GASTROENTEROLOGY 2005;128:1655–1667

Cholangiocarcinoma

KONSTANTINOS N. LAZARIDIS and GREGORY J. GORES

Division of Gastroenterology and Hepatology, Center for Basic Research in Digestive Diseases, Mayo Clinic College of Medicine, Rochester,
Minnesota

holangiocarcinoma (CCA) is the primary cancer of 100,000 in 1979 to 0.82 in 100,000 in 19982 (Figure 1).
C the bile ducts. Although it comprises only 10%–
15% of hepatobiliary neoplasms, its incidence is increas-
Below we present the epidemiology of intrahepatic and
extrahepatic CCA separately.
ing.1,2 CCA arises from malignant transformation of During the 1970s, the average age at diagnosis of
cholangiocytes, the epithelial cells that line the biliary intrahepatic CCA was the sixth decade of life. In the late
apparatus. Traditionally, CCA is divided into intrahe- 1980s and during the 1990s, however, the age at diag-
patic and extrahepatic disease according to its location nosis of intrahepatic CCA shifted toward the seventh
within the biliary tree. Intrahepatic CCA occurs within decade of life. This age change of the affected individuals
the hepatic parenchyma, forms classic mass lesions, and may reflect aging of the adult population, with subse-
often presents with advanced clinical features. Extrahe- quent development of CCA, and close follow-up and
patic CCA arises in large bile ducts (ie, left and right therapy of known risk factors (such as primary sclerosing
hepatic ducts and common hepatic and common bile cholangitis [PSC] and choledochal cysts) in young indi-
ducts). These tumors present with features of biliary viduals. In the United States, the male–female ratio for
obstruction. In epidemiological databases, tumors of the intrahepatic CCA is approximately 1.5. The age-adjusted
left and right hepatic ducts are often classified as intra- incidence of intrahepatic CCA in Caucasians and African
hepatic tumors because they can extend into the hepatic Americans is comparable; Asians, however, have twice
parenchyma; this classification should be discouraged as the incidence of Caucasians. The only ethnic group with
it is confusing. There are similarities between intrahe- a reported gradual increase in the age-adjusted incidence
patic and extrahepatic CCA; however, each entity has of intrahepatic CCA is Caucasian. The incidence of in-
distinct epidemiological and clinical features. More im-
trahepatic CCA varies across the world.2 It is highest in
portantly, the etiopathogenetic pathways of intrahepatic
northeast Thailand (96/100,000 in men and 38/100,000
vs extrahepatic CCA are probably independent.1 Because
in women), probably because of the high prevalence of
of scientific progress in the past decade, a better under-
liver fluke infestations.
standing of the pathobiology of CCA is emerging. As we
The mortality related to intrahepatic CCA is also
begin to shed light on the pathogenesis of CCA, we hope
increasing worldwide.5–7 In fact, the percentage of intra-
that its early detection and therapy will improve.3
hepatic CCA–increased mortality is greater than that
observed for hepatocellular carcinoma. In the United
Epidemiology States, the age-adjusted mortality rate for intrahepatic
Overall, CCA is a rare neoplasm. Nevertheless, CCA increased from 0.07 in 100,000 in 1973 to 0.69 in
during the past 3 decades, its incidence has increased.2,4 100,000 in 1997.6 The 5-year survival of patients with
In the United States, approximately 5000 new CCA cases intrahepatic CCA remains unacceptably low and virtu-
are diagnosed yearly.2 Two thirds of CCAs involve the ally unchanged over the past 20 years. Survival has not
extrahepatic bile ducts, whereas the remaining one third dramatically improved despite diagnosis at less advanced
affects the intrahepatic biliary tree.
Studies indicate that intrahepatic and extrahepatic Abbreviations used in this paper: CCA, cholangiocarcinoma; CT,
CCA have a distinct epidemiology. As a result, incorrect computerized tomography; DIA, digitized image analysis; ERCP, endo-
scopic retrograde cholangiopancreatography; EUS, endoscopic ultra-
classification between these 2 types may have affected the sonography; FISH, fluorescence in situ hybridization; 5-FU, 5-fluorou-
reported epidemiological observations of CCA. In the racil; MR, magnetic resonance; MRCP, magnetic resonance
United States, the age-adjusted incidence rates of intra- cholangiopancreatography; OLT, orthotopic liver transplantation; PDT,
hepatic CCA increased from 0.32 in 100,000 in 1975– photodynamic therapy; PSC, primary sclerosing cholangitis.
© 2005 by the American Gastroenterological Association
1979 to 0.85 in 100,000 in 1995–1999.2 Conversely, the 0016-5085/05/$30.00
incidence of extrahepatic CCA declined from 1.08 in doi:10.1053/j.gastro.2005.03.040
1656 LAZARIDIS AND GORES GASTROENTEROLOGY Vol. 128, No. 6

is frequent in Asia but sparse in Western Europe and the

United States. Hepatolithiasis is usually associated with
peripherally located intrahepatic CCA.13 Exposure to
Thorotrast (a colloidal suspension of 232ThO2) has been
linked to the development of CCA. Thorotrast was used
as radiology contrast agent in the early to mid 20th
century. Thorotrast likely causes microsatellite instabil-
ity and subsequently CCA, probably via clonal expansion
of cholangiocytes and inactivation of hMLH1.14 Environ-
mental exposures such as dioxin and vinyl chloride have
Figure 1. Incidence of intrahepatic and extrahepatic CCA in the
United States from 1975 to 1999. Note that the term intrahepatic
been suggested to cause CCA.15,16 Recently, hepatitis C
CCA includes hilar lesions (modified from Shaib et al2). virus infection has been proposed as an etiologic factor of
intrahepatic CCA.17
In conclusion, common features among many CCA
stages, better surgical approaches, and promising new risk factors are chronic biliary inflammation and cho-
therapies (eg, biliary stenting and photodynamic therapy lestasis. Both of these events contribute to malignant
[PDT]). transformation of the cholangiocyte.
The incidence of extrahepatic CCA also varies across
Molecular Pathogenesis
the globe. In the United States, the reported age-ad-
justed incidence of extrahepatic CCA is 1.2 in 100,000 During the last 10 years, considerable progress
for men and 0.8 in 100,000 for women.8 Nonetheless, has been made in beginning to understand the patho-
the overall incidence of extrahepatic CCA is declining. genesis of CCA. In general, malignant transformation of
Moreover, the age-adjusted mortality rate of extrahepatic the cholangiocyte occurs in an environment of chronic
CCA is also decreasing in the Western countries,5 with biliary inflammation, chronic cholestasis, or both. We
the exception of Italy and Japan. In the United States, now appreciate several risk factors that predispose to the
the age-adjusted mortality rates declined from 0.6 in development of CCA (Table 1). In fact, the milieu of
100,000 in 1979 to 0.3 in 100,000 in 1998.6 To this chronic biliary inflammation, and cholestasis, leads to the
end, current evidence indicates minor improvements in production of cytokines and reactive oxygen species, and
5-year survival rates of extrahepatic CCA from 11.7% in this causes protracted cellular (ie, cholangiocyte) stresses
1973–1977 to 15.1% in 1983–1987.8 and irreversible DNA damage.1 As a result, cholangio-
cytes attain subcellular and cellular phenotypes that re-
Risk Factors sult in malignant transformation. Molecular mechanisms
Table 1 reports the risk factors that thus far have responsible for bile duct carcinogenesis likely include the
been associated with the development of CCA. Still, for interaction of genetic variants and somatic cell alter-
most CCA cases the cause is unknown, and these indi- ations. Figure 2 illustrates the proposed pathways that
viduals lack exposure to or association with known risk partake in cholangiocarcinogenesis. These paths include
factors. PSC is a definite risk factor for CCA.9,10 In a PSC mechanisms that result in cholangiocyte: (1) self-suffi-
patient, the risk for developing CCA is approximately ciency and proliferation, (2) apoptosis resistance, (3) es-
1.5% per year after diagnosis of the cholestatic liver cape from senescence, and (4) tumor invasiveness and
disease.11 Among the patients with PSC who will acquire metastasis.
CCA, approximately 30% will be diagnosed with ma-
lignancy of the bile ducts within 2 years after diagnosis Table 1. CCA Risk Factors
of PSC.10,11 It is interesting to note that the risk of Age ⬎65 y
developing CCA is not associated with the duration of Primary sclerosing cholangitis (PSC)
Liver fluke infestation
PSC.11 Opisthorchis viverrini
Hepatobiliary flukes, namely, Opisthorchis viverrini and Clonorchis sinensis
Clonorchis sinensis, are strongly associated with CCA. Caroli disease
These worms inhabit the bile ducts and occasionally the Choledochal cysts
Bile duct adenoma and biliary papillomatosis
gallbladder after ingestion of undercooked fish. Patients Chronic intraductal stones (ie, hepatolithiasis)
with choledochal cysts (ie, congenital cystic dilatation of Liver cirrhosis
the ducts) have a 10%–15% lifetime risk of developing Surgical biliary/enteric drainage procedures
Chemicals/agents (ie, Thorotrast, dioxin, or vinyl chloride)
CCA.12 Hepatolithiasis (ie, intrahepatic bile duct stones)
May 2005 CHOLANGIOCARCINOMA 1657

Figure 4. Classification of hilar CCA by Bismuth–Corlette. There are

4 types of hilar CCA. Type I affects the common hepatic duct; type
II involves the common hepatic duct and the confluence of the right
and left hepatic ducts; types IIIa and IIIb occlude the common
Figure 2. Molecular alterations of cholangiocyte malignant transfor-
hepatic duct and either the right or left hepatic duct, respectively;
mation. CCA possess molecular mechanisms to manifest self-suffi-
and type IV involves the biliary confluence and extends to both the
ciency in growth and proliferation, avoid apoptosis, escape from
right and left hepatic ducts or refers to multifocal bile duct tumors.
senescence, and invade tissues and metastasize. The illustration
depicts the molecular pathways of malignant cells that likely interact
to produce the CCA phenotype. WISP1v, WNT1-inducible signaling
pathway protein 1; VEGF, vascular endothelial cell growth factor; two thirds of all CCA and is further divided into: (1)
WAF1, wild-type p53-activated fragment 1; Mdm-2, murine double hilar or Klatskin, (2) middle, and (3) distal tumors.
minute-2 gene; FLIP, flice-inhibitory protein; NO, nitric oxide; Bcl-2,
B-cell lymphoma/leukemia 2; Bcl-XL, BCL2-related protein, long iso-
Klatskin tumors represent approximately 60% of all
form; Mcl-1, myeloid cell leukemia 1; COX-2, cyclooxygenase 2; IL-6, extrahepatic CCA. Figure 4 illustrates the Bismuth–
interleukin 6; HGF, hepatocyte growth factor; EGF, epidermal growth Corlette classification of hilar CCA. Macroscopically, ex-
factor (modified from Gores1). (Numbers in brackets denote relevant
references).
trahepatic CCA presents as sclerosing, nodular, or pap-
illary phenotypes. The sclerosing type is the most
frequent and results in annular thickening of the bile
Pathologic Classification ducts because of infiltration and fibrosis of the periductal
According to its location in the biliary tree, CCA tissues. The intrahepatic CCA are classified into the
is classified into extrahepatic and intrahepatic types (Fig- following 4 growth types: (1) mass forming, (2) periduc-
ure 3). The extrahepatic type accounts for approximately tal infiltrating, (3) mass forming plus periductal infil-
trating, and (4) intraductal.18
Whether CCA is intrahepatic or extrahepatic, the
usual microscopic appearance is an adenocarcinoma. The
most common histological appearance is a well- to mod-
erately differentiated tubular adenocarcinoma within a
prominent, dense, desmoplastic stroma. Poorly differen-
tiated CCA is not uncommon and is characterized by
individual malignant cells dispersed in a fibrous stroma.
The tumor is often associated with prominent perineural
invasion. Other histological variants of CCA include
papillary adenocarcinoma, signet-ring carcinoma, squa-
mous cell or mucoepidermoid carcinoma, and a lympho-
epithelioma-like form.
Clinical Presentation, Laboratory Findings,
Figure 3. The term CCA refers to tumors involving the entire (ie, Imaging Studies, and Diagnosis
intrahepatic and extrahepatic) biliary tree. Intrahepatic CCA denotes
malignancy affecting the intrahepatic bile ducts. Extrahepatic CCAs Patients with extrahepatic CCA present with
are divided into hilar (ie, Klatskin), middle, and distal tumors. jaundice, dark urine, pale stools, pruritus, malaise, and
1658 LAZARIDIS AND GORES GASTROENTEROLOGY Vol. 128, No. 6

weight loss. Laboratory tests show obstructive cholestasis

with increased alkaline phosphatase and bilirubin. The
serum marker CA 19-9 may be increased. CA 19-9, the
most frequently tested marker for pancreatobiliary ma-
lignancies,19 detects circulating high-molecular-weight
mucin glycoproteins coated with sialylated blood group
epitopes (ie, sialyl Lewis).20 CA 19-9 blood levels are
dependent on the red blood cell Lewis phenotype.21
Approximately 7% of the population are Lewis negative,
and, thus, these individuals will have a nondetectable CA
19-9 even in the presence of malignancy.22 CA 19-9 is
not specific for CCA. Cancers of the pancreas, stomach,
and colon; bacterial cholangitis; smoking; and gyneco-
logic malignancies can cause increased CA 19-9. Bacte-
rial cholangitis is a frequent cause of increased CA 19-9
serum levels in this patient population.
Intrahepatic CCA presents with nonspecific symptoms
including abdominal pain, diminished appetite, weight
loss, malaise, and night sweats. Sometimes an incidental
abdominal mass detected during either a physical exam- Figure 5. Percutaneous transhepatic cholangiogram showing a hilar
CCA causing obstruction of the biliary bifurcation and the right and left
ination or imaging study is the only finding of CCA in hepatic ducts with extension into the left anterior and posterior seg-
asymptomatic patients. Laboratory tests usually show an ments. Note the presence of a metallic stent in the right hepatic duct
increased alkaline phosphatase with a normal bilirubin. and bifurcation.
Serum tumor markers, such as CA 19-9, may be in-
creased.
In extrahepatic CCA, imaging studies show dilatation (ie, dilation) and hepatic lobes (ie, atrophy/hypertrophy
of the biliary ducts and often define the anatomic level of complex; see below). On CT examination, intrahepatic
biliary obstruction. The position and extent of extrahe- CCA presents with delayed venous phase enhancement of
patic CCA along the biliary system is defined by endo- a hypodense lesion after contrast administration. On
scopic retrograde cholangiopancreatography (ERCP), cross-sectional MR imaging, CCA appears as a hypoin-
magnetic resonance cholangiopancreatography (MRCP), tense lesion on T1- weighted images and as a moderately
or percutaneous transhepatic cholangiography. Despite intense signal on T2-weighted images. In hilar CCA,
being invasive, ERCP permits brush cytology and biop- MRCP may show irregular thickening of the bile duct
sies of the bile ducts for histological assessment. Figure 5 wall if the lesion is ⬎5 mm, in addition to proximal
shows a percutaneous transhepatic cholangiogram of a biliary dilatation.23
patient with hilar CCA. Tissue-proven diagnosis of ex- Unilobular bile duct obstruction usually results in
trahepatic CCA can be daunting because it is a highly atrophy of the affected hepatic lobe followed by hyper-
desmoplastic tumor composed of aggregations of a few trophy of the nonaffected lobe. This phenomenon is
malignant cholangiocytes within excessive fibrous tissue. known as the atrophy-hypertrophy complex.24 On cross-sec-
The desmoplastic reaction surrounds the bile ducts and tional imaging the presence of an atrophied hepatic lobe
extends into the submucosa, and, thus, biliary cytology is alone without an atrophy-hypertrophy complex suggests
positive for CCA in only approximately one third of the vasculature encasement of the affected lobe by CCA.
cases.3 A combination of brush cytology and endoscopic New diagnostic methods, such as digitized image
biopsy may increase the yield of CCA-positive findings to analysis (DIA) and fluorescence in situ hybridization
40%–70%.3 (FISH), offer promise to evaluate extrahepatic bile duct
In CCA, computerized tomography (CT) scans and lesions for cellular aneuploidy and chromosomal aberra-
magnetic resonance (MR) imaging aid in the diagnosis tions.25 To perform DIA and FISH assays, bile duct
and evaluation for possible tumor resection. On abdom- brushing/aspirate specimens are collected at the time of
inal CT, the primary lesion of extrahepatic CCA is ERCP, and cells are fixed on a slide. DIA is a digital
usually not visible given its infiltrative nature, which camera–assisted image of the cell nucleus that captures
causes stricturing along the bile ducts. Nevertheless, CT pictures as light is transmitted through the slide speci-
can show the effect of CCA obstruction on the bile ducts men. This laboratory-based technique allows DNA con-
May 2005 CHOLANGIOCARCINOMA 1659

dard cytology for the discovery of malignant bile duct

strictures.
Endoscopic ultrasonography (EUS) has been used in
assessing the nature of biliary strictures. In a recent study
of 28 patients, EUS-guided fine-needle aspiration (FNA)
biopsy of suspected CCA has shown a specificity, sensi-
tivity, and positive predictive value of 86%, 100%, and
100%, respectively.29 From the same study, EUS with
FNA was reported to have a positive effect on the clinical
Figure 6. Fluorescence in situ hybridization of normal and malignant management of 84% of patients with CCA.29 These
cholangiocytes. Fluorescently labeled DNA probe decorating genomic studies were performed in non-PSC patients; therefore,
loci on 4 separate chromosomes are shown. The red color probe their role in identifying CCA in PSC patients is unclear.
indicates chromosome 3, the green color probe specifies chromo-
some 7, the gold color probe points to chromosome 9, and aqua Also, the risk of peritoneal seeding with this technique
identifies chromosome 17. Normal cholangiocytes (A) have 2 dupli- in patients with potentially resectable disease needs to be
cates of each probe, as anticipated for normal diploid cells. Malignant taken into account.
cholangiocytes (B) show gains of chromosomal probes, thus suggest-
ing polysomy.
The diagnosis of CCA is usually made by evaluating
the clinical examination and biochemical results and by
obtaining endoscopic and imaging procedures (ie, ERCP
tent quantification, assessment of chromatin distribu- and CT/MR imaging of abdomen) to delineate the biliary
tion, and nuclear morphology.26 In a recent prospective anatomy. In clinical practice, it is common to make the
study, DIA was compared with routine brush cytology diagnosis of CCA on the basis of clinical/laboratory/
for the detection of cancer in suspicious biliary tract imaging studies without tissue-proven evidence of tu-
strictures.27 This study showed that DIA was signifi- mor. However, it should be noted that 10% of malig-
cantly more sensitive (39.3%) than routine brush cytol- nant-appearing biliary strictures are benign in surgical
ogy (17.9%; P ⫽ .014). The specificity of DIA was series.30,31 In patients with PSC, the diagnosis of super-
77.3%, whereas the specificity of cytology was 97.7% (P imposed CCA can be very demanding. The patient may
⫽ .003). The lower DIA specificity was attributed to the have dominant biliary stricture, and it is not easy to
high proportion of patients with PSC in that study.27 differentiate whether it is a benign lesion or CCA. In a
Nevertheless, the sensitivity was greater, and the accu- patient with PSC, sudden and unexpected clinical dete-
racy of DIA was comparable to that of cytology (56% vs rioration, which is associated with progressive increases
53%). of alkaline phosphatase and serum CA 19-9 (⬎100
The FISH assay uses fluorescently labeled DNA-based U/mL), in the absence of bacterial cholangitis, indicates
probes to detect chromosomal aberrations in cholangio- probable development of CCA.32,33 Indeed, the sensitiv-
cytes. Bile duct cells with chromosomal gains suggest ity and specificity of a CA 19-9 value ⬎100 U/mL for
malignancy. To date, FISH has been performed with 4 detecting CCA in patients with PSC are 75%– 89% and
fluorescently labeled DNA probes hybridized to the cen- 80%– 86%, respectively.32,33
tromere of chromosomes 3, 7, and 17 and the p16 gene
on chromosome 9 (9p21). After hybridization, the cells Staging
are stained with the nuclear counterstain 4=,6-dia- CCA staging is clinically important to identify
midino-2-phenylindole, and fluorescence microscopy is candidates for surgical resection. Tables 2 and 3 describe
used to screen the slide for abnormal cholangiocytes the TNM classification for intrahepatic and extrahepatic
(Figure 6). A FISH assay is declared positive when ⱖ5 CCA, respectively. The worth of TNM system for extra-
cells display gains of ⱖ2 chromosomes or when ⱖ10 hepatic CCA is limited, because it relates to histopathol-
cells show a gain of a single chromosome. A positive ogy and not to the extent of disease. Knowing the disease
FISH study, however, cannot define the position of the extension is pivotal in making decisions regarding sur-
bile duct malignancy. In a recent comparison study of gical resection of the tumor. Table 4 shows proposed
FISH to standard cytology for detection of malignant preoperative T-stage criteria for hilar CCA. During the
bile duct strictures, the sensitivity of FISH and cytology clinical staging of extrahepatic CCA, the proximal and
were 34% and 15%, respectively (P ⬍ .01). The speci- distal margins of the tumor should be clearly identified.
ficity of FISH was 91%, and the specificity of cytology ERCP, MRCP, or, less common, percutaneous transhe-
was 98% (P ⫽ .06).28 It seems, therefore, that biliary patic cholangiography can be used to map the boundaries
FISH is more sensitive and virtually as specific as stan- of bile duct tumor. It is also important to exclude CCA
1660 LAZARIDIS AND GORES GASTROENTEROLOGY Vol. 128, No. 6

Table 2. TNM Pathologic Classification of Intrahepatic Table 4. Proposed Preoperative T-Stage Criteria for Hilar
CCA34 CCA35
Stage Tumor Node Metastasis Stage Criteria
I T1 N0 M0 T1 Tumor involving biliary confluence with or without unilateral
II T2 N0 M0 extension to second-order biliary radicles
IIIA T3 N0 M0 T2 Tumor involving biliary confluence with or without unilateral
IIIB T4 N0 M0 extension to second-order biliary radicles and ipsilateral
IIIC Any T N1 M0 portal vein involvement with or without ipsilateral
IV Any T Any N M1 hepatic lobar atrophy
T3 Tumor involving biliary confluence ⫹ bilateral extension to
T1, solitary tumor without vascular invasion; T2, solitary tumor with second-order biliary radicles or unilateral extension to
vascular invasion or multiple tumors, none ⬎5 cm; T3, multiple second-order biliary radicles with contralateral portal
tumors ⬎5 cm or tumor involving a major branch of the portal or vein involvement or unilateral extension to second-order
hepatic vein(s); T4, tumor(s) with direct invasion of adjacent organs biliary radicles with contralateral hepatic lobar atrophy
other than gallbladder or with perforation of visceral peritoneum; N0, or main or bilateral portal vein involvement
no regional lymph node metastasis; N1, regional lymph node metas-
tasis; M0, no distant metastasis; M1, distant metastasis.

vascular encasement of the contralateral (ie, nonaffected) CCA require major hepatectomy for complete surgical re-
liver lobe before committing to partial hepatectomy and moval of the malignancy. Thus, many patients are not
to verify vascular patency of the portal vein and hepatic deemed surgical candidates because of comorbidities or
artery. This usually can be accomplished by Doppler advanced age, despite evidence of resectable disease. Regret-
ultrasonography and/or CT or MR vascular studies. Fi- tably, more than half of CCA patients present with ad-
nally, regional metastases should be ruled out. Studies vanced, unresectable malignancy. In such cases, palliative
have shown that EUS is better compared with conven- therapies such as biliary stenting and photodynamic therapy
tional cross-sectional abdominal imaging (ie, CT or MR (PDT) provide symptom relief and may improve survival. A
imaging) to exclude metastatic disease. This is particu- clinical controversy is the presence of an intrahepatic mass
larly important for questionable regional lymph nodes, that may mimic intrahepatic CCA vs metastatic disease of
which can be biopsied during EUS to rule out metastasis. an unknown primary tumor. In our view, a patient who
Indeed, 15%–20% of CCA patients with unremarkable presents with a dominant liver mass and no evidence of
abdominal imaging studies have metastatic lymph node detectable extrahepatic disease by physical examination,
involvement according to EUS evaluation.25 laboratory studies, and body imaging most likely has intra-
hepatic CCA.
Therapy At present, chemotherapy and/or radiation therapies for
Surgical resection is the best available and poten- CCA have not been evaluated in the context of randomized,
tially curative therapy for both intrahepatic and extrahe- controlled trials, and, therefore, their efficacy remains du-
patic CCA. Almost all intrahepatic and most extrahepatic bious, including the use of adjuvant chemotherapy after
surgical resection to diminish the risk of CCA recurrence.3
Finally, at selected liver transplant centers, a small fraction
Table 3. TNM Pathologic Classification of Extrahepatic
of well-chosen patients with hilar CCA may undergo or-
CCA34
thotopic liver transplantation (OLT), with excellent survival
Stage Tumor Node Metastasis results. Figure 7 provides an algorithm for the overall
0 Tis N0 M0 management of CCA.
IA T1 N0 M0
IB T2 N0 M0 Surgical Therapy
IIA T3 N0 M0
IIB T1 to T3 N1 M0 Extrahepatic cholangiocarcinoma. Most extrahe-
III T4 Any N M0 patic CCAs are hilar tumors. CCA involving the distal
IV Any T Any N M1
bile ducts usually necessitates pancreaticoduodenectomy.
Tis, carcinoma in situ; T1, tumor confined to the bile duct histologi- Below we discuss the surgical therapy for hilar tumors.
cally; T2, tumor invades beyond the wall of the bile duct; T3, tumor
invades the liver, gallbladder, pancreas, and/or unilateral branches of
Staging of extrahepatic CCA is a critical step before
the portal vein (right or left) or hepatic artery (right or left); T4, tumor consideration for surgical resection. Clinical staging cri-
invades any of the following: main portal vein or its branches bilater- teria for hilar CCA have been proposed (Table 4). It is
ally, common hepatic artery, or other adjacent structures, such as the interesting to note that these criteria correlate with
colon, stomach, duodenum, or abdominal wall; N0, no regional lymph
node metastasis; N1, regional lymph node metastasis; M0, no dis- tumor resectability (ie, 60% in stage T1 and 0% in stage
tant metastasis; M1, distant metastasis. T3) and patient survival.35 The evaluation for resectabil-
May 2005 CHOLANGIOCARCINOMA 1661

may make determining the intraoperative disease extent

difficult. Patients with CCA limited to the common bile
duct likely do not benefit from preoperative stenting.
Patients who require hepatic resection may benefit from
stenting of the remaining liver if it is also obstructed.
Cholestasis may not resolve rapidly and can impair liver
regeneration.41
Resection of extrahepatic CCA is a major operation
with a 5%–10% mortality rate and notable morbidity
even at large medical centers with expertise in hepato-
biliary surgery.35 Infections are the dominant cause of
postoperative mortality.35 Regional lymph node metas-
tases are associated with reduced 3- and 5-year survival.42
However, it is unknown whether extended lymph node
resection during surgery improves patient survival. Be-
yond the curative intent, hepatobiliary surgery (ie, cho-
ledochojejunostomy or hepaticojejunostomy) may have a
palliative effect on obstructive jaundice in extrahepatic
CCA. However, current endoscopic modalities (ie, biliary
stenting, PDT, and brachytherapy) and the high cost,
morbidity, and mortality of operations have rendered the
Figure 7. An overview of clinical management for CCA. OLT, ortho-
topic liver transplantation. surgical approaches less favorable for palliation of
jaundice.
Intrahepatic cholangiocarcinoma. Surgery is also
ity requires careful patient selection and meticulous in- the best option for effective and potentially curative
terpretation of imaging studies. During preoperative therapy for intrahepatic CCA. Intrahepatic CCAs are
evaluation, approximately one third of CCA cases will be large tumors at the time of diagnosis that necessitate
deemed unresectable. Table 5 lists the criteria for non- major liver resection. Prognostic factors that indicate a
resectable CCA. Moreover, during laparoscopy, 25%– poor outcome after surgical resection of intrahepatic
30% of patients who were thought to be candidates for CCA are shown in Table 6. Metastasis of CCA to the
surgical resection will be found to have unresectable regional lymph nodes affects survival; however, the effect
CCA.35,37 Therefore, laparoscopy before resection of ex- of surgical node dissection on patient survival is unclear.
trahepatic CCA has become the standard surgical After surgical resection of intrahepatic CCA, the median
approach. and 5-year survival rates range from 12 to 28 months and
In general, patients with resectable extrahepatic CCA from 29% to 36%, respectively. The different types of
necessitate partial hepatectomy to achieve tumor-free intrahepatic CCA have now been defined and include a
margins. This is important because patients with posi- classic mass lesion, periductal infiltrating disease, mass
tive surgical margins have survival comparable to those lesion plus periductal infiltrating disease, and intraductal
receiving palliative therapy alone.35,38,39 Patients with papillary neoplasms. After surgical resection, patients
tumor-free margins have a 20%– 40% 5-year survival with a mass lesion plus periductal infiltrating disease had
rate.35,38 To accomplish a biopsy-proven negative surgi-
cal margin, the surgeon needs to perform resection of the
Table 5. Criteria of Nonresectable CCA36
tumor/extrahepatic bile ducts along with subhilar
lymphadenopathy. Data indicate that concurrent en bloc 1. Hepatic duct involvement up to secondary biliary radicals
bilaterally
partial hepatectomy is associated with a higher degree of 2. Encasement or occlusion of the main portal vein proximal to its
negative resection margins.35,39 Extrahepatic CCA involv- bifurcationa
ing the biliary confluence affects the main caudate duct and, 3. Atrophy of 1 hepatic lobe with encasement of contralateral
portal vein branch
thus, requires caudate lobe removal.35,40 An area of contro- 4. Atrophy of 1 hepatic lobe with contralateral involvement of
versy in patients with CCA who present with obstructive secondary biliary radicals
jaundice is the need for preoperative biliary stenting. Al- 5. Distant metastases (peritoneum, liver, lung)
though stenting of the bile ducts alleviates cholestasis in aRelative criterion. Portal vein resection and reconstruction may be

most cases, it poses the potential risk of infection and possible.

1662 LAZARIDIS AND GORES GASTROENTEROLOGY Vol. 128, No. 6

Table 6. Prognostic Factors Associated With Unfavorable consensus for biliary stenting of Bismuth type II, III, or
Outcome After Surgical Treatment of Intrahepatic IV hilar CCA. In a prospective, randomized controlled
CCA
trial of patients with hilar CCA (ie, Bismuth I to III), it
Preoperative CA 19-9 levels ⬎1000 U/mL was concluded that unilateral drainage was adequate to
Multifocal disease
Liver capsule invasion
alleviate biliary obstruction; moreover, endoscopic at-
Lack of cancer-free surgical margins tempts to place a second biliary stent were likely to result
Regional lymph node metastases in early complications (eg, bacterial cholangitis) without
Mass-forming or periductal infiltrating-type CCA growth
evidence of a patient survival benefit.46
Expression of MUC1 by CCA cells
Both plastic and metallic biliary stents have been used
to alleviate CCA. Plastic stents have a smaller diameter
and become occluded more easily compared with self-
a shorter survival than those with mass-forming expandable metal stents.43 The latter are also cost-effec-
tumors.18 tive in patients with CCA who survive for more than 3
months.
Palliative Therapeutic Approaches Photodynamic therapy. PDT has also been used
Many patients with intrahepatic or extrahepatic to alleviate malignant obstruction of the extrahepatic
CCA have unresectable disease at the time of diagnosis. bile ducts.47 During the first visit, the patient undergoes
In addition, because CCA affects usually the elderly, a systemic preadministration of a nontoxic photosensitiz-
significant percentage of surgical candidates have comor- ing drug, which accumulates mainly in the CCA. Two
bidities that preclude an operation for tumor resection. days later, the patient has ERCP to activate the photo-
Palliative therapies aim at improving or resolving ob- sensitizer intraductally via direct nonthermal laser appli-
structive jaundice and subsequently ameliorating patient cation. During activation, the photosensitizing agent
symptoms. Palliative therapies of obstructive jaundice reaches an excited reactive state (triplet state); subse-
include biliary stenting, PDT, and intraluminal brachy- quently, the energy is transferred from the triplet state of
therapy. Although the nonsurgical approaches to treat the photosensitizer to molecular oxygen. This process
obstructive jaundice can be performed endoscopically or creates singlet molecular oxygen (1O2), which results in
percutaneously, we discuss here only endoscopic-guided direct or indirect photodamage of the targeted tissue (ie,
therapies. CCA). Specifically, the photosensitizer accrues in the
Biliary stents. In patients with CCA, relief of mitochondria and causes apoptosis of malignant cholan-
obstructive jaundice improves symptoms and quality of giocytes and surrounding tissues, thus resulting in tumor
life, but not survival. Biliary stents are an effective regression (Figure 8).
modality in relieving malignant bile duct obstruction The most common photosensitizer in use is a deriva-
with subsequent amelioration of jaundice. Biliary drain- tive of hematoporphyrin (porfimer sodium; Photofrin;
age of only 25%–30% of the hepatic parenchyma is Axcan, Birmingham, AL). Sodium porfimer is given
required to achieve resolution of jaundice. Parameters to intravenously at 2 mg/kg body weight. In CCA, studies
be considered before palliative endoscopic therapy with have shown that porfimer enrichment is adequate for
biliary stents include location and extent of bile duct PDT between day 1 and 4 after intravenous administra-
obstruction, hepatic lobe atrophy, type of biliary stents tion. Intraluminal photoactivation is achieved by laser
(ie, plastic vs metallic) number of biliary stents (ie, single light (wavelength, 630 nm; light dose, 180 J/cm2). The
vs double drainage), and patient life expectancy.43 Ob- tumoricidal depth penetration of PDT with porfimer is
taining a cross-sectional imaging study before an ERCP approximately 4 – 6 mm. Parameters that determine the
provides guidance for the endoscopist to choose the depth and extent of tissue damage by PDT include the
optimal bile ducts for stenting, thus avoiding atrophied type and quantity of the photosensitizer used, the oxygen
hepatic segments and minimizing the risk of postproce- concentration in the affected tissue, and the intensity,
dural cholangitis.43,44 Having this information before an absorption, and distribution of laser light. The PDT
ERCP not only improves the success rate of stenting indications and contraindications are listed in Table 7.
obstructed bile ducts but also facilitates the resolution of Patients injected with porfimer can tolerate normal
jaundice.45 The type and number of biliary stents to treat artificial room light. Nevertheless, they must keep out of
obstructive jaundice in CCA should be individualized. bright, direct, or indirect sunlight because sunlight ex-
For instance, a patient jaundiced because of Bismuth posure can cause significant skin phototoxicity. Intraduc-
type I hilar CCA can receive successful palliative treat- tal PDT therapy can produce mild to moderate epigastric
ment with a single biliary stent. Nevertheless, there is no pain for up to 3 days after the endoscopic procedure. In
May 2005 CHOLANGIOCARCINOMA 1663

Figure 8. Photodynamic ther-

apy (PDT) in a patient with unre-
sectable hilar CCA. (A) ERCP be-
fore PDT. (B) ERCP 6 months
after PDT. Note the improve-
ment of the previously stric-
tured left hepatic duct.

the first 2–3 days after PDT, a transient increase of Nonrandomized pilot studies of PDT for unresectable
aspartate aminotransferase and leukocytosis is antici- CCA have shown promising results on: (1) improvement
pated. Biloma and hemobilia have been reported as PDT or resolution of cholestasis, (2) stabilization of the
complications. No biliary perforation has been reported Karnofsky performance status at almost normal or mod-
after PDT. The rate of cholangitis is not increased after erately diminished rates, and (3) improvement or pres-
PDT as compared with placement of biliary stents alone. ervation of quality of life.48 Recently, a multicenter
The reported 30-day mortality after PDT was approxi- prospective, randomized trial has evaluated the effect of
mately 2%47 because of 2 fatal episodes of pulmonary biliary stenting followed by PDT (group A) compared
embolism, which were probably related to paraneoplastic with biliary stenting alone (group B) in patients with
thromboses rather than to PDT. unresectable CCA.49 Group A showed prolonged survival
To “bleach out” the accumulated porfimer in the skin, (n ⫽ 20; median survival, 493 days; 95% confidence
patients should follow specific instructions. This goal can interval, 276 –710 days) compared with group B (n ⫽
be accomplished by short exposures (5–10 minutes) after
19; median survival, 98 days; 95% confidence interval,
day 4 to mild evening sunlight before sunset. If the
87–107 days; P ⬍ .0001).49 This study failed to improve
initial exposure is endured without skin sunburn, grad-
biliary obstruction and jaundice in group B (bile duct
ual re-exposure is recommended until bright sunlight is
stenting alone). Thus, it is likely that the survival benefit
tolerated.
reported in group A (biliary stenting and PDT) relates to
amelioration of cholestasis rather than tumor burden.50
Table 7. PDT Indications and Contraindications for Hilar Additional prospective, randomized trials of PDT for
CCA48 unresectable CCA are needed to further assess the utility
Indications of this promising approach.
Preliminary indication: nonresectable hilar CCA with unrelieved Intraluminal brachytherapy. During intralumi-
cholestasis
nal brachytherapy, premounted iridium-192 seeds are
Relative indications (ie, within clinical trials)
Nonresectable hilar CCA with successful biliary drainage deployed within a catheter and placed across malig-
Inoperable comorbid patient with resectable hilar CCA nant biliary strictures via ERCP or percutaneous
Borderline resectability of hilar CCA (neoadjuvant PDT for
purging of intrahepatic ducts from tumor cells beyond the
transhepatic cholangiography.51,52 It is expected that
tumor margins) brachytherapy provides focal, greater, and more effec-
Contraindications tive doses of radiation than external beam radiation
Porphyria (all genetic types)
Recent use of photosensitizing or dermatotoxic drugs (eg,
therapy. It is thus anticipated that intraluminal
bleomycin) brachytherapy extends the palliation of obstructive
Insertion of a coated metal stent jaundice while avoiding unnecessary radiation injury
Severe hepatic or renal failure
Relative contraindications
of the surrounding tissues/organs. Until now, the re-
Peritoneal carcinomatosis (cholestasis palliated) sults of intraluminal brachytherapy have varied.52,53
Karnofsky performance status ⬍30% More studies are required to evaluate the effectiveness
Biliary empyema or liver abscess
of this palliative treatment.
1664 LAZARIDIS AND GORES GASTROENTEROLOGY Vol. 128, No. 6

Liver Transplantation for Unresectable 5-FU or capecitabine (2000 mg/m2 daily) 2 out of 3
Cholangiocarcinoma weeks until OLT is performed.
The initial experience with OLT for CCA was After completing neoadjuvant chemoradiation ther-
apy, patients undergo staging laparotomy with biopsy of
unsatisfactory.54 – 60 Recurrence of CCA was common,
the regional hepatic lymph nodes, other intra-abdominal
and the 5-year survival rates were only 5%–15%. To this
lymph nodes, or nodules suggestive of tumor. Patients
end, most liver transplant centers consider CCA a con-
with negative staging operations proceed with OLT.
traindication for OLT. It is interesting to note that,
Cadaveric livers or living donor right liver grafts can be
however, a selected group of patients who underwent
used. After transplantation, patients receive standard
OLT and had negative surgical resection margins and
immunosuppression regimens.
negative regional lymph nodes had long-term survival.61
As of May 1, 2004, 56 patients with CCA were
Furthermore, in a small number of patients treated with
enrolled in this liver transplantation protocol at the
radiation, brachytherapy, and 5-fluorouracil (5-FU), the
Mayo Clinic in Rochester, Minnesota. Forty-eight pa-
observed 5-year survival rate was 22%.62 Because of these
tients underwent staging laparotomy, of whom 14 (29%)
favorable observations, an experimental liver transplan-
were found to have progression of CCA on laparotomy
tation protocol was developed at the Mayo Clinic that
and were excluded from the protocol. Between 1993 and
was aimed at treating selected patients with early-stage
2003, 28 patients underwent OLT; 6 died after OLT,
unresectable hilar CCA or CCA arising in the back- and 22 are currently alive. The actuarial patient survival
ground of PSC. after liver transplantation was 88% at 1 year and 82% at
To be eligible for this protocol, the diagnosis of CCA 5 years. The time from enrollment to OLT was approx-
needs to be confirmed by biopsy, brush cytology, or imately 4 and 10 months during the first 5 years and the
demonstration of cellular aneuploidy in the presence of second 5 years of the protocol, respectively. After the
malignant stricture based on cholangiographic studies. If described neoadjuvant chemoradiation protocol, the out-
a mass lesion is present in the perihilar region, then the come of OLT in selected patients with CCA was similar
diameter should be ⬍3 cm on cross-sectional imaging to that after liver transplantation for other chronic liver
studies. The CCA also has to be considered unresectable diseases. Of note, the outcome of our liver transplanta-
by the hepatobiliary team after meticulous clinical, lab- tion protocol for CCA surpasses the outcome of surgical
oratory, and imaging evaluation. Intrahepatic and extra- resection with curative intent—the gold standard ther-
hepatic CCA metastases have to be excluded by CT of the apy for CCA. A comparable liver transplantation proto-
abdomen and chest, abdominal ultrasound, EUS with col for CCA has been developed at the University of
FNA of the regional lymph nodes, and total body bone Nebraska and has also shown favorable patient survival
scan. Tumor vascular encasement causing an absence of outcome.63
blood flow without evidence of vessel invasion is not a
contraindication to enrollment. Future Directions
Qualified CCA patients have to be suitable for radia- In view of the increasing incidence of CCA, we
tion therapy, chemotherapy, and liver transplantation, as need better early detection methods and new, effective
determined by the interdisciplinary transplantation therapies to improve the survival of patients with this
team. Patients who meet eligibility criteria receive neo- distressing disease. To date, patients with CCA lack a
adjuvant chemotherapy and radiation therapy. External survival benefit if treated with chemotherapy or radiation
beam radiation therapy is administered (a total dose of therapy.3 Randomized, controlled clinical trials are nec-
4500 cGy in 30 sessions) over 3 weeks. 5-FU is given essary to evaluate novel chemotherapeutic agents in con-
intravenously at 500 mg/m2 daily as a bolus for 3 con- junction with radiation therapy. We hope that as the
secutive days during the initiation of external beam pathogenesis of CCA is elucidated, better pharmacolog-
radiation therapy. Complications of external beam radi- ical and other therapies will be devised to inhibit the
ation therapy include nausea, vomiting, leukopenia, critical pathways of carcinogenesis in these tumors.
cholangitis, gastrointestinal ulceration, and liver abscess.
After completion of external beam radiation therapy, References
patients receive brachytherapy for the tumor by using a 1. Gores GJ. Cholangiocarcinoma: current concepts and insights.
transcatheter loaded with iridium-192 seeds (total dose Hepatology 2003;37:961–969.
of 2000 –3000 cGy). During brachytherapy, patients are 2. Shaib YH, El-Serag HB. The epidemiology of cholangiocarcinoma.
Semin Liver Dis 2004;24:115–125.
treated with 5-FU 225 mg/m2 daily. Subsequently, pa- 3. Khan SA, Davidson BR, Goldin R, Pereira SP, Rosenberg WM,
tients continue to receive the previously described dose of Taylor-Robinson SD, Thillainayagam AV, Thomas HC, Thursz MR,
May 2005 CHOLANGIOCARCINOMA 1665

Wasan H. Guidelines for the diagnosis and treatment of 24. Hadjis NS, Adam A, Gibson R, Blenkharn JI, Benjamin IS, Blum-
cholangiocarcinoma: consensus document. Gut 2002;51 Suppl gart LH. Nonoperative approach to hilar cancer determined by the
6:VI1–VI9. atrophy-hypertrophy complex. Am J Surg 1989;157:395–399.
4. Jemal A, Murray T, Samuels A, Ghafoor A, Ward E, Thun MJ. 25. Gores GJ. Early detection and treatment of cholangiocarcinoma.
Cancer statistics, 2003. CA Cancer J Clin 2003;53:5–26. Liver Transpl 2000;6:S30 –S34.
5. Khan SA, Taylor-Robinson SD, Toledano MB, Beck A, Elliott P, 26. Rumalla A, Baron TH, Leontovich O, Burgart LJ, Yacavone RF,
Thomas HC. Changing international trends in mortality rates for Therneau TM, de Groen PC, Sebo TJ. Improved diagnostic yield of
liver, biliary and pancreatic tumours. J Hepatol 2002;37:806 – endoscopic biliary brush cytology by digital image analysis. Mayo
813. Clin Proc 2001;76:29 –33.
6. Patel T. Increasing incidence and mortality of primary intrahepatic 27. Baron TH, Harewood GC, Rumalla A, Pochron NL, Stadheim LM,
cholangiocarcinoma in the United States. Hepatology 2001;33: Gores GJ, Therneau TM, De Groen PC, Sebo TJ, Salomao Dr, Kipp
1353–1357. BR. A prospective comparison of digital image analysis and rou-
7. Taylor-Robinson SD, Toledano MB, Arora S, Keegan TJ, Har- tine cytology for the identification of malignancy in biliary tract
greaves S, Beck A, Khan SA, Elliott P, Thomas HC. Increase in strictures. Clin Gastroenterol Hepatol 2004;2:214 –219.
mortality rates from intrahepatic cholangiocarcinoma in England 28. Kipp BR, Stadheim LM, Halling SA, Pochron NL, Harmsen S,
and Wales 1968 –1998. Gut 2001;48:816 – 820. Nagorney D, Sebo TJ, Therneau TM, Gores GJ, de Groen PC,
8. Carriaga MT, Henson DE. Liver, gallbladder, extrahepatic bile Baron TH, Levy M, Halling KC, Roberts LR. A comparison of
ducts, and pancreas. Cancer 1995;75:171–190. routine cytology and fluorescence in situ hybridization for the
9. Broome U, Olsson R, Loof L, Bodemar G, Hultcrantz R, Daniels- detection of malignant bile duct strictures. Am J Gastroenterol
son A, Prytz H, Sandberg-Gertzen H, Wallerstedt S, Lindberg G. 2004;99:1675–1681.
Natural history and prognostic factors in 305 Swedish patients 29. Eloubeidi MA, Chen VK, Jhala NC, Eltoum IE, Jhala D, Chhieng
with primary sclerosing cholangitis. Gut 1996;38:610 – 615. DC, Syed SA, Vickers SM, Mel WC. Endoscopic ultrasound-guided
10. Bergquist A, Broome U. Hepatobiliary and extra-hepatic malignan- fine needle aspiration biopsy of suspected cholangiocarcinoma.
cies in primary sclerosing cholangitis. Best Pract Res Clin Gas- Clin Gastroenterol Hepatol 2004;2:209 –213.
troenterol 2001;15:643– 656. 30. Hadjis NS, Collier NA, Blumgart LH. Malignant masquerade at the
11. Bergquist A, Ekbom A, Olsson R, Kornfeldt D, Loof L, Danielsson hilum of the liver. British Journal of Surgery 1985;72:659 – 661.
A, Hultcrantz R, Lindgren S, Prytz H, Sandberg-Gertzen H, Almer 31. Albert MB, Steinberg W, Henry JP. Elevated serum levels of tumor
S, Granath F, Broome U. Hepatic and extrahepatic malignancies marker CA19-9 in acute cholangitis. Dig Dis 1988;33.
in primary sclerosing cholangitis. J Hepatol 2002;36:321–327. 32. Nichols JC, Gores GJ, LaRusso NF, Wiesner RH, Nagorney DM,
12. Simeone DM. Gallbladder and biliary tree: anatomy and structural Ritts RE, Jr. Diagnostic role of serum CA 19-9 for cholangiocar-
anomalies. Lippincott, Williams & Wilkins, 1999. cinoma in patients with primary sclerosing cholangitis. Mayo Clin
13. Chen MF, Jan YY, Wang CS, Hwang TL, Jeng LB, Chen SC, Chen Proc 1993;68:874 – 879.
TJ. A reappraisal of cholangiocarcinoma in patient with hepatoli- 33. Chalasani N, Baluyut A, Ismail A, Zaman A, Sood G, Ghalib R,
thiasis. Cancer 1993;71:2461–2465. McCashland TM, Reddy KR, Zervos X, Anbari MA, Hoen H. Chol-
14. Liu D, Momoi H, Li L, Ishikawa Y, Fukumoto M. Microsatellite angiocarcinoma in patients with primary sclerosing cholangitis: a
instability in thorotrast-induced human intrahepatic cholangiocar- multicenter case-control study. Hepatology 2000;31:7–11.
cinoma. Int J Cancer 2002;102:366 –371. 34. American Joint Committee on Cancer Staging. 6th Edition 2002;
15. Rall J, Chung R. Cholangiocarcinoma and tumors of the liver other Editor Greene FL et al., Springer-Verlag, NY.
than hepatocellular carcinoma. Lippincott-Raven, 1995. 35. Jarnagin WR, Fong Y, DeMatteo RP, Gonen M, Burke EC, Bod-
16. Wong O, Whorton M, Foliart D, Ragland D. An industry-wide niewicz BJ, Youssef BM, Klimstra D, Blumgart LH. Staging, re-
epidemiologic study of vinyl chloride workers, 1942–1982. Am J sectability, and outcome in 225 patients with hilar cholangiocar-
Ind Med 1991;20:317–334. cinoma. Ann Surg 2001;234:507–517; discussion 517–519.
17. Yamamoto S, Kubo S, Hai S, Uenishi T, Yamamoto T, Shuto T, 36. D’Angelica MI et al., Resectable hilar CCA surgical treatment and
Takemura S, Tanaka H, Yamazaki O, Hirohashi K, Tanaka T. long-term outcome. Surg Today 2004;34:885– 890.
Hepatitis C virus infection as a likely etiology of intrahepatic 37. Weber SM, DeMatteo RP, Fong Y, Blumgart LH, Jarnagin WR.
cholangiocarcinoma. Cancer Sci 2004;95:592–595. Staging laparoscopy in patients with extrahepatic biliary carci-
18. Hirohashi K, Uenishi T, Kubo S, Yamamoto T, Tanaka H, Shuto T, noma. Analysis of 100 patients. Ann Surg 2002;235:392–399.
Kinoshita H. Macroscopic types of intrahepatic cholangiocarci- 38. Rea DJ, Munoz-Juarez M, Farnell MB, Donohue JH, Que FG,
noma: clinicopathologic features and surgical outcomes. Hepa- Crownhart B, Larson D, Nagorney D. Major hepatic resection for
togastroenterology 2002;49:326 –329. hilar cholangiocarcinoma. Arch Surg 2004;139:514 –525.
19. Koprowski H, Steplewski Z, Mitchell K, Herlyn M, Herlyn D, Fuhrer 39. Burke EC, Jarnagin WR, Hochwald SN, Pisters PW, Fong Y, Blum-
P. Colorectal carcinoma antigens detected by hybridoma antibod- gart LH. Hilar cholangiocarcinoma: patterns of spread, the impor-
ies. Somatic Cell Genet 1979;5:957–972. tance of hepatic resection for curative operation, and a presur-
20. Magnani JL, Steplewski Z, Koprowski H, Ginsburg V. Identification gical clinical staging system. Ann Surg 1998;228:385–394.
of the gastrointestinal and pancreatic cancer-associated antigen 40. Mizumoto R, Suzuki H. Surgical anatomy of the hepatic hilum with
detected by monoclonal antibody 19-9 in the sera of patients as special reference to the caudate lobe. World J Surg 1988;12:2–
a mucin. Cancer Res 1983;43:5489 –5492. 10.
21. Vestergaard EM, Hein HO, Meyer H, Grunnet N, Jorgensen J, Wolf 41. Clavien PA, Emond J, Vauthey JN, Belghiti J, Chari RS, Strasberg
H, Orntoft TF. Reference values and biological variation for tumor SM. Protection of the liver during hepatic surgery. J Gastrointest
marker CA 19-9 in serum for different Lewis and secretor geno- Surg 2004;8:313–327.
types and evaluation of secretor and Lewis genotyping in a 42. Kitagawa Y, Nagino M, Kamiya J, Uesaka K, Sano T, Yamamoto
Caucasian population. Clin Chem 1999;45:54 – 61. H, Hayakawa N, Nimura Y. Lymph node metastasis from hilar
22. Steinberg W. The clinical utility of the CA 19-9 tumor-associated cholangiocarcinoma: audit of 110 patients who underwent re-
antigen. Am J Gastroenterol 1990;85:350 –355. gional and paraaortic node dissection. Ann Surg 2001;233:385–
23. Manfredi R, Barbaro B, Masselli G, Vecchioli A, Marano P. Mag- 392.
netic resonance imaging of cholangiocarcinoma. Semin Liver Dis 43. Levy MJ, Baron TH, Gostout CJ, Petersen BT, Farnell MB. Pallia-
2004;24:155–164. tion of malignant extrahepatic biliary obstruction with plastic
1666 LAZARIDIS AND GORES GASTROENTEROLOGY Vol. 128, No. 6

versus expandable metal stents: an evidence-based approach. giocellular carcinoma: analysis of a single-center experience and
Clin Gastroenterol Hepatol 2004;2:273–285. review of the literature. Liver Transpl 2001;7:1023–1033.
44. De Palma GD, Pezzullo A, Rega M, Persico M, Patrone F, Mas- 62. Foo ML, Gunderson LL, Bender CE, Buskirk SJ. External radiation
tantuono L, Persico G. Unilateral placement of metallic stents for therapy and transcatheter iridium in the treatment of extrahe-
malignant hilar obstruction: a prospective study. Gastrointest patic bile duct carcinoma. Int J Radiat Oncol Biol Phys 1997;39:
Endosc 2003;58:50 –53. 929 –935.
45. Hintze RE, Abou-Rebyeh H, Adler A, Veltzke-Schlieker W, Felix R, 63. Sudan D, DeRoover A, Chinnakotla S, Fox I, Shaw B, Jr., Mc-
Wiedenmann B. Magnetic resonance cholangiopancreatography- Cashland T, Sorrell M, Tempero M, Langnas A. Radiochemo-
guided unilateral endoscopic stent placement for Klatskin tu- therapy and transplantation allow long-term survival for nonre-
mors. Gastrointest Endosc 2001;53:40 – 46. sectable hilar cholangiocarcinoma. Am J Transplant 2002;
46. De Palma GD, Galloro G, Siciliano S, Iovino P, Catanzano C. 2:774 –779.
Unilateral versus bilateral endoscopic hepatic duct drainage in 64. Ashida K, Terada T, Kitamura Y, Kaibara N. Expression of E-
patients with malignant hilar biliary obstruction: results of a cadherin, alpha-catenin, beta-catenin, and CD44 (standard and
prospective, randomized, and controlled study. Gastrointest En- variant isoforms) in human cholangiocarcinoma: an immunohis-
dosc 2001;53:547–553. tochemical study. Hepatology 1998;27:974 –982.
47. Rumalla A, Baron TH, Wang KK, Gores GJ, Stadheim LM, de 65. Terada T, Okada Y, Nakanuma Y. Expression of immunoreactive
Groen PC. Endoscopic application of photodynamic therapy for matrix metalloproteinases and tissue inhibitors of matrix metal-
cholangiocarcinoma. Gastrointest Endosc 2001;53:500 –504. loproteinases in human normal livers and primary liver tumors.
48. Berr F. Photodynamic therapy for cholangiocarcinoma. Semin Hepatology 1996;23:1341–1344.
Liver Dis 2004;24:177–187. 66. Lavaissiere L, Jia S, Nishiyama M, de la Monte S, Stern AM,
49. Ortner M, Caca K, Berr F, Liebetruth J, Mansmann U, Huster D, Wands JR, Friedman PA. Overexpression of human aspartyl
Voderholzer W, Schachschal G, Mossner J, Lochs H. Successful (asparaginyl)beta-hydroxylase in hepatocellular carcinoma and
photodynamic therapy for nonresectable cholangiocarcinoma: a cholangiocarcinoma. J Clin Invest 1996;98:1313–1323.
randomized prospective study. Gastroenterology 2003;125: 67. Tanaka S, Sugimachi K, Kameyama T, Maehara S, Shirabe K,
1355–1363. Shimada M, Wands JR, Maehara Y. Human WISP1v, a member of
50. Gores GJ. A spotlight on cholangiocarcinoma. Gastroenterology the CCN family, is associated with invasive cholangiocarcinoma.
2003;125:1536 –1538. Hepatology 2003;37:1122–1129.
51. Montemaggi P, Costamagna G, Dobelbower RR, Cellini N, Mor- 68. Benckert C, Jonas S, Cramer T, Von Marschall Z, Schafer G,
ganti AG, Mutignani M, Perri V, Brizi G, Marano P. Intraluminal Peters M, Wagner K, Radke C, Wiedenmann B, Neuhaus P,
brachytherapy in the treatment of pancreas and bile duct carci- Hocker M, Rosewicz S. Transforming growth factor beta 1 stim-
noma. Int J Radiat Oncol Biol Phys 1995;32:437– 443. ulates vascular endothelial growth factor gene transcription in
52. Bruha R, Petrtyl J, Kubecova M, Marecek Z, Dufek V, Urbanek P, human cholangiocellular carcinoma cells. Cancer Res 2003;63:
Kodadova J, Chodounsky Z. Intraluminal brachytherapy and self- 1083–1092.
expandable stents in nonresectable biliary malignancies—the 69. Tannapfel A, Benicke M, Katalinic A, Uhlmann D, Kockerling F,
question of long-term palliation. Hepatogastroenterology 2001; Hauss J, Wittekind C. Frequency of p16-INK4A alterations and
48:631– 637. k-ras mutations in intrahepatic cholangiocarcinoma of the liver.
53. Gerhards MF, van Gulik TM, Gonzalez D, Rauws EA, Gouma DJ. Gut 2000;47:721–727.
Results of postoperative radiotherapy for resectable hilar cholan- 70. Kang YK, Kim WH, Lee HW, Lee HK, Kim YI. Mutation of p53 and
giocarcinoma. World J Surg 2003;27:173–179. K-ras and loss of heterozygosity of APC in intrahepatic cholangio-
54. Pichlmayr R, Weimann A, Klempnauer J, Oldhafer KJ, Maschek H, carcinoma. Lab Invest 1999;79:477– 483.
Tusch G, Ringe B. Surgical treatment in proximal bile duct cancer. 71. Lai GH, Radaeva S, Nakamura T, Sirica AE. Unique epithelial cell
A single-center experience. Ann Surg 1996;224:628 – 638. production of hepatocyte growth factor/scatter factor by putative
55. Goldstein RM, Stone M, Tillery GW, Senzer N, Levy M, Husberg precancerous intestinal metaplasias and associated “intestinal-
BS, Gonwa T, Klintmalm G. Is liver transplantation indicated for type” biliary cancer chemically induced in rat liver. Hepatology
cholangiocarcinoma? Am J Surg 1993;166:768 –771; discus- 2000;31:1257–1265.
sion 771–772. 72. Furubo S, Hadara K, Shimonishi T, Katayanagi K, Tsui W, Na-
56. Iwatsuki S, Todo S, Marsh JW, Madariaga JR, Lee RG, Dvorchik I, kanuma Y. Protein expression and genetic alterations of p53 and
Fung JJ, Starzl TE. Treatment of hilar cholangiocarcinoma ras in intrahepatic cholangiocarcinoma. Histopathology 1999;
(Klatskin tumors) with hepatic resection or transplantation. J Am 35:230 –240.
Coll Surg 1998;187:358 –364. 73. Itoi T, Shinohara Y, Takeda K, Takei K, Ohno H, Ohyashiki K,
57. Jeyarajah DR, Klintmalm GB. Is liver transplantation indicated for Yahata N, Ebihara Y, Saito T. Detection of telomerase activity in
cholangiocarcinoma? Pancreat Surg 1998;5:48 –51. biopsy specimens for diagnosis of biliary tract cancers. Gastro-
58. Loinaz C, Abradelo M, Gomez R, Colina F, Rey P, Ochando F, intest Endosc 2000;52:380 –386.
Canete AR, Gonzalez-Pinto I, Jimenez C, Garcia I, Gonzalez EM. 74. Itoi T, Shinohara Y, Takeda K, Nakamura K, Shimizu M, Ohyashiki
Liver transplantation and incidental primary liver tumors. Trans- K, Hisatomi H, Nakano H, Moriyasu F. Detection of telomerase
plant Proc 1998;30:3301–3302. reverse transcriptase mRNA in biopsy specimens and bile for
59. Bokemeyer C, Kollmannsberger C, Oettle H, Kanz L. [Current diagnosis of biliary tract cancers. Int J Mol Med 2001;7:281–
aspects of chemotherapy of metastatic pancreatic and biliary 287.
tract carcinomas]. Schweiz Rundsch Med Prax 2000;89:1545– 75. Sugawara H, Yasoshima M, Katayanagi K, Kono N, Watanabe Y,
1552. Harada K, Nakanuma Y. Relationship between interleukin-6 and
60. Jonas S, Kling N, Guckelberger O, Keck H, Bechstein WO, Neu- proliferation and differentiation in cholangiocarcinoma. Histopa-
haus P. Orthotopic liver transplantation after extended bile duct thology 1998;33:145–153.
resection as treatment of hilar cholangiocarcinoma. First long- 76. Yokomuro S, Tsuji H, Lunz JG, 3rd, Sakamoto T, Ezure T, Murase
term results. Transplant Int 1998;11:S206 –S208. N, Demetris AJ. Growth control of human biliary epithelial cells by
61. Shimoda M, Farmer DG, Colquhoun SD, Rosove M, Ghobrial RM, interleukin 6, hepatocyte growth factor, transforming growth fac-
Yersiz H, Chen P, Busuttil RW. Liver transplantation for cholan- tor beta1, and activin A: comparison of a cholangiocarcinoma cell
May 2005 CHOLANGIOCARCINOMA 1667

line with primary cultures and non-neoplastic biliary epithelial 83. Okaro AC, Deery AR, Hutchins RR, Davidson BR. The expression
cells. Hepatology 2000;32:26 –35. of antiapoptotic proteins Bcl-2, Bcl-X(L), and Mcl-1 in benign,
77. Aishima SI, Taguchi KI, Sugimachi K, Shimada M, Tsuneyoshi M. dysplastic, and malignant biliary epithelium. J Clin Pathol 2001;
c-erbB-2 and c-Met expression relates to cholangiocarcinogen- 54:927–932.
esis and progression of intrahepatic cholangiocarcinoma. Histo- 84. Yoon JH, Werneburg NW, Higuchi H, Canbay AE, Kaufmann SH,
pathology 2002;40:269 –278. Akgul C, Edwards SW, Gores GJ. Bile acids inhibit Mcl-1 protein
78. Chariyalertsak CS, Sirikulchayanonta V, Mayer D, Kopp-Schnei- turnover via an epidermal growth factor receptor/Raf-1-depen-
der A, Furstenberger G, Marks F, Muller-Decker K. Aberrant cy- dent mechanism. Cancer Res 2002;62:6500 – 6505.
clooxygenase isoenzyme expression in human intrahepatic chol- 85. Ince N, de la Monte SM, Wands JR. Overexpression of human
angiocarcinoma. Gut 2001;48:80 – 86. aspartyl (asparaginyl) beta-hydroxylase is associated with malig-
79. Endo K, Yoon BI, Pairojkul C, Demetris AJ, Sirica AE. ERBB-2 nant transformation. Cancer Res 2000;60:1261–1266.
overexpression and cyclooxygenase-2 up-regulation in human 86. Maeda T, Sepe P, Lahousse S, Tamaki S, Enjoji M, Wands JR, de
cholangiocarcinoma and risk conditions. Hepatology 2002;36: la Monte SM. Antisense oligodeoxynucleotides directed against
439 – 450. aspartyl (asparaginyl) beta-hydroxylase suppress migration of
80. Que FG, Phan VA, Phan VH, Celli A, Batts K, LaRusso NF, Gores cholangiocarcinoma cells. J Hepatol 2003;38:615– 622.
GJ. Cholangiocarcinomas express Fas ligand and disable the Fas
receptor. Hepatology 1999;30:1398 –1404.
81. Torok NJ, Higuchi H, Bronk S, Gores GJ. Nitric oxide inhibits
apoptosis downstream of cytochrome C release by nitrosylating Received December 9, 2004. Accepted February 15, 2005.
caspase 9. Cancer Res 2002;62:1648 –1653. Address requests for reprints to: Gregory J. Gores, MD, Center for
82. Harnois DM, Que FG, Celli A, LaRusso NF, Gores GJ. Bcl-2 is Basic Research in Digestive Diseases, Mayo Clinic College of Medicine,
overexpressed and alters the threshold for apoptosis in a chol- 200 First Street SW, Rochester, Minnesota 55905. e-mail:
angiocarcinoma cell line. Hepatology 1997;26:884 – 890. [email protected]; fax: (507) 284-0762.