Abdominal Imaging
ª Springer Science+Business Media, LLC 2008 Published online: 19 January 2008
Abdom Imaging (2009) 34:217–224 DOI: 10.1007/s00261-008-9363-0
Adenosquamous carcinoma of the extrahepatic bile duct: clinicopathologic and radiologic features Kyung Won Kim,1 Se Hyung Kim,1 Min A. Kim,2 Jeong Min Lee,1 Hee Sun Park,1 Ji Young Kim,1 Jae Young Lee,1 Joon Koo Han,1 Byung Ihn Choi1 1
Department of Radiology and Institute of Radiation Medicine, Seoul National University Hospital, 28 Yeongon-dong, Jongno-gu, Seoul 110-744, Korea 2 Department of Pathology, Seoul National University Hospital, 28 Yeongon-dong, Jongno-gu, Seoul 110-744, Korea
Abstract Background: To retrospectively analyze computed tomography (CT) features of adenosquamous carcinoma (ASC) of the extrahepatic bile duct and to correlate them with pathologic findings. Methods: Six patients who underwent radical surgical resection for ASC of extrahepatic bile duct were included. CT and pathologic findings were analyzed and correlated with each other. Results: Five ASCs were periductal infiltrative type and the remaining one was intraductal polypoid type. Histologically, ASC was composed of squamous and glandular components which mixed in an intermingled pattern and was characterized by a high local aggressiveness mainly due to the squamous component. The most common CT findings were a segmental wall thickening with delayed enhancement. Overall, CT features were well correlated with pathologic findings. However, CT tended to underestimate radial extent of the tumor compared with histologic findings. Conclusions: Histopathologically, ASC of the extrahepatic bile duct is characterized by greater local aggressiveness. However, their CT findings are not quite different from those of usual adenocarcinoma due to the unique geographic characteristic of being located at a very limited space. Key words: Adenosquamous carcinoma—Bile duct cancer—Extrahepatic bile duct—CT—Pathology
Correspondence to: Se Hyung Kim; email:
[email protected]
Extrahepatic bile duct cancer is a relatively rare and lethal entity of biliary tract malignancy of which the majority is adenocarcinoma. Adenosquamous carcinoma (ASC) of the biliary tract is a rarer histopathologic subtype, accounting for 3%–4.7% of all biliary tract malignancies [1, 2]. ASC of the bile duct has been described as an infiltrative bile duct cancer composed of a squamous component and a glandular component and is notorious for its poor prognosis [1, 3, 4]. However, previously published papers have mostly been in the form of case reports dealing with clinicopathologic features, and radiologic findings have been only generally mentioned [1, 3–8]. Moreover, in those papers, the number of patients who undergo resection at a single institution is limited [1]. For these reasons, the clinicopathologic and radiologic characteristics have not yet been fully clarified. The purpose of this study, therefore, is retrospectively to assess the radiologic findings of dynamic computed tomography (CT) and direct cholangiography and to evaluate the clinicopathologic features of ASC of the extrahepatic bile duct.
Materials and methods Patients This retrospective study did not require institutional review board approval or informed patient consent. From January 2000 to December 2006, a computerized search of our hospitalÕs pathology files revealed that 181 patients with extrahepatic bile duct cancer were surgically treated at our institution. Of these, histopathologic examination confirmed six cases (6/181, 3.3%) of ASC according to the World Health Organization classifica-
K. W. Kim et al.: Adenosquamous carcinoma of the bile duct
218
Table 1. Clinical findings in six patients with ASC of extrahepatic bile duct Case no.
Age (years)/sex
Symptoms
1 2 3 4 5 6
62/M 64/M 49/F 71/M 71/M 67/M
Jaundice Jaundice, anorexia Jaundice, abdominal pain Jaundice Jaundice, anorexia Jaundice
T-Bila (mg/dL)
TNM stage
AJCC stage
Treatment
Outcome (survival, months)b
3.7 5.3 14.2 3.8 22.0 14.9
pT3N0M0 pT3N0M0 pT3N0M0 pT3N0M0 pT3N1M0 pT3N0M0
IIa IIa IIa IIa IIb IIa
PPPD PPPD + CCRT PPPD + CCRT PPPD PPPD + CCRT Whipple + CCRT
NED (73) DOD (18) NED (62) DOD (18) NED (16) DOD (23)
a
T-Bil, total bilirubin level at initial manifestation Number in parenthesis is the interval between the time of initial manifestation and the last follow-up or death PPPD, pylorus-preserving pancreaticoduodenectomy; Whipple, WhippleÕs pancreaticoduodenectomy; CCRT, concurrent chemoradiotherapy; NED, (alive with) no evidence of disease; DOD, died of disease b
tion of tumors of the gallbladder and extrahepatic bile ducts [9]. Five patients were men and one was a woman, and the ages ranged from 49 to 71 years (mean age 64.0 years). All patients underwent preoperative dynamic abdominal CT scan and direct cholangiography. The mean interval between CT or direct cholangiography and surgery was 13.6 and 11.3 days, respectively.
Clinical features Clinical data regarding symptoms, evidence of laboratory abnormalities, type of surgical method, and history of adjuvant treatment were analyzed from the review of medical records and was tabulated for each case in Table 1. The treatment outcome such as survival period in months and the presence of tumor recurrence were evaluated by using the total clinical information of radiologic, pathologic, and laboratory findings.
was described as intrapancreatic common bile duct (CBD), suprapancreatic CBD, supra- and intrapancreatic CBD, or common hepatic duct. Wall thickness was measured by microscopic examination at the thickest portion of the involved bile duct on the transverse section. The coexistent pattern of the two components was defined and classified as follows: (a) collision pattern, containing separate and coincidental areas of adenocarcinoma, and squamous cell carcinoma (double primary cancer), (b) combined pattern, in which two components exist contiguously but independently, (c) intermingled pattern, where the two components are mixed and intermingled intimately. In addition, the TNM stage of the lesion according to the American Joint Committee on Cancer guideline was determined based on the operative findings and the results of histological examination [11].
CT acquisition Histopathologic findings Histopathologic records were available for all patients except for one. All histopathologic materials were rereviewed by an experienced hepatobiliary pathologist (M.A.K.). All surgical specimens were sectioned at approximately 5 mm intervals in the transverse plane and were stained with hematoxylin and eosin. ASC was defined as a tumor with two malignant components: squamous and glandular. The squamous component is characterized by whorls, cytoplasmic keratohyalin, keratin pearls, and/or intercellular bridges [9]. According to the macroscopic appearance, the ASC of the bile duct in each patient was classified into one of three types: mass forming, intraductal polypoid, and periductal infiltrating [10]. The location of the tumor and the length of the involved bile duct were evaluated on gross specimen. The slides were also assessed for the maximum transverse wall thickness of the involved bile duct, proportion of squamous component, coexistent pattern of the two histologic components, and extent of tumor spread such as periductal infiltration and invasion to the adjacent organ. The location of the tumor
In all patients, triple-phase CT examinations including precontrast, arterial, and portal venous phases were performed with a single-detector row CT (Somatom Plus 4: Siemens Medical Solutions, Erlangen, Germany) (n = 4) or an 8-row multidetector CT scanner (Lightspeed Ultra: GE Medical Systems, Milwaukee, WI, USA) (n = 2). The respective scanning parameters used for single-detector row CT and 8-MDCT scanners were as follows: detector configuration, 1 · 5 and 8 · 1.25 mm; slice thickness/reconstruction increment, 5/ 5 and 2.5 mm/2.5 mm; table speed, 7 and 13.5 mm/ rotation; rotation time, 1 and 0.5 s; 165 and 250 effective mAs; 120 kVp; and matrix, 512 · 512. Each patient received 120 mL of a non-ionic contrast material (iopromide, Ultravist 370; Schering Korea, Seoul, Korea) through the forearm vein at a rate of 3–4 mL/s using an automatic injector. For arterial phase, a delay of 11 or 13 s was used for single-detector row CT or 8-MDCT scanners, respectively, after the maximal attenuation of the aorta reached 100 HU using a bolus tracking technique. Portal venous phase scanning was performed 60 s after contrast agent administration.
K. W. Kim et al.: Adenosquamous carcinoma of the bile duct
Direct cholangiography Direct cholangiography was performed in all patients by an experienced endoscopist or abdominal interventional radiologist: endoscopic retrograde cholangiopancreatography (n = 3) and percutaneous transhepatic biliary drainage tube cholangiography (n = 3).
Image analysis Computed tomography scans were evaluated on a Picture Archiving and Communications System retrospectively and jointly by two radiologists (S.H.K. and K.W.K). Disagreement was minor and resolved through consensus. The analyzed parameters on CT are as follows: (a) location of the tumor, (b) length and maximum transverse wall thickness of the involved bile duct, (c) radial extent of tumor spread such as periductal infiltration or invasion to adjacent organs, (d) enhancement pattern of the involved ductal wall during hepatic arterial and portal venous phases. Location of the tumor was described as intrapancreatic, suprapancreatic, supra- and intrapancreatic CBD, or common hepatic duct. The lesion length was calculated by multiplying the slice thickness by the number of CT sections containing the lesion. Enhancement pattern was evaluated using subjective and objective methods. Subjectively, the relative enhancement degree of the lesion was classified as high, intermediate, or low in comparison to that of the upstream normal CBD wall. For an objective analysis, CT numbers were obtained by means of region-of-interest cursors placed on all wall portions of the bile duct lesion and on the dilated upstream CBD wall. These cursors were carefully placed to encompass as much of the lesion as possible and to avoid adjacent structures. Region of interest measurements were obtained after magnifying the images for both arterial and portal venous phase CT scans. A difference of more than 20 HU in mean CT numbers between the involved ductal wall and the upstream normal CBD wall was considered meaningful. Direct cholangiographic findings were interpreted with regard to the severity of biliary stenosis and the appearance of stenosis (abrupt narrowing with shouldering or smooth tapering), wall irregularity, intraluminal filling defect, and upstream bile duct dilatation. Based on CT and cholangiography findings, ASC of the bile duct in each patient was classified into one of three types: periductal infiltrative, exophytic, or intraductal polypoid [12].
Results Clinical features The clinical features of the six patients with ASC of the extrahepatic bile duct are presented in Table 1. All six patients presented with obstructive jaundice; their serum
219
total bilirubin was elevated and ranged from 3.7 to 22.0 mg/dL (mean 10.7 mg/dL). Serum AST, ALT, and alkaline phosphatase were also elevated in all patients. Serum CA 19–9 was elevated in five patients (mean 334.8 U/mL; range 38–900 U/mL) and CEA was abnormal in one patient (11.8 ng/mL). The staging of the lesion according to the American Joint Committee on Cancer guidelines revealed that all the patients were stage II: IIa in five patients and IIb in one. No evidence of distant metastasis was found in all patients. Five of six patients underwent pylorus-preserving pancreaticoduodenectomy (PPPD) with regional lymph node dissection, and the remaining patient underwent standard pancreaticoduodenectomy (WhippleÕs operation). In all patients, the immediate postoperative course was uneventful and four of the six patients underwent adjuvant concurrent chemoradiotherapy (CCRT) including external beam radiation therapy and chemotherapy with 5-fluorouracil after the surgical resection. With regard to the outcome, three of six patients are alive without evidence of disease at the time of writing, while the other three patients died of recurred tumors and consequent distant metastasis in the second year following resection. No recurrence occurred within 1 year in all patients. However, once the recurrence occurred, the patients died within 6 months.
Histopathologic features Histopathologic findings of six patients are summarized in Table 2. The cut surface of the resected gross specimens showed ill-defined grayish white (n = 5) and greenish white (n = 1) solid tumors. Except for one tumor of intraductal polypoid type (Case 2), all tumors were of the periductal infiltrating type. Location of the tumors was intrapancreatic CBD in two patients, supraand intrapancreatic CBD in three, and common hepatic duct in one. On histopathologic examination, the proportions of the squamous component were variable, ranging from 10% to 70%. In all five patients who had available pathologic specimens, both squamous and glandular components were mixed in an intermingled pattern (Fig. 1). In all five patients including the patient with intraductal polypoid tumor, periductal infiltration of the tumor cells into perimuscular fibrous and fatty tissue was shown. Perineural and serosal invasion were found in all patients and invasion into adjacent pancreatic parenchyma was noted in five patients. The main component showing periductal infiltration was squamous cell component in four patients and glandular adenomatous carcinoma component in one. In one patient, regional lymph node metastasis occurred from direct invasion by the squamous cell carcinoma component. In all patients, tumor cells were not found at the resection margin. No patients showed evidence of biliary stone disease, squamous metaplasia of the bile
K. W. Kim et al.: Adenosquamous carcinoma of the bile duct
220
Table 2. Histopathologic findings of six patients with ASC of extrahepatic bile duct Case no.
Type
Location
1
Periductal infiltrating Intraductal polypoid Periductal infiltrating Periductal infiltrating Periductal infiltrating Periductal infiltrating
Intrapancreatic CBD Intrapancreatic CBD Supra and intrapancreatic Supra and intrapancreatic Supra and intrapancreatic Common hepatic
2 3 4 5 6
Lesion length (cm)
Wall thickness (mm)
Proportion of squamous cell component (%)
Coexistent pattern
Invasion (N/S/P)
Component of periductal infiltration
1.8
5
40
Intermingled
+/+/+
Predominantly SQCC
1.5
5
60
Intermingled
+/+/+
Predominantly SQCC
2.5
7
70
Intermingled
+/+/+
Predominantly SQCC
4.0
4
10
Intermingled
+/+/+
3.0
6
30
Intermingled
+/+/+
2.0
NA
NA
NA
+/+/)
Mainly focal Mainly focal NA
CBD CBD CBD duct
ADC with SQCC SQCC with ADC
N, neural or perineural invasion; S, serosal invasion; P, pancreas invasion; SQCC, squamous carcinoma component; ADC, adenocarcinoma component; CBD, common bile duct; NA, not available
duct, or bile duct anomaly such as choledochal cyst or anomalous pancreatobiliary duct union.
Radiologic findings
Fig. 1. Microscopic feature of ASC of the bile duct. (A) On histopathologic examination (H&E stain, original magnification 100·), glandular component from the adenocarcinoma (arrows) and squamous carcinoma component (asterisk) are intimately admixed. However, a tendency toward the growth from the surface glandular differentiation and squamous differentiation in deeper areas is seen. (B) Area of squamous cell carcinoma at higher magnification (original magnification 400·) shows irregular nests of polygonal cells with no gland formation and characteristic cytoplasmic kerato-hyalin and keratin pearl (asterisk).
Computed tomography findings of six patients are summarized in Table 3. In five patients, the lesion was diagnosed as bile duct cancer of periductal infiltrative type seen as a segmental thickened ductal wall encircling the lumen with delayed enhancement (Figs. 2, 3). In the remaining one patient, the lesion was classified as intraductal polypoid type (Fig. 4). The involved ductal wall in all patients was thick (>1.5 mm). Periductal fat infiltration, manifested as a fatty strand in the periductal soft tissue on CT scan, was depicted in all six lesions even in the lesion of intraductal polypoid type (Fig. 2). Pancreas invasion was found in three patients (Figs. 2, 4). Hyperattenuation of the lesion on portal venous phase was shown in all patients (Figs. 2, 3), on the other hand, the enhancement of the lesion on hepatic arterial phase was variable (hypoattenuation in three patients and hyperattenuation in the remaining three) (Fig. 4). At direct cholangiography, all six patients showed severe segmental stenosis of the involved bile duct and dilatation of the upstream bile duct. With regard to the appearance of stenosis, five patients of periductal infiltrative type showed abrupt narrowing with shouldering as well as wall irregularity. One patient with intraductal polypoid type tumor showed intraluminal filling defect with segmental ductal wall irregularity.
CT-pathologic correlation The measured average length of the lesions was 26.7 mm on CT and 27.2 mm on gross pathologic specimen. The difference was not statistically significant (P = 0.85). The difference between the wall thickness measured on CT (5.3 mm) and pathology (5.4 mm) was also not sig-
221
K. W. Kim et al.: Adenosquamous carcinoma of the bile duct
Table 3. CT findings and CT-pathologic correlation
Type Location
Lesion length (mm)a Wall thickness (mm)a Periductal fat infiltration or serosal invasion Pancreas invasion Regional lymph node metastasis Enhancement pattern (arterial/portal)
Periductal infiltrative Intraductal polypoid Intrapancreatic CBD Suprapancreatic CBD Supra- and intraapancreatic CBD Common hepatic duct
Low/High High/High
CT findings
Pathologic findings
5 1 1 2 2 1 26.7 ± 5.1 5.3 ± 1.3b (4.9 ± 1.5)c 6 3 1 3 3
5 1 2 0 3 1 27.2 ± 10.2 5.4 ± 1.1b 6 5 1
Note: Except where indicated, data are number of patients CBD, common bile duct a Data are means ± standard deviations b Data obtained from Cases 1–5, because the histopathologic review was not available in Case 6 c Data obtained from Cases 1–6
nificant (P = 0.91). With respect to the location of the tumor, the accuracy of CT was 50% (three of six patients). Although periductal infiltration was depicted on both CT and histopathologic examination in all patients, the extent of infiltration on histopathologic examination tended to be more extensive than on CT because of the microscopic scattering of malignant cells in the periductal soft tissue. Indeed, two patients in whom pancreas invasion was diagnosed on histopathologic examination were misinterpreted as having no evidence of pancreas invasion on CT (Fig. 3).
Discussion Adenosquamous carcinoma of the extrahepatic bile duct consists of two distinct histological components: squamous and glandular [9]. Not much is known about the etiology of this tumor. Regarding the pathogenesis of the ASC of the liver and biliary tract, various theories have been suggested. Among them, the two following hypotheses are generally accepted as the most reliable theories: (a) preexistent adenocarcinoma undergoing squamous metaplasia or squamous transformation and (b) squamous metaplasia of bile duct epithelium caused by chronic inflammation due to infection, stones, or choledochal cyst undergoing a malignant transformation [6, 13–15]. In our series, neither squamous metaplasia of the bile duct nor its predisposing factors such as biliary stone, infection, and biliary tract anomaly were observed in any case. Furthermore, squamous and glandular components coexisted in an intermingled pattern. In addition, the proportion of the two components was variable and ranged from 10% to 70%. Based on the above three findings, our results seem to favor squamous transformation of preexistent adenocarcinoma as a pathogenesis of ASC of extrahepatic bile duct.
In several previous reports regarding ASC, the squamous component has been reported to have greater proliferation capacity and a tendency to be more locally aggressive compared to the glandular (adenocarcinoma) component [4, 16]. On histopathologic examinations in our study, all lesions were locally infiltrative with serosal, neural or perineural, and pancreatic invasion. As expected, local invasion was mainly caused by the squamous component of the tumor, which was much more frequently found or scattered in the periductal soft tissue compared to the glandular component. Even lymph node involvement, found in one patient, was caused by the direct spread of the squamous component, not by distant lymphangitic metastasis. These results partly coincide with previous reports on ASC of the gallbladder that the primary spread of squamous cell carcinoma of the gallbladder was due to direct invasion with few metastases to the lymph nodes or other organs, suggesting relatively lower metastatic potential of this rare ASC than adenocarcinoma [16]. Given that ASC of the gallbladder and liver frequently presents with a bulkier mass at an advanced stage compared to usual adenocarcinoma [16–20], we assumed that ASC of the extrahepatic bile duct also manifests itself as a bulky mass. However, contrast to our expectation, gross morphology and radiologic features of ASC of the extrahepatic bile duct were not bulkier or more massive but were similar to those of adenocarcinoma of the bile duct. It might be due to the fact that the extrahepatic bile duct is located at a very limited and narrow space unlike the liver and gallbladder. Consequently, clinical symptoms and signs such as obstructive jaundice can present at an earlier stage compared to other organs such as the liver or gallbladder. These unique features of the extrahepatic bile duct and its lesions might lead to their less aggressive or less bulky appearance at the time of initial presentation.
222
K. W. Kim et al.: Adenosquamous carcinoma of the bile duct
Fig. 2. A 71-year-old man with ASC of the infiltrative type at the extrahepatic bile duct (Case 5). (A) Portal phase CT scan obtained at suprapancreatic portion shows a strongly enhanced and thickened bile duct (arrows). Note subtle periductal infiltration suggesting periductal tumor infiltration. Right intrahepatic bile ducts (arrowheads) are dilated. (B) CT scan of the same phase obtained at the level of pancreatic head shows irregular thickening and enhancement of the intrapancreatic bile duct (arrows). The estimated length of the tumor on CT was 2.8 cm. The outer margin of the enhancing tumor shows a lobulating appearance, which indicates direct invasion to the pancreatic parenchyma by the tumor. (C) On CT scan obtained 0.75 cm cranial to (A), a 1.5 cm sized, low-attenuating lymph node (arrow) is visu-
alized at portocaval space. This necrotic lymph node is closely abutted against the thickened and enhanced ductal wall (arrowheads). (D) Gross specimen obtained after PPPD shows that the tumor (arrows) is located at both supra- and intrapancreatic common bile duct and measured 3 cm in length pathologically. Periductal and pancreatic parenchymal invasion were also found on microscopic examination (not shown). (E) On microscopic examination (H&E stain, 100·), squamous components (asterisk) in the perimuscular connective tissue of the bile duct directly infiltrate into the adjacent lymph node (arrows). This finding suggests that lymph node involvement is caused by direct spread of the tumor cells, not by lymphangitic metastasis. Finally, ASC of T3 and N1 stage was diagnosed.
K. W. Kim et al.: Adenosquamous carcinoma of the bile duct
223
Fig. 3. A 71-year-old man with ASC of the extrahepatic bile duct (Case 4). Oblique coronal, multiplanar reconstruction CT image obtained at portal venous phase shows enhancing wall thickening (arrows) of the suprapancreatic bile duct. The longitudinal length of the tumor on CT measured 3.5 cm. On pathologic examination, the longitudinal tumor extent was 4 cm in length. The tumor was located on both supra- and intrapancreatic bile duct and invaded into the pancreatic parenchyma as well as periductal connective tissue (not shown).
In this study, dynamic CT and direct cholangiographic findings of ASC of the extrahepatic bile duct were indistinguishable from those of usual adenocarcinoma of the bile duct. Adenocarcinoma of the bile duct, as documented in previous studies, typically manifests as one of three types: periductal infiltrative, exophytic, or intraductal polypoid [12]. Infiltrative type tumors, the most common type of extrahepatic bile duct carcinomas, manifest as focal wall thickening with early or delayed enhancement; and polypoid type tumors manifest as intraductal low attenuating, soft tissue masses on CT scan [12, 21]. This is consistent with the CT findings of ASC in our study. Furthermore, all the ASCs of the extrahepatic bile duct were detected at a relatively early stage (stage IIa or IIb) before the tumors were able to grow into a bulky mass. Preoperative determination of lesion extent is crucial in deciding further treatment plan, because complete surgical resection is considered to be the only curable treatment option for bile duct cancer [22]. Currently, CT is known to provide acceptable diagnostic accuracy for preoperative tumor staging in biliary malignancy [22]. Generally, CT findings in our study were well correlated with those of histopathologic features in terms of longitudinal and radial extent of the tumor, even though CT showed some weakness in the evaluation of the presence of pancreatic invasion and localization of the tumor. Indeed, even though the length of the lesion on CT was not quite different from that on pathology, tumor location was misinterpreted on CT in three patients. In two patients, CT incorrectly diagnosed the tumor location as
Fig. 4. A 64-year-old man with ASC of mainly intraductal polypoid type (Case 2). (A) CT scan obtained at hepatic arterial phase shows a low-attenuating intraductal soft-tissue mass (arrows) filling in the intrapancreatic common bile duct. Intrapancreatic bile duct is also thickened and the outer margin of the thickened ductal wall is not distinguishable to the pancreatic parenchyma, therefore, pancreatic invasion is suspected on CT. (B) On CT scan 1.5 cm cranial to (A), suprapancreatic bile duct is also thickened and enhanced (arrows). However, pathologic examination revealed that the tumor was confined only to the intrapancreatic duct (not shown). Such overestimation of tumor involvement at upstream bile duct on CT might be responsible for upstream cholangitis caused by malignant obstruction due to the tumor.
suprapancreatic CBD lesions for supra- and intrapancreatic CBD lesions. For the remaining one patient, CT incorrectly diagnosed the tumor location as supra- and intrapancreatic CBD lesions for the intrapancreatic CBD lesion. From that result, we found that CT scan had a tendency to overestimate the extent of a lesion superiorly and to underestimate it inferiorly. This tendency is not surprising because upstream cholangitis frequently occurs at the upper bile duct of the stenotic lesion and is responsible for the overestimation of bile duct cancer. On the contrary, intrapancreatic CBD involvement of the tumor, especially microscopic involvement can sometimes be underestimated because tumoral enhancement
224
K. W. Kim et al.: Adenosquamous carcinoma of the bile duct
on portal venous phase can be obscured by strong enhancement of adjacent normal pancreatic parenchyma. Based on their locally aggressive nature with less frequent distant metastasis, ASC of the extrahepatic bile duct appears more suitable for resection than adenocarcinoma. Besides, the tendency of ASC of the extrahepatic bile duct to be detected at an earlier stage may guarantee earlier resection and provide better surgical outcomes. However, higher tendency of ASC toward periductal, perineural infiltration, and pancreatic invasion may worsen the prognosis. Therefore, emphasis on aggressive local control such as radical curative resection could be made for better tumor clearance and a better outcome. Our study in which all patients underwent radical resection (PPPD or WhippleÕs operation) also suggests the better outcome of radical resection. Three of six patients died within 2 years due to local recurrence and consequent distant metastasis, on the other hand, three patients are still alive in a disease-free state (two are alive after more than 5 years and one is alive 16 months after surgery). If the patient who was followed up for only 16 months is not considered, the 5-year survival rate in our study would be 40% (2/5), which is better than 17.7% in a previous report on ASC of the extrahepatic bile duct [1]. Our study has several limitations. First, only a small number of patients were included due to the rarity of ASC. Even though our study is one of the largest series addressing ASC of the bile duct in a single institution and we extensively reviewed clinicopathologic and radiologic findings, our results could not be generalized. Second, a variety of CT scanners and protocols was used because of the retrospective nature of data collection. In conclusion, even though ASC is histopathologically characterized by their greater local aggressiveness mainly due to the squamous component and is different from usual adenocarcinoma of the bile duct; CT findings of ASC in the extrahepatic bile duct are quite similar to those of usual adenocarcinoma. This may be due to the unique geographic characteristic of being located at a very limited space. References 1. Okabayashi T, Kobayashi M, Nishimori I, et al. (2005) Adenosquamous carcinoma of the extrahepatic biliary tract: clinicopathological analysis of Japanese cases of this uncommon disease. J Gastroenterol 40:192–199
2. Cao SH (1992) Extrahepatic bile duct cancer. Report of 106 cases. Chin Med J (Engl) 105:622–629 3. Lantsberg L, Khodadadi J, Goldstein J (1986) Adenosquamous carcinoma of the common bile duct: a case report. J Surg Oncol 33:140–142 4. Ueno N, Sano T, Kanamaru T, et al. (2002) Adenosquamous cell carcinoma arising from the papilla major. Oncol Rep 9:317–320 5. Hayashi N, Yamaguchi Y, Ogawa M (1993) Concomitant adenosquamous carcinoma of the common bile duct and early adenocarcinoma of the gall-bladder. J Gastroenterol Hepatol 8:607–612 6. Okamura K, Hayakawa H, Kuze M, et al. (2000) Triple carcinomas of the biliary tract associated with congenital choledochal dilatation and pancreaticobiliary maljunction. J Gastroenterol 35:465– 471 7. Yavuz E, Kapran Y, Ozden I, et al. (2000) Pancreatobiliary adenosquamous carcinoma (report of two cases). Pathologica 92:323– 326 8. Sewkani A, Kapoor S, Sharma S, et al. (2005) Squamous cell carcinoma of the distal common bile duct. JOP 6:162–165 9. Albores-Saavedra J, Scoazec JC, Wittekind C, et al. (2000) Carcinoma of the gallbladder and extrahepatic bile ducts. In: Hamilton (ed) World Health Organization Classification of Tumours. Pathology and Genetics. Tumours of the Digestive System. IARC Press: Lyon, pp 203–217 10. Liver Cancer Study Group of Japan(2000). The General Rules for the Clinical and Pathological Study of Primary Liver Cancer4 Tokyo: Kanehra 11. Greene FL, Page DL, Fleming ID, et al. (2002). AJCC Cancer Staging Manual6 New York: Springer 12. Han JK, Choi BI, Kim AY, et al. (2002) Cholangiocarcinoma: pictorial essay of CT and cholangiographic findings. Radiographics 22:173–187 13. Yokota H, Matoba M, Tonami H, et al. (2007) Imaging findings in primary adenosquamous carcinoma of the liver: a case report. Clin Imaging 31:279–282 14. Iemura A, Yano H, Mizoguchi A, et al. (1992) A cholangiocellular carcinoma nude mouse strain showing histologic alteration from adenocarcinoma to squamous cell carcinoma. Cancer 70:415–422 15. Takahashi H, Hayakawa H, Tanaka M, et al. (1997) Primary adenosquamous carcinoma of liver resected by right trisegmentectomy: report of a case and review of the literature. J Gastroenterol 32:843–847 16. Oohashi Y, Shirai Y, Wakai T, et al. (2002) Adenosquamous carcinoma of the gallbladder warrants resection only if curative resection is feasible. Cancer 94:3000–3005 17. Suzuki E, Hirai R, Ota T, et al. (2002) Primary adenosquamous carcinoma of the liver: case report. J Hepatobiliary Pancreat Surg 9:769–773 18. Yeh CN, Jan YY, Chen MF (2003) Adenosquamous carcinoma of the liver: clinicopathologic study of 10 surgically treated cases. World J Surg 27:168–172 19. Wagreich JM, Shapiro RS, Glajchen N, et al. (1998) MRI findings in adenosquamous carcinoma of the gallbladder. Clin Imaging 22:130–133 20. Chan KM, Yu MC, Lee WC, et al. (2007) Adenosquamous/squamous cell carcinoma of the gallbladder. J Surg Oncol 95:129–134 21. Choi SH, Han JK, Lee JM, et al. (2005) Differentiating malignant from benign common bile duct stricture with multiphasic helical CT. Radiology 236:178–183 22. Khan SA, Thomas HC, Davidson BR, et al. (2005) Cholangiocarcinoma. Lancet 366:1303–1314
