Increased KIT signalling with up-regulation of cyclin D correlates to accelerated proliferation and shorter disease-free survival in gastrointestinal stromal tumours (GISTs) with KIT exon 11 deletions.

Gastrointestinal stromal tumours (GISTs) with deletions in KIT exon 11 are characterized by higher proliferation rates and shorter disease-free survival times, compared to GISTs with KIT exon 11 point mutations. Up-regulation of cyclin D is a crucial event for entry into the G1 phase of the cell cycle, and links mitogenic signalling to cell proliferation. Signalling from activated KIT to cyclin D is directed through the RAS/RAF/ERK, PI3K/AKT/mTOR/EIF4E, and JAK/STATs cascades. ERK and STATs initiate mRNA transcription of cyclin D, whereas EIF4E activation leads to increased translation efficiency and reduced degradation of cyclin D protein. The aim of the current study was to analyse the mRNA and protein expression as well as protein phosphorylation of central hubs of these signalling cascades in primary GISTs, to evaluate whether tumours with KIT exon 11 deletions and point mutations differently utilize these pathways. GISTs with KIT exon 11 deletions had significantly higher mitotic counts, higher proliferation rates, and shorter disease-free survival times. In line with this, they had significantly higher expression of cyclin D on the mRNA and protein level. Furthermore, there was a significantly higher amount of phosphorylated ERK1/2, and a higher protein amount of STAT3, mTOR, and EIF4E. PI3K and phosphorylated AKT were also up-regulated, but this was not significant. Ultimately, GISTs with KIT exon 11 deletions had significantly higher phosphorylation of the central negative cell-cycle regulator RB. Phosphorylation of RB is accomplished by activated cyclin D/CDK4/6 complex, and marks a central event in the release of the cell cycle. Altogether, these observations suggest increased KIT signalling with up-regulation of cyclin D as the basis for the unfavourable clinical course in GISTs with KIT exon 11 deletions.

An oncogenetic tree model in gastrointestinal stromal tumours (GISTs) identifies different pathways of cytogenetic evolution with prognostic implications.

To model the cytogenetic evolution in gastrointestinal stromal tumour (GIST), an oncogenetic tree model was reconstructed using comparative genomic hybridization data from 203 primary GISTs (116 gastric and 87 intestinal GISTs, including 151 newly analysed cases), with follow-up available in 173 cases (mean 40 months; maximum 133 months). The oncogenetic tree model identified three major cytogenetic pathways: one initiated by -14q, one by -1p, and another by -22q. The -14q pathway mainly characterized gastric tumours with predominantly stable karyotypes and more favourable clinical course. On the other hand, the -1p pathway was more characteristic of intestinal GISTs, with an increased capacity for cytogenetic complexity and more aggressive clinical course. Loss of 22q, more closely associated with -1p than -14q, appeared to initiate the critical transition to an unfavourable cytogenetic subpathway. This -22q pathway included accumulation of +8q, -9p, and -9q, which could all predict disease-free survival in addition to tumour site. Thus, insights into the cytogenetic evolution obtained from oncogenetic tree models may eventually help to gain a better understanding of the heterogeneous site-dependent biological behaviour of GISTs.

Nonsurgical treatment of the primary tumor in four consecutive cases of metastasized colorectal carcinoma.

BACKGROUND AND STUDY AIMS: Surgical resection of the primary tumor is standard treatment in stage IV colorectal cancer, but palliative surgery is associated with high morbidity and mortality and with uncertain benefit. The wisdom of surgical resection of asymptomatic or oligosymptomatic primary tumors is therefore questionable. By studying a small series of such patients, we aimed to assess whether endoscopic techniques can offer an effective alternative form of nonsurgical palliative treatment for the prevention of local complications caused by a primary colorectal tumor. PATIENTS AND METHODS: We treated four consecutive patients who had stage IV colorectal cancer by endoscopic tumor debulking, either using a standard polypectomy snare technique alone or by argon plasma coagulation ablation followed by snare debulking of the primary tumor. RESULTS: Palliation was achieved in all patients, demonstrated by regression of the primary tumor and absence of symptoms related to the colonic tumor during the observation period of up to 24 months. No procedure-associated complications were observed and it was possible to commence systemic chemotherapy immediately after the endoscopic treatment in all four patients. CONCLUSIONS: We believe that surgical resection of the primary tumor is not appropriate in all patients with stage IV colorectal cancer, and that this form of treatment should be reserved for patients with signs of complete obstruction in whom local ablative procedures are not possible. Simple endoscopic techniques for treatment of the primary tumor, in conjunction with systemic chemotherapy, may be the most suitable form of management for patients with stage IV colorectal tumors.

Gastrointestinal stromal tumor (GIST) -- single center experience of prolonged treatment with imatinib.

BACKGROUND: The tyrosine kinase inhibitor imatinib has been introduced into the treatment of gastrointestinal stromal tumors (GIST). Here we report our results of prolonged treatment in comparison to a similar group of GIST patients who had died before imatinib became available. METHODS: Fourteen patients with recurrent or metastatic GIST were treated with imatinib. Clinical data and tumor samples of ten patients from the pre-imatinib era were available for comparison. Comparative genomic hybridisation (CGH) was performed on tumors to identify changes that may predict response to treatment. RESULTS: Fourteen patients were treated, mean treatment time 22.3 months (1 non-response, 2 progression after initial response, 2 stable diseases, 8 partial responses, 1 complete response). Adverse side effects were mild in general. Survival was higher in the treated group (41.1 months vs. 34.8 months in the historical group). Eleven treated patients are alive. CGH analysis showed comparable numbers of chromosomal aberations in both groups. CONCLUSION: Prolonged treatment with imatinib is safe and effective in patients with recurrent or metastatic GIST.

Ultrasound-guided laser interstitial thermo therapy for treatment of non-resectable primary and secondary liver tumours--a feasibility study.

AIM: Therapeutic options for primary and secondary liver tumours not suitable for resection or transplantation are limited. In this palliative situation, the scope of ablative therapeutic procedures has improved. Laser interstitial thermotherapy is a minimal invasive procedure for local tumour destruction within solid organs. This pilot study reports initial clinical experience using ultrasound-guided percutaneous laser interstitial thermotherapy. METHODS: Sixty patients between the ages of 34 and 78 years with non-resectable primary and secondary liver tumours were treated palliatively with Nd:YAG laser interstitial thermotherapy. High resolution abdominal ultrasound with power duplex was used to control the placement and coagulation procedure. RESULTS: In all cases, sonographic imaging allowed exact placement of the laser probe and verification of thermocoagulation by a resulting hyperechogenic signal enhancement. The maximum diameter of laser-induced destruction measured 5 cm. Ultrasound with power duplex and echo enhancer, CT or MRI scans indicated necrosis of treated tumour lesions. No serious adverse event occurred and 30-day-mortality was zero. CONCLUSIONS: Ultrasound-guided laser interstitial thermotherapy is safe and reliably ablates primary and secondary liver tumours. The combination of high resolution ultrasound and laser therapy facilitates a minimally invasive but elaborate treatment. Besides chemotherapy, this procedure could be a useful palliative treatment to control the mass of liver tumours unsuitable for resection or transplantation.

Cytokines in the liver.

Cytokines comprise a group of small proteins released from cells in order to influence the function of other cells. By binding to highly specific cell-surface receptors, they trigger a vast array of intracellular signalling cascades. Cytokines have been described as interleukins, growth factors, interferons and chemokines. Unlike hormones, which act in a similar way, cytokines are produced by many different types of cell and act on many other types. Most of them are produced only after certain stimuli. The most intense field of cytokine activity is without doubt host defence. The liver resembles a central organ of cytokine activity due to the fact that it hosts hepatocytes, which are highly susceptible to the activity of cytokines in a variety of physiological and pathophysiological processes. Moreover, the non-parenchymal cells of the liver, in particular Kupffer cells (KCs), the resident tissue macrophages of the liver, are able to synthesize a variety of cytokines that may act systemically on any other organ of the body, or in a paracrine manner on hepatocytes and other non-parenchymal liver cells. A classic example of how cytokines act can be observed during the acute phase reaction discussed in this article. The role of cytokines in liver development, acute liver injury, liver regeneration, liver fibrosis and liver metastasis is also discussed.

C1Q synthesis by tissue mononuclear phagocytes from normal and from damaged rat liver: up-regulation by dexamethasone, down-regulation by interferon gamma, and lipopolysaccharide.

The subcomponent of complement C1, C1q, mediates complement activation via the classical pathway, and therefore may play an important role in the inflammatory processes in which complement activation is involved. The aim of our study was to investigate C1q synthesis by macrophages of normal and of acutely damaged livers. The localization of C1q in liver tissue was studied by immunohistochemistry. Rat liver tissue macrophages were isolated from normal as well as from acutely damaged (carbon tetrachloride model) liver, and were separated into small, monocyte-like phagocytes and large, mature tissue macrophages, as revealed by immunocytochemistry. C1q gene expression was studied by endogeneous labeling of newly synthesized proteins, immunoprecipitation, and sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and by reverse-transcription polymerase chain reaction (RT-PCR) of C1qB messenger RNA (mRNA). Semiquantitative analysis was performed by Northern blotting of total RNA and hybridization with the radioactively labeled RT-PCR product. C1 esterase inhibitor synthesis was studied in parallel. For comparison, C1q and C1-inhibitor synthesis were also investigated in blood monocytes and peritoneal macrophages. C1q was weakly detectable in sinusoidal cells of the normal liver. C1qB mRNA, as well as constitutive synthesis and secretion of C1q, was clearly detected in freshly isolated and cultured Kupffer cells from normal rat liver. In comparison, newly recruited "inflammatory" macrophages from damaged rat liver synthesized considerably lower amounts of the protein, similar to what was found in the monocyte-like macrophages of normal liver and in peritoneal macrophages. Monocyte C1qB mRNA was not detected even by RT-PCR, and remained undetectable during the time in culture. Similar behavior was observed for C1-inhibitor synthesis. Treatment of the cultures with interferon gamma (IFN-gamma) or lipopolysaccharide (LPS) strongly decreased, whereas treatment with dexamethasone strongly increased C1q gene expression in the macrophage populations, and induced C1qB mRNA in cultured monocytes, as revealed by RT-PCR. Kupffer cells of normal liver may produce considerable amounts of C1q, whereas the inflammatory macrophages of the acutely damaged liver may not be so important for the synthesis of C1q.

Mast cells distribution in human liver disease and experimental rat liver fibrosis. Indications for mast cell participation in development of liver fibrosis.

BACKGROUND/AIMS: The development of liver fibrosis due to chronic liver diseases is thought to be mediated by inflammatory cells releasing fibrogenic mediators that activate fat-storing cells (Ito-cells). Recently, the involvement of mast cells in fibrogenesis has been suggested. We studied the distribution of these cells in normal human liver and human nonfibrotic and fibrotic liver disease as well as in normal rat liver and acutely and chronically injured rat liver (CCl4 model). METHODS: Mast cells were identified by histochemical and immunohistochemical methods. The immunoreactivity of liver and comparatively of rat peritoneal mast cells to the serpins alpha1-antitrypsin, alpha1-antichymotrypsin and antithrombin III was also studied. RESULTS: In normal human and rat liver, mast cells were rarely found in portal tracts, and there was no change in cell numbers in nonfibrotic human or acutely injured rat livers. In contrast, cirrhotic human and rat livers contained numerous mast cells in the portal tracts and the fibrous septa. They exhibited strong immunoreactivity to the serpins, as did rat peritoneal mast cells. CONCLUSIONS: The results indicate that in the late stages of liver fibrogenesis, mast cells may be involved by displaying protease inhibitory activity in the fibrotic septa.

Subcellular distribution of glucocorticoid receptor in cultured rat and human liver-derived cells and cell lines: influence of dexamethasone.

Glucocorticoid receptor (GR) distribution in isolated rat hepatocytes and nonparenchymal hepatic stellate cells, Kupffer cells, and liver fibroblasts with and without dexamethasone treatment was investigated by immunostaining and confocal laser scanning microscopy. In addition, human liver fibroblasts, Hep3B and HepG2 cells were investigated. Subcellular distribution of GR immunostaining was assessed semiquantitatively by digital image analysis. Short-term incubation of rat liver cells with dexamethasone resulted in an increase of nuclear staining. The same was true for human liver fibroblasts. In contrast, predominant nuclear staining could be observed in untreated as well as in dexamethasone-treated Hep3B and HepG2 cells. By means of reverse-transcription polymerase chain reaction, it could be shown that messenger RNA of both known human GR isoforms, hGR alpha and nonhormone-binding hGR beta, are present in human cells. Furthermore, dexamethasone binding indicates that hGR alpha protein is expressed in all human cells investigated. The data of this study show that GR is present in all cells investigated. Rat liver cells and human liver fibroblasts contain a translocating GR, suggesting that glucocorticoid action is receptor mediated in these cells. Nuclear localization of unliganded GR in Hep3B and HepG2 indicates that factors other than glucocorticoids may direct subcellular GR distribution.

Functional characterization of two different Kupffer cell populations of normal rat liver.

BACKGROUND: Kupffer cells of the liver represent the largest population of tissue macrophages. Small and large Kupffer cells were distinguished in normal liver, leading to the suggestion that they have different functions. This study intends to further characterize small and large Kupffer cells of normal rat liver in vivo and in vitro. METHODS: Sections of rat liver were investigated by double-staining immunofluorescence with the monoclonal antibodies ED1 and ED2. Isolated nonparenchymal liver cells were separated according to size to obtain small and large Kupffer cells. In culture, phagocytosis was studied by zymosan ingestion and cell proliferation by incorporation of 3H-thymidine. Synthesis of the proteins C1-inhibitor, apolipoprotein E and interleukin-1 was studied by endogenous labeling of newly synthesized proteins, immunoprecipitation and sodium dodecylsulfate-polyacrylamide gel electrophoresis. RESULTS: ED1+ ED2+ Kupffer cells were located in the liver along the sinusoids. ED1+ ED2+ cells were found mainly located around the central vein and portal vessels. By counterflow elution, small ED1+ ED2- cells were separated from larger ED1+ ED2+ cells and cultured. The larger cells abundantly synthesized C1-inhibitor and apolipoprotein E, while the small cells synthesized only trace amounts of these proteins. Interferon-gamma increased C1-inhibitor synthesis in small (5-fold) and large cells (1.5-fold). 3H-thymidine incorporation was 11-fold higher in small than in large cells. However, lipopolysaccharide-induced pro-interleukin-1 alpha and pro-interleukin-1 beta synthesis and phagocytic activity were similar in both populations. CONCLUSIONS: The data demonstrate two different populations of mononuclear phagocytes in normal rat liver well distinguished by immunocytochemical and functional markers.

Expression of extracellular matrix proteoglycans perlecan and decorin in carbon-tetrachloride-injured rat liver and in isolated liver cells.

Proteoglycans are important components of the extracellular matrix. They are involved in liver regeneration as well as in liver fibrosis. The distribution and cellular source of proteoglycans under normal as well as pathological conditions is still under debate. Localization of decorin and perlecan was studied in normal, acutely damaged, and cirrhotic liver by histochemistry. Furthermore, their synthesis was analyzed in different liver cell populations isolated from normal rat liver. In normal liver, decorin positivity was observed in the perisinusoidal space and in the portal area. Perlecan was clearly detectable in the portal area (blood vessels and bile ducts); a moderate reaction was also seen along the sinusoids. Strong positivity for both proteoglycans was detectable in the necrotic areas of acutely damaged liver. Chronic liver damage was characterized by the deposition of decorin and perlecan in the fibrotic septa. Immunocytochemical reactions were positive for perlecan and decorin in cultured Ito and endothelial cells but not in hepatocytes and Kupffer cells. Northern hybridization confirmed the capacity of Ito cells and endothelial cells to express the two genes. Interestingly, although rat skin fibroblasts expressed strong messages for both proteoglycans, rat aortic smooth muscle cells did not synthesize decorin.

Organization of photosystem I polypeptides examined by chemical cross-linking.

Photosystem I from the cyanobacterium Synechocystis sp. PCC 6803 was examined using the chemical cross-linkers glutaraldehyde and N-ethyl-1-3-[3-(dimethylamino)propyl]carbodiimide to investigate the organization of the polypeptide subunits. Thylakoid membranes and photosystem I, which was isolated by Triton X-100 fractionation, were treated with cross-linking reagents and were resolved using a Tricine/urea low-molecular-weight resolution gel system. Subunit-specific antibodies and western blotting analysis were used to identify the components of cross-linked species. These analyses identified glutaraldehyde-dependent cross-linking products composed of small amounts of PsaD and PsaC, PsaC and PsaE, and PsaE and PsaF. The novel cross-link between PsaE and PsaF was also observed following treatment with N-ethyl-1-3-[3-(dimethylamino)propyl]carbodiimide. These cross-linking results suggest a structural interaction between PsaE and PsaF and predict a transmembrane topology for PsaF.

Mononuclear phagocytes of acutely injured rat liver abundantly synthesize and secrete fibronectin in contrast to Kupffer cells of normal liver.

Fibronectin (FN) is a multifunctional glycoprotein involved in wound healing. It is early deposited after liver injury in necrotic areas. The cellular source of this FN remained undetermined. For this reason, mononuclear phagocytes (MNP) of normal and acutely CCl4-injured rat liver were isolated, characterized immunocytochemically and kept in culture. Synthesis of FN was studied by biosynthetic labelling, immunoprecipitation, and SDS-PAGE and on RNA level by Northern blotting and hybridization with 32-P-labeled, FN-specific cDNAs. In contrast to normal liver, MNP of acutely injured livers synthesized and secreted FN in abundant amounts. Synthesis was higher in small than in large MNP and higher in MNP isolated 36 h after the injury (vs 60 h) and decreased during the time in culture. MNP could be an important cellular source of FN deposited in the early stage of liver injury.

Accumulation and cellular localization of fibrinogen/fibrin during short-term and long-term rat liver injury.

BACKGROUND/AIMS: During liver fibrosis, there is a putative pacemaker role of fibronectin. Fibrinogen is closely linked to fibronectin during clotting processes. The aim of this study was to show fibrinogen gene expression during liver damage. METHODS: Fibrinogen/fibrin deposition in damaged livers was studied by immunohistology. Fibrinogen gene expression was analyzed in vivo in a model of CCl4-induced rat liver damage and in vitro in isolated liver cells by means of Northern blot analysis and in situ hybridization. RESULTS: Immunohistology showed striking amounts of fibrinogen and fibrin deposits in pericentral necrotic areas (short-term damage) and within fibrotic septa (long-term damage). Total RNA extracted from short-term-damaged livers contained an increased fibrinogen messenger RNA level. By in situ hybridization, fibrinogen transcripts were localized in cells of the nonnecrotic areas (short-term damage) and outside fibrotic septa (long-term damage). In vitro studies showed fibrinogen de novo synthesis restricted to hepatocytes. CONCLUSIONS: The results show fibrinogen/fibrin deposition during short-term liver injury and liver fibrogenesis, which may suggest the involvement of a "clotting-like process" in short-term liver damage and liver fibrosis. The data might indicate that fibrin/fibronectin constitute a "provisional matrix," which affects the attraction and proliferation of inflammatory and matrix-producing cells.

Viable rat Kupffer cells synthesize but do not secrete interleukin-1: indications for necrosis-induced maturation of interleukin-1 alpha, but not of interleukin-1 beta.

Interleukin-1 is involved in host defense to infection and injury. In this work synthesis and secretion of IL-1 by cultured rat liver macrophages (Kupffer cells) in response to lipopolysaccharide were investigated. IL-1 was found only intracellularly as 31-kD (pro-IL-1 alpha) and 34-kD (pro-IL-1 beta) proteins. No mature 17-kD IL-1 alpha or IL-1 beta was detected in cell lysates or supernatants. Pulse-chase experiments showed that there was no release of IL-1 even after 24h, although pro-IL-1 continuously disappeared from the cells. Cultures were lysed by freezing-thawing. Pro-IL-1 beta was then found in the supernatants, but pro-IL-1 alpha was processed into several smaller fragments. A 17-kD protein could represent the mature IL-1 alpha. The results indicate that Kupffer cells are not able to secrete biologically active mature IL-1 proteins. By lysis of the cells. IL-1 is released. Pro-IL-1 alpha is processed, but pro-IL-1 beta seems to need further activation processes.

Expression of von Willebrand factor in normal and diseased rat livers and in cultivated liver cells.

Von Willebrand factor (vWf) is an adhesive glycoprotein known to play an important role in hemostasis and in tissue injury. Because the latter process resembles hepatic fibrogenesis, we studied the tissue distribution of vWf in diseased livers. In normal rat liver vWf was strongly expressed in the vascular endothelium and as small spots or fiber-like structures in the hepatic parenchyma. During acute liver injury, pronounced staining was observed within the area of necrosis. In fibrotic livers vWf deposits were distributed predominantly at the scar-parenchyma interface but also within the septum and in sinusoidal lining cells. Testing different liver cell populations in vitro demonstrated that vWf gene expression was limited to endothelial cells (ECs) and, therefore, the latter cell population might represent the vWf-positive cells detected in vivo. The distribution of vWf within fibrotic septa suggests that vWf becomes a component of the extracellular matrix (ECM) in fibrotic livers.

Organization of Photosystem I Polypeptides (A Structural Interaction between the PsaD and PsaL Subunits).

The wild-type, PsaD-less, and PsaL-less strains of the cyanobacterium Synechocystis sp. PCC 6803 were used to study subunit interactions in photosystem I (PSI). When the membranes of a PsaD-less strain were solubilized with Triton X-100 and PSI was purified using ion-exchange chromatography and sucrose-gradient ultracentrifugation, the PsaL subunit was substantially removed from the core of PSI, whereas other subunits, such as PsaE and PsaF, were quantitatively retained during purification. When the wild-type PSI was exposed to increasing concentrations of NaI, the PsaE, PsaD, and PsaC subunits were gradually removed, whereas PsaF, PsaL, PsaK, and PsaJ resisted removal by up to 3 M NaI. The absence of PsaL enhanced the accessibility of PsaD to removal by NaI. Treatment of the wild-type PSI complexes with glutaraldehyde at 4[deg] C resulted in a 29-kD cross-linked product between PsaD and PsaL. The formation of such cross-linked species was independent of PSI concentrations, suggesting an intracomplex cross-linking between PsaD and PsaL. Taken together, these results demonstrate a structural interaction between PsaD and PsaL that plays a role in their association with the PSI core.

Syndecan-1 gene expression in isolated rat liver cells (hepatocytes, Kupffer cells, endothelial and Ito cells).

Normal liver contains mainly heparan sulfate proteoglycans. To determine which liver cells are able to express syndecan-1 RNA, hepatocytes, endothelial, Kupffer and Ito cells were isolated from normal rat liver and kept in culture. Rat fibroblasts, monocytes and peritoneal macrophages were also studied. Immediately after isolation the steady state level of syndecan messages in hepatocytes was comparable to that of the normal liver. The expression increased at day one and then remained constant until day five. Ito and endothelial cells contained a low amount of syndecan message. Freshly isolated Kupffer cells failed to express syndecan, but strong upregulation was found on the first day in culture which gradually decreased by day 5. Syndecan transcripts were dose and time dependently upregulated by endotoxin and gamma interferon in Kupffer cells. Endotoxin had no effect on fibroblasts.

Interferon-alpha 2a increases serum concentration of hyaluronic acid and type III procollagen aminoterminal propeptide in patients with chronic hepatitis B virus infection.

Interferon-alpha (IFN-alpha) has become an important drug for the treatment of chronic viral liver diseases. However, the action of IFN-alpha remains unclear. We investigated whether human recombinant IFN-alpha modulates serum concentrations of hyaluronic acid (HA) and type III procollagen aminoterminal propeptide (P-III-NP) in 56 patients with chronic hepatitis-B under IFN-alpha therapy. IFN-alpha increased the HA serum level in 44 of 46 patients and, after cessation of treatment, HA serum levels returned to the pretherapy levels. The increase of HA serum level was higher in patients with active cirrhosis (aC) than in patients with chronic persistent hepatitis (CPH) and in patients with severe inflammation compared to those with moderate inflammation. Interestingly, HA serum concentration was unrelated to IFN dose and was of no predictive value for therapy response. In contrast, IFN-alpha increased significantly the P-III-NP serum level in patients with aC only. During follow-up, P-III-NP serum level decreased late in responders in parallel to the decrease of serum level of liver enzymes, in non-responders it was without significant change. The first dose of IFN induced a significant increase in HA serum level in each of 10 patients but in none of four healthy volunteers. In contrast, P-III-NP serum concentrations were not influenced by the first IFN-alpha dose. We conclude that: (1) immunstimulation with IFN-alpha induces a rapid increase of HA serum level in patients with chronic hepatitis B but not in normal persons; (2) IFN-alpha increases P-III-NP serum level only in patients with active liver cirrhosis; (3) measurement of HA and P-III-NP serum levels does not help predict response to IFN-alpha, and (4) HA serum level may be used as a compliance indicator.