Hepatology highlights


  • Potential conflict of interest: Nothing to report.

Interferon λongo

Ironically, as we are phasing out interferon (IFN)-based combinations to treat chronic hepatitis C, IFN signaling in hepatocytes is getting more attention. In this issue of Hepatology, two groups present their results comparing response to different types of IFNs. They exposed Huh7 cells and primary human hepatocytes to IFNs and performed gene expression profiling with a microarray. Both groups found that type I and III IFNs induced the same set of IFN-stimulated genes, but that the strength of this response differed. Type I IFNs induce a strong response, which is transient with IFN-α and sustained with IFN-β. Type III IFN-λs induce a much weaker, but sustained, response. Bolen et al. correlate gene expression pattern with different phosphorylation of the transcription factor, signal transducer and activator of transcription 1. Jilg et al. investigated the transcriptomic response in Huh7 cells infected with hepatitis C virus (HCV). They found that the presence of HCV blunted the transient response to IFN-α, which became similar to the response to IFN-λ. Because polymorphisms in IFN-λ3 (interleukin-28B) are associated with clinical outcomes, further details on its signaling are relevant. (Hepatology 2014;1250-1261. Hepatology 2014;59:1262-1272)

HBV: Overruling Stopping Rules

Nucleotide analogs are capable of changing the natural history of chronic hepatitis B. If hard endpoints are the occurrence of clinical complications of cirrhosis and death, rendering hepatitis B virus (HBV) viremia negative is a prerequisite for beneficial effects. Therefore, a lack of decline of the HBV viremia has been proposed as a measure of a lack of efficacy of nucleotide analogs and primary nonresponse, based on the HBV viremia kinetic, a reason to stop their administration. Yang et al. investigate how the American Association for the Study of Liver Diseases and European Association for the Study of the Liver stopping rules perform for entecavir. They examined 1,254 treatment-naïve patients who were treated with entecavir. Whether the patients had a primary nonresponse or not did not affect their chances of having a negative viremia 54 months after introduction of the treatment, only that the time to achieve virologic response was delayed. This work illustrates again how we should be cautious in translating to HBV concepts that have proven solid for HCV. (Hepatology 2014;59:1303-1310.)

Glomerular Filtration Rate in Patients With Cirrhosis

Renal function is the Achilles' heel of patients with cirrhosis. Correct assessment of renal function is essential in the management of patients with cirrhosis not only to decide whether a patient should receive a combined liver-kidney transplant, but simply in the daily adjustment of diuretics. The exact determination of the glomerular filtration rate (GFR) by insulin clearance is too cumbersome to be used routinely. In practice, decisions are made based on estimation of GFR by equations. Several have been proposed, but they may not perform equally in patients with cirrhosis. Francoz et al. compared the accuracy of these equations in 300 patients with cirrhosis evaluated for liver transplantation; De Souza did the same in 202 patients and Mindikoglu in 72 patients. These three articles deliver a similar message: The accuracy of equations to estimate GFR declines with progression of cirrhosis and worsening of renal function. If Modification of Diet in Renal Disease (MDRD)-6 performs better than MDRD-4, equations based on cystatin-C determination are more accurate, in particular, the Chronic Kidney Disease Epidemiology Collaboration equations. Two lessons emerge from these works: Equations to estimate GFR accurately in patients with cirrhosis should be developed specifically for this population and determination of cystatin-C should become more widely available. (Hepatology 1514-1521. Hepatology 2014;59:1522-1531. Hepatology 2014;59:1532-1542.)

Time to Relaxin

Relaxin is a peptide hormone that plays a role during pregnancy to soften pubic symphysis and increase arterial compliance. It is already known that relaxin has antifibrotic properties. In a series of detailed experiments, Fallowfield et al. demonstrate its therapeutic potential in experimental liver cirrhosis. Relaxin acts through a receptor, which is not present on quiescent hepatic stellate cells, but becomes highly expressed after their activation in myofibroblasts. Relaxin stimulates production of cyclic guanosine monophosphate and nitric oxide and decreases myofibroblast contractility and induces an antifibrogenic phenotype. In two models (bile duct ligation and carbon tetrachloride intoxication), relaxin reduced hepatic fibrosis and selectively improved portal pressure without altering mean arterial pressure. Relaxin combines many interesting features that make it a prime candidate for further clinical investigations. (Hepatology 2014;59:1492-1504.)

Intrahepatic Cholestasis of Pregnancy: Measure Bile Acids

Intrahepatic cholestasis of pregnancy (ICP) is not rare. It can be severe and less benign than thought. To elucidate the risks associated with ICP, Geenes et al. used the UK Obstetric Surveillance System to identify patients with severe ICP. The researchers defined severe ICP with a serum bile acid level above 40 μmol/L. A total of 713 cases of severe ICP were identified, corresponding to an incidence of just less than 1 per 10,000 pregnancies. Severe ICP was associated with adverse fetal outcomes, including stillbirth (1.5% vs. 0.5% in the control group). There was a significant positive correlation between nonfasting maternal serum bile acid levels and perinatal complications, such as preterm delivery and meconium-stained amniotic fluid. The researchers speculate that bile acids may affect the contractility of muscle cells. In any case, this study, which is unique for its large size, emphasizes the importance of measuring maternal bile acid levels in ICP. (Hepatology 2014;59:1482-1491.)

Nonalcoholic Steatohepatitis: Genders Are Not Equal

In nonalcoholic steatohepatitis (NASH), is fibrosis risk the same in men and women? This sounds like a simple question, and thus far the literature has not been unambiguous. Yang et al. tried to answer this question with a group of 541 patients with histologically proven NASH. They analyzed their cohort comparing men to pre- and postmenopausal women. They found that men and postmenopausal women have comparable severity of fibrosis, whereas premenopausal women have a lower risk for severe fibrosis. The researchers took age into account and tried to adjust for several confounders, but it remains possible that behavior characteristics differ (degree of physical activity, alcohol intake), which could, in part, explain the difference. Nevertheless, the researchers interpret their results by suggesting that premenopausal women might be at lower risk for fibrosis: They mention estrogen, and we can now add relaxin. (Hepatology 2014;59:1406-1414.)

Caffeine Is mTORiffic

Often, patients with liver disease think they should renounce alcohol, fast food, and coffee. If there is one piece of dietary advice we can give them, it is not to stop drinking coffee. According to numerous epidemiologic studies, coffee is good for the liver. But how? This is the question. Sinha et al. describe how caffeine decreases steatosis. They found that caffeine induces the formation of autophagosomes in HepG2 cells. Knocking down ATG5 or blocking lysosomes with chloroquine prevented caffeine-induced reduction in intracellular lipids. They complement these in vitro studies with in vivo experiments administering caffeine to mice. Caffeine induced hepatic β-oxidation, increased autophagy, and, interestingly, decreased mammalian target of rapamycin (mTOR) signaling. Then, the researchers fed the mice a high-fat diet. Caffeine decreased weight gain and prevented intrahepatic lipid accumulation. These effects were obtained at concentrations that are reached after drinking coffee. No reason for our patients to give up this “drug,” on the contrary! (Hepatology 2014;59:1366-1380.)

Cholangiocarcinoma With (ef)fusion

Intrahepatic cholangiocarcinoma (CCC) shares a rising incidence and poor prognosis with hepatocellular carcinoma (HCC). Systemic targeted therapy with the potential to prolong the survival of patients affected by this type of tumor is urgently needed. This requires the identification of molecular targets. Arai et al. performed whole transcriptome sequencing in patients affected by intrahepatic CCC without KRAS/BRAF/ROS1 alterations. They used a strategy to identify fusion proteins, then identified two fusion kinase genes involving fibroblast growth factor receptor 2 (FGFR2). Extending their finding in a group of 66 patients with intrahepatic CCC, they detected FGFR2 fusion in 13% of the cases. These patients did not present with peculiar clinical features, except an association with viral hepatitis, but the numbers are small. FGFR2 fusion kinase seems to be a very rare event in HCC (1%) and could not be detected in extrahepatic CCC cholangiocarcinoma. Expression of the fusion kinase transformed NIH3T3 cells and FGFR kinase inhibitors reversed these effects. This work illustrates the importance of stratifying cancer patients for driver alterations. A trial testing FGFR2 kinase inhibitor in intrahepatic CCC is very likely to be negative if its population is not selected for FGFR2 activation. (Hepatology 2014;59:1427-1434.)

HCC Screening: Ask the Model

Evidence-based medicine requires demonstration of benefits with randomized control trials. However, some clinical practices establish themselves nevertheless without such evidence. This is the case for liver transplantation. More controversial is the case of enrolling patients at risk for HCC in a surveillance program. American and European guidelines recommend screening with ultrasonography every 6 months. There is one randomized trial that has the merit to address this issue, but shortcomings in its realization have limited its importance. Because performing such a trial would currently face major ethical concerns, one of the best ways to address this issue is by mathematical modeling. Mourad et al. developed a sophisticated Markov model in patients with compensated HCV-related liver cirrhosis. They determined life expectancy in a fictitious cohort of 700 patients according to several scenarios. They were first cautious to confirm that the assumptions they had to make result in outcomes that closely fit reality. Then, with this model, they were able to show the complementary importance of access to screening and effectiveness of screening. The researchers artfully took into account lead time bias. These results clearly show a survival benefit with regular screening and add arguments in favor of HCC surveillance. (Hepatology 2014;59:1471-1481.)