Notching up on the cellular origins of intrahepatic cholangiocarcinoma


  • Alphonse E. Sirica Ph.D., M.S., A.G.A.F.

    1. Division of Cellular and Molecular Pathogenesis Department of Pathology, Virginia Commonwealth University School of Medicine, Richmond, VA
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

Fan B, Malato Y, Calvisi DF, Naqvi S, Razumilava N, Ribback S, et al. Cholangiocarcinomas can originate from hepatocytes in mice. J Clin Invest 2012;122:2911-2915. (Reprinted with permission.)

Abstract 1

Intrahepatic cholangiocarcinomas (ICCs) are primary liver tumors with a poor prognosis. The development of effective therapies has been hampered by a limited understanding of the biology of ICCs. Although ICCs exhibit heterogeneity in location, histology, and marker expression, they are currently thought to derive invariably from the cells lining the bile ducts, biliary epithelial cells (BECs), or liver progenitor cells (LPCs). Despite lack of experimental evidence establishing BECs or LPCs as the origin of ICCs, other liver cell types have not been considered. Here we show that ICCs can originate from fully differentiated hepatocytes. Using a mouse model of hepatocyte fate tracing, we found that activated NOTCH and AKT signaling cooperate to convert normal hepatocytes into biliary cells that act as precursors of rapidly progressing, lethal ICCs. Our findings suggest a previously overlooked mechanism of human ICC formation that may be targetable for anti-ICC therapy.

Sekiya S, Suzuki A. Intrahepatic cholangiocarcinoma can arise from Notch-mediated conversion of hepatocytes. J Clin Invest 2012;122:3914-3918. (Reprinted with permission.)

Abstract 2

Intrahepatic cholangiocarcinoma (ICC) is the second most common primary malignancy in the liver. ICC has been classified as a malignant tumor arising from cholangiocytes; however, the co-occurrence of ICC and viral hepatitis suggests that ICC originates in hepatocytes. In order to determine the cellular origin of ICC, we used a mouse model of ICC in which hepatocytes and cholangiocytes were labeled with heritable, cell type–specific reporters. Our studies reveal that ICC is generated by biliary lineage cells derived from hepatocytes, rather than cholangiocytes. Additionally, we found that Notch activation is critical for hepatocyte conversion into biliary lineage cells during the onset of ICC and its subsequent malignancy and progression. These findings will help to elucidate the pathogenic mechanism of ICC and to develop therapeutic strategies for this refractory disease.


Intrahepatic cholangiocarcinoma (ICC) denotes a histologically diverse group of hepatobiliary tract cancers that exhibit characteristics of cholangiocyte differentiation. Although rare in most regions of the world, because of increased incidence and mortality rates and a still incompletely understood cellular and molecular pathogenesis, ICC is currently being viewed as a cancer of rising importance1 and one that presents worthy biological and therapeutic challenges.2 Highlighting these challenges is the remarkable degree of heterogeneity characterizing ICCs in terms of their epidemiology, cellular, and molecular phenotypes, genomic differences, pathobiological behaviors, and clinicopathological features.

ICCs are macroscopically and microscopically diverse. The Liver Cancer Study Group of Japan classified ICCs as the mass-forming (MF) type, periductular infiltrating (PI) type, intraductal growth (IG) type, and MF plus PI type. The MF type, which has been increasing in incidence, is the most frequent among the macroscopic subtypes,3 followed by the MF plus PI type, which has the worst prognosis for all ICC patients.3, 4 The PI and IG types are the least common of the macroscopic ICC subtypes,3 with the IG type having the most favorable long-term surgical outcome, if curative hepatectomy can be performed.

Conventional small duct ICCs formed in the liver (peripheral ICC) are usually of the MF subtype, whereas those that develop anywhere within the larger second-order intrahepatic bile ducts (perihilar ICC) can be of the PI, MF, PI plus MF, or IG subtypes.5 The vast majority of cases of ICCs are usually diagnosed as well- to moderately differentiated adenocarcinomas,6 with varying degrees of desmoplasia. The histological diversity characterizing ICCs is exemplified in Fig. 1.

Figure 1.

Histological diversity in human ICC. (A) Large duct papillary ICC. (B) Well-differentiated small duct ICC. (C) Well-differentiated large duct tubular ICC. (D) Large duct “intestinal-type” ICC with prominent goblet cell metaplasia. (E) Cholangiolocellular-type ICC. (F) Biliary cystadenocarcinoma with mesnchymal/ovarian stroma. (G) HCC-CCA. (H) Poorly differentiated ICC. H&E, hematoxylin and eosin; CK19, cytokeratin 19 immunostaining. (A), (E), (F), and (G) were kindly provided by Dr. Eizaburo Sasatomi of the Department of Pathology, University of Pittsburgh Medical Center (Pittsburgh, PA).

Nakanuma et al.5 have proposed a new classification of ICCs that reflects their diverse clinical features, genotypes, and biological behavior. This classification takes into consideration gross classification, hepatic progenitor/stem cell phenotypes, and pathological similarities between biliary and pancreatic neoplasms. Under this novel concept, ICCs, which previously have been largely classified into adenocarcinomas and rare variants, were subdivided into the conventional type (small and large bile duct types), bile ductular type, intraductal neoplasm type, and rare variants (e.g., nonclassical types, such as combined hepatocellular and cholangiocarcinoma [HCC-CCA], undifferentiated ICC, and squamous/adenosquamous type), together with some other extremely uncommon forms.5

More recently, Cardinale et al. and Carpino et al.7, 8 have proposed another new classification of cholangiocarcinomas based on cell lineage. Under their classification scheme, which is compatible with the pathological classification of ICC proposed by Nakanuma et al.,5 it is suggested that there are multiple cells of origin in cholangiocarcinoma, including hepatic stem/progenitor cells postulated to be located within the canals of Hering (hepatic stem/progenitor cell lineage) or peribiliary glands (biliary tree stem/progenitor cell lineage), as well as immature or more mature cholangiocyte derivatives, that underlie biological, epidemiological, and clinical heterogeneity in small versus large duct ICCs and extrahepatic bile duct cancer.

Hepatic stem/progenitor cell “biomarkers,” such as neural cell adhesion molecule (NCAM), have been demonstrated to be selectively expressed in combined HCC-CCA9 and in the bile ductular (cholangiolocellular) type.5, 10 Small bile duct type ICCs have also been suggested to originate from interlobular bile ducts.11 Conversely, large duct or perihilar ICCs have been suggested to arise from biliary tree stem/progenitor cells or from more mature descendents.7, 8 A multistep carcinogenesis process indicative of a hyperplasia- dysplasia-carcinoma sequence is also currently recognized.12-15 In this context, malignant progression of precancerous precursor lesions, notably biliary intraepithelial neoplasia (BilIN) without12, 13 or with14 intestinal metaplasia, as well as intraductal papillary neoplasm of the bile ducts exhibiting various phenotypes (e.g., intestinal type, gastric type, and oncocytic type),5 are consistent with different and distinct cell lineage pathways in the cytohistogenesis of ICC variants (i.e., conventional ICC versus less common subtypes, such as intestinal-type ICC or biliary cystic mucinous neoplasm with ovarian stroma5, 14).

Although it has been generally considered that ICCs are derived from either cholangiocytes or, possibly, hepatic and/or biliary stem/progenitor cells, Fan et al.16 and Sekiya and Suzuki17 have now independently demonstrated, with unique mouse models and eloquent hepatocyte fate tracing methods, a compelling alternative to the cellular origin of ICC, namely, through transdifferentiation and neoplastic conversion of normal hepatocytes into malignant cholangiocytes by a mechanism mediated, in part, by overexpression of activated Notch. In the model described by Fan et al., ICCs induced in liver after hydrodynamic tail vein injection of the intracellular domain of Notch1 receptor plasmid, combined with concomitant injection of an Akt-overexpressing plasmid, were of the cystadenocarcinoma type, which formed in noncirrhotic liver. In contrast, in the mouse ICC model of Sekiya and Suzuki, chronically administered thioacetamide to Alb-CreERT2;R26RNotch/+ mice, in which hepatocytes overexpressed an intracellular fragment of activated mouse Notch1, induced macronodular cirrhosis and hepatic tumors exhibiting histological features of classical well-differentiated, small duct tubular-type ICCs with desmoplasia. Thus, it seems reasonable to propose that the microenvironment (cirrhotic versus noncirrhotic) may be an important determinant of ICC histogenesis in these models. Of further interest, in both the Fan et al. and Sekiya and Suzuki studies, ICCs were observed to originate from transdifferentiated hepatocytes in the central areas of the liver lobule and not in the periportal areas, where the hepatic stem/progenitor cell niche is localized. However, the mechanisms underlying the centrilobular origin of transdifferented hepatocytes and subsequent ICC development in these mouse models still need to be addressed.

Although these findings are intriguing, it remains to be determined whether the development of ICCs from transdifferentiated hepatocytes has human clinical relevance. Phenotypic biomarker studies performed as far back as the early 1980s have provided evidence of hepatocyte transdifferentiation into biliary epithelium in human livers under conditions of chronic hepatic injury and cholestasis,15 including established risk conditions for ICC, such as primary sclerosing cholangitis, as well as plausible ICC risk conditions associated with chronic hepatic injury and inflammation (e.g., chronic hepatitis C infection, alchoholic hepatitis, and cirrhosis). However, it remains uncertain whether hepatocyte transdifferentiation to cholangiocytes plays a major role in ICC development in patients with chronic hepatitis C (hepatitis C virus; HCV) or alcoholic liver disease. In this regard, it should be noted that BilIN, a recognized premalignant biliary lesion for human ICC in established risk conditions for ICC, has also been described in intrahepatic bile ducts of patients with chronic HCV and/or alcoholic-related cirrhosis.12, 13 Moreover, it is somewhat surprising that no HCC-CCA tumors were observed in livers of thioacetamide-treated mice genetically engineered to overexpress activated Notch in their hepatocytes, particularly because, in humans, this rare subtype is increasingly being reported in patients with chronic liver injury and cirrhosis.6, 18 Last, though it is currently appreciated that Notch signaling plays a critical role in biliary differentiation and morphogenesis, and is aberrantly overexpressed in human ICCs, Notch signaling has recently been reported to occur at a frequency of 30%-35% in analyzed cases of human HCCs, as well as to promote HCC development in genetically engineered mice.19 Interestingly, the HCCs that formed in these mice were mixed cell-type tumors containing both biliary and hepatocytic phenotypic features. Thus, it would be useful to determine whether overexpression of Notch in diagnosed cases of human HCCs indicates tumor-expressing genes of both hepatocellular and biliary cell differentiation, particularly because tumors with the HCC-CCA phenotype are known to be more aggressive than HCCs without biliary cell differentiation markers.