Persisting terminal arterioles are more common in patients with steatohepatitis who have advanced fibrosis, suggesting that high pressure may have a direct impact on sinusoidal homeostasis.
      • Gill R.M.
      • Belt P.
      • Wilson L.
      • Bass N.M.
      • Ferrell L.D.
      Centrizonal arteries and microvessels in nonalcoholic steatohepatitis.
      LSECs are the only endothelial cells in the human body with open fenestrae and no underlying basal lamina, making sinusoid microcirculation the most porous of all endothelial barriers.
      • Hammoutene A.
      • Rautou P.E.
      Role of liver sinusoidal endothelial cells in non-alcoholic fatty liver disease.
      Loss of fenestration and formation of a basal membrane (capillarisation) represent morphological features of endothelial dysfunction.
      • Poisson J.
      • Lemoinne S.
      • Boulanger C.
      • Durand F.
      • Moreau R.
      • Valla D.
      • et al.
      Liver sinusoidal endothelial cells: physiology and role in liver diseases.
      Loss of sinusoidal fenestration begins early in NAFLD in response to excessive lipid exposure.
      • Hammoutene A.
      • Rautou P.E.
      Role of liver sinusoidal endothelial cells in non-alcoholic fatty liver disease.
      Capillarised LSECs secrete proinflammatory, profibrogenic, prothrombotic and proangiogenic mediators, generating a wide web of cell-cell interactions.
      • DeLeve L.D.
      Liver sinusoidal endothelial cells in hepatic fibrosis.
      At physiological perfusion pressures, sinusoidal fenestrae act as a mechanical sieve that allow certain macromolecules and cellular projections to come into contact with the microvilli of hepatocytes across the space of Disse. As recently reported, loss of fenestration in LSECs is more commonly seen in isolated steatosis compared to steatohepatitis.
      • Verhaegh P.
      • Wisse E.
      • de Munck T.
      • Greve J.W.
      • Verheij J.
      • Riedl R.
      • et al.
      Electron microscopic observations in perfusion-fixed human non-alcoholic fatty liver disease biopsies.
      These surprising observations suggest that in some situations the loss of fenestration could be a protective mechanism intended to prevent exposure of hepatocytes to excess large lipid molecules or albumin-bound small molecules, such as toxic substances or microbial metabolites from the portal circulation.
      • Verhaegh P.
      • Wisse E.
      • de Munck T.
      • Greve J.W.
      • Verheij J.
      • Riedl R.
      • et al.
      Electron microscopic observations in perfusion-fixed human non-alcoholic fatty liver disease biopsies.
      Characterisation of liver cell subpopulations by single-cell RNA sequencing and immunofluorescence methods may shed more light on the role of LSECs in early NAFLD. Recent work identified 2 major subsets of LSECs in the human liver with distinct phenotypes and physiological functions.
      • MacParland S.A.
      • Liu J.C.
      • Ma X.Z.
      • Innes B.T.
      • Bartczak A.M.
      • Gage B.K.
      • et al.
      Single cell RNA sequencing of human liver reveals distinct intrahepatic macrophage populations.
      ,
      • Strauss O.
      • Phillips A.
      • Ruggiero K.
      • Bartlett A.
      • Dunbar P.R.
      Immunofluorescence identifies distinct subsets of endothelial cells in the human liver.
      In periportal areas, LSECs weakly express the immunomodulatory cell surface receptor CD32b and are rich in pathways related to angiogenesis and cell-cell junctions, while pericentral LSECs abundantly express CD32b and engage in immune pathways including phagocytosis.
      • MacParland S.A.
      • Liu J.C.
      • Ma X.Z.
      • Innes B.T.
      • Bartczak A.M.
      • Gage B.K.
      • et al.
      Single cell RNA sequencing of human liver reveals distinct intrahepatic macrophage populations.
      Interestingly, early-stage steatohepatitis induced by high-fat glucose-fructose diet in rats is associated with a shift to CD32b-negative LSECs that display capillarisation markers such as endothelin-1 and CD31.
      • Bravo M.
      • Raurell I.
      • Hide D.
      • Fernandez-Iglesias A.
      • Gil M.
      • Barbera A.
      • et al.
      Restoration of liver sinusoidal cell phenotypes by statins improves portal hypertension and histology in rats with NASH.
      These phenotypic changes in LSECs are associated with activation of hepatic stellate cells and increased portal pressure, which can be reverted by statin administration in this model.
      • Bravo M.
      • Raurell I.
      • Hide D.
      • Fernandez-Iglesias A.
      • Gil M.
      • Barbera A.
      • et al.
      Restoration of liver sinusoidal cell phenotypes by statins improves portal hypertension and histology in rats with NASH.

      What is the role of mechanotransduction in exacerbating disease progression and portal hypertension?

      Recent studies support the concept that even mild increases in portal pressure have biological consequences. Due to their anatomical location, LSECs are plausibly the first among different liver cell types affected by the mechanical forces resulting from rheological changes associated with increased portal pressure (Fig. 2).
      Mechanotransduction has been defined as a biochemical response to physical stimuli such as hydrostatic pressure, shear stress or tension.
      • Wells R.G.
      Tissue mechanics and fibrosis.
      Resistance to these forces is quantified in pascals (Pa) and can be used to measure liver tissue stiffness with various elastography techniques (widely used for non-invasive fibrosis assessment). The molecular network of mechanosensing (‘mechanosome’) links extracellular physical signals to the cytoskeleton and nucleus.
      • Kang N.
      Mechanotransduction in liver diseases.
      Mechanical forces may cause deformation of plasma membrane lipid rafts and caveolae, expose new protein binding sites, alter the activity of transmembrane receptors and adhesion molecules, or change ion channel conductivity. These structural changes may activate intracellular signalling cascades such as MAP and Rho family kinases, and stimulate the release of intracellular calcium, nitric oxide and reactive oxygen species. Flow-responsive transcriptional regulators include the Kruppel-like factor family, the Hippo-regulated YAP/TAZ pathway and the pleiotropic NF-κB that modulate endothelial responses such as barrier function, contractility, haemostasis and angiogenesis.
      • Chatterjee S.
      Endothelial mechanotransduction, redox signaling and the regulation of vascular inflammatory pathways.
      ,
      • Tanaka K.
      • Joshi D.
      • Timalsina S.
      • Schwartz M.A.
      Early events in endothelial flow sensing.
      Aberrant activation of mechanosignalling pathways and their associated genetic programmes may occur in LSECs, hepatic stellate cells and parenchymal liver cells due to increased stiffness of the cell culture matrix in vitro or the liver tissue in vivo, profoundly altering the function of these cells and their interactions with each other.
      • Kang N.
      Mechanotransduction in liver diseases.
      ,
      • Dou C.
      • Liu Z.
      • Tu K.
      • Zhang H.
      • Chen C.
      • Yaqoob U.
      • et al.
      P300 acetyltransferase mediates stiffness-induced activation of hepatic stellate cells into tumor-promoting myofibroblasts.
      ,
      • Guixe-Muntet S.
      • Ortega-Ribera M.
      • Wang C.
      • Selicean S.
      • Andreu I.
      • Kechagia J.Z.
      • et al.
      Nuclear deformation mediates liver cell mechanosensing in cirrhosis.
      Experimental observations indicate that accumulation of lipid droplets and nuclear dislocation is associated with YAP activation in cultured hepatocytes, revealing a molecular mechanism that may be relevant to early mechanotransduction events in NAFLD.
      • Chin L.
      • Theise N.D.
      • Loneker A.E.
      • Janmey P.A.
      • Wells R.G.
      Lipid droplets disrupt mechanosensing in human hepatocytes.
      Novel methods may be needed to elucidate the role of subclinical portal hypertension in NAFLD progression.
      LSECs are constantly subjected to shear stress caused by sinusoidal blood flow and tensile stress from mechanical distortion.
      • Sun X.
      • Harris E.N.
      New aspects of hepatic endothelial cells in physiology and nonalcoholic fatty liver disease.
      LSECs reside in a low-pressure zone, indicating that an increase in fluid pressure by a few mmHg from baseline represents a major change that could evoke pathological responses. Moreover, feed-forward cycles may exist as LSEC capillarisation and increasingly abundant extracellular matrix components add to sinusoidal rigidity and shear stress, further stimulating mechanocrine pathways and leading to additional cell injury and fibrosis.
      • Kang N.
      Mechanotransduction in liver diseases.
      ,
      • Sun X.
      • Harris E.N.
      New aspects of hepatic endothelial cells in physiology and nonalcoholic fatty liver disease.
      The mechanical stretching of LSECs in animal models of portal hypertension elicited by partial inferior vena cava ligation and bile duct ligation activates flow-responsive PIEZO cation channels, integrin-dependent Notch signalling and calcium influx, concluding in increased secretion of the neutrophil chemotactic chemokine CXCL1.
      • Hilscher M.B.
      • Sehrawat T.
      • Arab J.P.
      • Zeng Z.
      • Gao J.
      • Liu M.
      • et al.
      Mechanical stretch increases expression of CXCL1 in liver sinusoidal endothelial cells to recruit neutrophils, generate sinusoidal microthombi, and promote portal hypertension.
      Excess CXCL1 promotes leukocyte recruitment and formation of prothrombotic neutrophil extracellular traps that induce microvascular thrombosis, elevated portal pressure and fibrosis.
      • Hilscher M.B.
      • Sehrawat T.
      • Arab J.P.
      • Zeng Z.
      • Gao J.
      • Liu M.
      • et al.
      Mechanical stretch increases expression of CXCL1 in liver sinusoidal endothelial cells to recruit neutrophils, generate sinusoidal microthombi, and promote portal hypertension.
      These burgeoning observations link rheology and molecular biology in the research of sinusoidal homeostasis, providing insights into the role of subclinical portal hypertension in NAFLD pathophysiology.

      Perspectives

      NAFLD has a commanding presence in our life. There have been great advances in our understanding of its pathogenesis, the development of diagnostic tools and the identification of therapeutic targets for this disease. However, accurately predicting individual patient risk in the context of the highly diverse clinical outcomes associated with NAFLD remains one of the great challenges in current hepatology. The controversies discussed above raise concerns but also create opportunities. First, HVPG may not accurately detect subclinical changes in portal pressure, at least in some patients, which has led to the consideration of direct pressure measurements in the portal and hepatic vein, to eliminate the pitfalls of retrograde access. Endoscopic ultrasound-guided PPG measurement holds promise but will need to be evaluated against HVPG in prospective studies to determine reproducibility, clinical significance and associated costs before recommending routine use, and it also faces the problem of being conducted in non-physiologic conditions, with the inherent risk of obtaining unreliable results. Moreover, innovative diagnostic tools that allow for non-invasive ‘real-world’ monitoring of portal hypertension are necessary. Second, we have multiple lines of evidence that portal hypertension develops in non-cirrhotic NAFLD. Some data suggest that it may begin even in the absence of fibrosis. We need to explore the association of subclinical portal hypertension with other clinical and laboratory parameters to determine the utility of portal pressure as a potential biomarker in NAFLD prognostication. It is likely that subclinical portal hypertension could be divided into separate risk categories with relevant therapeutic implications. We may also need to revise our canonical schemes of disease evolution and consider portal pressure as a potential driver of liver fibrosis (rather than solely the other way around), which will have ramifications for our risk assessment strategies. Finally, emerging work provides important insights into mechanocrine signalling pathways that link the physical impact of increased portal pressure to the pathophysiology of liver diseases in general, and NAFLD in particular. This exciting research may identify novel molecular targets for the prevention and treatment of portal hypertension.

      Abbreviations

      CSPH, clinically significant portal hypertension; CXCL1, CXC motif chemokine ligand 1; HVPG, hepatic venous pressure gradient; LSEC, liver sinusoidal endothelial cell; MCD, methionine/choline-deficient; NAFLD, non-alcoholic fatty liver disease; NASH, non-alcoholic steatohepatitis; PPG, portal pressure gradient; TAZ, transcriptional coactivator with PDZ-binding motif; WHVP, wedged hepatic venous pressure; YAP, yes-associated protein.

      Financial support

      The authors received no financial support to produce this manuscript.

      Authors’ contributions

      GB conceived and wrote the manuscript. JB critically revised the manuscript. Both authors edited and reviewed the final manuscript.

      Conflict of interest

      The authors declare no conflicts of interest that pertain to this work.
      Please refer to the accompanying ICMJE disclosure forms for further details.

      Supplementary data

      The following is the supplementary data to this article:

      References

        • Younossi Z.M.
        Non-alcoholic fatty liver disease - a global public health perspective.
        J Hepatol. 2019; 70: 531-544
        • Powell E.E.
        • Wong V.W.
        • Rinella M.
        Non-alcoholic fatty liver disease.
        Lancet. 2021; 397: 2212-2224
        • Vonghia L.
        • Magrone T.
        • Verrijken A.
        • Michielsen P.
        • Van Gaal L.
        • Jirillo E.
        • et al.
        Peripheral and hepatic vein cytokine levels in correlation with non-alcoholic fatty liver disease (NAFLD)-related metabolic, histological, and haemodynamic features.
        PLoS One. 2015; 10e0143380
        • Francque S.
        • Verrijken A.
        • Mertens I.
        • Hubens G.
        • Van Marck E.
        • Pelckmans P.
        • et al.
        Visceral adiposity and insulin resistance are independent predictors of the presence of non-cirrhotic NAFLD-related portal hypertension.
        Int J Obes (Lond). 2011; 35: 270-278
        • Rodrigues S.G.
        • Montani M.
        • Guixe-Muntet S.
        • De Gottardi A.
        • Berzigotti A.
        • Bosch J.
        Patients with signs of advanced liver disease and clinically significant portal hypertension do not necessarily have cirrhosis.
        Clin Gastroenterol Hepatol. 2019; 17: 2101-2109 e2101
        • Mendes F.D.
        • Suzuki A.
        • Sanderson S.O.
        • Lindor K.D.
        • Angulo P.
        Prevalence and indicators of portal hypertension in patients with nonalcoholic fatty liver disease.
        Clin Gastroenterol Hepatol. 2012; 10: 1028-1033 e1022
        • Sourianarayanane A.
        • Talluri J.
        • Humar A.
        • McCullough A.J.
        Stage of fibrosis and portal pressure correlation in nonalcoholic steatohepatitis.
        Eur J Gastroenterol Hepatol. 2017; 29: 516-523
        • Ferrusquia-Acosta J.
        • Bassegoda O.
        • Turco L.
        • Reverter E.
        • Pellone M.
        • Bianchini M.
        • et al.
        Agreement between wedged hepatic venous pressure and portal pressure in non-alcoholic steatohepatitis-related cirrhosis.
        J Hepatol. 2021; 74: 811-818
        • Bravo M.
        • Raurell I.
        • Hide D.
        • Fernandez-Iglesias A.
        • Gil M.
        • Barbera A.
        • et al.
        Restoration of liver sinusoidal cell phenotypes by statins improves portal hypertension and histology in rats with NASH.
        Sci Rep. 2019; 9: 20183
        • Hilscher M.B.
        • Sehrawat T.
        • Arab J.P.
        • Zeng Z.
        • Gao J.
        • Liu M.
        • et al.
        Mechanical stretch increases expression of CXCL1 in liver sinusoidal endothelial cells to recruit neutrophils, generate sinusoidal microthombi, and promote portal hypertension.
        Gastroenterology. 2019; 157: 193-209 e199
        • Bosch J.
        • Iwakiri Y.
        The portal hypertension syndrome: etiology, classification, relevance, and animal models.
        Hepatol Int. 2018; 12: 1-10
        • Silva-Junior G.
        • Baiges A.
        • Turon F.
        • Torres F.
        • Hernandez-Gea V.
        • Bosch J.
        • et al.
        The prognostic value of hepatic venous pressure gradient in patients with cirrhosis is highly dependent on the accuracy of the technique.
        Hepatology. 2015; 62: 1584-1592
        • Leung J.C.
        • Loong T.C.
        • Pang J.
        • Wei J.L.
        • Wong V.W.
        Invasive and non-invasive assessment of portal hypertension.
        Hepatol Int. 2018; 12: S44-S55
        • Bosch J.
        • Abraldes J.G.
        • Berzigotti A.
        • Garcia-Pagan J.C.
        The clinical use of HVPG measurements in chronic liver disease.
        Nat Rev Gastroenterol Hepatol. 2009; 6: 573-582
        • Pomier-Layrargues G.
        • Kusielewicz D.
        • Willems B.
        • Villeneuve J.P.
        • Marleau D.
        • Cote J.
        • et al.
        Presinusoidal portal hypertension in non-alcoholic cirrhosis.
        Hepatology. 1985; 5: 415-418
        • Charlotte F.
        • Le Naour G.
        • Bernhardt C.
        • Poynard T.
        • Ratziu V.
        • Group L.S.
        A comparison of the fibrotic potential of nonalcoholic fatty liver disease and chronic hepatitis C.
        Hum Pathol. 2010; 41: 1178-1185
        • Reverter E.
        • Blasi A.
        • Abraldes J.G.
        • Martinez-Palli G.
        • Seijo S.
        • Turon F.
        • et al.
        Impact of deep sedation on the accuracy of hepatic and portal venous pressure measurements in patients with cirrhosis.
        Liver Int. 2014; 34: 16-25
        • Sanyal A.J.
        • Harrison S.A.
        • Ratziu V.
        • Abdelmalek M.F.
        • Diehl A.M.
        • Caldwell S.
        • et al.
        The natural history of advanced fibrosis due to nonalcoholic steatohepatitis: data from the simtuzumab trials.
        Hepatology. 2019; 70: 1913-1927
        • Ripoll C.
        • Groszmann R.
        • Garcia-Tsao G.
        • Grace N.
        • Burroughs A.
        • Planas R.
        • et al.
        Hepatic venous pressure gradient predicts clinical decompensation in patients with compensated cirrhosis.
        Gastroenterology. 2007; 133: 481-488
        • De Gottardi A.
        • Rautou P.E.
        • Schouten J.
        • Rubbia-Brandt L.
        • Leebeek F.
        • Trebicka J.
        • et al.
        Porto-sinusoidal vascular disease: proposal and description of a novel entity.
        Lancet Gastroenterol Hepatol. 2019; 4: 399-411
        • Bazarbashi A.N.
        • Ryou M.
        Portal pressure measurement: have we come full circle?.
        Gastrointest Endosc. 2021; 93: 573-576
        • Dinani A.M.
        • Kowdley K.V.
        • Noureddin M.
        Application of artificial intelligence for diagnosis and risk stratification in NAFLD and NASH- the state of the art.
        Hepatology. 2021; 74: 2233-2240
        • McCuskey R.S.
        • Ito Y.
        • Robertson G.R.
        • McCuskey M.K.
        • Perry M.
        • Farrell G.C.
        Hepatic microvascular dysfunction during evolution of dietary steatohepatitis in mice.
        Hepatology. 2004; 40: 386-393
        • Francque S.
        • Laleman W.
        • Verbeke L.
        • Van Steenkiste C.
        • Casteleyn C.
        • Kwanten W.
        • et al.
        Increased intrahepatic resistance in severe steatosis: endothelial dysfunction, vasoconstrictor overproduction and altered microvascular architecture.
        Lab Invest. 2012; 92: 1428-1439
        • Pasarin M.
        • La Mura V.
        • Gracia-Sancho J.
        • Garcia-Caldero H.
        • Rodriguez-Vilarrupla A.
        • Garcia-Pagan J.C.
        • et al.
        Sinusoidal endothelial dysfunction precedes inflammation and fibrosis in a model of NAFLD.
        PLoS One. 2012; 7e32785
        • Garcia-Lezana T.
        • Raurell I.
        • Bravo M.
        • Torres-Arauz M.
        • Salcedo M.T.
        • Santiago A.
        • et al.
        Restoration of a healthy intestinal microbiota normalizes portal hypertension in a rat model of nonalcoholic steatohepatitis.
        Hepatology. 2018; 67: 1485-1498
        • Mueller S.
        Does pressure cause liver cirrhosis? The sinusoidal pressure hypothesis.
        World J Gastroenterol. 2016; 22: 10482-10501
        • Ripoll C.
        Noninvasive predictors of fibrosis in NASH with and without cirrhosis, just as good as histology (and hepatic venous pressure gradient?).
        Hepatology. 2016; 63: 660-661
        • Poisson J.
        • Lemoinne S.
        • Boulanger C.
        • Durand F.
        • Moreau R.
        • Valla D.
        • et al.
        Liver sinusoidal endothelial cells: physiology and role in liver diseases.
        J Hepatol. 2017; 66: 212-227
        • Hammoutene A.
        • Rautou P.E.
        Role of liver sinusoidal endothelial cells in non-alcoholic fatty liver disease.
        J Hepatol. 2019; 70: 1278-1291
        • Gill R.M.
        • Belt P.
        • Wilson L.
        • Bass N.M.
        • Ferrell L.D.
        Centrizonal arteries and microvessels in nonalcoholic steatohepatitis.
        Am J Surg Pathol. 2011; 35: 1400-1404
        • DeLeve L.D.
        Liver sinusoidal endothelial cells in hepatic fibrosis.
        Hepatology. 2015; 61: 1740-1746
        • Verhaegh P.
        • Wisse E.
        • de Munck T.
        • Greve J.W.
        • Verheij J.
        • Riedl R.
        • et al.
        Electron microscopic observations in perfusion-fixed human non-alcoholic fatty liver disease biopsies.
        Pathology (Phila). 2021; 53: 220-228
        • MacParland S.A.
        • Liu J.C.
        • Ma X.Z.
        • Innes B.T.
        • Bartczak A.M.
        • Gage B.K.
        • et al.
        Single cell RNA sequencing of human liver reveals distinct intrahepatic macrophage populations.
        Nat Commun. 2018; 9: 4383
        • Strauss O.
        • Phillips A.
        • Ruggiero K.
        • Bartlett A.
        • Dunbar P.R.
        Immunofluorescence identifies distinct subsets of endothelial cells in the human liver.
        Sci Rep. 2017; 7: 44356
        • Wells R.G.
        Tissue mechanics and fibrosis.
        Biochim Biophys Acta. 2013; 1832: 884-890
        • Kang N.
        Mechanotransduction in liver diseases.
        Semin Liver Dis. 2020; 40: 84-90
        • Chatterjee S.
        Endothelial mechanotransduction, redox signaling and the regulation of vascular inflammatory pathways.
        Front Physiol. 2018; 9: 524
        • Tanaka K.
        • Joshi D.
        • Timalsina S.
        • Schwartz M.A.
        Early events in endothelial flow sensing.
        Cytoskeleton (Hoboken). 2021; https://doi.org/10.1002/cm.21652
        • Dou C.
        • Liu Z.
        • Tu K.
        • Zhang H.
        • Chen C.
        • Yaqoob U.
        • et al.
        P300 acetyltransferase mediates stiffness-induced activation of hepatic stellate cells into tumor-promoting myofibroblasts.
        Gastroenterology. 2018; 154: 2209-2221 e2214
        • Guixe-Muntet S.
        • Ortega-Ribera M.
        • Wang C.
        • Selicean S.
        • Andreu I.
        • Kechagia J.Z.
        • et al.
        Nuclear deformation mediates liver cell mechanosensing in cirrhosis.
        JHEP Rep. 2020; 2: 100145
        • Chin L.
        • Theise N.D.
        • Loneker A.E.
        • Janmey P.A.
        • Wells R.G.
        Lipid droplets disrupt mechanosensing in human hepatocytes.
        Am J Physiol Gastrointest Liver Physiol. 2020; 319: G11-G22
        • Sun X.
        • Harris E.N.
        New aspects of hepatic endothelial cells in physiology and nonalcoholic fatty liver disease.
        Am J Physiol Cell Physiol. 2020; 318: C1200-C1213