42 Malignant Biliary Disease
Ming-Ming Xu, Nikhil A. Kumta, Michel Kahaleh
Cholangiocarcinoma (CCA) is a heterogeneous group of epithelial tumors arising within the biliary tract. These are conventionally divided into distal, hilar, and intrahepatic cancers by their longitudinal extent along the biliary tract and also demonstrate differences in their pathogenesis, molecular signatures, and treatment. They share a dismal prognosis due to their aggressive natural history and often late-stage disease at diagnosis. Surgical resection and in select cases, liver transplantation, offer the only durable chance of cure when CCA is diagnosed at an early stage. We will review the evolution of imaging and endoscopic tools for the diagnosis of malignant biliary disease and the available treatment options from surgery, liver transplantation, and endoscopic palliation to advances in locoregional and adjuvant therapy.
42.2 Diagnostic Approach
The early diagnosis of cholangiocarcinoma (CCA) is the most essential determinate of prognosis as curative surgical resection and liver transplantation are only possible in the small subset of patients who have resectable disease. The majority of patients present at late stage when they develop clinical symptoms due to biliary obstruction, pain, or weight loss. One of the challenges in the diagnosis of malignant biliary disease is that these symptoms can also be seen in benign causes of biliary strictures and the differentiation between malignant and benign disease can often be challenging despite the obvious difference in their clinical consequence (▶Table 42.1). Laboratory tests often show an obstructive pattern of liver chemistries, and although tumor markers such as CA 19–9 and carcinoembryonic antigen (CEA) are often used to aid in the diagnosis, current data do not support their use to independently make the diagnosis of CCA. Serum CA 19–9 cutoff values of greater than 37 U/mL has been shown have a sensitivity and specificity for malignant stricture of 73 and 63%, respectively, but can also be elevated in benign causes of cholestasis. 1
Benign causes of biliary stricture
Malignant causes of biliary stricture
Post liver transplant stricture
Primary sclerosing cholangitis
Metastatic disease or lymphadenopathy
Benign fibrostenotic stricture
HIV, Human Immunodeficiency Virus; IgG4, immunoglobulin G4.
42.2.1 Radiologic Imaging
Cross-sectional imaging with computed tomography (CT) or magnetic resonance imaging (MRI)/magnetic resonance cholangiopancreatography (MRCP) have become the standard imaging tools in the evaluation of suspected malignant biliary obstruction. 2 Multidetector CT (MDCT) classically shows a mass or hypoattenuating ductal thickening in the portovenous phase with or without proximal biliary ductal dilation 2 (▶Fig. 42.1). CT imaging has been shown to have high accuracy in detecting vascular involvement and distant metastases but is less accurate in the evaluation of the longitudinal extent of the tumor and local lymph node invasion. 3 The overall accuracy of CT in assessing resectability of CCA has been reported to be 60 to 88% compared to the accuracy of MRI/MRCP in assessing tumor extent and resectability of 95%. 4 , 5 , 6 MRCP has the advantage of avoiding intravenous contrast and radiation while providing similar sensitivity and specificity to endoscopic retrograde cholangiopancreatography (ERCP) for determining the level or location of obstruction and a sensitivity of 88%, specificity of 95% for determining malignancy in one large meta-analysis. 7
42.2.2 Endoscopic Retrograde Cholangiopancreatography
Despite the high accuracy of MRCP in delineating the level of biliary obstruction and a suspected cause of biliary stricture, it ultimately cannot provide definitive tissue diagnosis. ERCP with brush cytology and endobiliary biopsy is the initial standard approach for tissue sampling in suspected biliary malignancy. The limitation of ERCP is the known poor sensitivity of brush cytology for malignancy of between 23 and 56% despite high specificity of nearly 100%. 8 , 9 , 10 , 11 This is likely the result of multiple factors including the desmoplastic reaction caused by the tumor, the anatomical location of the stricture, and cellular loss during specimen processing. 12 Attempts at stricture dilation prior to brushing increase sensitivity to only 34 from 27%. 12 When endobiliary biopsy is combined with brushing, the sensitivity increases to 70%. 13 The use of a protocolized method for repeated tissue sampling with on-site cytopathology and careful specimen processing, the “smash protocol,” achieved a 72% on-site pathologic diagnosis for biliary malignancy. 14
42.2.3 Fluorescence In-Situ Hybridization
A newer, adjunctive technique to increase the diagnostic yield from brushing cytology is fluorescence in situ hybridiza tion (FISH), which uses fluorescently labeled DNA probes to detect chromosomal aneuploidy or polysomy, which has been seen in up to 80% of CCA 15 (▶Fig. 42.2). The commercially available probes target chromosomes 3, 7, 17, and the 9p21 locus of chromosome 9. When all four probes are used in indeterminate biliary strictures, the sensitivity of FISH for malignancy is 84% with specificity of 97%. 16 FISH can significantly improve the diagnostic yield of routine cytology from ERCP without the need for additional procedures with high specificity. However, in patients with primary sclerosing cholangitis (PSC), FISH is less reliable with sensitivity for CCA of only 47%, specificity of 100%, positive predictive value of 100%, and negative predictive value of 88%. 17
The single-operator cholangioscopy (SOC) system, commonly known as the Spyglass Direct Visualization System (Boston Scientific, Natick, Massachusetts, United States) allows both direct visualization of the biliary tract and directed tissue sampling with a mini biopsy forceps (SpyBite). The system consists of a 10-F access and delivery catheter (SpyScope) through which the fiber optic probe is inserted, providing 6,000-pixel images with four-way tip maneuverability and a 30-degree view in each direction. The entire system is introduced into the biliary tree with guidewire assistance after traditional ERCP-based biliary access and fits through the working channel of the duodenoscope. A disposable 3-F SpyBite forceps can be inserted into the SpyScope working channel for visually directed biopsies. Direct cholangioscopy offers both the ability to visually characterize an indeterminate stricture as benign or malignant and to perform visually targeted biopsies of suspicious lesions (▶Fig. 42.3). Chen et al published the largest prospective, multicentered, observational study of the operating characteristics of the SOC system with 226 patients and reported a sensitivity of 78%, specificity of 82% for the visual impression of malignancy using SOC. 18 Visual impression had a higher sensitivity compared to SpyBite biopsy, which had a sensitivity of 47% and specificity of 98% for malignancy. 18 The main limitation of using cholangioscopy-based visual impression for diagnosing malignant strictures is the lack of interobserver agreement on the criteria that should be used. The tumor vessel sign has been shown to be a specific feature of malignancy but has low sensitivity of 61%. 19 One retrospective study involving multiple blinded expert endoscopists evaluating cholangioscopy videos of undifferentiated biliary strictures showed only fair agreement on the ultimate diagnosis of benign versus malignant etiology. 20 Significant complications of SOC includes a higher rate of cholangitis compared to ERCP alone (7 vs. 3%). 21 A new digital system was recently commercialized (Digital Spyglass, Boston Scientific) and provide better intraductal imaging. We, however, need further data to confirm early finding regarding its accuracy. 22
42.2.5 Endoscopic Ultrasound-Fine Needle Aspiration
Endoscopic ultrasound (EUS) has an important complementary role to ERCP in the evaluation and staging of CCA. In the absence of a visible mass on cross-sectional imaging, EUS can identify suspicious bile duct thickening and assess for local lymphovascular invasion of the lesion 23 (▶Fig. 42.4). The sensitivity of EUS-FNA for CCA is reported to be 53 to 89% across studies with better performance in distal CCA compared to proximal tumors. 24 , 25 , 26 , 27 , 28 EUS-FNA sensitivity is also lower in patients with previously placed biliary stents, which can cause acoustic shadowing. 23 Lastly, there is a theoretical risk of peritoneal seeding of malignant cells via the needle tract during FNA of proximal CCA, which has prompted some liver transplant centers to consider this technique a contraindication for transplantation. 29 Thus, it is important to consider resectability and transplant potential of the patient prior to any attempt at EUS-FNA of a proximal CCA.
42.2.6 Intraductal Ultrasound
Along with EUS, intraductal ultrasound (IDUS) was developed to further enhance endobiliary imaging with a high-frequency probe that is guided directly into the biliary tree during ERCP. IDUS criteria have been developed to help with the differentiation of benign from malignant strictures (▶Table 42.2). 30 Use of IDUS can improve the diagnostic accuracy of ERCP from 58 to 83% with improved sensitivity of 80 to 90% and specificity of 83% in diagnosing a malignant stricture. 31 , 32 In addition, IDUS can improve the accuracy of locoregional staging of hilar CCA compared to standard EUS because of the maneuverability of the probe to be in close proximity to the targeted tumor. 33 , 34
42.2.7 Probe-based Confocal Laser Endomicroscopy
Probe-based confocal laser endomicroscopy (pCLE, Cellvizio; Mauna Kea Technologies, Paris, France) is another advanced imaging technology which uses laser light to provide in vivo microscopic-level images of the biliary epithelium in real time during ERCP. A confocal miniprobe is passed into the working channel of the duodenoscope and applied directly onto the biliary tissue. Recently the criteria to differentiate between benign and malignant strictures were refined to improve its specificity for malignancy and provide better descriptions of benign inflammatory changes in strictures (Paris classification) (▶Fig. 42.5). 35 , 36 A prospective, multi-centered validation study of the Paris classification showed the combination of pCLE with ERCP had 89% sensitivity, 71% specificity, and 82% diagnostic accuracy for on-site diagnosis of malignant strictures. 36 A major limitation of pCLE is the lack of interobserver agreement in applying the criteria even among expert users of pCLE. 37
42.3 Classification Systems
Cholangiocarcinomas are classified as extrahepatic and intrahepatic based on their anatomical location. Extrahepatic CCA include both hilar and distal common bile duct tumors. Hilar tumors are further subcategorized into types I to IV based on the Bismuth classification, which describes tumors according to their longitudinal extension along the biliary tree (▶Fig. 42.6). 38 , 39 Type I refers to tumors limited to the common bile duct before the confluence, type II tumors involve the biliary confluence, type III cancers involve both the confluence and either the right (IIIa) or left (IIIb) hepatic ducts, and type IV refers to multifocal tumors involving both the confluence and both branches of the hepatic ducts. It is important to note that the Bismuth classification is an anatomical system of describing hilar CCA and is not a prognostic staging system. Clinical staging of CCA is based on the tumor nodal metastasis (TNM) system. Histologically, adenocarcinoma is the most common pathologic type of CCA, other rarer pathologic subtypes include papillary adenocarcinoma, intestinal type adenocarcinoma, clear cell adenocarcinoma, signet-ring cell carcinoma, and squamous cell carcinoma. 40
42.4 Guidelines and Systematic Reviews
In 2013, the American Society of Gastrointestinal Endoscopy (ASGE) published guidelines on the evaluation of biliary neoplasm. 41 All of the modalities discussed above were reviewed and cited in the guidelines as complementary tools in the diagnostic work-up of an indeterminate biliary stricture. No algorithmic flowchart for the order or preference of use was specifically noted in the ASGE guidelines. The British Society of Gastroenterology also published updated guidelines for diagnosing CCA in 2012, which are summarized in ▶Table 42.3. 42 Our suggested algorithm for the stepwise use of these modalities in the work-up of suspected biliary malignancy is shown in ▶Fig. 42.7.