Plasma concentration of transforming growth factor-β1 and hepatic fibrosis in dogs

Abstract

Liver fibrosis is a morphologic alteration that accompanies chronic liver diseases. Apart from analysis of liver biopsy specimens, there has been no means of diagnosing and evaluating the course of liver fibrosis in the dog. Several plasma markers, including transforming growth factor beta-1 (TGF-β1), are used to indicate liver fibrosis in humans, but none has been validated for use in dogs. There is a significant correlation between the presence and severity of hepatic fibrosis and the plasma concentration of TGF-β1 in humans with hepatic fibrosis and cirrhosis. The feasibility of using TGF-β1 as a marker for hepatic fibrosis in dogs was evaluated by comparing plasma concentrations in 29 healthy dogs and 18 dogs with liver disease. The plasma concentrations of TGF-β1, were 193 to 598 pg/mL in the healthy dogs, 143 to 475 pg/mL in the 7 dogs with mild hepatic fibrosis or none at all, and 427 to 1289 pg/mL in 11 dogs with moderate to severe hepatic fibrosis. The plasma concentrations of TGF-β1 in the dogs with moderate to severe fibrosis differed significantly (P < 0.001) from those in the other 2 groups, whereas the concentrations in the dogs with mild or no fibrosis did not differ significantly from those in the healthy dogs (P > 0.05). It was concluded that TGF-β1 is a potential plasma marker for hepatic fibrosis in dogs.

Introduction

Fibrosis is defined as an increase in extracellular collagen. The pathophysiological consequences of liver fibrosis may be a disruption of hemodynamics through constriction of the sinusoid and the terminal veins, an increase in the circulatory resistance, and consequent portal hypertension or arteriovenous shunting (1). Furthermore, disruptions of hepatocyte metabolism occur through a decreased exchange of material across the modified space of Disse and disturbances in gene expression in the hepatocytes (2). Hence, liver fibrosis represents an important pathological alteration of the diseased liver, and it is a therapeutically modifiable intermediate stage on the route to terminal liver cirrhosis.

Hepatic fibrosis cannot unequivocally be diagnosed clinically or in the laboratory. Currently, the gold standard for diagnosis is subjective assessment via biopsy, an invasive procedure that may yield nonrepresentative samples. The fibrosing process is regulated by cytokines, which are released from inflammatory cells, necrotic hepatocytes, active Kupffer cells, and Ito cells. Therefore, it is seems likely that the concentration of cytokines will change in the course of chronic liver disease.

One cytokine that is assumed to play an important role in the genesis of fibrosis is transforming growth factor beta-1 (TGF-β1). In blood, TGF-β1 is present as an inactive protein that can be activated by, among others, plasmin, thrombospondin, integrins, and oxygen radicals (3,4). Activated TGF-β1 has endocrine and paracrine functions, binding to several receptors on the target cells and acting via the “small mothers against decapentaplegic” (SMAD) and “mitogen-activated protein kinase” (MAPK) pathways on the cell nucleus (5). Generally, the biologic functions of TGF-β1 include termination of cell proliferation, tumor suppression, and immunosuppression; in the liver, TGF-β1 leads to hepatocyte apoptosis, fibrogenesis, and inhibition of liver regeneration (6).

The aim of the present study was to validate for dogs, a TGF-β1 assay used for humans, and to determine if the plasma concentration is related to hepatic fibrosis in dogs.

Materials and methods

Animals

The study included 18 dogs with liver diseases, patients of the Small Animal Clinic of Göttingen University. The diagnosis of liver disease was done by histologic evaluation of liver biopsy specimens. The liver diseases were classified according to standards of the liver standardization group of the World Small Animal Veterinary Association (7) into acute hepatitis, chronic hepatitis, or cirrhosis. A control group consisted of 29 dogs determined to be healthy by physical examination, hematologic and clinical chemistry studies, and ultrasonography of the abdominal organs.

Clinical chemistry and TGF-β1 studies

Blood samples (5 mL) were collected from the cephalic vein with minimal trauma and a wide-gauge (19-gauge) needle to minimize platelet activation. The first 2 mL of blood was allowed to clot and was used to determine levels of alanine and aspartate aminotransferase, glutamate dehdrogenase, albumin, bile acids, urea, creatinine, phosphorus, calcium, amylase, and lipase; all were measured by recognized techniques with the use of commercial test kits and an automated analyzer (Randox Daytona; Randox Laboratories, Crumlin, Northern Ireland). The remaining 3 mL was placed in a tube containing ethylenediamine tetraacetic acid (EDTA), cooled, and centrifuged (for 15 min at 1000 × g) within 30 min. To eliminate thrombocytes completely, the plasma was centrifuged again (for 10 min at 10 000 × g) at 2–6°C. It was then stored at −80°C until assayed for TGF-β1.

The Quantikine Human TGF-β1 kit (catalog number DB100B; R&D Systems, Minneapolis, Minnesota, USA) was used, because the amino acid sequence of TGF-β1 in dogs is 100% identical to that of human TGF-β1 (8). Latent TGF-β1 was activated and quantified using the Quantikine ELISA immunoassay.

Liver biopsy

The diagnosis of each liver disease and the estimation of degree of fibrosis were carried out by means of liver biopsy. Visually abnormal tissue was sampled with the use of ultrasound (16 dogs) or laparoscopy (2 dogs) and Trucut biopsy needles (Surgivet, Waukesha, Wisconsin, USA); in each case, 2 or 3 samples about ~1 cm long and 2 mm thick were taken. The tissue was fixed in 10% formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin and elastica van Gieson stain. The degree of fibrosis was estimated semiquantitatively and classified with reference to the Scheuer fibrosis score (9): 0 — no fibrosis; 1 — fibrotic portal tracts; 2 — periportal or portal septa but intact architecture; 3 — fibrosis with architectural distortion but no obvious cirrhosis; 4 — probable or definite cirrhosis.

Statistical analysis

For statistical procedures, the SAS System program, version 8.1 (SAS Institute, Cary, North Carolina, USA) was used. First, the patients were assigned to groups: group 1, hepatopathy with moderate to severe fibrosis (scores 2 to 4); group 2, hepatopathy with mild fibrosis (score 1), or none at all; and group 3, healthy. For each group, the distribution of TGF-β1 plasma concentrations was checked with the Kolmogorov–Smirnow test. In addition, the concentrations for the 3 groups were compared by means of the Mann-Whitney U-test. The significance level was taken to be P < 0.05. To check intra- and interspecific precision, the coefficient of variation was calculated for samples from 6 dogs.

Results

Table I shows the TGF-β1 plasma concentrations for the individual dogs with liver disease. Group 1, 11 dogs of various breeds with moderate to severe fibrosis or even cirrhosis, were given histological diagnoses of chronic hepatitis (9 dogs) or liver cirrhosis (2 dogs). Their plasma TGF-β1 concentrations ranged from 427 to 1289 pg/mL, with a median of 770 pg/mL (Figure 1). The diagnoses in group 2, 7 dogs of various breeds, were acute hepatitis in 1 dog and chronic hepatitis in the other 6 dogs. The plasma TGF-β1 concentrations ranged from 143 to 475 pg/mL, with a median of 380 pg/mL. The plasma TGF-β1 concentrations in group 3, 29 healthy dogs of various breeds, ranged from 193 to 598 pg/mL, with a median of 352 pg/mL.

Table I.

Breed, diagnosis, degree of fibrosis, and TGF-β1 plasma concentration of transforming growth factor beta-1 (TGF-β1) in 2 groups of dogs with liver disease

Dog number	Breed	Diagnosis	Degree of fibrosis	TGF-β1 level (pg/mL)
1	Doberman pinscher	Chronic hepatitis	Moderate/severe	556
2	German shepherd	Chronic hepatitis	Moderate/severe	953
3	Dachshund	Chronic hepatitis	Moderate/severe	836
4	Doberman pinscher	Chronic hepatitis	Moderate/severe	735
5	Pointer	Chronic hepatitis	Moderate/severe	1289
6	Mixed breed	Chronic hepatitis	Moderate/severe	843
7	Mixed breed	Chronic hepatitis	Moderate/severe	770
8	Pointer	Chronic hepatitis	Moderate/severe	594
9	Airedale terrier	Chronic hepatitis	Moderate/severe	479
10	Mixed breed	Chronic hepatitis with cirrhosis	Moderate/severe	1192
11	Cocker spaniel	Cirrhosis	Moderate/severe	427
12	Dachshund	Acute hepatitis	None/mild	143
13	Dachshund	Chronic hepatitis	None/mild	426
14	Mixed breed	Chronic hepatitis	None/mild	380
15	Mixed breed	Chronic hepatitis	None/mild	409
16	Mixed breed	Chronic hepatitis	None/mild	209
17	Rottweiler	Chronic hepatitis	None/mild	354
18	Mixed breed	Chronic hepatitis	None/mild	475

Figure 1.

Box-Whisker-Plot with minimum, median, maximum, and quartile 1 and 3 concentrations of transforming growth factor beta-1 (TGF-β1) in dogs with moderate or severe hepatic fibrosis (group 1), dogs with no or mild hepatic fibrosis or none at all (group 2), and healthy dogs (group 3). The boxes indicate 50% of the data.

The data for the dogs in all 3 groups were not normally distributed. The Mann-Whitney U-test showed a significant difference in plasma TGF-β1 concentrations between the dogs in group 1 and those in group 2 (P < 0.001), and between the dogs in group 1 and those in group 3 (P < 0.001). The difference in concentrations between groups 2 and 3 was not significant (P > 0.05). The intra- and interspecific precision was 3.1% and 5.5%, respectively.

Discussion

Hepatic fibrosis is the increased deposition of extracellular collagen as a result of chronic liver damage (10). Fibrotic changes begin in the pericentral region. As the disorder progresses, deposition appears in the space of Disse and, finally, periportally (11,12). This histological transformation results from an increased expression of extracellular matrix molecules through activated Ito cells in the perisinusoidal space (13,14). The activated Ito cells metamorphose to myofibroblasts and fibroblasts (15,16). The activation of resting Ito cells is the result of a cytokine-mediated interaction between sessile liver cells (hepatocytes, Kupffer cells, and endothelial cells) and infiltrated inflammatory cells (17). Numerous cytokines, including TGF-β1, connective tissue growth factor, platelet-derived growth factor, tumour necrosis factor, and endothelin-1 (18), participate in the process of activation. In the hierarchy of fibrogenic cytokines, TGF-β1 probably plays a decisive role (19).

At present, the diagnosis of liver fibrosis is possible only through biopsy (11). Consequences of this invasive investigative technique include those of anesthesia and poor representativeness as a result of the small size of the tissue sample. Hence, the development of plasma markers for the evaluation of fibrosis is desirable. In humans, several parameters can now provide evidence of fibrotic changes in the liver. These include hyaluronic acid, and laminin, C-terminal ends of types I, III, and IV collagen (11). For the dog, however, these parameters have not been validated.

Given the importance of TGF-β1 in the pathogenesis of liver fibrosis, the evidence that it is a marker of fibrosis in humans appears to be of diagnostic and prognostic significance. Some studies have investigated the correlation between TGF-β plasma concentration and degree of liver fibrosis; in all cases a significant correlation was found (20–22).

In the present study, the results in humans were confirmed for dogs. In the dogs with marked liver fibrosis, the median plasma concentration of TGF-β1 was more than twice that of the healthy dogs and the dogs with liver disease but either mild fibrosis or no fibrosis at all. This difference was statistically significant (P < 0.001). There was not, however, a statistically significant difference between the healthy dogs and those with only mild liver fibrosis. This shows the limitations of TGF-β1 measurement in diagnosing the onset of liver fibrosis in the dog. Because of the small sample sizes, patients without fibrosis and those that had mild fibrosis were grouped, and those with moderate and severe fibrosis were grouped. Further studies should be carried out with greater numbers of patients, to confirm the results that are observed herein.

On the basis of the present results, the inclusion of a TGF-β1 assay in the diagnosis and management of dogs with liver fibrosis appears to be both possible and potentially significant. In humans, no effect of age or sex on plasma TGF-β1 concentration has been detected (23), although degranulation of thrombocytes does have an effect. Since thrombocytes accumulate TGF-β in their cytoplasm, an increased concentration of TGF-β1 can be expected when blood sample collection involves degranulation of thrombocytes. This was taken into account in the present study by extracting and preparing the plasma by a method resulting in minimal thrombocyte damage (24).

It can be concluded from this study that TGF-β1 is quantifiable in the plasma of dogs, that hepatic fibrosis is associated with elevated plasma concentrations of TGF-β1, and that this association is especially great in dogs with moderate to severe fibrosis. The measurement of TGF-β1 cannot replace the histologic investigation of a biopsy specimen, because the former is unable to differentiate among the types of underlying hepatic disorders that lead to fibrosis, but it can help in estimating the degree of fibrosis and in observing the course of the disease. With the measurement of TGF-β1, a new tool in the evaluation of hepatopathies in the dog is at our disposal. Further studies should be focussed on the possibilities of diagnosing different types of liver disease or assessing the clinical course by measuring the plasma TGF-β1 concentration.