Gautam Mehta

Inflammation and portal hypertension – The undiscovered country

Portal hypertension has traditionally been viewed as a progressive process, involving ultrastructural changes including fibrosis, nodule formation, and vascular thrombosis, leading to increased intrahepatic resistance to flow. However, it is increasingly recognized that a significant component of this vascular resistance results from a dynamic process, regulated by complex interactions between the injured hepatocyte, the sinusoidal endothelial cell, the Kupffer cell and the hepatic stellate cell, which impact on sinusoidal calibre. Recent findings suggest these haemodynamic findings are most marked in patients with superimposed inflammation. The precise mechanisms for vascular dysfunction in cirrhosis with superimposed inflammation remain to be fully elucidated but several studies over the past decade have started to generate the hypothesis that inflammation may be a key mediator of the pathogenesis and severity of portal hypertension in this context. This review provides a comprehensive overview of the biological mechanisms for inflammation playing a key role in the severity of portal hypertension, and illustrates potential novel therapies that act by modifying these processes.

Hepatic dimethylarginine-dimethylaminohydrolase1 is reduced in cirrhosis and is a target for therapy in portal hypertension

BACKGROUND & AIMS Portal hypertension is characterized by reduced hepatic eNOS activity. Asymmetric-dimethylarginine (ADMA), an eNOS inhibitor, is elevated in cirrhosis and correlates with the severity of portal hypertension. Dimethylarginine dimethylaminohydrolase-1 (DDAH-1) is the key enzyme metabolizing hepatic ADMA. This study characterized DDAH-1 in cirrhosis, and explored hepatic DDAH-1 reconstitution through farnesoid X receptor (FXR) agonism and DDAH-1 gene therapy. METHODS DDAH-1 immunohistochemistry was conducted on human cirrhosis and healthy liver tissue. Subsequently, sham-operated or bile-duct-ligated (BDL) cirrhosis rats were treated with the FXR agonist obeticholic acid (OA, 5 mg/kg) or vehicle for 5 days. Further, animals underwent hydrodynamic injection with DDAH-1-expressing plasmid or saline control, which resulted in the following groups: sham+saline, BDL+saline, BDL+DDAH-1-plasmid. Portal pressure (PP) measurements were performed. Plasma ALT was measured by COBAS INTEGRA, DDAH-1 expression by qPCR and Western blot, eNOS activity by radiometric assay. RESULTS Immunohistochemistry and Western-blotting confirmed hepatic DDAH-1 was restricted to hepatocytes, and expression decreased significantly in cirrhosis. In BDL rats, reduced DDAH-1 expression was associated with elevated hepatic ADMA, reduced eNOS activity and high PP. OA treatment significantly increased DDAH-1 expression, reduced hepatic tissue ADMA, and increased liver NO generation. PP was significantly reduced in BDL+OA vs. BDL+vehicle (8±1 vs. 13.5±0.6 mmHg; p<0.01) with no change in the mean arterial pressure (MAP). Similarly, DDAH-1 hydrodynamic injection significantly increased hepatic DDAH-1 gene and protein expression, and significantly reduced PP in BDL+DDAH-1 vs. BDL+saline (p<0.01). CONCLUSIONS This study demonstrates DDAH-1 is a specific molecular target for portal pressure reduction, through actions on ADMA-mediated regulation of eNOS activity. Our data support translational studies, targeting DDAH-1 in cirrhosis and portal hypertension.

Developments and controversies in the management of oesophageal and gastric varices

Portal hypertension is a milestone in the progression of cirrhosis and heralds the onset of the fatal complications of liver disease. Over the last two decades significant progress has been made in the diagnosis and treatment of portal hypertension and of variceal haemorrhage (VH) in particular. As Bertrand Russell, the British philosopher, said, ‘the most savage controversies are those about matters as to which there is no good evidence either way’. This article will focus on developments in screening for oesophageal varices (OVs), for the prophylaxis of VH from OVs and gastric varices (GVs), and the remaining controversies that will shape treatment strategies for portal hypertension in the coming decade. VH is a catastrophic event for the patient with cirrhosis, with 1-year mortality of up to 40% following an episode of VH.1 2 Since one-third of patients with OVs will develop VH, screening for OVs in the cirrhotic population is mandatory on the basis of the availability of effective diagnostic and prophylactic measures. Oesophagogastroduodenoscopy (OGD) is considered the primary modality for the detection and surveillance of OVs, and is recommended at the time of diagnosis of cirrhosis.3 4 With this approach, OVs are found to be present in ∼40% of patients with compensated cirrhosis.5 The interval for endoscopic surveillance for OVs following an initial negative examination depends on the rate of progression of OVs. The overall incidence of OVs is 5–10% per year,5 although this is modified by the severity of portal hypertension. Current guidance is for OGD to be repeated at 1–3 yearly intervals depending on the severity of liver disease (box 1).3 4 ### Box 1 Recommendations for the diagnosis and management of OV #### Recommendations for endoscopic screening for OVs (modified from Garcia-Tsao et al 4): 1. Compensated cirrhosis without varices: 2–3 years. 2. Compensated cirrhosis with clinically significant portal hypertension (HVPG >10 mm Hg) without varices: 1–2 years. 3. Compensated cirrhosis and small varices: 1–2 years.* 4. Decompensated cirrhosis: 1 year. …

Cracking the ENCODE: from transcription to therapeutics.

The human genome encodes the blueprint of life, but the function of the vast majority of its nearly three billion bases is unknown. The Encyclopedia of DNA Elements (ENCODE) project has systematically mapped regions of transcription, transcription factor association, chromatin structure and histone modification. These data enabled us to assign biochemical functions for 80% of the genome, in particular outside of the well-studied protein-coding regions. Many discovered candidate regulatory elements are physically associated with one another and with expressed genes, providing new insights into the mechanisms of gene regulation. The newly identified elements also show a statistical correspondence to sequence variants linked to human disease, and can thereby guide interpretation of this variation. Overall, the project provides new insights into the organization and regulation of our genes and genome, and is an expansive resource of functional annotations for biomedical research.

Breaking Bad – the two sides of gut microbiota in portal hypertension

The main protagonist of the Emmy award-winning television series ‘Breaking Bad’, chemistry teacher Walter White, is famously noted for his ambiguous nature as both ‘hero’ and ‘anti-hero’. In a similar vein, the ambiguous influence of microbiota on liver disease has been an active area of discussion in hepatology circles, not least at the recent EASL meeting. Various authors have described the gut microbiota as ‘the forgotten organ’(1) and ‘the hottest subject in medicine’(2). In parallel, the pathobiology of portal hypertension, and the role of gut microbes and innate immune signalling in exacerbating intrahepatic resistance, has also been the subject of a renaissance of interest (3). In this issue of Liver International, Rincon and colleagues bring these threads together by exploring the effects of modulating gut flora on portal hypertension in the context of decompensated cirrhosis. An association between bacterial infection and portal hypertension in cirrhosis has been known for some years. Indeed, bacterial infection is an independent predictor of the occurrence of variceal haemorrhage (VH), and is also the strongest independent factor associated with failure to control VH, earlier re-bleeding, coagulation abnormalities and mortality (4, 5). A broader role for gut microbiota in the development of certain complications of cirrhosis, such as hepatic encephalopathy and spontaneous bacterial peritonitis (SBP), has been recognized for some time. Indeed, it was appreciated some years ago that the small intestine of cirrhotic patients is frequently colonized with coliform bacteria that may translocate and cause infection (6). However, the gut microbiota, and the complex effects of host– microbiome interactions, have only recently become the target of intense scrutiny with novel genomic, microbiological and immunological techniques. With regard to bacterial translocation, this has been shown to be a common event in advanced cirrhosis, and may therefore be a key event in several complications of cirrhosis (7). Previous studies using antibiotics to modulate gut flora and reduce portal pressure yielded mixed results. Two early studies demonstrated a beneficial effect on portal pressure, although neither was placebo controlled (8, 9). The subsequent two controlled studies did not reproduce a beneficial effect, although both showed a trend towards reduction in HVPG, which may reflect that they were inadequately powered to demonstrate an effect (10, 11). From a mechanistic perspective, there is evidence to support the hypothesis that downstream pro-inflammatory signals following bacterial translocation may lead to increased intrahepatic resistance. Serum bacterial DNA levels, as a surrogate marker of bacterial translocation, are correlated with severity of inflammation and portal hypertension in cirrhosis (12). Moreover, in patients with SBP, elevated levels of catecholamines and TNF-a are associated with higher HVPG (13). Thus, although a causal relationship remains to be fully established, the basis for modifying the gut microflora to prevent bacterial translocation and consequent complications of portal hypertension is founded on consistent and robust observations. The alternative strategies for modifying this process with antibiotics or anti-inflammatory agents would be limited by bacterial resistance or toxicity concerns. Rincon and colleagues have, instead, taken an approach to modify the gut microbiota using probiotics, hence producing a less ‘toxic’ flora. Of course a ‘healthy’ microbiome is not yet fully defined, and by implication, the constituents of ‘therapeutic’ probiotics also remain to be defined. Nevertheless, probiotics have been shown in varying settings, such as infectious diarrhoea and necrotizing enterocolitis, to improve intestinal integrity, and modulate gut-derived inflammatory responses (14, 15). The mechanisms of these beneficial effects remain to be ascertained, but proposed pathways include suppressing growth of pathogenic bacteria, blocking epithelial attachment by pathogens, enhancing mucosal function and directly modulating host immune responses (16). One of the most commonly used probiotic preparations is VSL#3, a multispecies probiotic consisting of eight strains of bacteria. This preparation has been used in rodent models and in humans with liver disease, with some favourable results in non-alcoholic steatohepatitis, and hepatic encephalopathy. Moreover, VSL#3 has also been shown to improve liver function tests, pro-inflammatory cytokines and degree of oxidative stress in cirrhotic patients. In the context of portal hypertension, two groups have evaluated its use in differing stages of cirrhosis. The Alberta group previously showed no significant benefit of VSL#3 supplementation on HVPG in small placebo-controlled studies comprising compensated and early decompensated patients, although a trend towards HVPG reduction was noted (17, 18). By contrast, Gupta et al. demonstrated a significant beneficial effect of VSL#3 when added to propranolol, in a larger double-blind,

ADMA and hepatic endothelial dysfunction in cirrhosis – the DDAH isoform is the key

To the Editor: We read with interest the recent article by Yang et al. (1) reporting improved hepatic endothelial dysfunction (HED) with vitamin E in cirrhotic rats. The authors have performed a rigorous assessment of liver haemodynamics confirming prior observations of portal pressure lowering effects with anti-oxidant therapy (2), whilst also demonstrating that beneficial haemodynamic effects occur without changes in hepatic inducible nitric oxide synthase (NOS). This affirms notions that largely endothelial NOS (eNOS) maintains intrahepatic resistance (3). However, their observations contrast markedly with the literature in two key areas that significantly impacts on the interpretation of their data. Firstly, Yang et al. describe their bile duct ligated rats (BDL) cirrhotic model as having low eNOS and Phospho-eNOS expression. This is at variance with established literature demonstrating increased eNOS mRNA and protein in BDL rats and other cirrhosis models (4). Indeed, it was these observations, in part, that have moved the field towards assessing eNOS activity and its post-translational regulation and inhibition, as eNOS expression itself is not rate limiting. Secondly, Yang et al. (1) describe dimethylarginine dimethylamino hydrolase 2 (DDAH2) as the predominant DDAH liver subtype. This is also at odds with previous studies demonstrating DDAH1 as the most prominent liver subtype at both gene and protein levels (5,6). Moreover, DDAH1 has several-fold more potent hydrolase activity than DDAH2, and is therefore the major pathway for hepatic asymmetric dimethylarginine (ADMA) degradation (6). It is interesting, therefore, to note that following vitamin E therapy, the authors show increased eNOS and DDAH-2 mRNA but ‘no’ change in DDAH1 gene expression. Protein expression is not described in the study by Yang et al. By contrast, we have shown that DDAH1 protein levels are reduced in BDL rats and these levels are restored with anti-TNFa therapy, associated with reduction in ADMA levels and portal pressure (7). It is most likely that reduced oxidant stress and potential nitrosative inhibition of DDAH-1 activity might explain the findings of Yang et al. (8). It is also important to note that the authors have interpreted their data as largely reflecting improved ADMA metabolism and not reduced hepatic ADMA generation, despite showing decreased PRMT levels after vitamin E therapy, implying potentially less arginine methylation. In summary, although modulation of redox state shows promise as a therapeutic option in liver disease, further studies with specific targeting of DDAH subtypes, along with mechanistic evaluation of NOS and DDAH activity is required to definitively show the importance of the ADMA-DDAH axis as a target in portal hypertension.

Hepatitis C treatment and quality of life – You can’t always get what you want, but you might get what you need

Chronic hepatitis C causes progressive hepatic fibrosis, and, in a proportion of infected patients, leads to cirrhosis and hepatocellular carcinoma. Severe extrahepatic manifestations are well recognised, including mixed type 2 cryoglobulinemia, non-Hodgkins lymphoma, porphyria cutanea tarda, and possibly type 2 diabetes. In addition to the major extrahepatic syndromes associated with chronic hepatitis C, several studies have pointed to a decrement in health-related quality of life (HRQL) in hepatitis C, in comparison to the general population [1–3]. HRQL is defined as a person’s subjective assessment of a range of conditions that can affect that person’s perception of their state of health [3]. The most common symptoms cited by patients with hepatitis C and that impact upon their quality of life are fatigue, depression, anxiety, cognitive impairment and painful muscle and joint symptoms. The biological mechanisms and pathogenesis of these symptoms remain uncertain. Some symptoms may be due to the release of inflammatory cytokines, or the direct presence of HCV in the central nervous system. Those with greater histological hepatic inflammatory activity may have worse fatigue, but symptoms tend to be present irrespective of the degree of hepatic fibrosis. New DAA’s have resulted in striking SVR rates, even in patients with advanced fibrosis due to hepatitis C [4]. What then are the effects of interferon-free treatments on HRQL? Younossi and co-authors have examined quality of life in patients with chronic hepatitis C in prospective trials of sofosbuvir and ledipasvir, to assess alterations in HRQL before, during and after treatment [5]. Their data provides an important opportunity to assess patient-reported outcomes (PROMs) during treatment and following an SVR, without the confounding effects of interferon that markedly reduced ‘on-treatment’ quality of life indices in previous studies [6–9]. The authors’ provide a detailed and comprehensive analysis of HRQL and work productivity in patients with different stages of hepatic fibrosis treated with sofosbuvir and ledipasvir, or

Obeticholic acid, a Farensoid-X receptor agonist, improves portal hypertension by two distinct pathways in cirrhotic rats: HEPATOLOGY, Vol. XX, No. X, 2014 CORRESPONDENCE

majority of the hepatic events, which could have been prevented by adequate endoscopic surveillance and treatment, either pharmacologically or endoscopically. In general esophageal variceal bleeding occurs at an annual rate of 5%-15% in patients that did not receive treatment for varices, which was significantly higher than the 3-year incident rate of 5.2% and 8.7% for variceal bleeding in the entecavir and control cohort, respectively. Furthermore, the surveillance provided to two cohorts of our patients was comparable or even better in the entecavir cohort because of advances in endoscopic therapy. Therefore, our observations still supported the conclusion that entecavir reduces hepatic events. Because of the nature of real-life cohort studies, there were significant differences in baseline characteristics between two cohorts of patients. We minimized bias by adjusting all clinical outcomes for the Model for End-Stage Liver Disease score, together with maintained viral suppression, another important difference, among patients with liver cirrhosis using Cox’s proportional hazard model. Risk reduction remained robust. For example, the adjusted hazard ratio for hepatic events was 0.51 (95% confidence interval: 0.340.78; P 5 0.002). Subsequent sensitivity analyses further highlighted the benefits of entecavir among patients at higher risk of events. The issue of inequality of follow-up duration in two cohorts was taken into account during the whole process of analysis. All clinical outcomes were estimated by survival analysis to allow for censoring of patients who had shorter follow-up durations. Besides, the cumulative probability of clinical events at 3 years was also lower in patients with cirrhosis receiving entecavir treatment. Finally, though it is true that the clinical benefit of antiviral therapy can only be firmly established by randomized, controlled trials, withholding treatment in patients with active liver injury or cirrhosis is no longer ethical. In such circumstances, one has to make clinical decisions based on the best available data. As such, wellcharacterized cohort studies offer valuable insights into the benefits and limitations of antiviral therapy.

All beta-blockers are created equal, but some beta-blockers are more equal than others.

Portal hypertensive-associated variceal bleeding in advanced cirrhosis carries a high mortality, and thus the landmark observation by Lebrec et al. (1) that betablockade with propranolol could significantly reduce the incidence of first variceal bleeding has defined the standard of care for the last three decades. Numerous mechanistic studies and clinical trials have followed showing that administration of non-selective betablockers (NSBB) results in both a decrease in cardiac output, because of b1 adrenoceptor blockade, and in addition a blockade of vasodilatory b2 receptors in the splanchnic circulation, resulting in decreased splanchnic blood flow, collectively lowering portal pressure. However, the lack of receptor selectivity with these agents leads to adverse events limiting the tolerance to this class of agent, and also partially accounts for the reported limited efficacy of NSBB, which is around 40 –50 per cent. Moreover, experimental data in rodent models show that b1 receptors are relatively down-regulated in cirrhosis with a greater predominance of b2 (2), perhaps also suggesting that the dose of NSBB is likely to be important and that in clinical studies, lack of clarity of achieving optimal dose may further cloud the interpretation of efficacy data. As our knowledge of the pathobiology of cirrhosis has advanced over recent decades, so have our therapeutic aims in patients with portal hypertension. Elegant studies by several groups have showed that as well as increased splanchnic and portal inflow, increased intrahepatic resistance is central to the development of portal hypertension (3, 4). Moreover, nitric oxide (NO) has been shown to be a key mediator of intrahepatic vascular tone, and that intrahepatic resistance in cirrhosis is partly because of endothelial dysfunction and decreased NO bioavailability. Indeed, several lines of investigation suggest that increased intrahepatic resistance may be the principal event signalling to the splanchnic and mesenteric vasculature, and may, therefore, precede the development of splanchnic vasodilatation in cirrhosis (5). More recent studies suggest that the presence of inflammation or/and infection exacerbates portal hypertension, in part, through further impact on endothelial dysfunction and increased intrahepatic resistance and underpins the rationale for use of antibiotics incorporated into guidelines for the management of variceal bleeding (6–9). Collectively, informed by this greater mechanistic knowledge and with recent advances in cardiovascular pharmacology, attention has turned to new targets for therapy. Whilst the first and second generation b-blockers have no significant ancillary properties, more recent generations of b-blockers may be selective or nonselective for b1 or b2 adrenoceptors, but in addition possess important further properties. Amongst these, Nebivolol and Carvedilol have been studied most intensively. Nebivolol is a highly selective b1-blocker with additional vasodilatory properties, acting through several pathways, which may include action on the b3 receptor. b3 adrenoceptor expression and its associated signalling G-protein alpha-s subunit have been noted to be increased in hepatic and mesenteric arteries of cirrhotic BDL rats, with consequent moderation of cAMP and calcium mobilization, resulting in vasodilation, in addition to promoting increased cGMP(2). This mechanism is thought to be most relevant in advanced stages of disease, when catecholamine levels are at their highest, perhaps perpetuating splanchnic vasodilatation. Traditional NSBB such as propranolol are believed to have little effect on b3 receptors. In contrast, carvedilol is a non-selective B-blocker with additional a1 adrenoceptor-blocking properties, counteracting intrahepatic tonic vasoconstriction and augmenting the portal pressure lowering effect of NSBBs, as well as reducing portal-collateral blood flow (as with all other NSBBSs). However, these vasodilatory effects are not liver-specific, and therefore the potential exists for systemic hypotension, activation of renin–angiotensin and exacerbation of sodium and fluid retention in cirrhosis. The importance of unchecked tonic activities of the alpha adrenergic pathway in portal hypertension are highlighted by the observation of Villaneuva et al. that addition of the a1 blocking agent, Prazosin, to Nadolol has a greater effect on hepatic venous pressure gradient (HVPG) lowering than Nadolol plus nitrate donation through isosorbide mononitrate (10). This said, the efficacy of carvedilol in the management of portal hypertension related bleeding has demonstrated a HVPG reduction between 16 and 43% vs. 12–13% with propanolol. Subsequently, two small RCTs have shown some efficacy in the use of carvedilol to prevent recurrent variceal haemorrhage (11, 12). However, both of these studies have some limitations, especially in relation to the excess re-bleeding rates

PMO-123 Gene transfer of dimethylarginine dimethylaminohydrolase-1 reduces portal pressure in a rodent model of cirrhosis

Introduction Portal hypertensive bleeding is a grave complication of cirrhosis. Asymmetric dimethylarginine (ADMA), an endogenous eNOS inhibitor, is elevated in cirrhosis, relates to degree of portal hypertension, and is prognostic in acute-on-chronic liver failure. Dimethylarginine-dimethylaminohydrolase-1 (DDAH-1), a key enzyme metabolising hepatic ADMA, is reduced in cirrhosis. Therapies which indirectly increase hepatic DDAH-1, such as anti-TNFα therapy or FXR agonists, lead to reduced hepatic ADMA, increased NO and lowered portal pressure. Therefore, there is accumulating evidence for DDAH-1 as a therapeutic target, but specific evidence for DDAH-1 reconstitution is lacking. The aim of this study was to adopt a DDAH-1 gene therapy strategy in portal hypertension. Methods Hydrodynamic injection leads to hepatic gene transduction by causing turbulent, retrograde venous flow, permeation of hepatic parenchymal cells and consequent plasmid expression. Human DDAH-1 cDNA was cloned into the pCMV-Sport6 expression plasmid. Sprague-dawley rats (n=9) underwent bile duct-ligation, and after 4 weeks were injected with 800 μg of either pCMV-Sport6-DDAH1 (n=5) or non-expressing control plasmid (n=4). After 72 h, rats underwent direct portal pressure assessment under anaesthesia and were then sacrificed. Plasma ALT was measured by Cobas-Integra analyser. Transgene expression was measured by quantitative PCR, with Taqman probes specific for human DDAH-1 to distinguish rodent DDAH-1. Protein expression was measured by western blot. Results Highly effective gene transfer (between 20 and 120-fold increase) of human DDAH-1 was seen in 3 out of 5 treated animals. None of the animals treated with control plasmid expressed human DDAH-1. The three “responders” to gene therapy also had highly significantly increased DDAH-1 protein expression compared with “non-responders” or controls (p<0.05). There was no difference in ALT between the groups. Portal pressure was significantly lower in “responders” to gene therapy than “non-responders” or animals treated with control plasmid (p<0.05). Conclusion This study demonstrates that DDAH-1 is a specific molecular target for portal pressure reduction. Hydrodynamic injection is variable in efficiency of gene delivery due to the nature of the technique. However, despite these limitations, this study clearly shows proof of concept for efficient vector-based DDAH-1 gene therapy in lowering portal pressure and preventing bleeding in cirrhosis. Competing interests None declared.