Endoscopic Retrograde Cholangioscopy and Advanced Biliary Imaging

Developments in endoscopic retrograde cholangioscopy provide multiple new advanced methods of biliary imaging. Cholangioscopy provides direct visualization of epithelium with white light as well as advanced modalities, such as narrow band imaging and autofluorescence. In vivo histologic images can be achieved with confocal endomicroscopy. Cross-sectional imaging is also possible with intraductal ultrasonography and optical coherence tomography. This article describes these advanced imaging techniques, which can be used together to assist in the diagnosis of biliary strictures and lesions.

•

Advanced biliary imaging with catheter and probe-based instruments provide us with multiple modalities to enhance evaluation of biliary pathology.
•

Fiber optic imaging allows for basic cholangioscopic imaging and characterization of mucosa. More advanced cholangioscopies provide enhanced features, such as narrow band imaging. Although these modalities theoretically offer promise, further improvements need to made in defining the role of such imaging.
•

Probe-based confocal endomicroscopy offers possibilities of in vivo histologic imaging, and thus far validation studies suggest a promising diagnostic performance. Further study is needed, however, in interpreting imaging.
•

Intraductal ultrasonography and optical coherence tomography offer promising diagnostic capabilities that rely on ductal wall patterns, which may be particularly beneficial when evaluating infiltrating lesions.
•

It is unlikely that any one of these modalities will solve the dilemma of distinguishing malignancy in indeterminate strictures alone, but rather will need to be used in combination and in tandem with sampling techniques.

Key Points

Introduction

Indeterminate biliary strictures remain an overwhelming challenge for the biliary endoscopist, and major advances have been made in biliary diagnostics. Over the past 10 to 15 years, biliary endoscopy has seen the emergence of multiple techniques that allow us to move beyond the interpretation of fluoroscopic images to the direct visualization of the bile duct. This visualization is in the form of white light; contrast-enhanced imaging, such as narrow band imaging (NBI); or in vivo histologic imaging with confocal endomicroscopy.

The above-mentioned imaging modalities, available during endoscopic retrograde cholangioscopy (ERCP), are discussed in the article. In addition, their indications and current status of diagnostic performance are reviewed.

Cholangioscopy

Cholangioscopy has been available for the past few decades, through the earlier advent of mother-daughter systems to the more recent, widely used peroral cholangioscopy (POCS) systems such as SpyGlass (Boston Scientific, Natick, MA, USA) to the less commercially available, charge-coupled device technology–based video cholangioscopes (Olympus, Center Valley, PA, USA). Significant work is also being done using upper endoscopes for direct POCS.

Using a catheter-type cholangioscope (Opticscope, Clinical Supply, Ltd, Tokyo, Japan), Fukuda and colleagues described malignant strictures as having irregularly dilated and tortuous vessels (known as tumor vessels), ooze easily, and have irregular surfaces. Benign strictures were classified based on the presence of a smooth mucosa without neovascularization and papillogranular mucosa with no obvious mass. In addition, using these criteria during POCS increased the accuracy of diagnosing malignant strictures from 78.1% to 93.4% compared with ERCP with tissue sampling. The investigators did note that caution needed to be taken when visualizing inflammation because 5 of 38 patients received false positive cholangioscopic diagnoses. Kim and colleagues reported that the presence of tumor vessels had a sensitivity of 61% in diagnosing malignancy, which increased to 96% when combined with biopsies. Tischendorf and colleagues classified cholangioscopic findings in 12 patients with primary sclerosing cholangitis as suspicious for malignancy if there was an associated polypoid or villous mass, or an irregularly shaped ulceration. Using these criteria, cholangioscopy had a sensitivity of 92%, specificity of 93%, and negative predictive value of 97%. Benign strictures caused by scarring were described as smooth with no surface irregularities, whereas inflammatory strictures were described to have erythematous mucosa with erosive and ulcerous mucosa. Itoi and colleagues described traditional white light cholangioscopy findings that were thought to represent normal mucosa, inflammation, scar, and cancer. Normal epithelium was described as having a flat surface (with or without pseudodiverticula) and a flat network of vessels compared with a bumpy surface with thin tortuous vessels that characterizes inflammation. Features suggestive of cancer included irregular papillary or granular lesions, and thin to thick tortuous vessels.

Recently, long-term data have been reported describing single operator of cholangioscopy (SOC) using the SpyGlass system (Boston Scientific Natick, MA, USA) ( Figs. 1 and 2 ). This cholangioscope is 3.4 mm in diameter, with a 1.2-mm accessory channel, and has a 2200-mm working length and can be passed through the working channel of standard duodenoscopes over a 0.035-mm guidewire. A 0.77-mm fiberoptic probe is passed through the 0.9-mm optic channel. Miniature biopsy forceps can be passed through the accessory channel. Irrigation can be provided through a 0.6-mm channel, which allows for improved visualization and clearance of debris. In a large, multicenter prospective study, including 297 patients, 95 patients (47% with malignancy) were analyzed for the diagnostic performance of a SOC-based impression. SOC impression had a sensitivity of 78% (84% for intrinsic lesions vs 67% for extrinsic lesions), specificity of 82%, positive predictive value (PPV) of 80%, and negative predictive value (NPV) of 80%. Investigators thought that cholangioscopic findings changed management in 64% of patients. Criteria used to classify malignant and benign cholangioscopic diagnosis were not included in this study. In a single center study using the same system, 36 patients with indeterminate strictures were evaluated with SOC. Malignancy was diagnosed by SOC impression in 21 of the 22 (95%) patients with a final diagnosis of malignancy. Three patients with benign strictures were incorrectly diagnosed with malignant strictures by SOC, and 79% were correctly classified as benign. The overall accuracy of differentiating malignant from benign disease based on SOC visual impression was 89% (sensitivity 95%, specificity 79%, PPV 88%, NPV 92%).

Cholangioscopy

Cholangioscopy has been available for the past few decades, through the earlier advent of mother-daughter systems to the more recent, widely used peroral cholangioscopy (POCS) systems such as SpyGlass (Boston Scientific, Natick, MA, USA) to the less commercially available, charge-coupled device technology–based video cholangioscopes (Olympus, Center Valley, PA, USA). Significant work is also being done using upper endoscopes for direct POCS.

Using a catheter-type cholangioscope (Opticscope, Clinical Supply, Ltd, Tokyo, Japan), Fukuda and colleagues described malignant strictures as having irregularly dilated and tortuous vessels (known as tumor vessels), ooze easily, and have irregular surfaces. Benign strictures were classified based on the presence of a smooth mucosa without neovascularization and papillogranular mucosa with no obvious mass. In addition, using these criteria during POCS increased the accuracy of diagnosing malignant strictures from 78.1% to 93.4% compared with ERCP with tissue sampling. The investigators did note that caution needed to be taken when visualizing inflammation because 5 of 38 patients received false positive cholangioscopic diagnoses. Kim and colleagues reported that the presence of tumor vessels had a sensitivity of 61% in diagnosing malignancy, which increased to 96% when combined with biopsies. Tischendorf and colleagues classified cholangioscopic findings in 12 patients with primary sclerosing cholangitis as suspicious for malignancy if there was an associated polypoid or villous mass, or an irregularly shaped ulceration. Using these criteria, cholangioscopy had a sensitivity of 92%, specificity of 93%, and negative predictive value of 97%. Benign strictures caused by scarring were described as smooth with no surface irregularities, whereas inflammatory strictures were described to have erythematous mucosa with erosive and ulcerous mucosa. Itoi and colleagues described traditional white light cholangioscopy findings that were thought to represent normal mucosa, inflammation, scar, and cancer. Normal epithelium was described as having a flat surface (with or without pseudodiverticula) and a flat network of vessels compared with a bumpy surface with thin tortuous vessels that characterizes inflammation. Features suggestive of cancer included irregular papillary or granular lesions, and thin to thick tortuous vessels.

Recently, long-term data have been reported describing single operator of cholangioscopy (SOC) using the SpyGlass system (Boston Scientific Natick, MA, USA) ( Figs. 1 and 2 ). This cholangioscope is 3.4 mm in diameter, with a 1.2-mm accessory channel, and has a 2200-mm working length and can be passed through the working channel of standard duodenoscopes over a 0.035-mm guidewire. A 0.77-mm fiberoptic probe is passed through the 0.9-mm optic channel. Miniature biopsy forceps can be passed through the accessory channel. Irrigation can be provided through a 0.6-mm channel, which allows for improved visualization and clearance of debris. In a large, multicenter prospective study, including 297 patients, 95 patients (47% with malignancy) were analyzed for the diagnostic performance of a SOC-based impression. SOC impression had a sensitivity of 78% (84% for intrinsic lesions vs 67% for extrinsic lesions), specificity of 82%, positive predictive value (PPV) of 80%, and negative predictive value (NPV) of 80%. Investigators thought that cholangioscopic findings changed management in 64% of patients. Criteria used to classify malignant and benign cholangioscopic diagnosis were not included in this study. In a single center study using the same system, 36 patients with indeterminate strictures were evaluated with SOC. Malignancy was diagnosed by SOC impression in 21 of the 22 (95%) patients with a final diagnosis of malignancy. Three patients with benign strictures were incorrectly diagnosed with malignant strictures by SOC, and 79% were correctly classified as benign. The overall accuracy of differentiating malignant from benign disease based on SOC visual impression was 89% (sensitivity 95%, specificity 79%, PPV 88%, NPV 92%).

Cholangioscopy with visual enhancement

Contrast-enhanced endoscopy is widely used throughout the gastrointestinal (GI) tract to help delineate lesion margins, determine vascularization of lesions, and describe mucosal surface patterns. Enhanced image of the bile duct has been described with NBI, chromoendoscopy with methylene blue, and autofluorescence.

Chromocholangioscopy of the Bile Duct

Endoscopy using various dyes including crystal violet, indigo carmine, and methylene blue has been widely performed in the esophagus, stomach, and colon. Data regarding the use of methylene blue in the bile ducts suggest that it is feasible but may be limited by the presence of mucin, bile, and contrast. Although mucosal patterns were enhanced, blood vessel delineation was unclear. Thus, chromocholangioscopy seems to have a limited role based on the available literature.

NBI

The value of visualization with NBI over white light endoscopy in the GI lumen is well known. Two video cholangioscopes are currently available (CHF-B260 and CHF-BP260; Olympus Medical Systems, Tokyo, Japan) ( Figs. 3–5 ), which can be used to provide NBI in the bile duct. Restricting light to 2 wavelengths, 415 nm and 540 nm, provides enhanced visualization of superficial mucosal capillary as well as pit patterns and thicker capillaries of the deeper tissues, respectively. Itoi and colleagues demonstrated the feasibility of NBI in the bile duct using a video cholangioscope (CHF-B260; Olympus Medical Systems Tokyo, Japan) with an NBI system (CV-260SL processor, CVL-260Sl light source; Olympus Tokyo, Japan) in determining lesion margins and identifying surface vessels. Twenty-one lesions in 12 patients were evaluated with NBI system and compared with those with conventional POCS. The investigators found that visualization of the surface structure and vessels with NBI was good or better than conventional imaging and excellent in a significantly greater number of lesions (57.4% vs 9.5%). Four lesions were detected by NBI that were not detected by white light endoscopy. Limitations to the use of NBI include the interference by the dark red appearance of blood and bile, as well as the lack of additional magnification systems on current cholangioscopes, a feature that is used to enhance NBI imaging of GI luminal epithelium.