• cirrhosis;
  • cytokeratin 7;
  • cytokeratin 19;
  • HepPar1;
  • liver disease;
  • oval cells;
  • progenitor cells;
  • stem cells;
  • submassive hepatic necrosis


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References

Abstract:  Background/Aim: Proliferative bile ductular reactions occur in a variety of liver diseases in humans. It is a matter of debate whether such reactions result from progenitor cell proliferation with biliary and hepatocytic differentiation, versus biliary metaplasia of damaged hepatocytes. We investigated bile ductular reactions in liver diseases, paying particular attention to the presence of cells with intermediate (hepatocytic/biliary) features (oval-like cells).

Methods: Five specimens each were selected of submassive hepatic necrosis and cirrhosis due to hepatitis B, hepatitis C, autoimmune hepatitis, alcohol injury, primary biliary cirrhosis and primary sclerosing cholangitis. Immunohistochemical stains were performed for biliary markers (cytokeratins [CKs] 7 and 19), as well as hepatocytic markers (HepParl and alpha-fetoprotein[AFP]) in sequential sections. The degree of staining of each cell type (biliary, hepatocytic, intermediate) was graded semiquantitatively.

Results: Hepatocytes always stained diffusely for HepParl, occasionally for CK7, and rarely for CK19. Biliary cells were always diffusely positive for CK7 and CK19, and rarely for HepParl. Intermediate cells were identified in all cases and showed widespread staining for both HepParl and CK7, and less commonly for CK19. AFP was not expressed in any cell type. The morphologic and immunohistochemical features of bile ductular reactions were similar in the different diseases.

Conclusions: Proliferating hepatic parenchymal cells with intermediate (hepatocytic/biliary) morphologic features and combined immunophenotype can be identified in a variety of acute and chronic liver diseases. The similarity of bile ductular reactions among chronic hepatitic, alcoholic and biliary diseases suggests that they result from proliferation of oval-like progenitor cells.

Proliferative bile ductular reactions are identified in various human liver diseases (1). In massive or submassive hepatic necrosis of viral or toxic origin, there is proliferation of duct-like structures, which interface between portal tracts or neighbouring lobular parenchyma and the areas of necrosis (2–5). Cells sharing characteristics of hepatocytes and biliary epithelium are present in these structures. Such cells have been termed ‘ductular hepatocytes’ (5). Depending on the plane of section, duct-like structures are often seen to be lined by cells with biliary morphology and immunophenotype at one end, more hepatocyte-like features at the other, and intermediate characteristics in between (4–7). Immunohistochemical studies of these structures with a variety of hepatocytic and biliary markers have demonstrated similarities to progenitor cell populations in animal models of hepatic regeneration and carcinogenesis, often referred to as ‘oval cells’(8–15). Moreover, we have recently provided evidence that ductular hepatocytes actually arise from the canals of Hering (16). Therefore, it is very likely that they represent an attempt for hepatic regeneration originating from progenitor cells lying within the canals.

In cirrhosis following a variety of chronic disease processes, including viral and autoimmune hepatitides, bile duct destructive diseases, such as primary biliary cirrhosis (PBC) and primary sclerosing cholangitis (PSC), and toxic injury from chronic alcohol use, a similar ductular reaction takes place (4,17–21). These ductules proliferate along the fibrous septa at the margins of regenerating nodules. They are lined by cuboidal cells and usually have inconspicuous lumens. In chronic biliary diseases, as well as in chronic hepatitis B, several investigators have demonstrated the presence of cells with similar immunophenotypic characteristics as those of oval cells and ductular hepatocytes (18–20).

Previous studies of ductular reactions have largely focused on chronic biliary diseases and therefore the proliferating ductules have been thought to arise from biliary metaplasia of hepatocytes (21–25). However, in diseases where chronic cholestasis is not a predominant feature, such as the chronic hepatitides, pronounced ductular reactions are difficult to explain on the basis of such a mechanism. Alternatively, if activation of a bipotent progenitor cell proliferation underlied these ductular reactions, similar morphologies and immunophenotypes would be expected across a range of chronic hepatitic and toxic injuries, without significant differences from endstage cholestatic diseases.

In this study, we have sought to extend the immunophenotypic observations concerning possible progenitor cells in chronic liver disease to include a variety of hepatitic processes, as well as the chronic toxic injury of alcoholic liver disease. To this end, we stained a variety of specimens for hepatocytic and biliary markers, and evaluated the cell types involved in bile ductular reactions semiquantitatively.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References

Formalin-fixed, paraffin-embedded tissues were randomly selected for study from archival explanted livers obtained on orthotopic liver transplantation at New York University Medical Center. One representative block of liver tissue was selected from each of the five cases in the following disease categories – massive hepatic necrosis (acetaminophen toxicity, n = 2, hepatitis A, n = 2, hepatitis B, n = l), and cirrhosis due to PBC, PSC, hepatitis B, hepatitis C, autoimmune hepatitis, and alcoholic liver injury. Diagnoses were made in all cases on the basis of clinical and pathologic findings. No specimens were included where more than one etiologic process was suspected.

Routine haematoxylin and eosin (H&E)-stained slides were available in all specimens. Immunohistochemical staining was performed in four serial 5-micron thick sections from each tissue block. Immunostaining was performed for biliary markers (cytokeratins [CKs] 7 and 19) and hepatocyte markers (HepParl (26) and alpha-fetoprotein [AFP]). The order of the serial sections was maintained and sequential levels were stained for CK7, HepParl, CK19, and AFP.

Immunohistochemical staining for all markers was carried out as follows. Sections were placed on poly-L-lysine coated slides. Deparaffinizing, rehydration, and blocking of endogenous peroxidase activity with absolute methanol/13% hydrogen peroxide (1.4) (Fisher Scientific, Springfield, NJ) was performed. This was followed by heat- induced epitope retrieval with 0.01 mol/L citrate buffer pH 6.0 (Sigma, St. Louis, MO) in a benchtop Napco autoclave at 15 Lbs/in2 pressure, 120 °C, for 6 min. The slides were allowed to cool, then washed three times with ‘automation buffer’ (Biomeda, Foster City, CA) and loaded on an automated immunostainer (Code-On Stainer, Fisher Scientific). After avidin-biotin blocking (Vector Laboratories, Burlingame, CA), they were treated with ‘tissue conditioner' (Biomeda) according to the manufacturer's instructions, and then incubated with the appropriate dilution of primary mouse monoclonal antibody, as follows: CK7 (OV-TL 12130, DAKO, Carpinteria, CA), HepParl (kind gift of M. Nalesnik, MD, Department of Pathology, University of Pittsburgh Medical Center, Pittsburgh, PA), CK19 (RCK108, DAKO), and AFP (C3, Biogenex, Burlingame, CA).

Incubations lasted 30 min at 37 °C. The slides were then washed three times with phosphate buffered saline, and incubated with biotinylated horse antimouse immunoglobulin G (Vector Laboratories) 1 : 100 for 15 min at 37 °C. After three more buffer washes, there was a 15-minute incubation with LAB probe (Biomeda) at 37 °C. 3-aminoethylcarbazole (Biomeda) was the chromogen used, followed by a light haematoxylin counterstain (Biomeda).

Immunopositivity was assessed semiquantitatively in three parenchymal cell types, which were defined as follows on the basis of morphologic features for the purposes of this study: (i) Biliary-type cells: small oval to cuboidal cells with basophilic, round to oval nuclei and nuclear : cytoplasmic ratio greater than 1 : 2, which proliferated at the margins of the fibrous tissue in portal tracts and fibrous septa. (ii) Hepatocytes: polygonal cells with a large nucleus, usually in central position, often with prominent nucleoli and otherwise dispersed chromatin, and a nuclear : cytoplasmic ratio less than 1 : 4. (iii) Intermediate cells: cells with a range of features intermediate between the above two cell types.

Semiquantitation was carried out on a 0 to 4 + scale, as previously described (27). Briefly, the scale was defined as follows: 0 = negative; 1 + = positive cells found after careful search under medium power (4x), and confirmed at higher power; 2 + = positive cells easily identified at medium power (4x); 3 + = positive cells easily identified at low power (2x); 4 + = diffusely present positive cells seen under low power (2x). A 10x eyepiece was used.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References

On examination of H&E-stained sections, cells with intermediate features were seen, in addition to hepatocytes and biliary-type cells, in the ductular reactions of submassive hepatic necrosis (Fig. 1A) and cirrhosis of all etiologies (Fig. 2). The immunohistochemical features of the three cell types were as follows.


Figure 1. Bile ductular reactions in submassive hepatic necrosis. A: H&E stain;B: CK7 immunostain;C: CK19 immunostain;D: HepParl immunostain. Bile ductular cells (arrows) and intermediate cells (arrowheads) are identified by morphologic and immunohistochemical features. (A 200X, B-D-100X)

Download figure to PowerPoint


Figure 2. Bile ductular reactions in cirrhosis (H&E stains). Etiology of cirrhosis: A: primary biliary cirrhosis; B: chronic hepatitis B; C: autoimmune hepatitis; D: alcoholic liver disease. Evidently, bile ductular reactions are similar in cirrhosis of different etiologies. Arrows: bile ductular cells; arrowheads: intermediate cells. (A-D-400X)

Download figure to PowerPoint

Biliary-type cells: In all cases of submassive hepatic necrosis and cirrhosis, biliary-type cells showed strong cytoplasmic staining for both CK7 and CK19 (Figs 1B, C and 3A, C, E). Maintenance of intense immunostaining for CK19 was found to be dependent on careful re-titration of the antibody when batches of slides were stained at different times. The staining for CK7 was found to be more stable over different batches. Nonetheless, controlling for variations in anti-CK19 antibody sensitivity over time, CK7 was often positive in more biliary-type cells than CK19. The hepatocyte marker HepParl was rarely, but distinctly positive in biliary-type cells in three of five cases of submassive hepatic necrosis, and in most cases of cirrhosis of all etiologies (all cases of chronic hepatitis and alcohol injury, as well as three of five cases each of PSC and PBC). AFP was never positive in biliary-type cells in any specimen. The findings of the semiquantitative evaluation of immunohistochemical stains are presented by disease type in Tables 1–7. Evidently, similar grades of immunopositivity were present in biliary-type cells of all different disease entities.


Figure 3. Bile ductular reactions in cirrhosis (immunohistochemical stains).A: CK19 stain in primary biliary cirrhosis;B: HepParl stain in primary sclerosing cholangitis;C: CK7 stain in chronic hepatitis C;D: HepParl stain in chronic hepatitis B;E: CK7 stain in alcoholic liver disease;F: HepParl stain in alcoholic liver disease. Arrows: bile ductular cells; arrowheads: intermediate cells. (A–F 100X)

Download figure to PowerPoint

Table 1.  Semiquantitative immunohistochemical findings in ductular reactions of submassive hepatic necrosis by cell type
Biliary-type cellsIntermediate cellsHepatocytes
CK19 4 +1 + to 2 +0 to 1 +
CK7 4 +1 + to 3 +0 to 2 +
HepPar1 1 +2 + to 4 +4 +
Table 2.  Semiquantitative immunohistochemical findings in ductular reactions of primary biliary cirrhosis, stage IV, by cell type
Biliary-type cellsIntermediate cellsHepatocytes
CK19 2 + to 4 +0 to 1 +0
CK7 3 + to 4 +1 + to 3 +0 to 2 +
HepPar1 0 to 1 +1 + to 2 +4 +
Table 3.  Semiquantitative immunohistochemical findings in ductular reactions of cirrhosis due to primary sclerosing cholangitis by cell type
Biliary-type cellsIntermediate cellsHepatocytes
CK19 3 + to 4 +0 to 1 +0
CK7 3 + to 4 +2 + to 3 +0 to 1 +
HepPar1 0 to 1 +1 + to 2 +3 + to 4 +
Table 4.  Semiquantitative immunohistochemical findings in ductular reactions of cirrhosis due to chronic hepatitis B by cell type
Biliary-type cellsIntermediate cellsHepatocytes
CK19 3 + to 4 +0 to 3 +0
CK7 4 +1 + to 3 +0 to 2 +
HepPar1 0 to 1 +1 + to 3 +4 +
Table 5.  Semiquantitative immunohistochemical findings in ductular reactions of cirrhosis due to chronic hepatitis C by cell type
Biliary-type cellsIntermediate cellsHepatocytes
CK19 3 + to 4 +0 to 1 +0
CK7 4 +1 + to 3 +0 to 2 +
HepPar1 0 to 3 +1 + to 3 +4 +
Table 6.  Semiquantitative immunohistochemical findings in ductular reactions of cirrhosis due to autoimmune hepatitis by cell type
Biliary-type cellsIntermediate cellsHepatocytes
CK19 2 + to 4 +0 to 1 +0
CK7 4 +2 + to 3 +0 to 2 +
HepPar1 0 to 1 +2 + to 3 +4 +
Table 7.  Semiquantitative immunohistochemical findings in ductular reactions of cirrhosis due to alcoholic liver disease by cell type
Biliary-type cellsIntermediate cellsHepatocytes
CK19 3 + to 4 +0 to 1 +0
CK7 4 +1 + to 3 +0 to 2 +
HepPar1 0 to 1 +1 + to 3 +4 +

Hepatocytes: Hepatocytes rarely stained for CK19 in submassive hepatic necrosis, and only in four of five cases (Fig. 1C). Hepatocytes were never found to be CK19 positive in cirrhosis of any etiology. CK7 staining was identified in a minority of hepatocytes in all cases of submassive hepatic necrosis (Fig. 1B), and in most cases of cirrhosis (all cases of chronic biliary tract disease and alcohol injury, as well as four of five cases each of chronic hepatitis B, chronic hepatitis C and autoimmune hepatitis) (Figs 3C, E). CK7 positive hepatocytes often formed large clusters. In both submassive hepatic necrosis and cirrhosis, CK7 staining of hepatocytes was usually faint in the cytoplasm, and stronger along the cell periphery or cell membrane. HepParl was diffusely positive in hepatocytes in nearly every specimen, one case of PSC only excepted, which showed focal absence of hepatocyte positivity (Figs 1D and 3B, D, F). AFP was never identified in hepatocytes in any specimen. As shown in Tables 1–7, the staining grade of hepatocytes with the antibodies utilized was similar in all diseases.

Intermediate cells: Both CK7 and CK19 were more commonly present in intermediate cells than in hepatocytes and less so than in biliary-type cells (Figs 1B, C and 3A, C, E). This was true both for cases of submassive hepatic necrosis and cirrhosis. CK7 was more readily identified in intermediate cells compared to CK19. The CK cytoplasmic staining in intermediate cells showed a range of intensities and distributions from the biliary to the hepatocytic type. Similarly, staining of intermediate cells with HepParl varied in intensity and distribution from that of hepatocytes to the one of biliary-type cells (Figs 1D and 3B, D, F). Moreover, comparison between sequential levels often demonstrated that individual intermediate cells were positive for both biliary and hepatocyte markers (CK19/HepParl and/or CK7/HepParl) in both submassive hepatic necrosis and cirrhosis of all types. AFP was not identified in intermediate cells in any specimen. As shown in Tables 1–7, the staining grade of intermediate cells with the antibodies utilized was similar in the different diseases.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References

Ductular hepatocytes have been considered to represent evidence for the existence of a facultative progenitor cell (stem cell) compartment in human liver, similar to those seen in animal models. These cells form duct-like structures in massive or submassive hepatic necrosis, demonstrating biliary morphology at one end, hepatocytic at the other, and intermediate in between. However, the origin of these cells has been debated. Those investigators who have favored the existence of stem cells in humans, suggest that both hepatocytes and biliary cells are derived from proliferating progenitor cells (2–7). On the other hand, investigators who have not been convinced of the existence of stem cells, suggest that the biliary end represents metaplasia of damaged hepatocytes and is a mark of injury rather than of regeneration, per se(21–25).

Using 3-dimensional reconstruction of CK19-stained sections of a liver with massive hepatic necrosis, we have recently demonstrated that the duct-like structures harboring ductular hepatocytes originated from the canals of Hering (16). This finding strongly suggests that the canals of Hering contain stem cells. The immunohistochemical stains reported in the present study show a pattern of expression that accords with such a concept. The biliary-type cells in our cases of submassive hepatic necrosis are positive for CK7 and CK19. In addition to these markers, other investigators have found biliary cells to be positive for chromogranin A and neural cell adhesion molecule (NCAM), and to react with antibodies to OV-6 and HEA-125 (3–5, 19, 20). We also show that cells with hepatocyte morphology are positive for HepParl and only occasionally for biliary markers; this is in line with the findings of other authors also (3, 7). Finally, and most interestingly, we demonstrate that intermediate cells are positive for markers of both biliary and hepatocytic lineage. Comparison of immunohistochemical staining in adjacent tissue sections confirms that overlapping markers are in the same cell population.

Turning to the ductular reactions of chronic cholestatic diseases, PBC and PSC, we again demonstrate an intermediate cell population as has been seen by others (4, 19, 20). Immunophenotyping again matches the morphology: cells of biliary morphology are positive for biliary markers (CK7 and CK19 in this study, additionally, HEA-125, OV-6, NCAM, chromogranin A in other studies), cells of hepatocytic morphology for HepParl, and intermediate cells for both. In our opinion, these findings suggest that a progenitor cell compartment is activated in chronic biliary tract disease and gives rise to the ductular proliferation. However, since chronic cholestasis is certainly present in these diseases, it is difficult to say how much of the ductular reaction might in fact be derived from biliary metaplasia of hepatocytes.

However, we have now demonstrated identical results in cirrhosis caused by infection with hepatitis B and C viruses, autoimmune hepatitis, and alcoholic liver injury. In all of these diseases, the proliferation of biliary-type and intermediate cells is similar. Since chronic cholestasis is a comparatively less important and only late component of the liver damage in these other forms of chronic liver disease, we may suggest that biliary metaplasia is much less likely to play a role in producing the ductular reactions. Yet we find no significant difference in the ductular reactions of these other diseases. This agrees with the earlier finding that the ductular reactions of biliary and hepatitic processes have an identical three-dimensional structure (28). Therefore, a stem cell compartment is a likely source for these reactions.

Interestingly, the primary sites of injury in these diseases are different. In cases of hepatitis, activation of the stem cell compartment primarily comes from hepatocyte damage and loss. In PBC and PSC, stem cell activation presumably is a response to bile duct damage with only a minor component of primary hepatocyte damage. In chronic alcohol toxicity, scarring arises from activation of fibrogenesis with or without significant inflammatory cell infiltrates. However, the relative proportion of biliary-type cells to intermediate cells and of intermediate cells to hepatocytes appears similar in all the diseases summarized here. Thus, it would seem that, in chronic liver disease culminating in cirrhosis, activation of a facultative bipotent stem cell population by any mechanism results in bi-directional, biliary and hepatocytic regeneration.

Of course, three-dimensional reconstruction is required to confirm that the exceedingly complex structures represented two-dimensionally as ductular reactions actually connect to the biliary tree proper. Without such studies, we cannot say for certain that there is not some component of biliary metaplasia of hepatocytes in these endstage livers. In one three-dimensional reconstruction of ductular reactions in PBC, some ductular reactions appeared to be separate from the biliary tree proper and thus were more likely to represent a metaplastic process (29). However, others seemed to extend from the biliary tree directly and thus, again, suggested a facultative stem cell proliferation.

Finally, we must explain the differences in extent of CK7 versus CK19 staining in these cases. In recent studies of the changes affecting the canals of Hering in early stages of PBC (30) and methotrexate toxicity (31), CK7 staining of hepatocytes was common and often diffuse in PBC, but not significantly noted in methotrexate toxicity at similar stages of duct proliferation and loss of canal of Hering profiles. This suggests that biliary metaplasia of hepatocytes, demonstrated by CK7 expression and sometimes rosette formation, does take place as well. In the current study, we found CK7 to always give a more diffuse pattern of staining than did CK19, with biliary-type cells, intermediate cells and even hepatocytes staining for CK7, while CK19 was restricted to intermediate and biliary-type cells. Furthermore, in many cases, CK19 did not completely stain all cells of a biliary phenotype, let alone cells of intermediate morphology. This suggests that CK19 may be a more specific marker of progenitor cells than CK7, while CK7 expression may reflect both progenitor cell origin and cholestasis-induced metaplasia of hepatocytes. Use of these antibodies in future studies should take these differences into account.

Finally, the absence of staining for alpha-fetoprotein should be noted. Examination of progenitor cell populations in the livers of mice and rats usually demonstrates significant levels of alpha-fetoprotein expression (10–15). In humans, however, alpha-fetoprotein seems to largely remain an oncofetal antigen, being strongly and widely demonstrable by immunohistochemistry only in fetal liver or malignant hepatocellular neoplasms; repeatedly ductular reactions and regenerating hepatocytes at best only show focal, mild immunoreactivity (7–11).

In conclusion, we confirm the reports of other investigators regarding ductular reactions in human livers with submassive hepatic necrosis (from toxic and viral injury) or endstage chronic biliary tract disease (PBC, PSC). Furthermore, we extend these observations to include cirrhosis following chronic hepatitis (hepatitis B and C, autoimmune hepatitis) and chronic alcoholic liver disease. Ductular reactions include proliferation of cells with a range of morphologies, from the biliary to the hepatocytic, with some cells demonstrating intermediate morphology. Immunophenotyping with hepatocytic and biliary markers mirrors the morphologic features. The similarities between these features and those seen in experimental models of hepatic regeneration suggest that ductular reactions represent proliferation and bi-directional differentiation of facultative hepatic stem cells. In massive hepatic necrosis, we have demonstrated that these proliferations are not merely representative of metaplastic changes (16). In addition, temporal studies (20) support a similar conclusion in chronic biliary tract disease. However, further studies of chronic hepatitis and alcoholic liver injury, extending these observations into three dimensions and over time, are needed to confirm that this is indeed the case in a wider array of chronic liver diseases.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  • 1
    Roskams T, Desmet V. Ductular reaction and its diagnostic significance. Semin Diagn Pathol 1998; 15: 25969.
  • 2
    Sirica A E. Ductular hepatocytes. Histol Histopathol 1995; 10: 43356.
  • 3
    Demetris A J, Seaberg E C, Wennerberg A, Ionellie J, Michalopoulos G. Ductular reaction after submassive necrosis in humans. Special emphasis on analysis of ductular hepatocytes. Am J Pathol 1996; 149: 43948.
  • 4
    Thung S N. The development of proliferating ductular structures in liver disease. An immunohistochemical study. Arch Pathol Lab Med 1990; 114: 40711.
  • 5
    Gerber M A, Thung S N, Shen S, Stromeyer F W, Ishak K. Phenotypic characterization of hepatic proliferation. Antigenic expression by proliferating epithelial cells in fetal liver, massive hepatic necrosis, and nodular transformation of the liver. Am J Pathol 1983; 110: 704.
  • 6
    Fiel M I, Antonio L B, Nalesnik M A, Thung S N, Gerber M A. Characterization of ductular hepatocytes in primary liver allograft failure. Mod Pathol 1997; 10: 34853.
  • 7
    Haque S, Haruna Y, Saito K, et al. Identification of bipotential progenitor cells in human liver regeneration. Lab Invest 1996; 75: 699705.
  • 8
    Sell S. Liver stem cells. Mod Pathol 1994; 7: 10512.
  • 9
    Park Y N, Brody R I, Sigal S HH, Thung S N, Theise N D. OV-6 positive, HepParl negative oval-like cells in human livers with hepatitis C cirrhosis or with acetaminophen toxicity. Hepatology 1995; 22: 230A.
  • 10
    Germain L, Goyette R, Marceau N. Differential cytokeratin and alphafetoprotein expression in morphologically distinct epithelial cells emerging at the early stage of rat hepatocarcinogenesis. Cancer Res 1985; 45: 67381.
  • 11
    Lemire J M, Shiwiri N, Fausto N. Oval cell proliferation and the origin of small hepatocytes in liver injury induced by d-galactosamine. Am J Pathol 1991; 139: 53552.
  • 12
    Evarts R P, Nagy P, Marsden E. In situ hybridization studies on expression of albumin and alpha-fetoprotein during the early stage of neoplastic transformation in rat liver. Cancer Res 1987; 47: 546975.
  • 13
    Evarts R P, Nakatsukasa H, Marsden E R. Cellular and molecular changes in the early stages of chemical hepatocarcinogenesis in the rat. Cancer Res 1990; 50: 343944.
  • 14
    Grisham J W. Cell types in long-term propagable cultures of rat liver. Ann NY Acad Sci 1980; 349: 12837.
  • 15
    Evarts R P, Nagy P, Marsden E. A precursor-product relationship exists between oval cells and hepatocytes in rat liver. Carcinogenesis 1987; 8: 173740.
  • 16
    Theise N D, Saxena R, Portmann B C, et al. The canals of Hering and hepatic stem cells in humans. Hepatology 1999; 30: 142533.
  • 17
    Su Q, Liu Y-F, Zhang J-F, Zhang S-X, Li D-F, Yang J-J. Expresion of insulin-like growth factor II in hepatitis B, cirrhosis, and hepatocellular carcinoma: Its relationship with hepatitis B virus antigen expression. Hepatology 1994; 20: 78899.
  • 18
    Hsia C C, Evarts R P, Nakatsukasa H, Marsden E R, Thorgeirsson S S. Occurrence of oval-type cells in hepatitis B virus-associated human hepatocarcinogenesis. Hepatology 1992; 16: 132733.
  • 19
    Crosby H A, Hubscher S, Fabris L, et al. Immunolocalization of putative human liver progenitor cells in livers from patients with end-stage primary biliary cirrhosis and sclerosing cholangitis using the monoclonal antibody OV-6. Am J Pathol 1998; 152: 7719.
  • 20
    Roskams T, De Vos R, Van Eyken P. Hepatic OV-6 expression in human liver disease and rat experiments. Evidence for hepatic progenitor cells in man. J Hepatol 1998; 29: 45563.
  • 21
    Uchida T, Peters R L. The nature and origin of proliferated bile ductules in alcoholic liver disease. Am J Clin Pathol 1983; 79: 32633.
  • 22
    Scarpelli D G. Multipotent developmental capacity of cells in the adult animal. Lab Invest 1985; 52: 3313.
  • 23
    Desmet W. Modulation of biliary epithelium. In: ReutterW, PopperH, AriasI M, eds. Modulation of Liver Cell Expression. Lancaster: MTP Press 1987: 195214.
  • 24
    Nagore N, Howe S, Boxer L. Liver cell rosettes. Structural differences in cholestasis and hepatitis. Liver 1989; 9: 4351.
  • 25
    Roskams T, Van Den Oord J J, De Vos R. Neuroendocrine features of reactive bile ductules in cholestatic liver disease. Am J Pathol 1990; 137: 101925.
  • 26
    Wennerberg A E, Nalesnik M A, Coleman W B. Hepatocyte paraffin 1: a monoclonal antibody that reacts with hepatocytes and can be used for differential diagnosis of hepatic tumors. Am J Pathol 1993; 143: 15.
  • 27
    Theise N D, Conn M, Thung S N. Localization of cytomegalovirus antigens in liver allografts over time. Hum Pathol 1993; 24: 1038.
  • 28
    Jorgensen M. A stereological study of intrahepatic bile ducts: 2. Bile duct proliferation in some pathological conditions. Acta Path Microbiol Scand Section A 1973; 81: 6639.
  • 29
    Yamada S, Howe S, Scheuer P J. Three-dimensional reconstruction of biliary pathways in primary biliary cirrhosis: a computer-assisted study. J Pathol 1987; 152: 31723.
  • 30
    Saxena R, Hytiroglou P, Thung S N, Peralta N, Theise N D. Expression of HLA-DR and loss of canals of Hering in early stage primary biliary cirrhosis. Hepatology 1998; 728: 415A.
  • 31
    Hytiroglou P, Tobias H, Abramidou M, Saxena R, Papadimitriou C S, Theise N D. The canals of Hering may represent the primary target of methotrexate toxicity. Hepatology 1998; 28: 603A.