Strictures at the hepatic hilum do not always represent malignancy. Wetter et al. ^[6] published their series of 98 patients with localized common hepatic duct stricture 8 of which were proved to be benign on pathology examination. In a series of 14 proximal bile duct strictures reported from Japan ^[7], all these patients had fibrosis of the bile duct with an inflammatory infiltrate. These benign strictures represents a spectrum of disease process with some may represent idiopathic fibrosis and others a form fruste of sclerosing cholangitis. There has been a report of eosinophilic cholangitis presenting as malignant masquerade ^[8] as well as tuberculosis stricture mimicking hilar cholangiocarcinoma ^[9]. Very often patients with suspicious biliary tract obstruction undergo a series of imaging studies including not only CT scan but MRI-MRCP, endoscopic retrograde cholangiopancreatography (ERCP), percutaneous transhepatic cholangiography (PTC) and positron emission tomography (PET). These tests may show imaging pattern highly suggestive of malignancy, none of these are specific to accurately differentiate benign disease from malignant pathology.

More than 85% of cholangiocarcinoma can have raised CA 19-9 but elevated CA 19-9 are found in  multiple causes of benign stricture as well as malignant stricture ^[10]. Moreover, CA 19-9 cannot differentiate between pancreatic malignancy, gastric malignancy, or cholangiocarcinoma and has also been shown to be elevated in primary sclerosing cholangitis (PSC).

Carcinoembryonic antigen (CEA) has been found to be elevated in approximately 40-50% of patients of cholangiocarcinoma but they are found elevated in benign diseases like inflammatory bowel disease and severe liver injury. Many other markers like CA-125, CA-195, CA-242, DU-PAN-2, interleukin-6, and trypsinogen-2, have been tested but they all lack sensitivity as well as specificity to accurately predict benign versus malignant cause of underlying pathology.

Ultrasonography (USG) lacks sensitivity and specificity in differentiating benign versus malignant disease but helpful for ruling out stone disease as a possible cause of biliary although the presence of cholelithiasis does not necessarily exclude the possibility of concomitant cholangiocarcinoma. Garber and co-workers ^[11] found that the sensitivity of USG for defining a mass in the biliary tree is as low as 21% to 47% and positive predictive value is 94% and were also able to identify extensive regional lymphadenopathy, liver lesions, and portal vein invasion, all of which are associated with malignant disease and much less likely with benign strictures. However it should be strongly noted that liver abscesses can appear like metastatic disease and adenopathy as well as portal vein thrombosis have been seen with benign inflammatory processes. In the series of 171 patients with hilar obstruction from Memorial Sloan-Kettering Cancer Centre, the presence of vascular invasion was strongly associated with a final diagnosis of cholangiocarcinoma as compared with benign disease ^[12]. The sensitivity of vascular invasion was highest when associated with lobar atrophy, which can also be seen on US, but which is very operator dependent.

CT scan can provide tremendous information about extra and intra hepatic biliary tree, liver parenchyma, regional lymph nodes, vascular involvement and metastatic disease. Are and co-workers ^[12] demonstrated that vascular invasion occurs much more frequently in patients with malignancy, 58% versus 16% for patients with benign disease. Lobar atrophy was found in 41% of malignant stricture compared with only 6% of benign biliary stricture. Choi and co-workers ^[13] identified some findings observed significantly more in malignant stricture and may help in differentiating it from benign stricture: a stricture wall thicker than 1.5 mm; a longer involved segment of duct (17.9 mm +/- 6.6 mm for malignancy versus benign at 8.9 mm +/- 6.8 mm [P<0.0001]); and a more dilated duct proximal to the obstruction (22 mm +/- 5.4 mm for malignancy versus benign at 17.8 mm +/- 4.6 mm [P<0.033]). The hyperenhancement pattern of the involved bile duct wall during portal venous phase seen more commonly in malignant stricture compared with benign stricture. (P<0.0001) Both ductal dilation and lymphadenopathy are not specific for malignancy. Lymphadenopathy can be present in inflammatory processes, such as PSC. Triple phase CT scan can identify vascular invasion and assess the resulting atrophy-hypertrophy.

If indicated, ERCP is chosen over PTC because most patients with obstructive jaundice have either choledocholithiasis or a periampullary neoplasm. PTC is often used when ERCP is unable to be performed, and is preferred in proximal biliary obstruction, especially the future liver remnant drainage is needed for surgery planning.

MRI-MRCP is highly sensitive for detecting biliary obstruction (72%–98%) but has less sensitivity in differentiating benign from malignant disease, with a wide range of sensitivities in the literature (30%–98%) ^[14]. Recently MRCP has been shown to be equivalent to both ERCP and PTC in the evaluation of benign versus malignant disease. Park and co-workers ^[14] and Rosch and co-workers ^[15] found out that all three modalities are equivalent at determining benign or malignant disease but lacking specificity. ERCP may provide a tissue diagnosis but often biopsy does not provide a tissue diagnosis either, especially for more proximal lesions. When a malignant diagnosis is being considered, contrast-enhanced MRI with MRCP can provide excellent information regarding the liver parenchyma (including atrophy-hypertrophy, and liver metastases), vascular invasion, and lymphadenopathy, and can be complementary to CT scan.

PET-CT scan can evaluate regional lymphadenopathy and distant metastasis more accurately and can upgrade some to unresectable or downgrade others to resectable disease compared with only conventional imaging, allowing curative resection for some and sparing others from an unwanted laparotomy ^[16]. PET currently lacks the sensitivity for diagnosing malignancy versus benign disease, especially mucinous cholangiocarcinoma, which lacks FDG avidity.

Most biliary strictures, especially hilar stricture in patients without previous biliary tree manipulation and without stone disease, are mostly adenocarcinoma. Cytologically negative biopsy is more often false negative biopsy rather than a positive finding of benign biliary disease ^[12]. Previous studies demonstrate that the yield for brushings from ERCP or PTC have been positive in only 30% of patients with diagnosed cholangiocarcinoma ^[12,17]. The sensitivity was only 36% for ERCP with biopsy and 46% when brushings were performed in diagnosed cases of malignancy ^[18]. Some recent study showed higher sensitivity (70%-100%) and specificity (80%-95%) with biliary brush cytology for the early diagnosis of cholangiocarcinoma in PSC ^[19] .

Corvera and co-workers showed that pathological causes can be classified into five general categories ^[3] (1) lymphoplasmacytic sclerosing pancreatitis and cholangitis (LPSPC), (2) PSC, (3) nonspecific fibroinflammatory process (NFIP), (4) granulomatous disease, and (5) stone disease. One more category should be considered as iatrogenic-idiopathic-autoimmune. Multiple case reports in the literature have shown benign biliary stricture development after an injection of sclerosant for duodenal ulceration, ^[5] as a consequence of non-Hodgkin’s lymphoma, ^[20] pyogenic cholangitis, ^[21] atherosclerosis. ^[22] Several autoimmune processes such as systemic lupus erythematosus, Wegner’s granulomatosis, and Sjogren’s syndrome have been shown to cause biliary stricture ^[3,23]. In our patient, there were benign cysts with localised organised abscess formation at base of segment IVb masquerading as malignant mass necessitating an operative resection in the form of left hepatectomy to provide a cure.