Hepatology highlights


  • Potential conflict of interest: Nothing to report.

FLIP for the Diagnosis of NASH

As long as we are missing noninvasive histologic biomarkers, the diagnosis of nonalcoholic steatohepatitis (NASH) will remain a histologic diagnosis. Rather complex histologic scores have been proposed that are useful for clinical trials, although a simple, reproducible, diagnostic algorithm based on a few well-characterized features is needed. Bedossa with colleagues of the European Fatty Liver Inhibition of Progression (FLIP) Pathology Consortium propose an elegant three-step histological diagnosis. It considers steatosis, ballooning, and lobular inflammation. The diagnosis of NASH requires the presence of ballooning and lobular inflammation on top of steatosis. Without lobular inflammation or without ballooning, the diagnosis of NASH cannot be made. Each feature is graded semiquantitatively with a three-level scale. To demonstrate the utility of this approach, two groups of pathologists had to categorize 40 liver biopsies before and after training. Diagnosis concordance increased significantly. Similar to the Metavir score for chronic hepatitis C (CHC), the FLIP algorithm will become standard practice. (Hepatology 2014;60:565-575.)

FLOP for the Secretion of Phospholipids

Bile is an unusual fluid containing a mixture of lipids in an aqueous environment. Secretion of cholesterol, bile acids, and phospholipids is mediated by specialized transporters located in the canalicular membrane. The balance between these compounds is essential for effective biliary secretion. ABCB4 flops phospholipids from the inner to the outer leaflet of the canalicular membrane. This is a particularly relevant transporter in hepatic pathophysiology; several diseases have been linked to mutations in its gene, in particular, the low phospholipid-associated cholelithiasis. Gautherot et al. investigated the effects of two missense mutations of the N-terminal domain, which have been associated with this disease. They report that the T34M- and R47G-mutated ABCB4 proteins are normally expressed, but have defective phosphatidylcholine secretion activity as a result of impaired phosphorylation by protein kinase A or C. Because protein kinase C can be activated by bile acids, this work underlies an important mechanism by which bile acids promote ABCB4-mediated phospholipid biliary secretion. (Hepatology 2014;60:610-621.)

Better Than Passive and Active

Passive and active vaccination at birth of infants born to hepatitis B viremic mothers drastically decreases the risk of vertical transmission. Nevertheless, a residual ∼10% risk remains, especially if the mother has a high viremia. Theoretically, hepatitis B virus (HBV) viremia can be reduced by administration of an antiviral treatment during the third trimester of pregnancy. In order to test the efficacy and safety of this approach, Zhang et al. enrolled 700 pregnant women with a HBV viremia of >1 million copies/mL to receive lamivudine, telbivudine, or no treatment from the 28th week of pregnancy up to 4 weeks after delivery. This was not a randomized trial. Approximately half of the women received a treatment, and telbivudine was prescribed 5 times more frequently than lamivudine. On-treatment analysis found no transmission in the case of treatment and 2.8% in the case of no treatment. Treatment appeared safe for the infants, but was associated with some alanine aminotransferase (ALT) flares in the mothers, none of whom had ALT above 10× the upper limit of normal. Despite methodological shortcomings in its design, this clinical study delivers a strong signal in favor of this strategy. (Hepatology 2014;60:468-476.)

Stats on Statins

Statins are among the most widely prescribed drugs. Not infrequently, the treatment is interrupted as a result of an elevation of aminotransferase levels. Such an elevation occurs in approximately 3% of patients. But, how frequently are the statins actually hepatotoxic? Russo et al. used the registry of the Drug Induced Liver Injury Network to answer this question. This registry contains prospective information on 1,188 cases. Only 22 cases were the result of statins, but these cases were all severe, with 4 patients developing hepatic failure, and there was 1 death. The latency to onset of liver injury was strikingly long, with a median of >5 months. Some patients had an autoimmune phenotype, and these patients were particularly at risk of developing a chronic liver injury. Importantly, this appears to be a class effect, with no particular statin being more dangerous than another. The investigators conclude that prospective monitoring for statin-induced liver injury is not warranted. Another important message is that a statin can be hepatotoxic years after its introduction and should not be discarded as a cause of liver injury based on a long latency. (Hepatology 2014;60:679-686.)

The Content of the Cup Matters

Patients with liver diseases are aware that their habits play a role. Except for alcohol, it is difficult to answer their questions regarding diet. There is experimental and epidemiological evidence to suggest that coffee might be hepatoprotective. Goh et al. add an important piece of evidence to the literature. In a prospective, population-based cohort of >63,000 Chinese, they investigated drinking habits and cirrhosis mortality over a 14-year follow-up period. As expected, alcohol had a dose-dependent association with cirrhosis mortality. Coffee intake had a negative dose-dependent inverse association with nonviral hepatitis cirrhosis mortality. What else? There was no association between consumption of black tea, green tea, fruit juices, and soft drinks and the risk of cirrhosis mortality. The content of the cup matters, and this is not only about caffeine, because tea was not hepatoprotective. (Hepatology 2014;60:661-669.)

The Content of the Plate Matters

What about food? Do eating habits have an effect on the risk of developing a liver disease? This kind of research faces major challenges. To be representative and based on a large number of subjects, it cannot be interventional. To reach reasonable conclusions, it has to be prospective with a long observation time and it needs to capture the start of the relevant information. The work of Li et al. is, in this sense, exemplary. Between 1995 and 1996, 494,942 U.S. residents filled out a food-frequency questionnaire. This information served to calculate scores representing dietary patterns. Association with the development of hepatocellular carcinoma and occurrence of liver-related death up to the year 2011 were determined and adjusted for alcohol intake, body mass index, and diabetes. Healthy dietary pattern and a high Mediterranean diet score were significantly hepatoprotective. Even if confounding factors linked to lifestyles are likely, the message of this work can be implemented in our daily practice. (Hepatology 2014;60:588-597.)

HIV: Must HBV Viremia Be Zero?

Patients coinfected with human immunodeficiency virus (HIV) and HBV are frequently treated with tenofovir (TNV). This nucleotide analog is expected to bring the HBV viremia to undetectable levels. This is not always the case, despite anamnestic adherence to the treatment. The magnitude and significance of this situation are poorly known. Boyd et al. investigated 111 coinfected patients receiving a TNV-containing antiretroviral therapy. Seventy-seven percent of patients reached and remained with a negative HBV viremia. In 3% of patients, the HBV viremia was occasionally above 2,000 IU/mL, and these patients were nonadherent. In 20% of patients, low levels of HBV viremia were repeatedly observed, despite detectable plasma levels of TNV, which suggests adherence to the treatment. No specific mutations in the HBV polymerase gene could be found. These patients did not have worse clinical outcome, but none seroconverted. Suboptimal HBV control is not rare in coinfected patients receiving nucleotide analogs, suggesting a role for the immune system. (Hepatology 2014;60:497-507.)

Truncated Acute Hepatitis C

Nearly one quarter of individuals acutely infected with hepatitis C virus can clear the virus spontaneously. Understanding the mechanisms at play would allow us to address why they fail in the majority of the population. C-X-C chemokine 10 (CXCL10) attracts antiviral T and natural killer (NK) cells. The protease, dipeptidylpeptidase 4 (DPP-4), cleaves CXCL10, and truncated CXCL10 acts as a chemokine antagonist. Riva et al. studied in detail 16 patients with acute hepatitis C. The 5 patients who spontaneously cleared the virus had less DPP-4 activity and lower concentrations of CXCL10, but it was predominantly untruncated, in contrast to the 11 who developed CHC. This was associated with higher frequency of cytotoxic NK cells and interferon-gamma-producing T cells. This elegant work suggests that inhibition of DPP-4 could favor viral clearance. Such inhibitors are already marketed as antihyperglycemic drugs. We eagerly await more published work on this approach. (Hepatology 2014;60:487-496.)

If Less Is Bad, More Is Not Necessarily Good

Nuclear factor (erythroid-derived 2)-like-2 factor (Nrf2) is a transcription factor activated by reactive oxygen species, which governs the expression of antioxidant proteins and detoxifying enzymes. Liver regeneration is impaired in mice lacking Nrf2. Because pharmacological activation of Nrf2 may have chemopreventive and anti-inflammatory properties, it appears interesting to investigate whether Nrf2 activation may promote liver regeneration. Using mice with a constitutively active Nrf2, Köhler et al. found the opposite to be the case. They observed a delayed hepatocyte proliferation and enhanced apoptosis after partial hepatectomy. They explain these findings by an increased expression of the cyclin-dependent kinase inhibitor, p15, and the proapoptotic protein, Bcl2l11, which are targets of Nrf2. Optimal liver regeneration requires this sensor to be present and not activated. (Hepatology 2014;60:670-678.)

After Vasopressin Comes Vasohibin

Cirrhosis requires the formation of new blood vessels. Angiogenesis is closely linked to fibrosis in the disruption of the normal liver architecture and plays an essential role in the progression of portal hypertension. Vasohibin-1 is a newly identified endogenous inhibitor of angiogenesis, which has the peculiar property of being induced by vascular endothelial growth factor as a negative feedback mechanism. Coch et al. found that vasohibin-1 is overexpressed in the mesentery and liver during cirrhosis. Adenoviral-mediated vasohibin-1 gene transfer attenuated mesenteric and intrahepatic pathologic neovascularization, inhibited hepatic stellate cell activation, and ameliorated portal hypertension in the bile duct ligation model. Importantly, vasohibin-1 seems to have no effect on normal vasculature. This remarkable work suggests that identification of vasohibin-1 receptors would pave the way for pharmacologic manipulation of this pathway. (Hepatology 2014;60:633-647.)