—  SPECIALTY CONFERENCE  —

Liver Pathology

Case 2 - Oncocytic Variant of Cholangiocarcinoma

Jason Daniels
bbott Northwestern Hospital and Virginia Piper Cancer Institute
Minneapolis, MN





Click on slide thumbnail images for an enlarged view.

If you have any difficulties viewing these slides, email the webmaster.



Case History:
A 69-year-old female presents with obstructive jaundice. Her past medical history is unremarkable with no known liver disease.

Recent laboratory values are as follows:
  • AST (SGPT): 141 (10-42 IU/L)

  • ALT (SGOT): 212 (10-40 IU/L)

  • Alkaline Phosphatase: 571 (34-104 IU/L)

  • Total protein: 7.4 (6.0-8.0 g/dL)

  • Albumin: 4.3 (3.5-5.0 g/dL)

  • Total bilirubin: 4.9 (<1.6 mg/dL)

  • Platelets: 256,000/mm3 (140,000-440,000/mm3).
She underwent an endoscopic ultrasound which demonstrated a normal pancreas and distal bile duct. The pancreatic parenchyma appeared normal. In the liver there was a left lobe and porta hepatis mass. An ERCP demonstrated a narrowing of the common bile duct at the right duct. The left duct was completely obliterated. A cholangiogram revealed a normal CBD with complete obstruction of the left hepatic duct and stenosis involving the bifurcation extending 3 to 4 cm to the right biliary system.

A CT scan revealed a 10 x 8.5 x 8.5 cm hypervascular central mass arising from the medial left hepatic lobe. The focus was primarily in segment 4B. Clinically it appeared to be hepatocellular carcinoma (HCC). Serum tumor markers include: AFP: 3.2 ng/ml, CA 19-9: 19 (0-36 U/ml), and CEA: 1.0 ng/ml. A biopsy was obtained (see microscopic images).

She then underwent a left trisegmentectomy with radical bile duct excision. The tumor was found to be invading into the bile duct. She was reconstructed with a Roux-en-Y hepaticojejunostomy.


Case 2 - Figure 1
H&E (4x)

Case 2 - Figure 2
H&E (10x)

Case 2 - Figure 3
H&E (20x)

Case 2 - Figure 4
H&E (40x)

Case 2 - Figure 5
Hep Par 1 (40x)

Case 2 - Figure 6
CD10 (40x)

Case 2 - Figure 7
Cytokeratin 7 (40x)

Case 2 - Figure 8
Cytokeratin 19 (40x)


Differential Diagnosis:
The differential diagnosis in this case was primarily to evaluate for a primary liver tumor, as the patient presented with a solitary mass in the liver without any other signs of malignancy elsewhere. The clinicians in this case were suspecting hepatocellular carcinoma. Some features that were unusual for hepatocellular carcinoma included her lack of history of any background liver disease, and no evidence of cirrhosis either clinically or radiographically.

To make this case more difficult, the histologic appearance mimicked hepatocellular carcinoma. These histologic findings included trabeculae and nests of tumor cells with abundant eosinophilic granular cytoplasm, a relatively monotonous appearance, and rare mitotic activity. Background liver parenchyma showed no evidence of chronic or progressive liver disease, as there was no evidence of advanced fibrosis or cirrhosis. Due to the apparent lack of pre-existing liver disease, we felt it worthwhile to try to confirm hepatocellular origin by immunohistochemical staining.

Immunohistochemical stains showed the tumor cells were positive for cytokeratin AE1/AE3, cytokeratin 7, cytokeratin 19, polyclonal CEA, and were negative for cytokeratin 20, CD 10, Hep-Par 1, TTF-1, thyroglobulin, chromogranin, and synaptophysin. This immunohistochemical staining pattern suggested bile duct origin, and argued against hepatocellular carcinoma.

Final Diagnosis:
Oncocytic Variant of Cholangiocarcinoma

Case Discussion and Review of Literature:
Our case highlights an unusual variant of intrahepatic cholangiocarcinoma with oncocytic features. Intrahepatic cholangiocarcinoma typically occurs in a non-cirrhotic liver, though the patient may have various pre-existing conditions including fibrocystic disease of the liver (3%), primary sclerosing cholangitis, gallbladder/intrahepatic stones, parasitic infection (Clonorchis), radiation (historically Thorotrast), and other drugs, alcohol, or metabolic diseases (iron overload, A1AT deficiency). Many of these aforementioned conditions induces inflammation or cholestasis, which can lead to epithelial hyperplasia, dysplasia, and ultimately to invasive malignancy [7].

Studies from the AFIP and Japan show a similar male predominance (60-70%) with an average age around 62 years old. Laboratory studies may show elevated alkaline phosphatase or bilirubin levels, correlating with biliary inflammation/obstruction. Grossly, the majority of these lesions are solitary, though satellite lesions do occur. Microscopically, infiltrating glands typical of adenocarcinoma can be seen, often in a sclerotic background. However, other variants exist, including tumor cells found in trabeculae and nests, similar to our presented case [7]. Dysplasia of the background bile ducts, so-called biliary intraepithelial neoplasia (BilIN) may be seen

Immunohistochemically, cytokeratins 7 and 19 are usually expressed in cholangiocarcinoma, as is CEA. Cytokeratin 20 is usually negative (90% of the time), and Hep Par-1 antibody should be negative [7] Another useful antibody is CD10, which would lack the canulicular pattern more typical of hepatocellular carcinoma. Other noninvasive diagnostic modalities for cholangiocarcinoma include bile duct cytology and flourescent in-situ hybridization (FISH) [5].

Metastases occur in almost ¾ of cases, usually to the lymph nodes and lungs. The median survival is approximately 6 months in all patients, although those with surgically resectable tumors do better. Radiation, chemotherapy, and transplant have not been historically effective therapeutic options in the treatment of cholangiocarcinoma [7], although some institutions are now performing liver transplants for unresectable tumors with good results [10].

A close cousin of intrahepatic cholangiocarcinoma is pancreatic ductal adenocarcinoma. Embryologically, just as the biliary system grows into the primitive liver during development, buds of epithelium from the foregut grow into the pancreas to form the branching ductal system [6]. The common origin of the intrahepatic biliary tree and pancreatic ductal system provides a common link between tumors arising in these two closely related (embryologic) organs.

Just as with cholangiocarcinoma, pancreatic ductal adenocarcinoma is a tumor with a poor prognosis, with only a 4% five-year survival. They typically occur in patients aged 60-80 years old, and may develop in association with dysplasia (PanIN). Risk factors include cigarette smoking, radiation, chronic pancreatitis, and familial (genetic) causes [6].

Grossly, ductal adenocarcinoma usually arises in the head of the pancreas (60-70%), forming a firm, scirrhous mass. Microscopically, this is seen as individual malignant glands infiltrating a dense fibrous stroma. This dense fibrosis associated with the tumor is similar to that seen in intrahepatic cholangiocarcinoma. Perineural and angiolymphatic invasion are common in pancreatic ductal adenocarcinoma (as well as cholangiocarcinoma), and likely contribute to its poor prognosis [6].

Immunohistochemical findings are similar between intrahepatic cholangiocarcinoma and pancreatic ductal adenocarcinoma. Both tumors express cytokeratins 7 and 19, and CEA. In practice, if a malignant adenocarcinoma is seen on liver biopsy of a mass lesion with this immunoprofile, we would classify it as being from "pancreatobiliary origin, including primary cholangiocarcinoma," due to the similarities between the two malignancies. Radiographic examination may be helpful to the clinicians to further clarify this differential diagnosis. We often defer to the clinical and radiologic impression to determine whether a solitary mass in the liver is metastatic from the pancreas, or instead represents a primary cholangiocarcinoma. One immunostain that has proven useful in some institutions to confirm pancreatitc origin is DPC4, which is absent in approximately half the cases of pancreatic ductal carcinomas [6]. We do not have this antibody available in our practice, and given that DPC4 is also lost in 15% of proximal cholangiocarcinomas [2], we feel that this marker would be best used in the differential of pancreatobiliary versus other site of malignancy, rather than pancreatic ductal adenocarcinoma versus intrahepatic cholangiocarcinoma.

Intrahepatic cholangiocarcinoma can arise as a de novo tumor in the liver (from BilIN), or may be the result of malignant degeneration of a pre-existing neoplasm. One such precursor is biliary papillomatosis. Biliary papillomatosis typically occurs in patients in their 50s to 70s with a male predominance. Symptoms of duct obstruction are common ways for them to present. Secreted mucin (mucobilia) may be identified on cholangioscopy, and a patulous ampulla extruding mucin may be seen at time of upper endoscopy/ERCP in some cases of biliary papillomatosis [8].

Grossly and microscopically, the intrahepatic bile ducts are filled with papillary fronds with delicate fibrovascular stalks, and skip areas may be present [13]. The surface epithelial cells covering these papillae are pancreatobiliary, gastric, or intestinal type cells with varying degrees of dysplasia. Approximately 60% of these tumors are benign, which leaves the other 40% showing an invasive component of cholangiocarcinoma somewhere in the tumor [6]. This invasive tumor may be a conventional tubular type adenocarcinoma, or in 10-15% of cases, a mucinous (colloid) adenocarcinoma [8, 13]. Just as with primary cholangiocarcinoma, treatment is surgical, but lateral spread along the biliary system makes recurrences common.

Oncocytic variants of biliary papillomatosis have also been reported, sharing similar clinical features as their non-oncocytic counterparts. The invasive tumors associated with these oncocytic papillary lesions is also oncocytic, due to the presence of abundant cytoplasmic mitochondria [11].

The pancreatic counterpart to biliary papillomatosis is an intraductal papillary mucinous neoplasm (IPMN). Grossly visible, these mucin producing epithelial neoplasms show a papillary architecture within the pancreatic ductal system, similar to that seen in biliary papillomatosis [6].

IPMNs compose 3-5% of all pancreatic neoplasms, and that incidence rises to 20% if the lesion is cystic. The mean age is 63 with a male predominance (3:2 ratio). On CT scan, cystic dilatation of the duct system can be seen, and mucin can be seen protruding from the ampulla on ERCP (similar to some cases of biliary papillomatosis). Symptoms occur secondary to duct obstruction, with approximately 70% of these lesions occurring in the pancreatic head, although the entire pancreas can be affected [6].

Grossly, these tumors are in direct continuity with the pancreatic ductal system, whether it is the main pancreatic duct or a side-branch. Papillary fronds can be seen, which correlate microscopically to papillary fronds covered by epithelium of gastric, intestinal, or pancreatobiliary type with varying degrees of dysplasia. The surgical margin status does not necessarily correlate with risk of recurrence, as 20-30% of IPMNs can be multifocal with skip areas [6].

Just as cholangiocarcinoma can arise in biliary papillomatosis (approximately 40% of the time), invasive ductal adenocarcinoma can arise in IPMN (approximately 35% of the time), and can be of traditional or mucinous morphology. Typically, if cancer is seen in an IPMN, the patients tend to be 3-5 years older. The 5-year survival in a resected IPMN is 90-100%, but drops to 40% if an invasive component is present. Approximately 30% of IPMNs with an invasive component have lymph node metastasis at time of surgery, as compared to 75% in de novo pancreatic ductal adenocarcinoma [6]. This equates to improved survival that may be related to the earlier detection rate of cancers due to symptoms occurring secondary to the intraductal component. Interestingly, the incidence of lymph node metastasis in biliary papillomatosis and de novo cholangiocarcinoma mirrors that seen in IPMN and de novo pancreatic ductal adenocarcinoma. Unfortunately, cholangiocarcinoma in the liver shows no difference in survival between those tumors that arise from a papillary lesion, and those that arise de novo [13].

A unique variant of IPMN with similar clinical features is an intraductal oncocytic papillary neoplasm (IOPN). Microscopically, these tumors differ from IPMNs in that their epithelial cells have abundant eosinophilic granular cytoplasm, with interspersed mucin-producing cells. This oncocytic appearance is due to numerous mitochondria in the cytoplasm, similar to other oncocytic tumors [1]. Cancer can arise in IOPN, in about 18-25% of cases, and these invasive tumors can have oncocytic differentiation [1, 6].

There is another cystic neoplasm that may give rise to invasive adenocarcinoma, in both the liver and pancreas. The hepatic variant of this tumor is a biliary cystadenoma, and the largest study on these tumors was reported by Devaney et al. Almost all of these tumors (96%) occur in women, with a median age of 46 years (range 2-87 years) [4]. Symptoms may occur secondary to mass effect, or they may be found incidentally on radiographic examination for another issue.

Grossly, biliary cystadenomas may become large (2.5-28 cm), with a multicystic loculated appearance [4]. There is no communication with the intrahepatic biliary system. If solid areas are present, either radiographically or at time of gross inspection, then concern should be raised for an invasive component.

Microscopically, columnar mucinous epithelium lines the cyst wall, although dysplastic epithelium may also be seen. Underneath this epithelial lining is an ovarian type stroma, which is absent in cases arising in men [4]. This stroma stains with antibodies to ER and PR, 70% and 60% of the time, respectively, and growth can occur during hormone replacement therapy and pregnancy [3]. It is not known whether these lesions are secondary to some embryologic foregut rest, or other etiology. As dysplasia may be patchy and invasive tumors may arise from these biliary cystadenomas, treatment is by surgical excision rather than in-situ ablation [3, 4].

When an invasive component accompanies a biliary cystadenoma, it is best classified as a biliary cystadenocarcinoma. This term also applies to an invasive cholangiocarcinoma arising in a biliary cyst, and these may account for some cases reported in men, as the incidence of biliary cystadenocarcinoma is approximately equal between genders, with only a slight female bias. Cholangiocarcinoma arising in a biliary cystadenoma typically has a better prognosis than if it arises de novo [4]. Oncocytic variants of biliary cystadenocarcinoma have also been reported [12].

The correlate to a biliary cystadenoma of the liver is a mucinous cystic neoplasm of the pancreas. Just as with biliary cystadenomas, mucinous cystic neoplasms occur almost exclusively in women (20:1 ratio). Typically patients are in their 40-50s, and they may come to clinical attention if symptoms arise due to the lesion's size. However, a subset of these tumors may be discovered incidentally, just as the case with biliary cystadenoma. These lesions are not associated with cigarette smoking or any known genetic syndrome. The etiology of these tumors, as with the biliary counterpart, is unknown. Whether a remnant of a primitive gonad is incorporated into the pancreas early on in embryologic development, or if the stroma is derived from primitive pancreatic mesenchyme is uncertain [6].

Grossly, mucinous cystic neoplasms typically occur in the body or tail of the pancreas, and can be multiloculated, ranging in size from 2-36 cm. A thick fibrous pseudocapsule may be present, and just as with biliary cystadenomas in the liver, communication with the ductal system usually does not occur [6].

Microscopically, mucinous cystic neoplasms of the pancreas are lined by mucinous epithelium, and patchy areas of dysplasia may be seen. Underlying the epithelium, just as in the liver, an ovarian type stroma is present [6]. The ovarian stroma is positive for ER (range 25-80%) and PR (range 40-75%) [3, 6].

Approximately 1/3 of mucinous cystic neoplasms are associated with an invasive adenocarcinoma. Patients are approximately 5-10 years older if there is invasive adenocarcinoma present. Treatment is by complete surgical resection. If no invasive component is present, then the patients are cured. However, the 2-year survival drops to 67% and the 5-year survival drops to 50% if an invasive tumor is present. Although these survival rates are low, they are still much better than de novo pancreatic ductal adenocarcinoma [6].

In conclusion, our case highlighted an oncocytic variant of cholangiocarcinoma. No precursor lesion was identified, but cases of cholangiocarcinoma can arise from pre-existing biliary papillomatosis or biliary cystadenoma. Interestingly, pancreatic ductal adenocarcinoma shares many features with cholangiocarcinoma, both morphologically and immunohistochemically. In addition, pancreatic counterparts exist for both biliary papillomatosis (IPMN) and biliary cystadenoma (mucinous cystic neoplasm). It is not surprising then, that there has been literature supporting a unified nomenclature for some time [9].

References:
  1. Adsay NV, Adair CF, Heffess CS, et al.: Intraductal oncocytic papillary neoplasms of the pancreas. Am J Surg Pathol 1996, 20:980-94

  2. Argani P, Shaukat A, Kaushal M, et al.: Differing Rates of Loss of Dpc4 Expression and of P53 Overexpression among Carcinomas of the Proximal and Distal Bile Ducts: Evidence for a Biological Distinction. Cancer 2001, 91:1332-41

  3. Daniels JA, Coad JE, Payne WD, et al.: Biliary cystadenomas: hormone receptor expression and clinical management. Dig Dis Sci 2006, 51:623-8

  4. Devaney K, Goodman ZD, Ishak KG: Hepatobiliary cystadenoma and cystadenocarcinoma. A light microscopic and immunohistochemical study of 70 patients. Am J Surg Pathol 1994, 18:1078-91

  5. Fritcher EGB, Kipp BR, Halling KC, et al.: A Multivariate Model Using Advanced Cytologic Methods for the Evaluation of Indeterminate Pancreatobiliary Strictures. Gastroenterology 2009, 136:2180-6

  6. Hruban RH, Pitman MB, Klimstra DS: Tumors of the Pancreas. Washington D.C., AFIP, 2007

  7. Ishak KG, Goodman ZD, Stocker JT: Tumors of the Liver and Intrahepatic Bile Ducts. Washington D.C., AFIP, 2001

  8. Lee SS, Kim MH, Lee SK, et al.: Clinicopathologic review of 58 patients with biliary papillomatosis. Cancer 2004, 100:783-93

  9. Longnecker DS, Terhune PG: The case for parallel classification of biliary tract and pancreatic neoplasms. Mod Pathol 1996, 9:828-37

  10. Rea DJ, Rosen CB, Nagorney DM, et al.: Transplantation for Cholangiocarcinoma: When and for Whom? Surgical Oncology Clinics of North America 2009, 18:325-37

  11. Rouzbahman M, Serra S, Adsay NV, et al.: Oncocytic papillary neoplasms of the biliary tract: a clinicopathological, mucin core and Wnt pathway protein analysis of four cases. Pathology 2007, 39:413-8

  12. Wolf HK, Garcia JA, Bossen EH: Oncocytic differentiation in intrahepatic biliary cystadenocarcinoma. Mod Pathol 1992, 5:665-8

  13. Zen Y, Fujii T, Itatsu K, et al.: Biliary papillary tumors share pathological features with intraductal papillary mucinous neoplasm of the pancreas. Hepatology 2006, 44:1333-43