Hepatic fibrosis and cirrhosis in the Fontan circulation: a detailed morphological study.

AIMS: To describe the histological features of the liver in patients with a Fontan circulation. METHODS: Specimens from liver biopsies carried out as part of preoperative assessment prior to extracardiac cavopulmonary conversion of an older style Fontan were examined and scored semi-quantitatively for pertinent histological features. To support the use of the scoring, biopsy specimens were also ranked by eye for severity to allow correlation with assigned scores. RESULTS: Liver biopsy specimens from 18 patients with a Fontan circulation were assessed. All specimens showed sinusoidal fibrosis. In 17 cases there was at least fibrous spur formation, with 14 showing bridging fibrosis and 2 showing frank cirrhosis. In 17 cases at least some of the dense or sinusoidal fibrosis was orcein positive, although a larger proportion of the dense fibrous bands were orcein positive compared with the sinusoidal component. All specimens showed marked sinusoidal dilatation, and 14 showed bile ductular proliferation; 1 showed minimal iron deposition, and 1 showed mild lobular lymphocytic inflammation. There was no cholestasis or evidence of hepatocellular damage. Similar appearances were observed in 2 patients with severe tricuspid regurgitation. DISCUSSION: The histological features of the liver in patients with a Fontan circulation are similar to those described in cardiac sclerosis. Sinusoidal dilatation and sinusoidal fibrosis are marked in the Fontan series. The presence of a significant amount of orcein negative sinusoidal fibrosis suggests there may be a remediable component, although the dense fibrous bands are predominantly orcein positive, suggesting chronicity and permanence. No inflammation or hepatocellular damage is evident, suggesting that fibrosis may be mediated by a non-inflammatory mechanism.

Interleukin-10 expression and function in experimental murine liver inflammation and fibrosis.

Kupffer cells (KC) play a central role in the initiation and perpetuation of hepatic inflammation, which, if uncontrolled, can result in tissue damage, fibrosis, and cirrhosis. Interleukin-10 (IL-10) can inhibit a range of macrophage functions. We hypothesized that the transcription, synthesis, and release of IL-10 may influence the development of liver injury. Rat KC were activated in vitro with lipopolysaccharide (LPS), and expression of IL-10 mRNA compared with IL-13 and IL-1beta by reverse-transcription polymerase chain reaction (RT-PCR). The effects of pretreatment with recombinant IL-10 (rIL-10) on KC phagocytosis, production of superoxide (SO), and tumor necrosis factor (TNF-) were examined by fluorescent activated cell sorter (FACS), reduction of ferricytochrome C, and bioassay, respectively. Rats were administered intraperitoneal carbon tetrachloride (CCl4), and expression of IL-10 mRNA and protein in vivo compared with IL-13 and IL-1beta by RT-PCR and immunoblotting. Results were correlated with histological inflammatory changes. Finally, IL-10 gene-deleted (IL-10-/-) mice and wild-type (WT) controls were administered intraperitoneal CCl4 biweekly for up to 70 days, and the development of inflammation and fibrosis compared by scoring histological changes. IL-10 mRNA was up-regulated early, both in KC in vitro and in whole liver in vivo, concurrent with that of IL-1beta. IL-10 was able to inhibit KC production of both SO and TNF- in vitro, and this was achieved more effectively than IL-4 or IL-13; no such effects were seen on KC phagocytosis. After 70 days of treatment with CCl4, IL-10-/- mice showed significantly more severe fibrosis and exhibited higher hepatic TNF- levels than WT controls. These results suggest that IL-10 synthesized during the course of liver inflammation and fibrosis may modulate KC actions, and influence subsequent progression of fibrosis.

Evaluation of a transgenic mouse model for alpha-1-antitrypsin (AAT) related liver disease.

We have attempted to produce a transgenic mouse model of the neonatal liver disease associated with the human PIZ allele. Analysis of a number of transgenic mouse lines carrying either a normal human PIM gene construct or the mutant Z is reported. Using isoelectric focusing analysis of plasma from transgenic mice, we have shown that the human AAT proteins produced in mice are processed in a similar way to their counterparts in humans. By comparing the level of M and Z mRNA in liver with the levels of M and Z proteins in plasma we have inferred that, as in humans, the mutant protein tends to accumulate within the hepatocyte. Accumulation of Z protein has also been demonstrated by immunocytochemistry. Two of the M transgenic lines produce such high levels of the human protein that it, like the Z protein, accumulates as globules. Histological features of livers from 116 mice of different ages and genotypes were examined: 37 non-transgenic, 62 Z transgenic (23 low expressing and 39 high expressing) and 17 M transgenic mice, all high expressing. Cirrhosis or fibrosis was not seen in any animal and we were unable to find any evidence for neonatal liver disease. Some necrosis was seen in all genotypes and this increased significantly with age with one Z line showing significantly more frequent necrosis than any other group. This line, the highest expressing Z line, was back crossed onto 7 different genetic backgrounds but no major differences between the back crosses with respect to liver disease were observed. The mouse model we have developed is compared with other transgenic Z mouse models; none of these is representative of human neonatal liver disease. Our view is that the transgenic animals generated in these experiments may be most useful for investigating the liver manifestations that almost invariably occur in ZZ adults. Alteration of additional factors other than accumulation of Z protein, for example inactivation of the endogenous mouse genes or some environmental challenge, might produce a mouse model with more relevance to neonatal liver disease.

Hepatocyte inclusions of alpha 1-antichymotrypsin in a patient with partial deficiency of alpha 1-antichymotrypsin and chronic liver disease.

We present a case of chronic liver disease with selective and exclusive hepatocyte endoplasmic reticulum storage of alpha 1-antichymotrypsin in the form of granules, detected by specific immunohistochemistry at the light microscopy level and corresponding to material found in dilated endoplasmic reticulum of hepatocytes by electron microscopy. The patient had intermediate deficiency of alpha 1-antichymotrypsin. Thus, the hepatocyte accumulation of alpha 1-antichymotrypsin may indicate the presence of an export block resembling that of a closely-related protein, namely PiZ alpha 1-antitrypsin. It is proposed that hepatocyte storage of alpha 1-antichymotrypsin may be an expression of an inborn error of metabolism bearing the characteristics of endoplasmic reticulum storage diseases such as PiZ alpha 1-antitrypsin deficiency and hereditary hypofibrinogenaemia.

Production of human chorionic gonadotrophin by a hepatoblastoma resulting in precocious puberty.

A two year old boy presented with precocious puberty associated with hepatoblastoma. Serum concentrations of beta human chorionic gonadotrophin and alpha-fetoprotein were raised. An aggressive chemotherapeutic regimen resulted in useful palliation and interesting changes in the beta human chorionic gonadotrophin and alpha-fetoprotein concentrations. In contrast to a previous report, the ultrastructure of the tumour showed frequent Golgi apparatus but no other electron dense membrane bound vesicles.