Clinical Gastroenterology and Hepatology 2015;13:2152–2165

Primary Sclerosing Cholangitis as a Premalignant Biliary TractDisease: Surveillance and Management

Sumera Rizvi, John E. Eaton, and Gregory J. Gores

Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota

Primary sclerosing cholangitis (PSC) is a premalignantbiliary tract disease that confers a significant risk for thedevelopment of cholangiocarcinoma (CCA). The chronicbiliary tract inflammation of PSC promotes pro-oncogenicprocesses such as cellular proliferation, induction of DNAdamage, alterations of the extracellular matrix, andcholestasis. The diagnosis of malignancy in PSC can bechallenging because inflammation-related changes in PSCmay produce dominant biliary tract strictures mimickingCCA. Biomarkers such as detection of methylated genes inbiliary specimens represent noninvasive techniques thatmay discriminate malignant biliary ductal changes fromPSC strictures. However, conventional cytology andadvanced cytologic techniques such as fluorescence in situhybridization for polysomy remain the practice standardfor diagnosing CCA in PSC. Curative treatment options ofmalignancy arising in PSC are limited. For a subset of pa-tients selected by using stringent criteria, liver trans-plantation after neoadjuvant chemoradiation is a potentialcurative therapy. However, most patients have advancedmalignancy at the time of diagnosis. Advances directed atidentifying high-risk patients, early cancer detection, anddevelopment of chemopreventive strategies will be essen-tial to better manage the cancer risk in this premalignantdisease. A better understanding of dysplasia definition andespecially its natural history is also needed in this disease.Herein, we review recent developments in our under-standing of the risk factors, pathogenic mechanisms of PSCassociated with CCA, as well as advances in early detectionand therapies.

Keywords: Biliary Dysplasia; Chemoprevention; Interleukin-6;Perihilar Cholangiocarcinoma.

Chronic inflammation and cholestasis in primarysclerosing cholangitis (PSC) promote carcinogen-

esis of the biliary tract by fostering pro-survival signaling

Abbreviations used in this paper: AUC, area under the curve; CA 19-9,carbohydrate antigen 19-9; CCA, cholangiocarcinoma; COX-2, cyclo-oxygenase-2; CT, computed tomography; EGFR, epidermal growth factorreceptor; ERC, endoscopic retrograde cholangiography; ERCP, endo-scopic retrograde cholangiopancreatography; 18F-FDG, fluorine-18-fluorodeoxyglucose; FGFR, fibroblast growth factor receptor; FISH,fluorescence in situ hybridization; FUT, fucosyltransferase; IDUS, intra-ductal ultrasound; IL, interleukin; iNOS, inducible nitric oxide synthase;JAK, Janus kinase; MAPK, mitogen-activated protein kinase; MRC,magnetic resonance cholangiography; MRI, magnetic resonance imaging;NO, nitric oxide; pCCA, perihilar cholangiocarcinoma; PET, positron

pathways and development of genetic aberrations.1 In-flammatory pathways are not only strongly associatedwith carcinogenesis but are often retained by cholangio-carcinoma (CCA) cells to facilitate tumor invasion andmigration, an observation that is not surprising becauseinflammation is critical in tissue remodeling and cancermimics dysregulated tissue remodeling. The risk of CCAamong patients with PSC is increased 400-fold whencompared with the general population.2 CCA remainsone of the leading causes of liver-related mortality inthis population.3,4 Perihilar cholangiocarcinoma (pCCA)is the CCA subtype most often seen in the context ofPSC.5 Although there have been reports of a dysplasia-carcinoma sequence in PSC, the characteristics of prema-lignant lesions and prevalence of biliary dysplasia in PSCare incompletely understood.6–8 These facts result inseveral different questions for the clinician caring for thePSC patient. Is my PSC patient at risk for developingCCA, and if so, what is the surveillance strategy? If I detectdysplasia, what are the therapeutic strategies? If my pa-tient has CCA, what are the treatment options? We reviewrecent advances addressing these pertinent and difficultquestions. We also provide our perspectives reflectingon these advances and our vision for research-driven ad-vances to answer these questions. In addition to CCA, pa-tients with PSC remain at risk for other malignancies(Figure 1). PSC associated malignancies have been pub-lished elsewhere and will not be discussed in this review.9

Epidemiology and Risk Factors

Population-based studies suggest that the annual riskfor CCA is approximately 2%, with 10-year and 30-yearcumulative incidence of 6%–11% and 20%, respec-tively.2,3,10 The development of CCA can be the heralding

emission tomography; PSC, primary sclerosing cholangitis; STAT, signaltransducer and activator of transcription; SUV, standardized uptake value;TRAIL, tumor necrosis factor–associated apoptosis-inducing ligand; VOC,volatile organic compound; YAP, yes-associated protein.

Most current article

© 2015 by the AGA Institute1542-3565/$36.00

http://dx.doi.org/10.1016/j.cgh.2015.05.035

Figure 1.Malignancies associated with PSC.

November 2015 PSC and CCA 2153

event that brings patients with undiagnosed PSC to theattention of clinicians. For example, population-basedstudies have reported that 27%–37% of biliary cancersare detected within the first year of a diagnosis ofPSC.2,10 Consequently, it is important to have a high in-dex of suspicion for CCA around the time of the diagnosisof PSC. However, clinicians should maintain vigilancethroughout the disease course because the majority ofbiliary cancers will develop more than a year after theinitial PSC diagnosis.

Our knowledge concerning risk factors for CCA in PSCis limited. An older age at the time of PSC diagnosis,history of colorectal cancer or dysplasia, longer durationof inflammatory bowel disease, variceal hemorrhage,smoking, and alcohol consumption have been reported tohave a positive association with biliary cancer andPSC.2,4,10–15 An older age at the time of PSC diagnosis anda history of colorectal cancer have been implicated inmore than 1 study.2,4,10,15 However, the evidence isinsufficient to use these factors to risk-stratify PSC

2154 Rizvi et al Clinical Gastroenterology and Hepatology Vol. 13, No. 12

patients who are more likely to benefit from a screeningprogram.

There are several PSC subgroups that appear to havea lower risk of CCA. Among longitudinal studies of smallduct PSC, biliary cancer has not been reported unlessthere is progression to large duct PSC.16 Hence, we donot routinely screen for CCA among asymptomatic pa-tients with small duct PSC. In addition, the risk of CCAamong pediatric patients appears to be lower whencompared with their adult counterparts. For example, ina case series conducted during a 25-year period at alarge tertiary institution that is a referral center for CCAand PSC, there were no cases of biliary cancer amongpediatric PSC patients younger than 18 years old.17

Similarly, among 29 pediatric PSC patients from apopulation-based study in North America, only 1 person(3%) was diagnosed with CCA before the age of 18years.18 Consequently, routine CCA screening of pediatricPSC patients would likely be low-yield. There is emergingevidence to suggest that patients with PSC and a loweralkaline phosphatase level are more likely to haveimproved outcomes.19 Although a study found no casesof CCA among PSC patients with a persistent reduction inserum alkaline phosphatase less than 1.5 times the upperlimit of normal, this observation was not confirmed in asubsequent study.19,20 Therefore, there is insufficientevidence to suggest alkaline phosphatase could be usedas a serologic marker to risk-stratify PSC patients whoshould undergo routine CCA screening.

In our opinion, it is likely that various geneticallydriven subsets of PSC exist, some of which are suscep-tible to CCA and some that are not. Hopefully, futurestudies of PSC patients that use a precision/individual-ized medicine genetic approach will address this topic.

Mechanisms of Inflammation-inducedBiliary Tract Cancer

Although the precise etiology of PSC is ambiguous,recent studies have highlighted an immune-mediatedbasis. In approximately 25% of cases, PSC is seen inthe context of at least one other autoimmune diseaseoutside the gastrointestinal tract.21 PSC has robust as-sociations within the HLA complex, which providesfurther credence to the notion that immunity plays a rolein its pathogenesis.22 Haplotypes with well-establishedassociations with PSC include HLA-DRB1*1301-DQB1*0603, HLA-A1-B8-DRB1*0301-DQB1*0201, andHLA-DRB1*1501-DQB1*0602.23,24 Overall, there are 16known genome-wide significant loci in PSC, including theHLA complex on chromosome 6.25–27 Of these 16 loci, 12were recently identified by using the Immunochip, agenotyping array with marker coverage across a numberof loci from 12 immune-mediated diseases.25 A strongerassociation with PSC than with inflammatory boweldisease is seen with 6 of these 12 loci, which denotes thatalthough there is some overlap, the genetic architecture

for these 2 diseases is distinct.25 To date, no germlineoncogenic genetic mutations have been described in PSC.These findings reinforce the association between PSCand other immune-based diseases, highlight an immune-mediated pathophysiological basis for PSC, and suggestCCA development is a secondary event related toinflammation and not a primary genetic process.25

Biliary tract cancer is a prototype of malignanciesoccurring in the context of inflammation. PSC promoteschronic inflammation of the biliary tree, predisposing tothe development of CCA.28 Chronic inflammation facili-tates oncogenesis via induction of DNA damage, promo-tion of cellular proliferation, and inhibition of apoptosis.For example, inflammatory cytokines activate induciblenitric oxide (iNOS) with excess production of nitric oxide(NO) and consequent nitrosative stress.29 iNOS is notpresent in normal biliary epithelia, but its expression hasbeen demonstrated in PSC as well as CCA. Oxidative DNAlesions are the primary mechanism of DNA damage ininflammation, and the most abundant oxidative DNAlesion is 8-oxodeoxyguanine. These lesions are typicallyexcised by DNA repair processes. NO inhibits 8-oxodeoxyguanine base excision DNA repair processeswith resultant accumulation of this oxidative lesion inPSC.30 The failure to repair 8-oxodeoxyguanine is muta-genic and fosters cancer development and progression.30

Thus, NO has an integral role in mediating DNA damage inbiliary tract inflammation and carcinogenesis.30

Sublethal proapoptotic signaling has recently beenmechanistically linked to the genesis of chromosalinstability, a hallmark of cancer.31,32 The proapoptoticdeath receptor agonist, tumor necrosis factor–relatedapoptosis-inducing ligand (TRAIL), has been implicatedin both PSC and cancer development, presumably by thismechanism of inducing chromosomal instability.33,34 Therole of TRAIL and sublethal cell injury in the develop-ment of PSC-associated CCA merits further study.

Cholestasis occurring in the setting of PSC also con-fers an enhanced risk of CCA development. Bile acidsactivate receptor tyrosine kinases such as epidermalgrowth factor receptor (EGFR). Sustained activation ofEGFR in CCA mediates proliferation and inducesexpression of cyclooxygenase-2 (COX-2) via a mitogen-activated protein kinase (MAPK)-dependent mecha-nism.35,36 COX-2 is also induced by various inflammatorycytokines and contributes to carcinogenesis by promot-ing proliferation and angiogenesis and inhibitingapoptosis.35 Oxysterols, oxygenated derivatives ofcholesterol, are abundant in the bile of patients withbiliary tract inflammation.37,38 Stabilization of COX-2expression by oxysterols has been implicated in thegenesis and promotion of CCA.39 Oxysterols also serve asactivators of the hedgehog signaling pathway, a devel-opmental pathway implicated in CCA development.40–43

A recently developed oncogene-driven murine modelof CCA highlights the role of inflammatory cytokines inCCA oncogenesis.44 In this model, biliary transduction ofconstitutively active Akt and yes-associated protein (YAP)

November 2015 PSC and CCA 2155

coupled with lobar bile duct ligation and systemic inter-leukin (IL)-33 administration resulted in the developmentof CCA.44 IL33 promotes downstream activation of IL6signaling in this model. IL33, an IL1 family member, is aknown biliary mitogen that promotes inflammation andfibrosis in the biliary tract.45,46 Inflammatory stimulipromote production of IL6, an inflammatory cytokine, bycholangiocytes. Activation of IL6/signal transducer andactivator of transcription (STAT)3 signaling, in turn, pro-motes growth stimulation of malignant cholangiocytes byactivation of p38 or p44/42 MAPK signaling pathways inan autocrine or paracrine manner.47 A recent article alsorelates IL6 signaling to YAP activation inmucosal injury ofthe intestine, and perhaps this pathway is also relevant tothe biliary tree48; YAP-mediated epithelial regenerationcould also be an initiator of carcinogenesis if sustained andunrelenting.

These mechanistic insights linking chronic inflam-mation to carcinogenesis provide potential therapeuticavenues for chemoprevention in PSC. For example, in-hibitors of STAT3 and Janus kinases (JAKs), which acti-vate STAT3 downstream of IL6, are in clinicaldevelopment.1 JAK kinase inhibitors are currently beingtested for the treatment of inflammatory bowel disease,which coexists in 85% of PSC patients.49 Thus, the use ofJAK inhibitors in these patients may come to fruition,permitting an assessment of their chemopreventive ef-fects in this patient population. In addition to JAK-STATinhibitors, caspase, iNOS, COX-2, and Hippo pathwayantagonists may also be chemopreventive.

Cholangiocarcinoma Screeningand Surveillance

CCA is divided into pCCA, intrahepatic CCA, anddistal subtypes that are based on the anatomic locationof the tumor within the biliary tree.28 pCCA is not onlythe most common subtype overall but also the subtypeprimarily seen in the context of PSC.28 pCCA frequentlypresents with an obstructive biliary stricture withoutthe presence of a mass on cross-sectional imaging(Figure 2A). Inflammatory/fibrotic obstructive biliarystrictures in PSC, so-called dominant strictures, mimicmalignant strictures, and hence, distinguishing a benign

Figure 2. Imaging featuresof pCCA in PSC. (A) ERCimage of dominant com-mon hepatic duct stricture(white arrow) in PSC pa-tient with periductal infil-trating pCCA. (B) MRIimage of perihilar mass(white arrow) with resultantbiliary obstruction.

inflammatory/fibrotic stricture in PSC from a malignantstricture can be quite challenging and is not possible byusing noninvasive diagnostic modalities such as mag-netic resonance cholangiography (MRC). Endoscopicretrograde cholangiography (ERC) is essential in thissetting because it has diagnostic and therapeutic utility.

Radiologic Imaging

Imaging plays a central role in the detection of CCA,and abnormalities seen on imaging often trigger addi-tional investigations aimed at establishing a diagnosis ofbiliary cancer. Ultrasonography, computed tomography(CT), positron emission tomography (PET), and magneticresonance imaging (MRI)/MRC have been investigated asdiagnostic imaging modalities.

Although inexpensive and noninvasive, ultrasonog-raphy may only delineate intrahepatic ductal dilationwithout providing further detailed information in theabsence of a mass lesion.50 Although a CT scan canreadily characterize mass lesions and investigateinvasion into adjacent structures or metastases, itssensitivity and specificity for CCA detection in PSCare 75% and 85%, respectively.50,51 A meta-analysisexamined 23 studies and the ability of fluorine-18-fluorodeoxyglucose (18F-FDG)-PET or PET/CT todetect CCA. The pooled sensitivity and specificity forCCA among those with and without PSC were 81% and82%, respectively.52 Notably, false-positive results canoccur as a result of inflammation associated with PSC.53

A recent study examined the role of 18F-FDG-PET/CTand the 18FDG uptake values, normalized to the back-ground liver, at 180 minutes (standard uptake value[SUV]max/liver) among 70 patients with PSC and adominant stricture. There were 9 cases of CCA in thecohort, and 55% did not have definitive features ofcancer on MRI. An analysis of the 18FDG uptake showedthat SUVmax/liver quotient of 3.3 was able to distinguishbetween CCA and benign strictures with sensitivity andspecificity of 89% and 92%, respectively, whereas aquotient of less than 2.4 excluded CCA (sensitivity100%, specificity 78%).54 Although these findings war-rant confirmation in a larger cohort, these results sug-gest that the use of 18F-FDG-PET/CT may be helpful in a

2156 Rizvi et al Clinical Gastroenterology and Hepatology Vol. 13, No. 12

subset of patients, particularly if there is a persistentconcern for CCA despite a negative MRI/MRC andnegative biliary brushings. Although 18F-FDG-PET/CTand the evolving 18F-FDG-PET/MRI technologies all holdpromise for the diagnosis of PSC-associated CCA, manyconfirmed cases of CCA are 18F-FDG-PET negative. Theultimate role of this approach for the diagnosis of CCAremains to be clarified.

Of these imaging modalities, MRI/MRC is the diag-nostic imaging method of choice. A mass lesion withvenous phase enhancement is very specific for CCA(Figure 2B).52 However, such definitive features arefrequently absent. More commonly, CCA infiltrates alongthe biliary tree, leading to ductal narrowing, thickening,and dilation. When present, CCA typically has maximalenhancement on delayed phases.50 Such findings can besubtle, and distinguishing benign from malignant stric-tures in PSC is challenging. Radiographic findings thatshould raise concern are the development of thickened ornodular bile ducts or a new dominant stricture.52 Indeed,inflammatory/obstructive strictures have been reportedin up to 50% of patients with PSC, which leads tosymptoms in approximately 10%–30% of individuals.One-fourth of the so-called dominant strictures are ma-lignant.53,55 However, CCA may be detected in individualswithout an obstructive stricture, so their absence does notexclude malignancy.56 The presence of perihilar lymph-adenopathy by itself should not raise concern becausethis is commonly seen in patients with PSC.57

Carbohydrate Antigen 19-9

The most widely studied serum biomarker for CCA inPSC is carbohydrate antigen 19-9 (CA 19-9). The syn-thesis and ability to express CA 19-9 are dependent onfucosyltransferase-2 and 3 (FUT-2 and FUT-3) activity,and individuals who lack FUT-3 activity (Lewis antigennegative) are unable to express CA 19-9 (approximately7% of the population).9,58 By using a CA 19-9 cutoff of129 U/mL, the sensitivity and specificity for CCAdetection were 79% and 99%, respectively, and using athreshold of 100 U/mL yielded a similar diagnosticperformance. However, only advanced cases of CCA weredetected by either cutoff. Furthermore, examiningchanges in CA 19-9 over time did not add to the diag-nostic yield of a single CA 19-9 value.59 However, otherstudies have called into question the high specificity of aCA 19-9 value greater than 129 U/mL. For example, 2studies have reported that one-third of PSC patients witha CA 19-9 greater than 129 U/mL do not have under-lying CCA.60,61 Recognizing this assay’s limitations, au-thors have sought to improve the diagnosticperformance of CA 19-9 by using FUT2/3 genotypespecific CA 19-9 thresholds. This method has beenshown to improve the sensitivity and decrease thenumber of false-positive test results by 43% among alarge cohort of PSC patients.58

Biliary Cytology

Conventional biliary cytology can be classified into 5categories: nondiagnostic, normal, atypical, suspicious, orpositive for adenocarcinoma.62 The primary advantage ofbiliary cytology is its high specificity (97%–100%) whenadenocarcinoma is detected.63,64 However, suspiciouscytology also represents a concerning finding. Forexample, 34%–42% of PSC patients without a masslesion and suspicious cytology may ultimately be diag-nosed with CCA, and suspicious cytology is an indepen-dent predictor for the development of biliary cancer.56,65

Atypical cytology is frequently encountered in PSC. Thisis primarily due to the presence of biliary inflammationand by itself should not raise concern.56 The chiefdisadvantage of biliary cytology is its limited sensitivity(43%) and potential for false-negative results.63 This is aresult of the desmoplastic, paucicellular nature of CCA,which can reside in areas that are difficult to access.Therefore, the absence of a positive cytology does notrule out malignancy.

Overview of Fluorescence In Situ Hybridization

Because of the limitations of biliary cytology, fluo-rescence in situ hybridization (FISH) has been used as asecond-generation test to enhance clinicians’ ability torisk-stratify PSC patients. The FISH assay examined inpatients with PSC uses 3 centromeric probes that targetchromosomes 3, 7, and 17 and a locus specific probe to9p21 from samples obtained by biliary brushings. FISHdetects abnormal gains or losses of chromosomes(aneusomy), and the results can be categorized asnormal, trisomy (10 or more cells with 3 copies ofchromosome 7 and 2 or fewer copies of the other 3probes), tetrasomy (10 or more cells show 4 copies of allprobes), or polysomy (5 or more cells show gains of 2 ormore of the 4 probes).62 FISH polysomy indicatingduplication of more than 1 chromosome is a marker forchromosomal instability, a hallmark of cancer.

When compared with a normal FISH result, trisomy/tetrasomy are not independent predictors of CCA. Incontrast, polysomy is strongly associated with a diag-nosis of CCA.56,65–67 A meta-analysis examined the per-formance of FISH testing among 690 PSC patients, andthe pooled sensitivity, specificity, and positive andnegative likelihood ratios for polysomy and CCA were51%, 93%, 6.8, and 0.6, respectively.68 This study did notdistinguish between other factors that can increase theprobability of biliary cancer when polysomy is present,and it is important for FISH results to be interpreted inthe clinical context of each individual patient.

Polysomy associated with a dominant stricture orelevated carbohydrate antigen 19-9. Polysomy in thepresence of a dominant stricture may increase theprobability of cancer. For example, in a study thatincluded 235 PSC patients (with and without a mass

November 2015 PSC and CCA 2157

lesion), 55% of patients with polysomy were diagnosedwith biliary cancer. However, if a dominant stricture waspresent in the setting of polysomy, 73% of PSC patientswere diagnosed with CCA.66

An elevated CA 19-9 in the setting of polysomy is alsoconcerning for cancer. For example, among PSC patients(without a mass lesion at baseline) suspected of CCAwith equivocal cytology (atypical or suspicious) andpolysomy, 100% of patients with CA 19-9 �129 U/mLwere diagnosed with CCA within 3 years compared with24% of subjects with CA 19-9 <129 U/mL.65 In a sepa-rate study, 71% of patients with polysomy (regardless ofcytology) and CA 19-9 �129 U/mL who lacked a masslesion on imaging were diagnosed with CCA comparedwith 37% of individuals with CA 19-9 <129 U/mL.56

When polysomy was incorporated in the multivariableanalyses in both of these studies, CA 19-9 was no longerstatistically significant. However, because of the highproportion of individuals with elevated CA 19-9 andpolysomy who were diagnosed with cancer, an increasedCA 19-9 should raise concern.

Serial or multifocal polysomy. When compared withthe PSC population at large, those with polysomy in theabsence of definitive features of CCA at the time of theinitial assessment (ie, lack a positive cytology or defini-tive mass lesion) are likely rare (estimated to be lessthan 5% of total PSC population seen at 1 institutionduring a 7-year period).56 Although uncommon, thisimportant subgroup of patients represents a diagnosticchallenge. Consequently, a study examined the perfor-mance of polysomy when it is detected on subsequentexaminations (serial polysomy) among those withoutdefinitive radiographic features of biliary malignancy.The 3-year cumulative incidence of CCA was 75% amongthose with serial polysomy and 18% of subjects withnon-serial polysomy (polysomy detected only once atthe index examination). After 1 year of follow-up theincidence of CCA did not increase among those with non-serial polysomy.67 These results reinforce the impor-tance of repeating an ERC with brushings for cytologyand FISH if polysomy is detected in the absence ofdefinitive features of CCA.

The largest study that examined FISH in PSC included371 patients without a mass lesion on imaging andinvestigated the natural history of polysomy when it wasdetected in multiple areas of the biliary tree (multifocalpolysomy) when compared with unifocal polysomy, se-rial polysomy, and other FISH subtypes. In the adjustedanalysis, multifocal polysomy was the strongest predic-tor of a diagnosis of CCA (regardless of whether adominant stricture or serial polysomy was present), andthe 1-year and 3-year cumulative incidence of CCAamong those with multifocal polysomy was 65% and83%, respectively. Among those with polysomy andpositive cytology, 71% had polysomy detected at anotherregion of the biliary tree where adenocarcinoma was notdetected by routine cytology. These findings reinforcethe role of FISH and indirectly support the hypothesis

that CCA in PSC may arise from a field defect beyond theprimary site of malignancy. It also suggests that it may beimportant to brush multiple areas of the biliary tree,regardless of whether and where a dominant stricture islocated, and place the specimens in separate vials whenCCA is suspected. However, individuals with unifocalpolysomy should not be ignored because this was also arisk factor for CCA when compared with a normal FISHresult.56

Advanced Endoscopic BiliaryImaging Techniques

Because conventional cytology has limited sensitivityand propensity for false-negative results and the avail-ability of FISH is limited, advanced techniques forendoscopic biliary imaging have become a recentresearch focus. However, these techniques have largelybeen examined in non-PSC patients. These techniquesinclude Spyglass Spyscope (Boston Scientific, Marl-borough, MA), which enables direct endoscopic visuali-zation of bile ducts and directed biopsies of suspiciouslesions. In a retrospective analysis, Spyglass Spyscopehad 77% accuracy in detection of CCA in patients withdiagnostic uncertainty after conventional cytology ofendoscopically obtained biliary brushings and endo-scopic ultrasound–fine-needle aspiration.69 Intraductalultrasound (IDUS) of the biliary system provides evalu-ation of the periductal tissues during endoscopic retro-grade cholangiopancreatography (ERCP). In aprospective comparative analysis, ERCP supplementedwith IDUS allowed differentiation of benign from malig-nant biliary strictures in 88% of patients (n ¼ 33).70

Moreover, ERCP plus IDUS had significantly greater ac-curacy in differentiating between malignant and benignstrictures compared with magnetic resonance chol-angiopancreatography.70 Another promising diagnosticapproach that can supplement conventional ERCP isprobe-based confocal laser endomicroscopy. Mucosalimaging was conducted in 14 patients by using a confocallaser scanning miniprobe introduced via the accessorychannel of a conventional cholangioscope.71 Subse-quently, targeted biopsy specimens were obtained undervisual cholangioscopic guidance in all patients.71 Bydetecting a specific pattern of neovascularization,confocal laser microscopy enabled prediction of CCAwith sensitivity of 83% and specificity of 88%.71 Incomparison, standard histopathology had sensitivity andspecificity of 50% and 100%, respectively.71 Finally, wenote that narrow-band imaging in combination withcholangioscopy is disappointing in its ability to identifybiliary dysplasia in the setting of PSC.72 These advancedbiliary imaging techniques hold promise; however, theirutility in PSC patients, a patient population more com-plex than those with de novo CCA, remains unstudied.Moreover, in our own experience, technical failure withpoor visualization of the PSC biliary tract is common

2158 Rizvi et al Clinical Gastroenterology and Hepatology Vol. 13, No. 12

with Spyglass Spyscope technology, diminishing sensi-tivity and specificity on an “intent-to-diagnose”assessment.

Approach to Biliary High-grade Dysplasia

Presence of high-grade dysplasia of the biliary tractmay herald CCA development in PSC patients.6–8,73 Liverexplants from patients with concomitant PSC and CCAare more likely to harbor high-grade dysplasia comparedwith PSC explants without CCA.6 Moreover, cytogeneticassessment of liver explants with biliary dysplasia frompatients with PSC has demonstrated that patients withprevious or current CCA are more likely to have FISHpolysomy in dysplasia than patients without.7 In thisstudy, FISH polysomy was detected in 58% of the areaswith high-grade dysplasia compared with 11% of areaswith low-grade dysplasia.74 Thus, improved cytogenetictechniques can detect high-grade dysplasia before itsprogression to CCA.7 The presence of FISH polysomy inthe absence of other diagnostic features of CCA (cross-sectional imaging with mass or dominant stricture �positive cytology) may indicate underlying high-gradedysplasia.

Management of PSC patients with dysplasia remainscomplex and unclear. On one hand, definitions andstratification into low-grade and high-grade dysplasia arenot standardized. The natural history is also unknown.On the other hand, a dysplasia-carcinoma sequence hasbeen reported,6 and most patients with dysplasiadevelop CCA (personal observation). Hence, one canargue equally vociferously for careful observation andsurveillance or an aggressive approach with liver trans-plantation. In countries where organ allocation is favor-able to these patients, liver transplantation has beensuggested as the treatment of choice.73 In the UnitedStates, a conservative approach is the sole optionbecause only patients with overt CCA are eligible fortransplantation.

Such patients should be enrolled in intensive sur-veillance for CCA detection. PSC patients with biliarydysplasia would be ideal candidates for chemopreventivestrategies, once potential agents are identified.

A Rational Approach to CholangiocarcinomaScreening and Surveillance

Among adults older than the age of 20 with large ductPSC, it is our practice to obtain liver tests every 3–6months and an annual MRI/MRC and CA 19-9.52 If asuspected CCA is detected, a transperitoneal biopsy ofthe primary mass lesion via endoscopic ultrasound or apercutaneous approach should not be performedbecause of the high risk of peritoneal seeding, whichwould make patients ineligible for a liver transplant.75

Although the cost-effectiveness of this surveillancestrategy and its effect on long-term outcome are unclear,

patients and physicians desire a surveillance approachbecause of the disastrous consequences of diagnosingCCA at a time when it is symptomatic. In this regard,noninvasive imaging of the biliary tree and determina-tion of cancer-associated biomarkers are a pragmaticapproach until further information is available.52

An ERC with brushings for cytology and FISH istypically performed among individuals with suspiciousfeatures on imaging (a new dominant stricture or thedevelopment of focal bile duct thickening, irregularity, orenhancement), symptoms that suggest biliary obstruc-tion or worsening laboratory tests (including new ele-vations of CA 19-9 greater than 100 U/mL in the absenceof cholangitis). Subsequent follow-up is based on theresults of this initial assessment (Figure 3). Currently,there is no evidence to support or refute this practice,and it is unclear whether this strategy improves patientoutcomes or is cost-effective. However, because of theprevalence of CCA among PSC patients and the possi-bility to undergo a curative therapy if detected early, webelieve this strategy is rational and pragmatic.

Future Biomarkers

Contemporary methods used to establish a diagnosisof CCA are suboptimal. A better understanding of cancerbiology, bile acid composition, and key –omics (tumorgenomics, epigenomics, transcriptomics, proteomics, andlipidomics) has paved the way for a series of preliminarystudies that have examined the role of novel biomarkersin the diagnosis of CCA. Table 1 highlights key studiesthat have examined unique markers that also includedCCA associated with PSC and PSC controls without CCA.

Compared with non-neoplastic tissue, tumor cellsoften have a higher proportion of aberrant DNAmethylation, which can in turn serve as a usefulbiomarker for cancer detection.76,77 Among the studiesthat have examined new biomarkers in PSC-associatedCCA, a 4 methylated gene panel (CDO1, CNRIP1, SEPT9,and VIM) obtained from biliary brushings among pa-tients with and without PSC has the best diagnosticperformance reported to date (sensitivity 85% andspecificity 98%).74 However, among the cases of CCA inpatients with PSC, 83% had advanced cancers and wouldhave been ineligible for a liver transplant, and the role ofthis biomarker panel to detect early-stage CCA amongpatients with PSC who are under a longitudinal surveil-lance program is unclear but warrants further study.

In addition to DNA methylation, noncoding RNAshave been examined as markers in malignant condi-tions.78,79 Indeed, among patients with CCA (not associ-ated with PSC) and PSC controls, the measurement of U2small nuclear RNA fragments in bile was able to distin-guish CCA from PSC without cancer, area under the curve(AUC) 0.86.80 A subsequent study examined the role ofanother group of noncoding RNAs (microRNAs) andfound that a panel of biliary vesicle microRNAs among

Figure 3. Screening and surveillance in CCA. Approach to CCA screening and surveillance among adults with large ductprimary sclerosing cholangitis. Modified from Eaton et al.62

November 2015 PSC and CCA 2159

Table 1. Novel Biomarkers for CCA Detection Among Patients With PSC

Biomarker Source Study population Sensitivity (%) Specificity (%) AUCa

DNA methylation74

CDO1CNRIP1SEPT9VIM

Biliary brushings CCA, n ¼ 42(PSC, n¼24)Controls, n ¼ 50(PSC, n ¼ 49)

85 98 0.94

Noncoding RNA81

MicroRNAs in extracellular vesiclesBile CCA, n ¼ 46

(PSC, n ¼ 4)Controls, n ¼ 50(PSC, n ¼ 13)

67 97 —

Peptides85

Panel of 22 peptidesBile CCA, n ¼ 25

(PSC, n ¼ 10)Controls, n ¼ 18(PSC, n ¼ 18)

84 78 0.87b

VOCs83

Acrylonitrile 3-methyl hexane benzeneBile CCA, n ¼ 11

(PSC, n ¼ 11)Controls, n ¼ 21(PSC, n ¼ 21)

91 73 0.89

Peptides86

Panel of 42 peptidesUrine CCA, n ¼ 42

(PSC, n ¼ 10)Controls, n ¼ 81(PSC, n ¼ 45)

83 79 0.87c

Protein87

Angiopoietin-2 �5333 pg/mLSerum CCA, n ¼ 49

(PSC, n ¼ 7)Controls, n ¼ 48(PSC, n ¼ 34)

74 94 0.85

Protein88

Cytokeratin 19 fragments �3 ng/mLSerum CCA, n ¼ 66

(PSC, n ¼ 6)Controls, n ¼ 58(PSC, n ¼ 19)

30 97 —d

Protein89

Trypsinogen-2Serum CCA, n ¼ 41

(PSC, n ¼ 8)Controls, n ¼ 43(PSC, n ¼ 43)

— — 0.80e

aReflects diagnostic performance of entire cohort (ie, patients with and without PSC).bData from test cohort (discovery cohort not included). Eighty percent of PSC-CCA correctly identified.cData from test cohort (discovery cohort not included). One hundred percent of PSC-CCA correctly identified.dAmong those only with PSC (n ¼ 25), sensitivity and specificity were 17% and 95%, respectively.eOptimal cutoff for trypsinogen-2 concentration not reported. Among those only with PSC (n ¼ 51), AUC was 0.76.

2160 Rizvi et al Clinical Gastroenterology and Hepatology Vol. 13, No. 12

those with and without PSC had sensitivity of 67% andspecificity of 96% for the diagnosis of CCA.81

Volatile organic compounds (VOCs) represent a gas-phase biomarker that has been examined as a potentialdiagnostic modality in a variety of conditions.82 Exam-ining the presence of VOCs in bile or urine to distinguishbenign from malignant biliary strictures has also beeninvestigated.83 For example, among a dedicated PSCcohort, a model adjusted for age and gender plus VOClevels obtained from bile was able to identify CCA withsensitivity and specificity of 91% and 73%, respec-tively.83 In addition to VOCs, a small pilot study exam-ined the bile lipid profile among those with de novo CCAand benign biliary conditions including PSC and foundthat a combination of 2 phosphatidylcholines was able todistinguish benign from malignant strictures withsensitivity and specificity of 100% and 83%,respectively.84

To date, much of the published work to improvemethods of CCA detection have centered on proteomics

obtained from bile, urine, or serum. Indeed, there ap-pears to be a varying protein composition of bile be-tween those with benign and malignant strictures, and apanel of 22 peptides was able to identify 80% of CCAassociated with PSC.85 Similarly, a panel of 42 urinepeptides was able to accurately identify all of the PSCpatients with CCA.86 A variety of studies have examinedprotein markers in serum. One such study investigatedthe performance of circulating angiopoietin-2. This pro-tein was noted to have a better diagnostic accuracycompared with CA 19-9 (AUC, 0.85 versus 0.77) andsensitivity and specificity for the detection of CCA of 74%and 94%, respectively.87 An increase in serum cytoker-atin 19 fragments was associated with an increase inmortality, and it demonstrated excellent specificity(�95%) even when limited to a PSC only subgroup, butsensitivity remained poor (�30%) when cutoff of �3ng/mL was used.88 Trypsinogen-2, another serum pro-tein, is increased in CCA, and the use of trypsinogen-2was able to differentiate between PSC with and without

November 2015 PSC and CCA 2161

CCA (AUC, 0.76).89 Last, an additional study combined avariety of serum proteins (CA 19-9, leucine-rich a2-glycoprotein, and IL6) among patients with CCA andconditions associated with benign biliary strictures andreported AUC of 0.98. However, it did not appear thatPSC patients with CCA were examined.90 These studieshighlight the evolving methods that are being investi-gated to distinguish benign from malignant stricturesamong patients with PSC.

Management

Outcomes of surgical resection for pCCA rising in thesetting of PSC have been quite disappointing.91,92

Furthermore, curative surgical resection is often not anoption for pCCA occurring in the setting of PSC becauseof the underlying parenchymal liver disease, the fielddefect, and the predilection for skip lesions.5 These tu-mors are also usually unresectable because of eitherlocal tumor extent or the underlying disease itself.93

Liver transplantation appeared to be an optimal treat-ment for pCCA in PSC patients because it would addressthe underlying disease as well as hepatic or vascularinvasion.93 However, despite this logical basis the out-comes were subpar, with one study noting a 3-yearsurvival of 30% in PSC patients incidentally diagnosedwith CCA.94 Efficacy of palliative radiotherapy led to thedevelopment of a protocol combining chemoradiationfollowed by liver transplantation. Selection criteria forthis protocol are quite rigorous and include the followingprerequisites in a PSC patient: confirmed diagnosis ofpCCA, radial tumor diameter less than 3 cm, and absenceof intrahepatic or extrahepatic metastasis (Figure 4).28

According to a retrospective analysis of 191 patientsenrolled in neoadjuvant chemoradiotherapy followed byliver transplantation protocol, 16 patients underwenttransperitoneal fine-needle aspiration of the primarytumor (13 percutaneous, 3 endoscopic ultrasound).75 Atotal of 6 patients had biopsies demonstrating malig-nancy. Five of these patients (83%) were found to havedirect metastasis into the peritoneum at operative stag-ing, suggesting peritoneal seeding after transperitoneal

Figure 4. Criteria for liver transplantation in PSC patients withpCCA.

biopsy.75 Thus, patients with a prior transperitonealbiopsy of the primary tumor are excluded from thisprotocol. After judicious selection, patients undergo acombination of radiosensitizing chemotherapy with5-fluorouracil, external beam radiation therapy, brachy-therapy with endoscopically placed iridium-192 beads,and maintenance chemotherapy with capecitabine.95

Before orthotopic liver transplantation, patients un-dergo a staging laparotomy to assess for the presence ofmetastasis.95 The pCCA patients who underwent neo-adjuvant chemoradiation in anticipation of liver trans-plantation at 12 U.S. centers had 65% recurrence-freesurvival at 5 years.95 The overall 10-year survival for PSCpatients with pCCA who underwent neoadjuvant che-moradiation in anticipation of liver transplantation atour center is approximately 75% (Figure 5). Approxi-mately 30% of patients drop out of the protocol beforeliver transplantation, primarily because of cancer pro-gression.96 Small case series of combined orthotopic livertransplantation and en bloc Whipple procedure withchemoradiation alone97 or chemoradiation plus brachy-therapy98 have been reported for treatment of hilarcholangiocarcinoma. This is a rational approach forpotentially curative therapy of distal chol-angiocarcinoma. However, in the setting of hilar chol-angiocarcinoma without any involvement of the distalbile duct by malignant or premalignant lesions, thismethod may be overly aggressive.

For advanced stage pCCA, liver transplantation afterneoadjuvant chemoradiation is not an option. In thissetting, the combination of gemcitabine and cisplatin re-mains the pragmatic standard of care.99 Genetic analysis ofa tumor can identify potentially actionable events such asmutations that may be candidates for targeted therapy.Recent reports have highlighted evidence of diseaseregression with chemotherapy directed at the aberrantpathway.100 For instance, patients with fibroblast growthfactor receptor 2 (FGFR2) gene fusions had stable diseasewith ponatinib, an FGFR inhibitor.100 However, FGFR2gene fusions and other mutations such as isocitrate

Figure 5. Kaplan-Meier curve for overall survival in PSC pa-tients with pCCA undergoing neoadjuvant chemoradiationfollowed by liver transplantation.

2162 Rizvi et al Clinical Gastroenterology and Hepatology Vol. 13, No. 12

dehydrogenase 1 and 2 mutations tend to occur morefrequently in intrahepatic CCA than pCCA. The geneticaberrations of PSC-associated CCA have yet to be identi-fied. Hopefully, advances in genomic characterization oftumors will allow detection of driver mutations anddruggable targets associated with perihilar tumors.

Future Directions

The majority of patients with PSC do not develop CCA,and thus it remains unclear which patients have a higherpropensity of developing CCA. Recognizing this high-risksubset is imperative and will aid not only in earlierdetection but also in selecting patients for intensivescreening and potential chemopreventive strategies inthe future. Currently, there is a lack of chemopreventivetherapies, and further preclinical studies are needed toidentify such agents. Detecting early-stage CCA is anotherchallenge in PSC patients. Continued development oftumor biomarkers in biological specimens will be crucialin this regard. Curative nonsurgical treatment of malig-nancy arising in the context of PSC remains a therapeuticconundrum, partly because of the genetic heterogeneityof these tumors and rapid development of therapeuticresistance with genetic evolution of the tumor. Althoughsignificant progress has been made in recognizingoncogenic pathways and mutational changes in CCA,such studies focusing on CCA in PSC remain to be per-formed. We look forward to these and other advances sothat we can prevent and better manage this devastatingcomplication of PSC.

References

1. Rizvi S, Borad MJ, Patel T, et al. Cholangiocarcinoma: molecular

pathways and therapeutic opportunities. Semin Liver Dis 2014;34:456–464.

2. Boonstra K, Weersma RK, van Erpecum KJ, et al. Population-based epidemiology, malignancy risk, and outcome of primarysclerosing cholangitis. Hepatology 2013;58:2045–2055.

3. Kornfeld D, Ekbom A, Ihre T. Survival and risk of chol-angiocarcinoma in patients with primary sclerosing cholangitis:a population-based study. Scand J Gastroenterol 1997;32:1042–1045.

4. de Valle MB, Bjornsson E, Lindkvist B. Mortality and cancer riskrelated to primary sclerosing cholangitis in a Swedishpopulation-based cohort. Liver Int 2012;32:441–448.

5. Rizvi S, Gores GJ. Current diagnostic and management optionsin perihilar cholangiocarcinoma. Digestion 2014;89:216–224.

6. Lewis JT, Talwalkar JA, Rosen CB, et al. Precancerous bile ductpathology in end-stage primary sclerosing cholangitis, with andwithout cholangiocarcinoma. Am J Surg Pathol 2010;34:27–34.

7. Kerr SE, Barr Fritcher EG, Campion MB, et al. Biliary dysplasia inprimary sclerosing cholangitis harbors cytogenetic abnormal-ities similar to cholangiocarcinoma. Hum Pathol 2014;45:1797–1804.

8. Fleming KA, Boberg KM, Glaumann H, et al. Biliary dysplasia asa marker of cholangiocarcinoma in primary sclerosing chol-angitis. J Hepatol 2001;34:360–365.

9. Razumilava N, Gores GJ, Lindor KD. Cancer surveillance inpatients with primary sclerosing cholangitis. Hepatology 2011;54:1842–1852.

10. Bergquist A, Ekbom A, Olsson R, et al. Hepatic and extrahepaticmalignancies in primary sclerosing cholangitis. J Hepatol 2002;36:321–327.

11. Boberg KM, Bergquist A, Mitchell S, et al. Cholangiocarcinomain primary sclerosing cholangitis: risk factors and clinical pre-sentation. Scand J Gastroenterol 2002;37:1205–1211.

12. Burak K, Angulo P, Pasha TM, et al. Incidence and risk factorsfor cholangiocarcinoma in primary sclerosing cholangitis. Am JGastroenterol 2004;99:523–526.

13. Chalasani N, Baluyut A, Ismail A, et al. Cholangiocarcinoma inpatients with primary sclerosing cholangitis: a multicenter case-control study. Hepatology 2000;31:7–11.

14. Bergquist A, Glaumann H, Persson B, et al. Risk factors andclinical presentation of hepatobiliary carcinoma in patients withprimary sclerosing cholangitis: a case-control study. Hepatol-ogy 1998;27:311–316.

15. Broome U, Lofberg R, Veress B, et al. Primary sclerosing chol-angitis and ulcerative colitis: evidence for increased neoplasticpotential. Hepatology 1995;22:1404–1408.

16. Singal AK, Stanca CM, Clark V, et al. Natural history of smallduct primary sclerosing cholangitis: a case series with review ofthe literature. Hepatol Int 2011;5:808–813.

17. Bjornsson E, Angulo P. Cholangiocarcinoma in young in-dividuals with and without primary sclerosing cholangitis. Am JGastroenterol 2007;102:1677–1682.

18. Deneau M, Jensen MK, Holmen J, et al. Primary sclerosingcholangitis, autoimmune hepatitis, and overlap in Utah children:epidemiology and natural history. Hepatology 2013;58:1392–1400.

19. Rupp C, Rossler A, Halibasic E, et al. Reduction in alkalinephosphatase is associated with longer survival in primary scle-rosing cholangitis, independent of dominant stenosis. AlimentPharmacol Ther 2014;40:1292–1301.

20. Al Mamari S, Djordjevic J, Halliday JS, et al. Improvement ofserum alkaline phosphatase to <1.5 upper limit of normal pre-dicts better outcome and reduced risk of cholangiocarcinoma inprimary sclerosing cholangitis. J Hepatol 2013;58:329–334.

21. Saarinen S, Olerup O, Broome U. Increased frequency ofautoimmune diseases in patients with primary sclerosing chol-angitis. Am J Gastroenterol 2000;95:3195–3199.

22. Eaton JE, Talwalkar JA, Lazaridis KN, et al. Pathogenesis ofprimary sclerosing cholangitis and advances in diagnosis andmanagement. Gastroenterology 2013;145:521–536.

23. Olerup O, Olsson R, Hultcrantz R, et al. HLA-DR and HLA-DQare not markers for rapid disease progression in primary scle-rosing cholangitis. Gastroenterology 1995;108:870–878.

24. Spurkland A, Saarinen S, Boberg KM, et al. HLA class II hap-lotypes in primary sclerosing cholangitis patients from five Eu-ropean populations. Tissue Antigens 1999;53:459–469.

25. Liu JZ, Hov JR, Folseraas T, et al. Dense genotyping of immune-related disease regions identifies nine new risk loci for primarysclerosing cholangitis. Nat Genet 2013;45:670–675.

26. Karlsen TH, Franke A, Melum E, et al. Genome-wide associationanalysis in primary sclerosing cholangitis. Gastroenterology2010;138:1102–1111.

27. Melum E, Franke A, Schramm C, et al. Genome-wide associa-tion analysis in primary sclerosing cholangitis identifies two non-HLA susceptibility loci. Nat Genet 2011;43:17–19.

November 2015 PSC and CCA 2163

28. Rizvi S, Gores GJ. Pathogenesis, diagnosis, and manage-ment of cholangiocarcinoma. Gastroenterology 2013;145:1215–1229.

29. Jaiswal M, LaRusso NF, Burgart LJ, et al. Inflammatory cyto-kines induce DNA damage and inhibit DNA repair in chol-angiocarcinoma cells by a nitric oxide-dependent mechanism.Cancer Res 2000;60:184–190.

30. Jaiswal M, LaRusso NF, Shapiro RA, et al. Nitric oxide-mediatedinhibition of DNA repair potentiates oxidative DNA damage incholangiocytes. Gastroenterology 2001;120:190–199.

31. Ichim G, Lopez J, Ahmed SU, et al. Limited mitochondrial per-meabilization causes DNA damage and genomic instability inthe absence of cell death. Mol Cell 2015;57:860–872.

32. Liu X, He Y, Li F, et al. Caspase-3 promotes genetic instabilityand carcinogenesis. Mol Cell 2015;58:284–296.

33. Lovric MM, Hawkins CJ. TRAIL treatment provokes mutations insurviving cells. Oncogene 2010;29:5048–5060.

34. Takeda K, Kojima Y, Ikejima K, et al. Death receptor 5 mediated-apoptosis contributes to cholestatic liver disease. Proc NatlAcad Sci U S A 2008;105:10895–10900.

35. Yoon J-H, Higuchi H, Werneburg NW, et al. Bile acids inducecyclooxygenase-2 expression via the epidermal growth factorreceptor in a human cholangiocarcinoma cell line. Gastroen-terology 2002;122:985–993.

36. Yoon JH, Gwak GY, Lee HS, et al. Enhanced epidermal growthfactor receptor activation in human cholangiocarcinoma cells.J Hepatol 2004;41:808–814.

37. Kuver R. Mechanisms of oxysterol-induced disease: insightsfrom the biliary system. Clin Lipidol 2012;7:537–548.

38. Haigh WG, Lee SP. Identification of oxysterols in human bile andpigment gallstones. Gastroenterology 2001;121:118–123.

39. Yoon JH, Canbay AE, Werneburg NW, et al. Oxysterolsinduce cyclooxygenase-2 expression in cholangiocytes: im-plications for biliary tract carcinogenesis. Hepatology 2004;39:732–738.

40. Dwyer JR, Sever N, Carlson M, et al. Oxysterols are novel ac-tivators of the hedgehog signaling pathway in pluripotentmesenchymal cells. J Biol Chem 2007;282:8959–8968.

41. El Khatib M, Kalnytska A, Palagani V, et al. Inhibition ofhedgehog signaling attenuates carcinogenesis in vitro and in-creases necrosis of cholangiocellular carcinoma. Hepatology2013;57:1035–1045.

42. Fingas CD, Bronk SF, Werneburg NW, et al. Myofibroblast-derived PDGF-BB promotes Hedgehog survival signaling incholangiocarcinoma cells. Hepatology 2011;54:2076–2088.

43. Nachtergaele S, Mydock LK, Krishnan K, et al. Oxysterols areallosteric activators of the oncoprotein Smoothened. Nat ChemBiol 2012;8:211–220.

44. Yamada D, Rizvi S, Razumilava N, et al. IL-33 facilitates onco-gene driven cholangiocarcinoma in mice. Hepatology 2014;60:641a–642a.

45. Li J, Razumilava N, Gores GJ, et al. Biliary repair and carcino-genesis are mediated by IL-33-dependent cholangiocyte pro-liferation. J Clin Invest 2014;124:3241–3251.

46. Marvie P, Lisbonne M, L’Helgoualc’h A, et al. Interleukin-33overexpression is associated with liver fibrosis in mice andhumans. J Cell Mol Med 2010;14:1726–1739.

47. Park J, Tadlock L, Gores GJ, et al. Inhibition of interleukin6-mediated mitogen-activated protein kinase activation attenu-ates growth of a cholangiocarcinoma cell line. Hepatology 1999;30:1128–1133.

48. Taniguchi K, Wu LW, Grivennikov SI, et al. A gp130-Src-YAPmodule links inflammation to epithelial regeneration. Nature2015;519:57–62.

49. Sandborn WJ, Ghosh S, Panes J, et al. Tofacitinib, an oral Januskinase inhibitor, in active ulcerative colitis. N Engl J Med 2012;367:616–624.

50. Jhaveri KS, Hosseini-Nik H. MRI of cholangiocarcinoma. J MagnReson Imaging 2014 [Epub ahead of print].

51. Charatcharoenwitthaya P, Enders FB, Halling KC, et al. Utility ofserum tumor markers, imaging, and biliary cytology for detectingcholangiocarcinoma in primary sclerosing cholangitis. Hepatol-ogy 2008;48:1106–1117.

52. Chapman R, Fevery J, Kalloo A, et al. Diagnosis and manage-ment of primary sclerosing cholangitis. Hepatology 2010;51:660–678.

53. Stiehl A, Rudolph G, Kloters-Plachky P, et al. Development ofdominant bile duct stenoses in patients with primary sclerosingcholangitis treated with ursodeoxycholic acid: outcome afterendoscopic treatment. J Hepatol 2002;36:151–156.

54. Sangfelt P, Sundin A, Wanders A, et al. Monitoring dominantstrictures in primary sclerosing cholangitis with brush cytologyand FDG-PET. J Hepatol 2014;61:1352–1357.

55. Kaya M, Petersen BT, Angulo P, et al. Balloon dilation comparedto stenting of dominant strictures in primary sclerosing chol-angitis. Am J Gastroenterol 2001;96:1059–1066.

56. Eaton JE, Barr Fritcher EG, Gores GJ, et al. Biliary multifocalchromosomal polysomy and cholangiocarcinoma in primarysclerosing cholangitis. Am J Gastroenterol 2015;110:299–309.

57. Outwater E, Kaplan MM, Bankoff MS. Lymphadenopathy insclerosing cholangitis: pitfall in the diagnosis of malignant biliaryobstruction. Gastrointest Radiol 1992;17:157–160.

58. Wannhoff A, Hov JR, Folseraas T, et al. FUT2 and FUT3 geno-type determines CA19-9 cut-off values for detection of chol-angiocarcinoma in patients with primary sclerosing cholangitis.J Hepatol 2013;59:1278–1284.

59. Levy C, Lymp J, Angulo P, et al. The value of serum CA 19-9 inpredicting cholangiocarcinomas in patients with primary scle-rosing cholangitis. Dig Dis Sci 2005;50:1734–1740.

60. Sinakos E, Saenger AK, Keach J, et al. Many patients with pri-mary sclerosing cholangitis and increased serum levels of car-bohydrate antigen 19-9 do not have cholangiocarcinoma. ClinGastroenterol Hepatol 2011;9:434–439.

61. Venkatesh PG, Navaneethan U, Shen B, et al. Increased serumlevels of carbohydrate antigen 19-9 and outcomes in primarysclerosing cholangitis patients without cholangiocarcinoma. DigDis Sci 2013;58:850–857.

62. Eaton JE, Gossard AA, Talwalkar JA. Recall processes for biliarycytology in primary sclerosing cholangitis. Curr Opin Gastro-enterol 2014;30:287–294.

63. Trikudanathan G, Navaneethan U, Njei B, et al. Diagnostic yieldof bile duct brushings for cholangiocarcinoma in primary scle-rosing cholangitis: a systematic review and meta-analysis.Gastrointest Endosc 2014;79:783–789.

64. Furmanczyk PS, Grieco VS, Agoff SN. Biliary brush cytology andthe detection of cholangiocarcinoma in primary sclerosingcholangitis: evaluation of specific cytomorphologic features andCA19-9 levels. Am J Clin Pathol 2005;124:355–360.

65. Barr Fritcher EG, Voss JS, Jenkins SM, et al. Primary sclerosingcholangitis with equivocal cytology: fluorescence in situ hy-bridization and serum CA 19–9 predict risk of malignancy.Cancer Cytopathol 2013;121:708–717.

2164 Rizvi et al Clinical Gastroenterology and Hepatology Vol. 13, No. 12

66. Bangarulingam SY, Bjornsson E, Enders F, et al. Long-termoutcomes of positive fluorescence in situ hybridization testsin primary sclerosing cholangitis. Hepatology 2010;51:174–180.

67. Barr Fritcher EG, Kipp BR, Voss JS, et al. Primary scle-rosing cholangitis patients with serial polysomy fluores-cence in situ hybridization results are at increased risk ofcholangiocarcinoma. Am J Gastroenterol 2011;106:2023–2028.

68. Navaneethan U, Njei B, Venkatesh PG, et al. Fluorescence insitu hybridization for diagnosis of cholangiocarcinoma in primarysclerosing cholangitis: a systematic review and meta-analysis.Gastrointest Endosc 2014;79:943–950.

69. Siddiqui AA, Mehendiratta V, Jackson W, et al. Identifica-tion of cholangiocarcinoma by using the Spyglass Spy-scope system for peroral cholangioscopy and biopsycollection. Clin Gastroenterol Hepatol 2012;10:466–471,quiz e48.

70. Domagk D, Wessling J, Reimer P, et al. Endoscopic retrogradecholangiopancreatography, intraductal ultrasonography, andmagnetic resonance cholangiopancreatography in bile ductstrictures: a prospective comparison of imaging diagnosticswith histopathological correlation. Am J Gastroenterol 2004;99:1684–1689.

71. Meining A, Frimberger E, Becker V, et al. Detection of chol-angiocarcinoma in vivo using miniprobe-based confocal fluo-rescence microscopy. Clin Gastroenterol Hepatol 2008;6:1057–1060.

72. Azeem N, Gostout CJ, Knipschield M, et al. Cholangioscopywith narrow-band imaging in patients with primary sclerosingcholangitis undergoing ERCP. Gastrointest Endosc 2014;79:773–779.

73. Boberg KM, Jebsen P, Clausen OP, et al. Diagnostic benefit ofbiliary brush cytology in cholangiocarcinoma in primary scle-rosing cholangitis. J Hepatol 2006;45:568–574.

74. Andresen K, Boberg KM, Vedeld HM, et al. Four DNAmethylationbiomarkers in biliary brush samples accurately identify the pres-ence of cholangiocarcinoma. Hepatology 2015;61:1651–1659.

75. Heimbach JK, Sanchez W, Rosen CB, et al. Trans-peritonealfine needle aspiration biopsy of hilar cholangiocarcinoma isassociated with disease dissemination. HPB 2011;13:356–360.

76. Laird PW. The power and the promise of DNA methylationmarkers. Nat Rev Cancer 2003;3:253–266.

77. Shin SH, Lee K, Kim BH, et al. Bile-based detection ofextrahepatic cholangiocarcinoma with quantitative DNAmethylation markers and its high sensitivity. J Mol Diagn2012;14:256–263.

78. Shigehara K, Yokomuro S, Ishibashi O, et al. Real-time PCR-based analysis of the human bile microRNAome identifiesmiR-9 as a potential diagnostic biomarker for biliary tract can-cer. PLoS One 2011;6:e23584.

79. Baraniskin A, Nopel-Dunnebacke S, Ahrens M, et al. CirculatingU2 small nuclear RNA fragments as a novel diagnosticbiomarker for pancreatic and colorectal adenocarcinoma. Int JCancer 2013;132:E48–E57.

80. Baraniskin A, Nopel-Dunnebacke S, Schumacher B, et al.Analysis of U2 small nuclear RNA fragments in the bile differ-entiates cholangiocarcinoma from primary sclerosing chol-angitis and other benign biliary disorders. Dig Dis Sci 2014;59:1436–1441.

81. Li L, Masica D, Ishida M, et al. Human bile contains microRNA-laden extracellular vesicles that can be used for chol-angiocarcinoma diagnosis. Hepatology 2014;60:896–907.

82. ArasaradnamRP,CovingtonJA,HarmstonC, et al. Reviewarticle:next generation diagnostic modalities in gastroenterology—gasphase volatile compound biomarker detection. Aliment Pharma-col Ther 2014;39:780–789.

83. Navaneethan U, Parsi MA, Lourdusamy V, et al. Volatile organiccompounds in bile for early diagnosis of cholangiocarcinoma inpatients with primary sclerosing cholangitis: a pilot study.Gastrointest Endosc 2015;81:943–949.

84. Navaneethan U, Gutierrez NG, Venkatesh PG, et al. Lipidomicprofiling of bile in distinguishing benign from malignant biliarystrictures: a single-blinded pilot study. Am J Gastroenterol 2014;109:895–902.

85. Lankisch TO, Metzger J, Negm AA, et al. Bile proteomicprofiles differentiate cholangiocarcinoma from primary scle-rosing cholangitis and choledocholithiasis. Hepatology 2011;53:875–884.

86. Metzger J, Negm AA, Plentz RR, et al. Urine proteomic analysisdifferentiates cholangiocarcinoma from primary sclerosingcholangitis and other benign biliary disorders. Gut 2013;62:122–130.

87. Voigtlander T, David S, Thamm K, et al. Angiopoietin-2 andbiliary diseases: elevated serum, but not bile levels are associ-ated with cholangiocarcinoma. PLoS One 2014;9:e97046.

88. Chapman MH, Sandanayake NS, Andreola F, et al. CirculatingCYFRA 21-1 is a specific diagnostic and prognostic biomarkerin biliary tract cancer. J Clin Exp Hepatol 2011;1:6–12.

89. Lempinen M, Isoniemi H, Makisalo H, et al. Enhanced detectionof cholangiocarcinoma with serum trypsinogen-2 in patientswith severe bile duct strictures. J Hepatol 2007;47:677–683.

90. Sandanayake NS, Sinclair J, Andreola F, et al. A combination ofserum leucine-rich alpha-2-glycoprotein 1, CA19-9 andinterleukin-6 differentiate biliary tract cancer from benign biliarystrictures. Br J Cancer 2011;105:1370–1378.

91. Rosen CB, Nagorney DM. Cholangiocarcinoma complicatingprimary sclerosing cholangitis. Semin Liver Dis 1991;11:26–30.

92. Rosen CB, Nagorney DM, Wiesner RH, et al. Chol-angiocarcinoma complicating primary sclerosing cholangitis.Ann Surg 1991;213:21–25.

93. Rea DJ, Rosen CB, Nagorney DM, et al. Transplantation forcholangiocarcinoma: when and for whom? Surg Oncol Clin NAm 2009;18:325–337, ix.

94. Ghali P, Marotta PJ, Yoshida EM, et al. Liver transplantation forincidental cholangiocarcinoma: analysis of the Canadian expe-rience. Liver Transpl 2005;11:1412–1416.

95. Darwish Murad S, Kim WR, Harnois DM, et al. Efficacy of neo-adjuvant chemoradiation, followed by liver transplantation, forperihilar cholangiocarcinoma at 12 US centers. Gastroenter-ology 2012;143:88–98, quiz e14.

96. Darwish Murad S, Kim WR, Therneau T, et al. Predictors of pre-transplant dropout and posttransplant recurrence in patients withperihilar cholangiocarcinoma. Hepatology 2012;56:972–981.

97. Nikeghbalian S, Shamsaeefar A, Eshraghian A, et al. Livertransplantation and Whipple surgery combined with chemo-radiotherapy for treatment of hilar cholangiocarcinoma in pa-tients with primary sclerosing cholangitis. Liver Transpl 2015;21:696–699.

98. Wu Y, Johlin FC, Rayhill SC, et al. Long-term, tumor-free sur-vival after radiotherapy combining hepatectomy-Whipple en

November 2015 PSC and CCA 2165

bloc and orthotopic liver transplantation for early-stage hilarcholangiocarcinoma. Liver Transpl 2008;14:279–286.

99. Valle J, Wasan H, Palmer DH, et al. Cisplatin plus gemcitabineversus gemcitabine for biliary tract cancer. N Engl J Med 2010;362:1273–1281.

100. Borad MJ, Champion MD, Egan JB, et al. Integratedgenomic characterization reveals novel, therapeutically rele-vant drug targets in FGFR and EGFR pathways in sporadicintrahepatic cholangiocarcinoma. Plos Genet 2014;10:e1004135.

Reprint requestsAddress requests for reprints to: Gregory J. Gores, MD, Mayo Clinic, 200 FirstStreet SW, Rochester, Minnesota 55905. e-mail: gores.gregory@mayo.edu;fax: (507) 284-0762.

Conflicts of interestThe authors disclose no conflicts.

FundingSupported by National Institutes of Health grants DK59427 (G.J.G.),T32DK007198, the American Liver Foundation and International Liver CancerAssociation (S.R.), and the Mayo Foundation.

GASTROENTEROLOGY ARTICLE OF THE WEEK February 4, 2016

Rizvi S, Eaton JE, Gores GJ. Primary sclerosing cholangitis as a premalignant biliary tract disease: Surveillance and management. Clin Gastroenterol Hepatol 2015;13:2152-2165 1. MRI is the preferred modality for detecting CCA in PSC. Findings that suggest CCA include: a. Delayed enhancement of thickened bile duct walls or strictures b. Arterial phase enhancing mass lesion c. Thickened or nodular bile ducts d. Perihilar adenopathy True or False 2. The most common type of CCA seen in PSC is the intrahepatic type 3. Approximately 50% of dominant strictures will eventually prove to be malignant 4. Up to 1/3 of CCA’s associated with PSC are diagnosed within the first year after initial diagnosis of PSC 5. Standard biliary cytology results are only helpful if they confirm adenocarcinoma 6. Patients with small duct PSC have similar risk of CCA compared to large duct PSC 7. PET-CT is useful for detecting or excluding CCA in PSC, if the PET-CT is negative, it is highly unlikely that CCA is present 8. FISH analysis of biliary fluid cytology enhances the ability to detect CCA, the presence of polysomy is particularly important in detecting CCA 9. If a mass lesion is detected and is suspicious for CCA, a percutaneous biopsy should be obtained for confirmation if the lesion is accessible. 10. Patients with polysomy by FISH analysis of biliary fluid, but no other evidence of CCA should undergo repeat FISH at intervals, persistent positivity raises risk for CCA 11. Recommended surveillance for CCA in PSC includes MRI/MRC and CA 19-9 every 6 months 12. The treatment of choice of asymptomatic CCA is liver transplantation