Clinical and Pathological Features of Intraductal Papillary Neoplasm of the Biliary Tract and Gallbladder

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector DOI:10.1111/hpb.12460 HPB

ORIGINAL ARTICLE Clinical and pathological features of intraductal papillary neoplasm of the biliary tract and gallbladder

Sean Bennett1, E. Celia Marginean2, Melanie Paquin-Gobeil1, Jason Wasserman2, Joel Weaver1, Richard Mimeault1, Fady K. Balaa1 & Guillaume Martel1,3

1Liver and Pancreas Unit, Department of Surgery, University of Ottawa, Ottawa, ON, Canada, 2Department of Pathology and Laboratory Medicine, University of Ottawa, Ottawa, ON, Canada, and 3Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa, ON, Canada Abstract Background: Intraductal papillary neoplasms of the biliary tract (IPNB) and intracholecystic papillary neoplasms (ICPN) are rare tumours characterized by intraluminal papillary growth that can be associated with invasive carcinoma. Their natural history remains poorly understood. This study examines clinicopathological features and outcomes. Methods: Patients who underwent surgery for IPNB/ICPN (2008–2014) were identiﬁed. Descriptive statistics and survival data were generated. Results: Of 23 patients with IPNB/ICPN, 10 were male, and the mean age was 68 years. The most common presentations were abdominal pain (n = 10) and jaundice (n = 9). Tumour locations were: intrahepatic (n = 5), hilar (n = 3), the extrahepatic bile duct (n = 8) and the gallbladder (n = 7). Invasive cancer was found in 20/23 patients. Epithelial subtypes included pancreatobiliary (n = 15), intestinal (n = 7) and gastric (n = 1). The median follow-up was 30 months. The 5-year overall (OS) and disease- free survivals (DFS) were 51% and 57%, respectively. Decreased OS (P = 0.09) and DFS (P = 0.05) were seen in patients with tumours expressing MUC1 on immunohistochemistry (IHC). Conclusion: IPNB/ICPN are rare precursor lesions that can affect the entire biliary epithelium. At pathology, the majority of patients have invasive carcinoma, thus warranting a radical resection. Patients with tumours expressing MUC1 appear to have worse OS and DFSs.

Received 22 March 2015; accepted 25 May 2015

Correspondence Guillaume Martel, The Ottawa Hospital – General Campus, 501 Smyth Road, Ottawa, ON, Canada, K1H 8L6. Tel.: +1 613 737 8899 ext 71053. Fax: +1 613 739 6854. E-mail: [email protected]

Introduction dysplastic epithelium with frequent intestinal metaplasia and mucin hypersecretion. These lesions were previously called Intrahepatic and extrahepatic cholangiocarcinoma (CC) papillomas, papillary adenomas, papillomatosis (when multi- develop through a dysplasia-carcinoma sequence. Until ple) or mucin-secreting bile duct tumours. An intraductal pap- recently, the nomenclature and criteria for the classification of illary neoplasm of the biliary tract (IPNB) was recognized as a biliary precursor lesions were not well established. At least two distinct pathological entity by the World Health Organization major precursor lesions have been associated with the develop- – in 2010.3 8 IPNB has widely been considered to be analogous ment of invasive CC. The first is a microscopic lesion of flat or to an intraductal papillary mucinous neoplasm (IPMN) of the micropapillary dysplastic epithelium, biliary dysplasia or biliary – pancreas.4,5,7,9 11 Both IPNB and IPMN feature intraluminal intraepithelial neoplasia (BilIN).1,2 The second is an intraductal growth, an association with mucin hypersecretion, and the papillary neoplasm of the bile duct, which is a macroscopic same four histological subtypes have been described: pancre- lesion, single or multiple along the biliary tract, characterized atobiliary, intestinal, gastric and oncocytic.9,12 The same by intraluminal growth, prominent papillary proliferation of tumour type occurring in the gallbladder is commonly referred to as intracholecystic papillary neoplasm (ICPN).13 This study was presented at the Annual Meeting of the AHPBA, 11-15 While there is an increasing body of literature on IPNB and March 2015, Miami, Florida. ICPN emerging from Asia, the North American experience

HPB 2015, 17, 811–818 ª 2015 International Hepato-Pancreato-Biliary Association 812 HPB

remains limited. The largest North American series have iden- classified using the Clavien–Dindo grading system. Survival tified 39 and 23 patients, respectively,11,14 and did not include data were obtained from clinic notes and, in cases where fol- cases of ICPN. From the Asian literature, it is believed that low-up was conducted elsewhere, correspondence with the IPNB demonstrates a favourable prognosis compared with typ- treating oncologist. Overall survival was measured from the – ical cholangiocarcinoma.7,15 17 This may be related to its intra- date of surgical resection to the date of death. Patients still luminal growth rather than a periductal infiltrating growth. alive at the date of last clinical encounter were censored. Dis- Compared with the Asian experience18,19, the North American ease-free survival was measured from the date of surgical resec- literature seems to demonstrate a higher proportion of invasive tion to the date of first recurrence or death from the disease. carcinoma at the time of surgical resection.11,14 The objective Patients were assumed to have died of disease unless they had of this paper was to describe further the clinical and pathologi- clear evidence to the contrary. Patients still alive at the date of cal characteristics of a cohort of North American patients with last clinical encounter and without evidence of disease recur- IPNB and ICPN. rence were censored. Survival analysis was performed using the Kaplan–Meier estimate of survival, and comparisons were performed with the log-rank test. Statistical analysis was per- Patients and methods formed with SAS 9.3 (SAS Institute Inc., Carry, NC, USA). After approval from The Ottawa Health Science Research Eth- ics Board (20140648-01H), a computerized search was per- Results formed in the Anatomic Pathology archives. All cases between Patients and pre-operative workup March 2008 and September 2014 of resected gallbladder, intra- Sixteen patients with IPNB and seven patients with ICPN were hepatic, hilar and extrahepatic bile duct tumours, cross-refer- identified. Demographics and clinical presentations are detailed enced with the words ‘cholangiocarcinoma’, ‘adenocarcinoma’, in Table 1. Most patients with IPNB presented in a manner typi- ‘intraductal’, ‘papilloma’, ‘papillomatosis’ or ‘papillary architec- cal of cholangiocarcinoma. Over two-thirds of patients had evi- ture’, were retrieved from the surgical pathology files. Pancre- dence of biochemical biliary obstruction, whereas less than half atic intraductal papillary mucinous neoplasms, ampullary and of patients had abdominal pain. All IPNB patients underwent periampullary tumours of uncertain origin, and cases with in- either computed tomography (CT) or magnetic resonance imag- traductal epithelial dysplasia/neoplasia without compelling evi- ing (MRI) pre-operatively. The majority of intrahepatic lesions dence of intraductal mass formation were excluded. were found in the left hemiliver (four out of five). All patients All available tumour slides for the selected cases were with IPNB went on to radical resection after cross-sectional reviewed by an expert pathologist in the liver, biliary and pan- imaging and, occasionally, endoscopic cytology/histology results. creatic disease (ECM). Tumours were reclassified as IPNB/ In contrast, patients with ICPN were more likely to present ICPN with or without invasion according to the WHO classifi- incidentally or with symptoms suggestive of cholelithiasis cation.8 The tumours were evaluated for architecture (papil- (6/7 patients). Specifically, two patients presented with epigastric lary, tubular or solid), epithelial type (intestinal, gastric, pain suggestive of biliary colic, one with pancreatitis, and one oncocytic or pancreatobiliary), grade of dysplasia, the presence with acalculous cholecystitis. One of the patients with biliary and depth of invasion, lymphovascular invasion, perineural colic had a mass on imaging prior to surgery whereas the other invasion, lymph node metastasis and resection margin status. three patients had no indication of having a gallbladder mass or The most representative tumour block from each case was polyp. Two additional patients were asymptomatic and were subjected to immunohistochemical (IHC) staining with MUC1 found to have a gallbladder mass or polyp incidentally. One (clone Ma695, dilution 1:100; Leica Biosystems, Buffalo Grove, patient presented with an abdominal mass on examination. For IL, USA), MUC2 (clone B306-1, dilution 1:50; ABCAM, Cam- both IPNB and ICPN, serum carbohydrate antigen 19-9 (CA bridge, MA, USA) and MUC5AC (clone CLH2, dilution 1:100; 19-9) was not routinely assessed during the pre-operative Leica Biosystems, Buffalo Grove, IL, USA). Over 20 human work-up, but was used primarily for follow-up in select patients. mucins (MUC1-MUC20) have been identified. Mucins can be broadly subdivided into two groups: proteins that are secreted Surgical and pathological data and form extracellular gels (MUC2, MUC5AC, MUC5B and Surgical and pathological details are presented in Table 2. MUC6) and membrane-bound mucins (MUC1, MUC3 Pre-operative tissue diagnosis was attempted as part of the pre- and MUC4).20 In this cohort, the expression of MUC1, operative workup in 10 patients with IPNB, whether by endo- MUC2, MUC5AC was evaluated semiquantitatively. Tumours scopic retrograde cholangiopancreatography (ERCP) brushings, showing intense positive staining in more than 10% of tumour ERCP biopsy of a visible duodenal tumour or by endoscopic cells were considered positive, as previously reported.20 ultrasound (EUS) fine-needle aspiration. Despite the final Demographic, clinical, operative and imaging data were pathology results of microinvasive or invasive carcinoma in all retrieved retrospectively from patient records. Post-operative patients with IPNB, only 2/10 pre-operative histology/cytology complications were also retrieved from patient records and specimens were initially consistent with invasive malignancy.

HPB 2015, 17, 811–818 ª 2015 International Hepato-Pancreato-Biliary Association HPB 813

Table 1 Patient characteristics (n = 23) followed by an intestinal type (Fig. 1e) and a single tumour Variable IPNB (n = 16) ICPN (n = 7) with a gastric phenotype (Fig. 1f). No tumour showed onco- Age at diagnosis (mean Æ SD) 70.1 Æ 9.6 64.1 Æ 12.9 cytic features. Male 9 1 MUC1 was expressed in the luminal border and cytoplasm of cancer cells. MUC2 and MUC5AC were mainly expressed in Presentation the cytoplasm of tumour cells. MUC1 was expressed predomi- Jaundice 9 0 nantly in the pancreatobiliary subtype, whereas the intestinal Abdominal Pain 6 4 type showed constant expression of MUC2 (Table 3). Palpable mass 0 1 Incidental finding on imaging 1 2 Outcomes Elevated AST/ALT 10 1 Seven patients experienced a post-operative complication – Elevated ALP 11 0 (Clavien Dindo grade 2 or higher), of which 3 patients had Pre-operative imaging grade 3a complications. These included three cases of percuta- neous drain insertion for a pancreatic fistula (n = 1) and bile Ultrasound 5 7 leaks after a hepaticojejunostomy (n = 2). Grade 2 complica- CT scan 13 4 tions included three cases of antibiotics therapy for small MRI 12 2 undrained collections and one case of a blood transfusion. ERCP 11 0 There were no complications requiring intensive care, and no EUS 3 0 mortalities. AST, aspartate aminotransferase; ALT, alanine transaminase; ALP, Nine of the sixteen patients with IPNB underwent adjuvant alkaline phosphatase; CT, computed tomography; ERCP, endoscopic chemotherapy, most commonly gemcitabine and cisplatin. One retrograde cholangiopancreatography; EUS, endoscopic ultrasound; out of seven ICPN patients underwent adjuvant chemotherapy. ICNB, intracholecystic papillary neoplasm; IPNB, intraductal papillary neoplasm of the biliary tract; MRI, magnetic resonance imaging; n, The median follow-up of patients was 30 months. There number of patients; SD, standard deviation. were seven patients who developed a recurrence of cancer: four with intrahepatic metastases, one with distant metastases, and two with both local recurrence and distant metastases. Five of All three patients with ICPN who presented with pain and the recurrences occurred in patients with distal extrahepatic did not have a mass on imaging went on to have a cholecystec- primary cancers, and two in patients with gallbladder lesions. tomy. All three had carcinoma in situ on final pathology. One The furthest recurrence from the time of surgery was asymptomatic patient with a gallbladder polyp went on to lap- 22 months. Among patients with recurrence, six died, and one aroscopic cholecystectomy and was found to have a T3 carci- additional patient died without evidence of disease recurrence. noma with negative margins. He declined further liver surgery. The OS at 3 and 5 years was 71% and 51%, respectively. All three other ICPN patients who had either a palpable mass Disease-free survival was 57% at both 3 and 5 years. There was or mass on pre-operative imaging were considered to have no significant difference in OS and DFS for an R1 resection gallbladder cancer and went on to have a cholecystectomy and (P = 0.20 and 0.37) and epithelial subtype (P = 0.62 and 0.34). liver resection. Among cases of ICPN, invasive cancer was seen Similarly, there was no difference in OS or DFS when compar- in three of the five tumours with a pancreatobiliary subtype, ing micro-invasive and invasive tumours (P = 0.91 and 0.73). and the one with a gastric subtype. The ICPN tumour with an Patients with tumours expressing MUC1 demonstrated a intestinal subtype was non-invasive. decreased OS (P = 0.09), and DFS (P = 0.05). No other immu- An R0 resection was achieved for all patients with ICPN and nohistochemical markers were associated with prognosis. Fig- 12/16 patients with IPBN. Among the four patients with R1 ure 2 displays the Kaplan–Meier survival curves for OS and resections, two patients had false-negative intra-operative fro- DFS for all 23 patients, and for the comparison of patients zen section bile duct margin analyses, one was unfit for more with and without expression of MUC1. extensive radical resection, and one had intraductal tumour infiltration within the remnant liver. Discussion After pathological review for this work, 9/23 tumours were re-classified as IPNB/ICPN with or without invasion. Original The present study describes the clinical and pathological char- reports included terms such as ‘papillary adenoma‘, ‘villous acteristics of patients undergoing surgery for IPNB and ICPN. adenoma‘, ‘billiary papillomatosis‘ or ‘papillary differentiation‘. It adds to a relatively limited body of literature on this topic, All included tumours showed a visible intraluminal growth particularly in North America. (Fig. 1a) with papillary fronds showing delicate fibrovascular The current paper demonstrates that patients with IPNB or cores, lined by dysplastic epithelium (Fig. 1b, c). The most ICPN tend to present similarly to patients with typical cholan- common histological subtype was pancreatobiliary (Fig. 1d), giocarcinoma or gallbladder cancer, albeit with occasional

HPB 2015, 17, 811–818 ª 2015 International Hepato-Pancreato-Biliary Association 814 HPB

Table 2 Surgical and pathological characteristics Table 2 Continued Variable IPNB (n = 16) ICPN (n = 7) Variable IPNB (n = 16) ICPN (n = 7) = b = Anatomic location Lymphovascular 6(n 15) 3(n 6) Invasion present Left intrahepatic 4 N/A = = bile ducts Perineural Invasion 6(n 13) 0 (n 5) Present Right intrahepatic 1 N/A a bile ducts Combines T2, T2a, T2b (each anatomic location has unique T stag- ing). b Liver hilum 3 N/A not all pathology reports commented on lymphovascular or perineural Common bile duct 8 N/A invasion. cm, centimetre; mm, millimetre; n, number of patients; SD, standard = Pre-operative histology/cytology (n 10) deviation. Negative 2 N/A Cytologic atypia 1 N/A abdominal pain. Usually, the diagnosis of IPNB/ICPN was not Suspicious for 1 N/A malignancy established until the final surgical specimen is reviewed. The use of pre-operative endoscopic histology/cytology was not High grade dysplasia 4 N/A particularly reliable in establishing a diagnosis of malignancy Adenocarcinoma 2 N/A (2/10), possibly owing to limited foci of invasive disease. At Surgery performed final pathology, the rate of invasive cancer amongst patients Left hepatectomy 4 with intrahepatic, hilar, or distal extrahepatic IPNB was 100%, Segmental resection 2 whereas that with ICPN of the gallbladder was 57%. This high Pancreaticoduo- 5 rate of invasive disease among patients with IPNB is consistent denectomy with other North American papers but is higher than most 18,19 Choledochectomy 5 Asian series. The discrepancy in the rates of invasion Lap cholecystectomy 3 between the two continents is not entirely understood. Hepato- Open cholecystectomy 1 lithiasis and liver fluke infection have been identified as risk factors among Asian patients,4,9,10,18,21,22 which may hint at a Open cholecystectomy + 3 liver wedge difference in pathophysiology. It may also explain the overall R0 resection 12 7 increased burden of disease in Asia. Earlier diagnosis or genetic differences may also play a role. At least one other series has R1 resection 4 0 noted that ICPN was more common in Asia, whereas a gall- Epithelial subtype bladder lesion from flat dysplasia was more frequent in the Pancreatobiliary 10 5 West.23 Consistent with other series,4,7,14,19,24 the current study Intestinal 6 1 found the majority (80%) of intrahepatic IPNBs to be within Gastric 0 1 the left hemiliver. Mean tumour size 3.1 Æ 2.0 6.2 Æ 7.3 As described earlier, IPNB/ICPN is a precursor lesion to (cm Æ SD) invasive cholangiocarcinoma and gallbladder carcinoma that is 8 Non-invasive 0 3 now considered distinct from the more common BilIN. IPNB/ Microinvasion (<5 mm) 4 2 ICPN leads to macroscopically visible intraluminal lesions, Invasion (>5 mm) 12 2 whereas BilIN is typically ‘flat‘ and macroscopically non- tumour forming.8,25 When associated with BilIN, invasive car- T stage cinoma usually demonstrates a desmoplastic stromal reaction Tis 0 3 and tubular pattern.25 In contrast, IPNB/ICPN, which accounts 171for 10–15% of biliary tumours, lead to biliary or cholecystic 2a 71luminal papillary tumour growth.25 As a result of tumour fill- 322ing (with or without mucin), the latter leads to cystic or fusi- 25 N stage form dilation of the bile duct. These lesions appear to be X43more common in the extrahepatic biliary tree (2:1), although they can be found synchronously in both the intra- and extra- 084 hepatic bile ducts.25 IPNB/ICPN would also appear to be much 120 more commonly associated with early symptoms of intermit- 220tent pain, jaundice or cholangitis.

HPB 2015, 17, 811–818 ª 2015 International Hepato-Pancreato-Biliary Association HPB 815

(a) (b)

(d) (c)

(e) (f)

Figure 1 (a) Liver left hepatectomy surgical resection specimen, showing a large papillary tumour filling the dilated intrahepatic bile ducts (thick arrows). (b) Tumour depicted in (a), showing the intraductal papillary neoplasm (IPNB) that fills the bile duct lumen and spreads along the biliary tree (thick arrows). An area of invasive adenocarcinoma into the liver parenchyma is noted, tubular type, poorly differentiated (thin arrow). Hematoxylin and eosin (H&E) stain, 29 magnification. (c) IPNB, common bile duct (thick arrows). Invasive adenocarcinoma is present, tubular type, invading the duct wall into the periductal adipose tissue (thin arrow). H&E stain, 29 magnification. (d) IPNB, pancreatico-billiary type. Tumour cells resemble biliary or pancreatic epithelium, are composed of columnar cells or low cuboidal cells with eosinophilic cytoplasm and round nuclei. H&E stain (200x). (e) IPNB, intestinal type. Tumour cells resemble intestinal adenoma or adenocarcinoma and are characterized by stratified tall columnar cells with some goblet cells. H&E stain (200x). (f) IPNB, gastric type. Tumour cells resemble gastric epithelium and are composed of columnar cells with abundant intracytoplasmic mucin. H&E stain (200x)

Both IPNB and ICPN were included in the present study, as The current paper represents one of the first surgical series to the pathophysiological process of papillary tumour growth describe gallbladder lesions using this system. within the biliary lumen appears to be the same for both In terms of epithelial subtypes, IPNB tumours demonstrated tumour types.8,13 Clinically and anatomically, however, IPNB only pancreatobiliary and intestinal, with a slightly higher prev- and ICPN behave differently in terms of presentation and sur- alence of pancreatobiliary. This is similar to other published gical management. For this reason, both tumours were divided reports,11,14 although gastric and oncocytic subtypes were also when describing presentation, management and outcomes. Fur- seen in small numbers. ICPN tumours, which were not thermore, this is particularly relevant as the description of included in the other North American series, also demonstrate ICPN in the literature is in its infancy.13,23,26 The definition, mostly a pancreatobiliary subtype. This series found no cases described by Adsay et al.13, is of an exophytic intramucosal of an oncocytic epithelial subtype. gallbladder mass greater than 1 cm and composed of dysplastic The 5-year OS and DFS for this series were 51% and 57%, cells forming a lesion distinct from the neighbouring mucosa. respectively, and did not differ significantly between the patients

HPB 2015, 17, 811–818 ª 2015 International Hepato-Pancreato-Biliary Association 816 HPB

Overall survival (n = 23) Disease-free survival (n = 23) With number of subjects at risk With number of subjects at risk 1.0 Censored 1.0 Censored

0.8 0.8

0.6 0.6

0.4 0.4 Survival probability Survival probability

0.2 0.2

0.0 0.0 At risk 23 20 16 13 8 6 2 At risk 23 20 14 8 7 5 2 (a) 0 10 20 30 40 50 60 (b) 010203040 50 60 Months Months

Overall survival by MUC1 (n = 22) Disease-free survival by MUC1 (n = 22) With number of subjects at risk With number of subjects at risk

1.0 Censored 1.0 Censored Logrank P = 0.0941 Logrank P = 0.0515

0.8 0.8

0.6 0.6

0.4 0.4 Survival probability Survival probability

0.2 0.2

0.0 0.0 At risk 09 8 7 6 5 3 2 At risk 09 8 7 6 5 3 2 1 13 11 8 6 2 2 0 113 11 7 2 2 2 0 (c) 0 10 20 30 40 50 60 (d) 0102030405060 Months Months muc1 0 1 muc1 0 1

Figure 2 Kaplan–Meier survival curves for (a) overall survival, (b) disease-free survival, (c) overall survival with/without MUC1, and (d) disease-free survival with/without MUC1

Table 3 Immunohistochemical characteristics of IPNB/ICPN tumours truly have a better biological prognosis, stage for stage. (n = 22) Furthermore, the present study adds to previously published evi- Epithelial subtype MUC1 MUC2 MUC5AC dence that MUC1 immunohistochemical staining is a poor prog- 14 Pancreatobiliarya 11/14 2/14 8/14 nostic marker in IPNB. MUC1 was the only factor that showed Intestinal 2/7 7/7 4/7 a statistically significant difference when comparing survival curves. An increased depth of invasion and R1 resection, both Gastric 0/1 0/1 1/1 well-established negative prognostic factors in surgical oncology, Total 13/22 9/22 13/22 did not reach statistical significance for OS and DFS. a One specimen was not available for immunohistochemical profiling. Interestingly, all five patients with IPNB who experienced n, number of patients. disease recurrence were among those with distal extrahepatic bile duct primary tumours. As no patients in the intrahepatic with IPNB and ICPN. This is similar to the 62 month OS previ- or hilar groups had a recurrence, comparison of survival curves ously reported,14 and slightly higher than the 38% 5-year OS or univariate analysis is not appropriate as there were no reported by Barton et al.11 All of these data indicate that IPNB events. An increased risk of recurrence has not been associated may confer an improved prognosis when compared with typical with the anatomical location in other papers, and it is difficult cholangiocarcinoma, as suggested in the Asian literature. That to draw any conclusions from such a small sample. being said, it remains unclear whether this observation is related It has been suggested from the Asian literature that pre- to an earlier presentation for most patients or whether, these operative cholangioscopy and an intra-operative frozen section

HPB 2015, 17, 811–818 ª 2015 International Hepato-Pancreato-Biliary Association HPB 817

be more routinely used in patients with IPNB.4,10 The hypoth- cholangiocarcinoma: preliminary report based on interobserver agree- esis being that with IPNB’s intraluminal growth, these modali- ment. Pathol Int 55:180–188. ties may better delineate the extent of disease, and increase the 2. Zen Y, Adsay NV, Bardadin K, Colombari R, Ferrell L, Haga H et al. likelihood of an R0 resection. Given that the majority of (2007) Biliary intraepithelial neoplasia: an international interobserver agreement study and proposal for diagnostic criteria. Mod Pathol patients in the North American setting do not have an estab- 20:701–709. lished pre-operative diagnosis of IPNB, this may not prove fea- 3. Zen Y, Sasaki M, Fujii T, Chen TC, Chen MF, Yeh TS et al. (2006) Differ- sible. Further research on the utility of these tools in the ent expression patterns of mucin core proteins and cytokeratins during setting of IPNB is required. A pre-operative cholangioscopy intrahepatic cholangiocarcinogenesis from biliary intraepithelial neopla- was not performed on any of the patients in this series, and an sia and intraductal papillary neoplasm of the bile duct—an immunohis- intra-operative frozen section was used at the surgeon’s discre- tochemical study of 110 cases of hepatolithiasis. J Hepatol 44:350–358. tion. At least two patients had false-negative frozen sections, 4. Ohtsuka M, Shimizu H, Kato A, Yoshitomi H, Furukawa K, Tsuyuguchi T leading to R1 resection. et al. (2014) Intraductal papillary neoplasms of the bile duct. Int J Hepa- The major strength of this paper is the contribution of clini- tol 2014:459091. cal and pathological data about patients with a disease that is 5. Aishima S, Kubo Y, Tanaka Y, Oda Y. (2014) Histological features of poorly described in North America. To date, only two other precancerous and early cancerous lesions of biliary tract carcinoma. J Hepatobiliary Pancreat Sci 21:448–452. papers have described a series of patients this large. The med- 6. Bickenbach K, Galka E, Roggin KK. (2009) Molecular mechanisms of ian follow-up of 30 months is robust, and the patient popula- cholangiocarcinogenesis: are biliary intraepithelial neoplasia and intra- tion is recent. Furthermore, it is only the second North ductal papillary neoplasms of the bile duct precursors to cholangiocar- American paper to report the immunohistochemical features cinoma? Surg Oncol Clin N Am 18:215–224. 14 of these tumours. Limitations of this paper include its small 7. Nakanuma Y, Sato Y, Ojima H, Kanai Y, Aishima S, Yamamoto M et al. sample size, making it difficult to demonstrate statistically sig- (2014) Clinicopathological characterization of so-called “cholangiocarci- nificant results in comparative survival analyses. The retrospec- noma with intraductal papillary growth” with respect to “intraductal pap- tive nature of the study also introduces limitations and biases illary neoplasm of bile duct (IPNB)”. Int J Clin Exp Pathol 7:3112–3122. into the data collection process. For the specimens with inva- 8. Bosman FT, Carneiro F, Hruban RH, Theise ND. (2010) WHO Classifica- sive cancer, all were found to be tubular adenocarcinoma. It tion of Tumours of the Digestive System, 4th edn. Geneva:World Health has been shown that mucinous carcinoma demonstrates Organization. 9. Choi SC, Lee JK, Jung JH, Lee JS, Lee KH, Lee KT et al. (2010) The clini- improved survival compared with tubular,6 but this relation- copathological features of biliary intraductal papillary neoplasms accord- ship cannot be studied in this series. Compared with other ing to the location of tumors. J Gastroenterol Hepatol 25:725–730. studies, fewer IHC markers were utilized. The IHC markers 10. Wan XS, Xu YY, Qian JY, Yang XB, Wang AQ, He L et al. (2013) Intra- chosen were those typically performed at our institution. Also, ductal papillary neoplasm of the bile duct. World J Gastroenterol the majority of patients in this series did not undergo a pre- 19:8595–8604. operative serum CA 19-9 test, and, therefore, its utility as a 11. Barton JG, Barrett DA, Maricevich MA, Schnelldorfer T, Wood CM, prognostic tool cannot be studied. Smyrk TC et al. (2009) Intraductal papillary mucinous neoplasm of the biliary tract: a real disease? HPB 11:684–691. 12. Minagawa N, Sato N, Mori Y, Tamura T, Higure A, Yamaguchi K. (2013) Conclusion A comparison between intraductal papillary neoplasms of the biliary IPNB and ICPN are rare precursor lesions that can affect the tract (BT-IPMNs) and intraductal papillary mucinous neoplasms of the pancreas (P-IPMNs) reveals distinct clinical manifestations and out- entire biliary epithelium. At pathology, the majority of patients comes. Eur J Surg Oncol 39:554–558. have invasive carcinoma, thus warranting a radical resection. 13. Adsay V, Jang KT, Roa JC, Dursun N, Ohike N, Bagci P et al. (2012) In- Prognosis may be superior to traditional cholangiocarcinoma tracholecystic papillary-tubular neoplasms (ICPN) of the gallbladder and gallbladder cancer. Patients with tumours expressing (neoplastic polyps, adenomas, and papillary neoplasms that are ≥ MUC1 appear to have worse OS and DFSs. 1.0 cm): clinicopathologic and immunohistochemical analysis of 123 cases. Am J Surg Pathol 36:1279–1301. Funding sources 14. Rocha FG, Lee H, Katabi N, DeMatteo RP, Fong Y, D’Angelica MI et al. Self-funded. (2012) Intraductal papillary neoplasm of the bile duct: a biliary equiva- lent to intraductal papillary mucinous neoplasm of the pancreas? Hepa- Conflicts of interest tology 56:1352–1360. None to declare. 15. Chen TC, Nakanuma Y, Zen Y, Chen MF, Jan YY, Yeh TS et al. (2001) Intraductal papillary neoplasia of the liver associated with hepatolithia- References sis. Hepatology 34:651–658. 1. Zen Y, Aishima S, Ajioka Y, Haratake J, Kage M, Kondo F et al. (2005) 16. Suh KS, Roh HR, Koh YT, Lee KU, Park YH, Kim SW. (2000) Clinico- Proposal of histological criteria for intraepithelial atypical/proliferative pathologic features of the intraductal growth type of peripheral cholan- biliary epithelial lesions of the bile duct in hepatolithiasis with respect to giocarcinoma. Hepatology 31:12–17.

HPB 2015, 17, 811–818 ª 2015 International Hepato-Pancreato-Biliary Association 818 HPB

17. Takanami K, Yamada T, Tsuda M, Takase K, Ishida K, Nakamura Y et al. 22. Jung G, Park KM, Lee SS, Yu E, Hong SM, Kim J. (2012) Long-term (2011) Intraductal papillary mucininous neoplasm of the bile ducts: multimo- clinical outcome of the surgically resected intraductal papillary neo- dality assessment with pathologic correlation. Abdom Imaging 36:447–456. plasm of the bile duct. J Hepatol 57:787–793. 18. Kim KM, Lee JK, Shin JU, Lee KH, Lee KT, Sung JY et al. (2012) Clini- 23. Jang KT. (2014) Intracholecystic papillary-tubular neoplasm of the gall- copathologic features of intraductal papillary neoplasm of the bile duct bladder. Pathology 46(Suppl 2):S24. according to histologic subtype. Am J Gastroenterol 107:118–125. 24. Paik KY, Heo JS, Choi SH, Choi DW. (2008) Intraductal papillary neo- 19. Kubota K, Nakanuma Y, Kondo F, Hachiya H, Miyazaki M, Nagino M et al. plasm of the bile ducts: the clinical features and surgical outcome of 25 (2014) Clinicopathological features and prognosis of mucin-producing bile cases. J Surg Oncol 97:508–512. duct tumor and mucinous cystic tumor of the liver: a multi-institutional study 25. Kloppel G, Adsay V, Konukiewitz B, Kleef J, Schlitter AM, Esposito I. by the Japan Biliary Association. J Hepatobiliary Pancreat Sci 21:176–185. (2013) Precancerous lesions of the biliary tree. Best Pract Res Clin Gas- 20. Park SY, Roh SJ, Kim YN, Kim SZ, Park HS, Jang KY et al. (2009) troenterol 27:285–297. Expression of MUC1, MUC2, MUC5AC and MUC6 in cholangiocarcino- 26. Yamamoto K, Yamamoto F, Maeda A, Igimi H, Yamamoto M, Yamagu- ma: prognostic impact. Oncol Rep 22:649–657. chi R et al. (2014) Tubulopapillary adenoma of the gallbladder accom- 21. Yang J, Wang W, Yan L. (2012) The clinicopathological features of intra- panied by bile duct tumor thrombus. World J Gastroenterol 20:8736– ductal papillary neoplasms of the bile duct in a Chinese population. Dig 8739. Liver Dis 44:251–256.

HPB 2015, 17, 811–818 ª 2015 International Hepato-Pancreato-Biliary Association