Hepatology Highlights


AMA in PBC: A Pathogenetic Role at Last!

The work of Matsumura et al. may prove to be a major development in our understanding of the pathogenesis of primary biliary cirrhosis (PBC). Hitherto, despite being the hallmark of the disease, no plausible role in pathogenesis of PBC has been attributed to the antimitochondrial antibodies (AMA). Following its binding to an in vitro cell line transfected with human polymeric immunoglobulin (Ig) receptor, dimeric IgA with specificity for the PDC-E2 component of AMA induced a significant increase of caspase activation, a pro-apoptotic change. Caspase activation was significantly increased in response to IgA from 7/8 AMA positive PBC patients but not in controls (see figure). Caspase activation was demonstrated when the specificity of IgA was for PDC-E2 but not for BCOADC-E2 or OGDC-E2, other mitochondrial antigens that react with AMA. AMA of IgG subtype did not activate the caspase system. The slow course of caspase activation suggested that it was stress mediated rather than a response to binding of IgA to cell surface death receptors. The data are compatible with the hypothesis that IgA with specificity for PDC-E2, as it transits the biliary epithelial cell, binds newly synthesized PDC-E2 that is en route from endoplasmic reticulum to mitochondria, thus disrupting vital functions and threatening cell survival. The selective concentration of IgA in bile and other mucosal secretions may, equally plausibly, be the source of similar pathology in other secretory epithelia, thus accounting for the characteristic “sicca” dry gland problems found in association with PBC. (See HEPATOLOGY 2004;39:1415–1422.)

Illustration 1.

Adverse Outcome From High-Dose Infliximab and Steroids for Alcoholic Hepatitis

This study is commended not only for double-blind randomization but also for the effective means of monitoring that led to its timely, albeit premature, termination. Following upbeat reports based on smaller or unrandomized pilot studies, which it contradicts, the case for controlled trials is well represented. Patients with biopsy-proven severe acute alcoholic hepatitis (AAH) who met stringent inclusion criteria were given prednisolone 40 mg daily and allocated to receive Infliximab (monoclonal anti-TNFα) or placebo. The dose of infliximab was greater than in previous trials for Crohn's disease or rheumatoid arthritis because of the relatively higher blood levels of TNFα reported in AAH. The study was terminated prior to the first interim analysis because the primary endpoint of death within 2 months had unbalanced rates in the 2 groups (death was reported in 7 receiving infliximab and 3 receiving placebo) (see figure). Severe infections were more frequent with infliximab (10 episodes in 8 patients, 4 of whom died, compared with just 1 episode leading to the death of that patient in the placebo arm). When neutrophils were stimulated ex vivo with lipopolysaccharide or formyl methionyl leucyl phenylalanine, their ability to respond was dramatically reduced compared to the placebo/steroid-treated group and controls. It has generally been assumed that activated neutrophils are responsible for much of the injury in AAH; yet there was no apparent benefit on liver function from infliximab-induced neutrophil disarmament. The authors conclude that they have excluded the possibility that infliximab is better than placebo when combined with steroids in treatment of AAH, and they argue against any future evaluation of this combination in AAH, more so if infection and severe disease were to constitute criteria for exclusion from the study. (See HEPATOLOGY 2004;39:1390–1397.)

Illustration 2.

Unravelling the Tangles of Mallory Bodies— Conservationist Recycling

Mallory's bodies (MBs) are found in the cytoplasm of ballooned hepatocytes affected by alcohol (ASH), nonalcoholic steatohepatitis (NASH), and various other inflammatory or neoplastic liver diseases. They share many characteristics with the neurofibrillary tangles of Alzheimer's disease and the Lewy bodies of Parkinson's disease. Müller et al. report on the immunohistochemical characteristics of MBs within liver from patients with Wilson's disease and two other disorders in which an unknown genetic predisposition and excessive copper ingestion are postulated to combine to produce severe liver disease in childhood. As in ASH and NASH, most MBs were decorated by antibodies against keratin, ubiquitin, and p62. The overall similarity between the findings in these three copper toxicoses and in previous study of ASH and NASH permits the authors to put forward a unifying theory according to which the MB is a deposit of misfolded proteins resulting from genetic mutations, oxidative stress, and other threats to cellular viability. The MBs may be regarded as a sequestrosome for unfolded proteins that have exceeded the cell's capacity for their degradation by proteolysis or repair via the heat shock protein chaperone system. p62, whose production is enhanced in response to both oxidative stress and accumulation of ubiquinated proteins, is known to trigger a signalling cascade that is involved in activation of NFkB, a pathway that aids hepatocyte recovery from injury and inhibits hepatocyte apoptosis. The association of p62 with ubiquitin in MBs thus raises the intriguing possibility that sequestered components of MBs could be recycled by the cell after resolution of the stressful episode and thus aid hepatocyte survival via a vital two-way equilibrium responsive to ambient stress levels. (See HEPATOLOGY 2004;39:963–969.)

Weight Loss Works for NASH and Other Components of the Metabolic Syndrome

The report by Dixon et al. appears to give clear answers to some of the key questions surrounding nonalcoholic steatohepatitis (NASH). Twenty-three patients treated for obesity by laparoscopic adjustable gastric banding had NASH diagnosed on biopsy at the time of the initial laparoscopy. Repeat liver biopsies from these individuals were taken 26 ± 10 months later. Histological staging and grading were performed blindly and results compared for each paired specimen. Only 4 of 23 biopsies failed to show resolution of NASH. Remarkably, of 10 patients with stage 3 fibrosis in their index biopsies, 7 regressed to stage 0. Zone 3 fibrosis had resolved completely in 16 of 23. In assessing initial clinical features that predicted a better than median improvement in histological measures, the metabolic syndrome was found to be the predominant preoperative predictor for improvement by 2 or more in grade and stage of NASH; after controlling for the metabolic syndrome, no other variable was predictive. Of the cohort of 36 subjects with nonalcoholic fatty liver disease in the study, the number who fulfilled the Adult Treatment Panel III criteria for the metabolic syndrome reduced from 23 (64%) to 7 (19%). Therefore it appears that weight reduction, if it can be achieved safely, is a choice treatment for NASH, since it not only reverses precirrhotic pathology within the liver but also tends to ameliorate other aspects of the metabolic syndrome with which nonalcoholic fatty liver disease is strongly associated. (See HEPATOLOGY 2004;39:1647–1654.)

Progression of Fibrosis During Treatment of Autoimmune Hepatitis

The study by Czaja and Carpenter confirms that most patients treated conventionally for precirrhotic autoimmune hepatitis will not have progressive fibrosis. Of 73 patients in whom biopsies were analysed according to the Ishak scoring system, fibrosis scores were stable or regressed in 55 patients (75%). The remaining 25% in whom fibrosis progressed, including 5 who became cirrhotic, could not be distinguished by their clinical course when categorized as having relapse, a failed treatment, or an incomplete response, but patients in whom the fibrosis score increased tended to have longer follow-up. An increase of the Ishak hepatitis activity index of 2 points during therapy was significantly associated with a greater frequency of progressive fibrosis. The authors deduce that in most patients with autoimmune hepatitis there may be a threshold of ongoing inflammatory activity below which fibrosis remains stable or regresses. The single pretreatment characteristic significantly predictive of progressive fibrosis during treatment was heterozygosity for HLA DR3/DR4, found in 30% of those whose fibrosis scores increased by ≥2 points but in only 7% of the entire cohort. The study points to a future in which patients with autoimmune hepatitis may have their risk of progressive fibrosis better known at onset of treatment and in which noninvasive methods of detecting progressive fibrosis permit tailoring of treatment to a threshold which prevents cirrhosis whenever autoimmune hepatitis presents at a precirrhotic stage. (See HEPATOLOGY 2004;39:1631–1638.)

Hyponatremia Exacerbates Brain Osmolyte Depletion in Cirrhosis

The influence of dilutional hyponatremia on the brain in cirrhosis was investigated with 1H–magnetic resonance spectroscopy and magnetization transfer ratio. As previously reported, the brain glutamine/glutamate levels were increased in cirrhosis and correlated with plasma ammonia but did not correlate with serum sodium. In contrast, brain concentrations of myo-inositol and other organic osmolytes such as creatine/phosphocreatine were reduced, and their concentrations correlated strongly with serum sodium concentrations and serum osmolality (see figure). In multivariate analysis, hyponatremia was the only independent predictor of low brain myo-inositol levels. No evidence of cerebral edema was found in this relatively stable group of patients who had been hyponatremic and off all diuretic therapy for several days prior to the study. It is postulated that the reduction of brain osmolyte content may be an adaptive osmoregulatory response that compensates for the increased astrocyte content of glutamine/glutamate, which left unbalanced would result in cerebral edema. Osmolytes such as myo-inositol and phosphocreatine are likely to have many functions other than osmoregulation, and their depletion in hyponatremia may contribute to encephalopathy. In patients with cirrhosis hyponatremia should therefore be seen as a threat that compounds and exacerbates the osmoregulatory drive for depletion of brain osmolytes driven by sequestration of ammonia. (See HEPATOLOGY 2004;39:1613–1622.)

Illustration 3.