- Email: [email protected]
Role of single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions: a single-center, prospective study
ORIGINAL ARTICLE: Clinical Endoscopy
Role of single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions: a single-center, prospective study Mohan Ramchandani, MD, DM, D. Nageshwar Reddy, MD, DM, FRCP, Rajesh Gupta, MD, DM, Sandeep Lakhtakia, MD, DM, Manu Tandan, MD, DM, Santosh Darisetty, DA, Anuradha Sekaran, MD, DNB, Guduru Venkat Rao, MS, FRCS Hyderabad, India
Background: Currently available techniques to diagnose indeterminate biliary lesions have many limitations. Objective: To assess the accuracy of single-operator peroral cholangioscopy by using the SpyGlass system to differentiate malignant from benign disease in patients with indeterminate biliary lesions. Design: Prospective, single-arm, single-center study. Setting: Tertiary referral center. Patients: Thirty-six patients with indeterminate biliary strictures and filling defects who had inconclusive results on previous biliary ductal tissue sampling. Interventions: SpyGlass cholangioscopy with cholangioscopically guided intraductal biopsies. Main Outcome Measurements: Accuracy of SpyGlass visual impression and SpyBite biopsies for differentiating malignant from benign ductal lesions. Results: Thirty-six patients (22 men, mean age 48.3 years [range 27-68 years]) with indeterminate stricture and/or filling defects underwent SpyGlass cholangioscopy. Of the 22 patients with a final diagnosis of malignant lesion, cholangioscopic impression was malignant in 21 patients (95%) and benign in 1 patient (5%). Of the 14 patients with a final diagnosis of benign disease, including the 3 patients with common bile duct stones and no stricture, cholangioscopic impression was malignant in 3 patients (21%) and benign in 11 patients (79%). The overall accuracy of SpyGlass visual impression for differentiating malignant from benign ductal lesions was 89% (32/36). The accuracy of SpyBite biopsies for differentiating malignant from benign ductal lesions that were inconclusive on ERCP-guided brushing or biopsy was 82% (27/33) in an intent-to-treat analysis. Limitations: No randomized comparison with alternative diagnostic modalities for the nature of biliary strictures. Conclusions: SpyGlass cholangioscopy with SpyBite biopsies has a high accuracy with regard to confirming or excluding malignancy in patients with indeterminate biliary lesions. (Gastrointest Endosc 2011;74:511-9.)
Abbreviations: CBD, common bile duct; POC, peroral cholangioscopy; SOPOC, single-operator peroral cholangioscopy. DISCLOSURE: The authors disclosed no financial relationships relevant to this publication. Copyright © 2011 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 doi:10.1016/j.gie.2011.04.034 Received January 14, 2011. Accepted April 21, 2011. Current affiliations: Asian Institute of Gastroenterology, Hyderabad, India. Reprint requests: D. Nageshwar Reddy, MD, DM, Asian Institute of Gastroenterology, 6-3-661, Somajiguda, Hyderabad 500 082, India. If you would like to chat with an author of this article, you may contact Dr. Reddy at [email protected].
www.giejournal.org
Despite significant advances in pancreaticobiliary imaging and EUS-guided assessment of tumors, the characterization of intraductal biliary lesions remains a challenge, even at centers with significant ERCP and EUS expertise. Cholangioscopy, which facilitates direct visual assessment and visually guided tissue sampling, holds promise as an advanced technique in cases that elude successful diagnosis by conventional ERCP or other imaging modalities. Peroral cholangioscopy is less invasive than transcutaneous cholangioscopy and can be easily performed during ERCP.1-10 However, conventional peroral cholangioscopy (POC) has certain disadvantages. It requires 2 experienced endoscopists and has limited maneuverability, inadequate irrigation, and a small instrument channel. Single-operator POC (SOPOC) using the SpyGlass Direct Visualization Volume 74, No. 3 : 2011 GASTROINTESTINAL ENDOSCOPY 511
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
System (Microvasive Endoscopy; Boston Scientific Corp, Natick, Mass) for cholangioscopy has overcome several of these limitations.4,5 The SpyGlass system is a singleoperator system that allows not only optical viewing, but also optically guided biopsies. This study was done in a high-volume center in south Asia performing approximately 6000 ERCPs annually. The indeterminate biliary lesions pose a diagnostic challenge, especially in this region where infective biliary lesions11,12 are important differential diagnoses apart from biliary malignancies. This prospective study aimed to assess the accuracy of SOPOC and cholangioscopically guided biopsies using the SpyGlass system in patients with previously inconclusive results of ERCP-guided brushing and biopsies. Compared with a previous report5 on the use of SpyGlass cholangioscopy for the characterization of indeterminate biliary strictures, this study enrolled a larger number of patients.
METHODS Study design The study was a prospective, single-center cohort study of patients meeting inclusion and exclusion criteria. It was conducted at the Asian Institute of Gastroenterology, a tertiary referral center with patient enrollment between May 2009 and February 2010. Institutional review board approval was obtained for the prospective collection of patient data and follow-up.
Patients Patients older than 18 years of age who had undergone a previous ERCP for the evaluation of obstructive jaundice and were diagnosed as having an indeterminate stricture or filling defects on ERCP were considered for this study (Fig. 1). The indeterminate nature of the ERCP finding was defined as previous brush cytology and/or endoscopic biliary biopsy with inconclusive cytology/histology or benign cytology/histology findings in the presence of strong suspicion of malignancy. Exclusion criteria were the patients’ inability to give informed consent and patients medically unfit to undergo an ERCP.
Materials and interventions SOPOC was performed by using the SpyGlass Direct Visualization System (Microvasive Endoscopy, Boston Scientific Corp, Natick, Mass), which consists of capital equipment, including a pump, a light source, and a monitor, and of 3 disposable devices: (1) optical probe (SpyGlass), (2) access and delivery catheter (SpyScope), and (3) biopsy forceps (SpyBite). The SpyGlass probe is a 6000-pixel fiberoptic bundle that enters the biliary anatomy through the SpyScope catheter. It is 231 cm long and is designed to acquire and transmit endoscopic images and conduct light into the 512 GASTROINTESTINAL ENDOSCOPY Volume 74, No. 3 : 2011
Ramchandani et al
Take-home Message ●
●
In a single-center study, the accuracy of SpyGlass cholangioscopy in the diagnosis of indeterminate biliary lesions was 89%. The visual impression on SpyGlass cholangioscopy is most sensitive, and cholangioscopy-directed biopsy is the most specific method to diagnose indeterminate biliary lesions.
biliary anatomy, providing a 70-degree field of view. This is a multiple-use device that may be reprocessed after each use. In our experience, with careful handling, the SpyGlass optical probe can be reused approximately 20 times. The SpyScope access and delivery catheter is a singleuse, single-operator-controlled, 10F diameter, 230-cmlong device designed to provide a pathway into the biliary anatomy for diagnostic and therapeutic devices. This 4-lumen catheter has an optic channel that accommodates the SpyGlass optical probe, a 1.2-mm accessory channel, and 2 independent irrigation channels. The SpyScope incorporates a handle with 2 knobs controlling 4-way steering of the catheter, thus facilitating visualization of the entire circumference of the biliary tree. The SpyBite biopsy forceps are a single-use device that passes through the 1.2-mm minimum working channel of the SpyScope catheter. The jaws of the forceps open to 4.1 mm, obtaining tissue adequate for histology in the majority of cases. It is advisable to notify the pathologist that the intraductal SpyBite tissue samples are quite small. Cholangioscopy with the SpyGlass system is performed by a single operator; this technique is described by Chen and Pleskow.5
Definitions and study objectives The primary aim of the study was to assess the accuracy, sensitivity, and specificity of SpyGlass cholangioscopy in differentiating malignant from benign biliary lesions in patients with indeterminate biliary lesions. Secondary objectives included assessment of overall procedure success and complications observed during and after the procedure. A lesion was defined as definite malignant, suspected malignant, or benign based on the presence or absence of the following visual observations: (1) mass, (2) dilated tortuous vessels, (3) papillary or villous projections, and (4) intraductal nodules. A lesion was diagnosed as definite malignant if a mass with dilated tortuous vessels was seen (Fig. 2). The rest of the lesions were defined as suspected malignant lesion. Benign lesions were characterized by smooth surface mucosa without definite neovascularization and homogeneous granular mucosa without a primary mass. Biopsy samples were collected by using SpyBite forceps and subwww.giejournal.org
Ramchandani et al
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
Figure 1. Study profile.
malignancy was established by SpyGlass cholangioscopic biopsies or other definitive tissue sampling method, such as CT/EUS-guided biopsy, intraoperative biopsy, and surgical specimen collection. Patients who received a diagnosis of malignant disease based on cholangioscopic biopsies underwent either curative surgery or palliative biliary stenting. Patients who had inconclusive SpyGlass cholangioscopic biopsy findings underwent surgery.
Data collection and follow-up
Figure 2. Cholangioscopic image showing a mass lesion with dilated tortuous vessels suggestive of a definite malignant lesion.
mitted for histopathological evaluation. In each case, 3 to 4 biopsy samples were taken. The SpyGlass procedure was deemed successful when the target lesion could be visualized and biopsy samples were deemed adequate for histopathology. This assessment was made in all patients who underwent cholangioscopy-guided biopsies in which SpyBite forceps were used, ie, in all patients except those with biliary calculi and no bile duct strictures. The SpyBite biopsy samples were categorized as malignant or benign or as inadequate sample if the specimens were deemed inadequate for histological assessment. The final diagnosis of a benign condition was assumed in cases of negative histology of ductal tissue sampling and uneventful extended clinical follow-up of at least 6 months after the index SpyGlass procedure. The final diagnosis of www.giejournal.org
Patient demographics, stricture location, previous interventions, and indications for the SpyGlass procedure based on previous ERCP findings were noted, after obtaining informed consent. SpyGlass procedures were performed by 2 experienced ERCP endoscopists independently, and biopsy samples were examined by a pathologist who was blinded to cholangioscopic findings. At the time of the procedure, the SpyGlass visual impression was recorded and classified. All patients were followed for 1 month, whereas patients with benign disease on SpyBite biopsy were followed every 2 months for at least 6 months. During each visit, all patients underwent clinical examination, routine biochemical examination, and radiological imaging including an abdominal US/CT scan. Patients underwent ERCP and biliary stent exchange if clinically warranted.
Statistical analysis Data were analyzed by using SPSS 14.0 version (SPSS South Asia Pvt. Ltd, Bangalore, India). Descriptive statistics consisted of the mean, standard deviation, and range. Patients originally enrolled in the study, but found to have biliary calculi without a biliary stricture on cholangioscopy were excluded from subsequent analysis because no cholangioscopically guided biopsy samples were taken in such cases. Volume 74, No. 3 : 2011 GASTROINTESTINAL ENDOSCOPY 513
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
TABLE 1. Baseline characteristics Total no. of patients Age, y, mean ⫾ SD (range)
36 48.3 ⫾ 12.14 (27-68)
Male/female
22/14
Location of lesion (⬎1 may apply) Intrahepatic
5
Hilar
21
Mid CBD
13
Lower CBD
1
Procedures at the time of SpyGlass cholangioscopy (⬎1 may apply) Sphincterotomy
5
Extension of previous sphincterotomy
31
Biopsies
33
Stent placement
32
Balloon dilation
8
Procedure time, min, ⫾ SD (range)
36 ⫾ 10.5 (20-65)
SD, Standard deviation; CBD, common bile duct.
Preliminary sensitivity of SpyGlass cholangioscopy and SpyGlass-directed biopsy for evaluation of malignancy was calculated as the quotient of true positives divided by the sum of true positives and false negatives, and preliminary specificity as true negatives divided by the sum of true negatives and false positives.
RESULTS During the study period, 1100 patients underwent ERCP for evaluation of obstructive biliary pathology. Thirty-six of the 132 patients with indeterminate biliary lesions consented to undergo SpyGlass cholangioscopy.
Baseline characteristics and indications for SpyGlass cholangioscopy Thirty-six patients (22 men, mean age of 48.3 years [range 27-68 years]) with indeterminate stricture or mass or filling defect underwent SpyGlass cholangioscopy (Table 1). Indeterminate lesions were located mostly in the hilum and/or mid common bile duct (CBD). Indications for cholangioscopy were based on previous ERCP findings (Table 2).
Cholangioscopic findings In 3 cases, SpyGlass cholangioscopy revealed CBD calculi without a stricture. In the previous ERCP, these cases had been reported as indeterminate filling defects, and the 514 GASTROINTESTINAL ENDOSCOPY Volume 74, No. 3 : 2011
Ramchandani et al
diagnosis of CBD calculi was missed. They were included for assessment of diagnostic accuracy of cholangioscopic visual impression and of complications, but were excluded from assessment of diagnostic accuracy of cholangioscopically guided biopsy. Cholangioscopic findings or SpyGlass visual impressions are described in Table 2.
Histopathology of cholangioscopy-guided biopsies and procedural success Excluding the 3 patients with CBD calculi and no stricture, biopsy samples were obtained by using SpyBite forceps under direct visualization in 33 patients even if no suspicious lesions were identified on cholangioscopy. SpyBite biopsy specimens were deemed adequate for histology in 27 of 33 patients (Table 2), thus yielding a malignant or benign histological assessment in 27 of 33 patients and an inconclusive result in 6 of 33 patients, and cholangioscopy alone successfully determined the presence of CBD stones without a stricture in 3 patients. Hence, procedure success was attained in 30 of 36 (83%) patients.
Final diagnosis The final diagnosis of a malignant lesion was established by surgical resection (8 patients), diagnostic laparoscopy and biopsy (5 patients), clinical follow-up (3 patients), EUS-FNA cytology (2 patients), ascitic fluid cytology (2 patients), and CT-guided FNA cytology (2 patients). The final diagnosis of benign biliary stricture was made by surgical resection (6 patients) and clinical follow-up (5 patients). In 3 patients, cholangioscopy demonstrated CBD calculi without a stricture. In 11 patients who had benign SpyBite biopsy samples, the mean follow-up was 12.2 months (range 6-18 months). Final diagnoses (Table 2) were malignant in 22 patients (cholangiocarcinoma in 21 patients [Fig. 3] and gallbladder carcinoma in 1 patient) and corresponded to benign histology in 11 patients (oriental cholangiohepatitis [4 patients], autoimmune pancreatitis with biliary stricture [2 patients] (Fig. 4), primary sclerosing cholangitis with benign stricture [2 patients], idiopathic stricture [2 patients], and tubercular stricture [1 patient]).
Diagnostic accuracy of cholangioscopic impression in 36 patients The diagnostic accuracy of the SpyGlass visual impression as a predictor of a malignant or benign final diagnosis was assessed in all 36 enrolled patients. Definite malignant impression and suspected malignant impression were lumped together and termed malignant impression. Of 22 patients with a final malignant diagnosis, cholangioscopic impression was malignant in 21 patients (95%) and benign in 1 patient (5%). Of 14 patients with a final benign diagnosis including the 3 patients with CBD stones www.giejournal.org
Ramchandani et al
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
TABLE 2. ERCP findings, SpyGlass visual impressions, SpyBite biopsy results, final diagnosis ERCP findings (n ⴝ 36)
SpyGlass visual impression (n ⴝ 36)
SpyBite biopsy results (n ⴝ 33)
Final diagnosis (n ⴝ 36)
Indeterminate biliary stricture (n ⫽ 23)
Definite malignant (n ⫽ 7), suspected malignant (n ⫽ 12), benign (n ⫽ 4)
Malignant (n ⫽ 14), benign (n ⫽ 3), inadequate sample (n ⫽ 6)
Cholangiocarcinoma (n ⫽ 18), autoimmune pancreatitis with biliary stricture (n ⫽ 2), tubercular stricture (n ⫽ 1), idiopathic stricture (n ⫽ 2)
Indeterminate filling defects (n ⫽ 7)
Definite malignant (n ⫽ 4), CBD calculi without stricture* (n ⫽ 3)
Malignant (n ⫽ 4)
Carcinoma gallbladder (n ⫽ 1), cholangiocarcinoma (n ⫽ 3), choledocholithiasis* (n ⫽ 3)
Intrahepatic strictures and filling defects (n ⫽ 4)
Intrahepatic calculi with stricture: suspected malignant (n ⫽ 1), benign (n ⫽ 3)
Benign (n ⫽ 4)
Oriental cholangiohepatitis with benign strictures and intrahepatic calculi (n ⫽ 4)
PSC with dominant stricture (n ⫽ 2)
Benign (n ⫽ 2)
Benign (n ⫽ 2)
PSC with benign stricture (n ⫽ 2)
CBD, Common bile duct; PSC, primary sclerosing cholangitis. *No biopsy samples taken from 3 patients with CBD calculi and no strictures.
and no stricture, cholangioscopic impression was malignant in 3 patients (21%) and benign in 11 patients (79%). Eleven patients with cholangioscopic findings of a mass and dilated tortuous vessels were classified as definite malignant. All of these patients had a final diagnosis of malignant disease. In 13 patients with suspected malignant disease on visual impressions, the findings were villous or papillary projections and nodularity. Of these 13 patients, 3 had a final diagnosis of benign disease. Thus, the overall accuracy of SpyGlass visual impression for differentiating malignant from benign ductal lesions that were inconclusive on previous ERCP-guided brushing or biopsy was 89% (32/36) with follow-up until a definitive malignant diagnosis or at least 6 months, whichever came first. Table 3 provides detailed diagnostic measures.
Diagnostic accuracy of cholangioscopically guided biopsies in 33 patients The diagnostic accuracy of the SpyBite biopsies to determine the malignant or benign nature of intraductal lesions was assessed in all 33 patients in whom SpyBite biopsy samples were taken. The 6 patients with inconclusive histology were retained for the assessment of diagnostic accuracy of the SpyGlass-guided biopsies, thus yielding an intent-to-treat analysis. Of 22 patients with a final malignant diagnosis, SpyBite specimens yielded malignant histology in 18 patients (82%), benign histology in none, and inconclusive results in 4 patients (18%). Of 11 patients with a final benign diagnosis, SpyBite biopsies yielded malignant histology in none, benign histology in 9 patients (82%), and inconclusive results in 2 patients (18%). Thus, the overall accuracy of SpyBite biopsies for differentiating malignant from benign ductal lesions that www.giejournal.org
were inconclusive on previous ERCP-guided brushing or biopsy was 82% (27/33) in an intent-to-treat analysis with follow-up until definitive malignant diagnosis or at least 6 months, whichever came first. Table 4 provides detailed diagnostic measures.
Concordance of diagnostic measures Consideration of concordance or discordance of SpyGlass cholangioscopic impressions, histology of the SpyBite biopsies, and final diagnoses is of particular interest in cases of false-positive or false-negative diagnostic results. Three benign lesions at final diagnosis were false positives during cholangioscopic visual examination. These were wrongly characterized as malignant on the basis of nodularity. One of these patients had benign disease based on a SpyBite biopsy. Two patients with possible malignant bile duct strictures by cholangioscopy and inadequate biopsy subsequently underwent surgery. The diagnosis of benign stricture was confirmed on histology of the surgical specimen for both patients. One malignant lesion on a final diagnosis was diagnosed as benign on visual impression; however, the SpyBite biopsy sample was suggestive of malignant disease.
Complications In 36 enrolled patients, cholangitis that resolved with antibiotic therapy developed in 2 patients (5.6%), and mild pancreatitis developed in 1 patient (2.8%). No other complications were observed.
DISCUSSION Cholangioscopy has distinct advantages over other diagnostic modalities such as ERCP and EUS in allowing the Volume 74, No. 3 : 2011 GASTROINTESTINAL ENDOSCOPY 515
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
Ramchandani et al
Figure 3. A, Cholangiographic view of a indeterminate mass involving the left hepatic duct. B, C, Cholangiographic image showing SpyScope, SpyGlass, and SpyBite forceps inside the biliary tree. D, Endoscopic view of a left hepatic duct mass during SpyGlass examination. E, SpyGlass-directed forceps biopsy revealed cholangiocarcinoma. Histology demonstrated dysplastic ductal epithelial cells in glandular pattern (H&E, orig. mag. ⫻100). Inset, Histological appearance of glands lined by columnar cells with pleomorphic nuclei (H&E, orig. mag. ⫻400).
direct examination of the bile duct and in obtaining targeted biopsy samples. This is especially important in developing countries such as India where bile duct strictures have varied etiologies such as tuberculosis, parasitic cholangiopathy, and AIDS cholangiopathy, apart from bile duct malignancy. Methods of peroral cholangioscopy currently in clinical practice are conventional mother-baby cholangioscopy, SpyGlass cholangioscopy, and cholangioscopy by using ultraslim upper endoscopes. The major disadvantage of conventional POC is that it requires 2 operators, has poor 516 GASTROINTESTINAL ENDOSCOPY Volume 74, No. 3 : 2011
irrigation capability, is expensive, is easily damaged, and is difficult to use. SpyGlass cholangioscopy is designed to overcome these disadvantages. We wanted to assess the efficacy of SpyGlass cholangioscopy with cholangioscopically guided biopsies by using SpyBite forceps in the setting of our practice. In this prospective, single-center study in 36 patients, SpyGlass cholangioscopy without biopsy in 3 patients and with SpyBite biopsy in 33 patients was performed to evaluate filling defects and strictures that had inconclusive evaluations in previous ERCPguided brushing or biopsy. Six of 36 evaluations (17%) www.giejournal.org
Ramchandani et al
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
Figure 4. A, Cholangiographic view of hilar stricture in a patient with autoimmune pancreatitis. B, Cholangioscopic image revealing benign smooth stricture. C, Histology demonstrating stricture lined by bland columnar epithelium with moderate chronic inflammation in subepithelial tissue (H&E, orig. mag. ⫻100 and ⫻400 [inset]). D, Cholangiographic picture showing resolution of hilar stricture after steroid therapy.
remained inconclusive after SpyGlass cholangioscopy and SpyBite biopsy. The fewer biopsy samples (3-4 per case) and small size of the biopsy sample acquired with SpyBite forceps may be the reason for these inconclusive results. Further studies are needed to study the optimum number of biopsy samples to be taken during each cholangioscopic procedure when assessing a stricture.13 www.giejournal.org
Chen et al5 reported their data on the application of the SpyGlass Direct Visualization System in biliary diseases. Of 35 patients, 22 had indeterminate strictures and 5 had indeterminate filling defects. The procedure was successful in 91% of patients. SpyBite biopsy samples were taken from 20 patients in their study, with 71% sensitivity and 100% specificity in differentiating malignant versus benign pathologies. Volume 74, No. 3 : 2011 GASTROINTESTINAL ENDOSCOPY 517
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
TABLE 3. Diagnostic accuracy of SpyGlass cholangioscopic impression in 36 patients Final diagnosis Malignant
Benign
Malignant
21
3
Benign
1
11
Visual impression
Sensitivity
95% (21/22)
Specificity
79% (11/14)
Positive predictive value
88% (21/24), 95% CI, 68.9%-95.6%
Negative predictive value
92% (11/12), 95% CI, 64.6%-98.5%
Overall accuracy
89% (32/36)
CI, Confidence interval.
TABLE 4. Diagnostic accuracy of SpyBite cholangioscopic biopsy in 33 patients Final diagnosis Malignant
Benign
Malignant
18
0
Benign
0
9
Inconclusive
4
2
Biopsy
Sensitivity
82% (18/22)
Specificity
82% (9/11)
Positive predictive value
100% (18/18); 95% CI, 82.4%-100%
Negative predictive value
100% (9/9); 95% CI, 70%-100%
Overall accuracy
82% (27/33)
CI, Confidence interval.
One of the major advantages of cholangioscopy is to accurately localize the lesion and take targeted biopsy samples. Tissue sampling14-16 at ERCP is dependent on localization of pathology based on the fluoroscopic appearance of a stricture or an intraductal lesion. This might be particularly difficult in patients with primary sclerosing cholangitis17-19 in which there may be diffuse involvement, and targeted biopsy samples are difficult to obtain. Overall, the sensitivity of ERCP for detecting malignancy varies from 30% to 70% in various series. Shah et al20 reported that 92% of the cholangioscopic biopsy samples were adequate for diagnosis, but in their 518 GASTROINTESTINAL ENDOSCOPY Volume 74, No. 3 : 2011
Ramchandani et al
study, biopsy samples were obtained under direct visualization (cholangioscopy directed) or through the duodenoscope (cholangioscopy assisted). We did not perform any cholangioscopy-assisted biopsies in our series because the patients in this series had inconclusive pathology and/or cytology in previous biopsy and/or brushing tissue acquisition. In our study, SpyBite biopsy specimens were deemed adequate for histology in 81% of patients. The visual impression seems to be most sensitive to diagnose malignancy on cholangioscopy.21 In our series, the sensitivity of the SpyGlass visual impression was 95%, with a specificity of 79%. There are no reported trials comparing conventional POC and SpyGlass cholangioscopy. Fukuda et al9 reported an accuracy of 93% in distinguishing malignant from benign biliary strictures by using conventional POC, which is comparable to that of SpyGlass cholangioscopy (89%) in our study. In this series, we classified the lesions as definite and suspected malignant lesions. A lesion was diagnosed as definite malignant if a mass with dilated tortuous vessels was seen. The rest of the lesions, such as papillary and villous projections and intraductal nodularity, were defined as suspected malignant lesions. Benign lesions were characterized by smooth surface mucosa, without definite neovascularization, and homogeneous granular mucosa without a primary mass. Nimura22 and Nimura and Kamiya23 reported that irregularly dilated and tortuous vessels were not found in benign bile duct disease. Previous studies found that irregularly dilated and tortuous vessels were only seen in malignant biliary disorders, with both specificity and a positive predictive value of 100%.24 In our study, tortuous dilated vessels were seen in malignant disease only, whereas in benign inflammatory disease, none had such a finding. Another cholangioscopic feature that is highly suspicious of malignancy was granular or papillary projection. Wakai et al25 also demonstrated that a fine granular or fine papillary mucosal appearance on cholangioscopy is pathognomonic of superficial ductal spread of papillary cholangiocarcinoma. The limitations of this study are that it was not randomized or multicentric. In addition, definitions used for malignant and benign cholangioscopic visual impressions corresponded to previous published criteria for malignancy, but full validation of proposed classification criteria may require a future randomized study. In summary, we report a series of SpyGlass cholangioscopies and biopsies for the evaluation of patients from India with indeterminate biliary lesions. The results indicate that SpyGlass cholangioscopy is safe, allows adequate directed tissue sampling, and has an excellent predictive value for confirming or excluding malignancy in patients with indeterminate strictures or equivocal ERCP findings. Cholangioscopic confirmation of CBD stones without a stricture or confirmed malignant or benign histology of SpyBite biopsy specimens provide an accurate definitive www.giejournal.org
Ramchandani et al
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
diagnosis in the majority (83%; 30/36) of patients in whom previous ERCP-guided brushing and biopsies had yielded inconclusive diagnoses. A randomized trial comparing the diagnostic accuracy and complication profile of SpyGlass cholangioscopy with SpyBite biopsy with that of ERCPguided brushing and biopsy in patients with indeterminate strictures or indeterminate filling defects on previous imaging studies is warranted. REFERENCES 1. Röesch W, Koch H, Demling L. Peroral cholangioscopy. Endoscopy 1976; 8:172-5. 2. Siddique I, Galati J, Ankoma-Sey V, et al. The role of choledochoscopy in the diagnosis and management of biliary tract disease. Gastrointest Endosc 1999;50:67-73. 3. Urakami Y, Seifert E, Butke H. Peroral direct cholangioscopy (PDCS) using routine straight-view endoscope: first report. Endoscopy 1977;9: 27-30. 4. Chen YK, Parsi MA, Binmoeller KF, et al. Peroral cholangioscopy (PO) using a disposable steerable single operator catheter for biliary stone therapy and assessment of indeterminate strictures: a multicenter experience using Spyglass [abstract]. Gastrointest Endosc 2008;67:AB103. 5. Chen YK, Pleskow DK. SpyGlass single-operator peroral cholangiopancreatoscopy system for the diagnosis and therapy of bile-duct disorders: a clinical feasibility study (with video). Gastrointest Endosc 2007; 65:832-41. 6. Bogardus ST, Hanan I, Ruchim M, et al. “Mother-baby” biliary endoscopy: the University of Chicago experience. Am J Gastroenterol 1996;91: 105-10. 7. Soda K, Shitou K, Yoshida Y, et al. Peroral cholangioscopy using new fine-caliber flexible scope for detailed examination without papillotomy. Gastrointest Endosc 1996;43:233-8. 8. Sander R, Poesl H. Initial experience with a new babyscope for endoscopic retrograde cholangiopancreaticoscopy. Gastrointest Endosc 1996;44:191-4. 9. Fukuda Y, Tsuyuguchi T, Sakai Y, et al. Diagnostic utility of peroral cholangioscopy for various bile-duct lesions. Gastrointest Endosc 2005; 62:374-82. 10. Itoi T, Sofuni A, Itokawa F, et al. Peroral cholangioscopic diagnosis of biliary-tract diseases by using narrow-band imaging (with videos). Gastrointest Endosc 2007;66:730-6.
www.giejournal.org
11. Yeh TS, Chen NH, Jan YY, et al. Obstructive jaundice caused by biliary tuberculosis: spectrum of the diagnosis and management. Gastrointest Endosc 1999;50:105-8. 12. Behera A, Kochhar R, Dhavan S, et al. Isolated common bile duct tuberculosis mimicking malignant obstruction. Am J Gastroenterol 1997;92: 2122-3. 13. Tamada K, Kurihara K, Tomiyama T, et al. How many biopsies should be performed during percutaneous transhepatic cholangioscopy to diagnose biliary tract cancer? Gastrointest Endosc 1999;50:653-8. 14. Ferrari Júnior AP, Lichtenstein DR, et al. Brush cytology during ERCP for the diagnosis of biliary and pancreatic malignancies. Gastrointest Endosc 1994;40:140-5. 15. Howell DA, Parsons WG, Jones MA, et al. Complete tissue sampling of biliary strictures at ERCP using a new device. Gastrointest Endosc 1996; 43:498-502. 16. Farrell RJ, Jain AK, Brandwein SL, et al. The combination of stricture dilation, endoscopic needle aspiration, and biliary brushings significantly improves diagnostic yield from malignant bile duct strictures. Gastrointest Endosc 2001;54:587-94. 17. Siqueira E, Schoen RE, Silverman W, et al. Detecting cholangiocarcinoma in patients with primary sclerosing cholangitis. Gastrointest Endosc 2002;56:40-7. 18. Awadallah NS, Chen YK, Piraka C, et al. Is there a role for cholangioscopy in patients with primary sclerosing cholangitis? Am J Gastroenterol 2006;101:284-91. 19. Tischendorf JJ, Krüger M, Trautwein C, et al. Cholangioscopic characterization of dominant bile duct stenoses in patients with primary sclerosing cholangitis. Endoscopy 2006;38:665-9. Erratum in: Endoscopy 2006; 38:852. 20. Shah RJ, Langer DA, Antillon MR, et al. Cholangioscopy and cholangioscopic forceps biopsy in patients with indeterminate pancreaticobiliary pathology. Clin Gastroenterol Hepatol 2006;4:219-25. 21. Pleskow D, Parsi MA, Chen YK, et al. Biopsy of indeterminate biliary strictures-does direct visualization help? A multicenter experience [abstract]. Gastrointest Endosc 2008;67:AB103. 22. Nimura Y. Staging of biliary carcinoma: cholangiography and cholangioscopy. Endoscopy 1993;25:76-80. 23. Nimura Y, Kamiya J. Cholangioscopy. Endoscopy 1998;30:182-8. 24. Kim HJ, Kim MH, Lee SK, et al. Tumor vessel: a valuable cholangioscopic clue of malignant biliary stricture. Gastrointest Endosc 2000;52:635-8. 25. Wakai T, Shirai Y, Hatakeyama K. Peroral cholangioscopy for noninvasive papillary cholangiocarcinoma with extensive superficial ductal spread. World J Gastroenterol 2005;11:6554-6.
Volume 74, No. 3 : 2011 GASTROINTESTINAL ENDOSCOPY 519
Role of single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions: a single-center, prospective study Mohan Ramchandani, MD, DM, D. Nageshwar Reddy, MD, DM, FRCP, Rajesh Gupta, MD, DM, Sandeep Lakhtakia, MD, DM, Manu Tandan, MD, DM, Santosh Darisetty, DA, Anuradha Sekaran, MD, DNB, Guduru Venkat Rao, MS, FRCS Hyderabad, India
Background: Currently available techniques to diagnose indeterminate biliary lesions have many limitations. Objective: To assess the accuracy of single-operator peroral cholangioscopy by using the SpyGlass system to differentiate malignant from benign disease in patients with indeterminate biliary lesions. Design: Prospective, single-arm, single-center study. Setting: Tertiary referral center. Patients: Thirty-six patients with indeterminate biliary strictures and filling defects who had inconclusive results on previous biliary ductal tissue sampling. Interventions: SpyGlass cholangioscopy with cholangioscopically guided intraductal biopsies. Main Outcome Measurements: Accuracy of SpyGlass visual impression and SpyBite biopsies for differentiating malignant from benign ductal lesions. Results: Thirty-six patients (22 men, mean age 48.3 years [range 27-68 years]) with indeterminate stricture and/or filling defects underwent SpyGlass cholangioscopy. Of the 22 patients with a final diagnosis of malignant lesion, cholangioscopic impression was malignant in 21 patients (95%) and benign in 1 patient (5%). Of the 14 patients with a final diagnosis of benign disease, including the 3 patients with common bile duct stones and no stricture, cholangioscopic impression was malignant in 3 patients (21%) and benign in 11 patients (79%). The overall accuracy of SpyGlass visual impression for differentiating malignant from benign ductal lesions was 89% (32/36). The accuracy of SpyBite biopsies for differentiating malignant from benign ductal lesions that were inconclusive on ERCP-guided brushing or biopsy was 82% (27/33) in an intent-to-treat analysis. Limitations: No randomized comparison with alternative diagnostic modalities for the nature of biliary strictures. Conclusions: SpyGlass cholangioscopy with SpyBite biopsies has a high accuracy with regard to confirming or excluding malignancy in patients with indeterminate biliary lesions. (Gastrointest Endosc 2011;74:511-9.)
Abbreviations: CBD, common bile duct; POC, peroral cholangioscopy; SOPOC, single-operator peroral cholangioscopy. DISCLOSURE: The authors disclosed no financial relationships relevant to this publication. Copyright © 2011 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 doi:10.1016/j.gie.2011.04.034 Received January 14, 2011. Accepted April 21, 2011. Current affiliations: Asian Institute of Gastroenterology, Hyderabad, India. Reprint requests: D. Nageshwar Reddy, MD, DM, Asian Institute of Gastroenterology, 6-3-661, Somajiguda, Hyderabad 500 082, India. If you would like to chat with an author of this article, you may contact Dr. Reddy at [email protected].
www.giejournal.org
Despite significant advances in pancreaticobiliary imaging and EUS-guided assessment of tumors, the characterization of intraductal biliary lesions remains a challenge, even at centers with significant ERCP and EUS expertise. Cholangioscopy, which facilitates direct visual assessment and visually guided tissue sampling, holds promise as an advanced technique in cases that elude successful diagnosis by conventional ERCP or other imaging modalities. Peroral cholangioscopy is less invasive than transcutaneous cholangioscopy and can be easily performed during ERCP.1-10 However, conventional peroral cholangioscopy (POC) has certain disadvantages. It requires 2 experienced endoscopists and has limited maneuverability, inadequate irrigation, and a small instrument channel. Single-operator POC (SOPOC) using the SpyGlass Direct Visualization Volume 74, No. 3 : 2011 GASTROINTESTINAL ENDOSCOPY 511
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
System (Microvasive Endoscopy; Boston Scientific Corp, Natick, Mass) for cholangioscopy has overcome several of these limitations.4,5 The SpyGlass system is a singleoperator system that allows not only optical viewing, but also optically guided biopsies. This study was done in a high-volume center in south Asia performing approximately 6000 ERCPs annually. The indeterminate biliary lesions pose a diagnostic challenge, especially in this region where infective biliary lesions11,12 are important differential diagnoses apart from biliary malignancies. This prospective study aimed to assess the accuracy of SOPOC and cholangioscopically guided biopsies using the SpyGlass system in patients with previously inconclusive results of ERCP-guided brushing and biopsies. Compared with a previous report5 on the use of SpyGlass cholangioscopy for the characterization of indeterminate biliary strictures, this study enrolled a larger number of patients.
METHODS Study design The study was a prospective, single-center cohort study of patients meeting inclusion and exclusion criteria. It was conducted at the Asian Institute of Gastroenterology, a tertiary referral center with patient enrollment between May 2009 and February 2010. Institutional review board approval was obtained for the prospective collection of patient data and follow-up.
Patients Patients older than 18 years of age who had undergone a previous ERCP for the evaluation of obstructive jaundice and were diagnosed as having an indeterminate stricture or filling defects on ERCP were considered for this study (Fig. 1). The indeterminate nature of the ERCP finding was defined as previous brush cytology and/or endoscopic biliary biopsy with inconclusive cytology/histology or benign cytology/histology findings in the presence of strong suspicion of malignancy. Exclusion criteria were the patients’ inability to give informed consent and patients medically unfit to undergo an ERCP.
Materials and interventions SOPOC was performed by using the SpyGlass Direct Visualization System (Microvasive Endoscopy, Boston Scientific Corp, Natick, Mass), which consists of capital equipment, including a pump, a light source, and a monitor, and of 3 disposable devices: (1) optical probe (SpyGlass), (2) access and delivery catheter (SpyScope), and (3) biopsy forceps (SpyBite). The SpyGlass probe is a 6000-pixel fiberoptic bundle that enters the biliary anatomy through the SpyScope catheter. It is 231 cm long and is designed to acquire and transmit endoscopic images and conduct light into the 512 GASTROINTESTINAL ENDOSCOPY Volume 74, No. 3 : 2011
Ramchandani et al
Take-home Message ●
●
In a single-center study, the accuracy of SpyGlass cholangioscopy in the diagnosis of indeterminate biliary lesions was 89%. The visual impression on SpyGlass cholangioscopy is most sensitive, and cholangioscopy-directed biopsy is the most specific method to diagnose indeterminate biliary lesions.
biliary anatomy, providing a 70-degree field of view. This is a multiple-use device that may be reprocessed after each use. In our experience, with careful handling, the SpyGlass optical probe can be reused approximately 20 times. The SpyScope access and delivery catheter is a singleuse, single-operator-controlled, 10F diameter, 230-cmlong device designed to provide a pathway into the biliary anatomy for diagnostic and therapeutic devices. This 4-lumen catheter has an optic channel that accommodates the SpyGlass optical probe, a 1.2-mm accessory channel, and 2 independent irrigation channels. The SpyScope incorporates a handle with 2 knobs controlling 4-way steering of the catheter, thus facilitating visualization of the entire circumference of the biliary tree. The SpyBite biopsy forceps are a single-use device that passes through the 1.2-mm minimum working channel of the SpyScope catheter. The jaws of the forceps open to 4.1 mm, obtaining tissue adequate for histology in the majority of cases. It is advisable to notify the pathologist that the intraductal SpyBite tissue samples are quite small. Cholangioscopy with the SpyGlass system is performed by a single operator; this technique is described by Chen and Pleskow.5
Definitions and study objectives The primary aim of the study was to assess the accuracy, sensitivity, and specificity of SpyGlass cholangioscopy in differentiating malignant from benign biliary lesions in patients with indeterminate biliary lesions. Secondary objectives included assessment of overall procedure success and complications observed during and after the procedure. A lesion was defined as definite malignant, suspected malignant, or benign based on the presence or absence of the following visual observations: (1) mass, (2) dilated tortuous vessels, (3) papillary or villous projections, and (4) intraductal nodules. A lesion was diagnosed as definite malignant if a mass with dilated tortuous vessels was seen (Fig. 2). The rest of the lesions were defined as suspected malignant lesion. Benign lesions were characterized by smooth surface mucosa without definite neovascularization and homogeneous granular mucosa without a primary mass. Biopsy samples were collected by using SpyBite forceps and subwww.giejournal.org
Ramchandani et al
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
Figure 1. Study profile.
malignancy was established by SpyGlass cholangioscopic biopsies or other definitive tissue sampling method, such as CT/EUS-guided biopsy, intraoperative biopsy, and surgical specimen collection. Patients who received a diagnosis of malignant disease based on cholangioscopic biopsies underwent either curative surgery or palliative biliary stenting. Patients who had inconclusive SpyGlass cholangioscopic biopsy findings underwent surgery.
Data collection and follow-up
Figure 2. Cholangioscopic image showing a mass lesion with dilated tortuous vessels suggestive of a definite malignant lesion.
mitted for histopathological evaluation. In each case, 3 to 4 biopsy samples were taken. The SpyGlass procedure was deemed successful when the target lesion could be visualized and biopsy samples were deemed adequate for histopathology. This assessment was made in all patients who underwent cholangioscopy-guided biopsies in which SpyBite forceps were used, ie, in all patients except those with biliary calculi and no bile duct strictures. The SpyBite biopsy samples were categorized as malignant or benign or as inadequate sample if the specimens were deemed inadequate for histological assessment. The final diagnosis of a benign condition was assumed in cases of negative histology of ductal tissue sampling and uneventful extended clinical follow-up of at least 6 months after the index SpyGlass procedure. The final diagnosis of www.giejournal.org
Patient demographics, stricture location, previous interventions, and indications for the SpyGlass procedure based on previous ERCP findings were noted, after obtaining informed consent. SpyGlass procedures were performed by 2 experienced ERCP endoscopists independently, and biopsy samples were examined by a pathologist who was blinded to cholangioscopic findings. At the time of the procedure, the SpyGlass visual impression was recorded and classified. All patients were followed for 1 month, whereas patients with benign disease on SpyBite biopsy were followed every 2 months for at least 6 months. During each visit, all patients underwent clinical examination, routine biochemical examination, and radiological imaging including an abdominal US/CT scan. Patients underwent ERCP and biliary stent exchange if clinically warranted.
Statistical analysis Data were analyzed by using SPSS 14.0 version (SPSS South Asia Pvt. Ltd, Bangalore, India). Descriptive statistics consisted of the mean, standard deviation, and range. Patients originally enrolled in the study, but found to have biliary calculi without a biliary stricture on cholangioscopy were excluded from subsequent analysis because no cholangioscopically guided biopsy samples were taken in such cases. Volume 74, No. 3 : 2011 GASTROINTESTINAL ENDOSCOPY 513
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
TABLE 1. Baseline characteristics Total no. of patients Age, y, mean ⫾ SD (range)
36 48.3 ⫾ 12.14 (27-68)
Male/female
22/14
Location of lesion (⬎1 may apply) Intrahepatic
5
Hilar
21
Mid CBD
13
Lower CBD
1
Procedures at the time of SpyGlass cholangioscopy (⬎1 may apply) Sphincterotomy
5
Extension of previous sphincterotomy
31
Biopsies
33
Stent placement
32
Balloon dilation
8
Procedure time, min, ⫾ SD (range)
36 ⫾ 10.5 (20-65)
SD, Standard deviation; CBD, common bile duct.
Preliminary sensitivity of SpyGlass cholangioscopy and SpyGlass-directed biopsy for evaluation of malignancy was calculated as the quotient of true positives divided by the sum of true positives and false negatives, and preliminary specificity as true negatives divided by the sum of true negatives and false positives.
RESULTS During the study period, 1100 patients underwent ERCP for evaluation of obstructive biliary pathology. Thirty-six of the 132 patients with indeterminate biliary lesions consented to undergo SpyGlass cholangioscopy.
Baseline characteristics and indications for SpyGlass cholangioscopy Thirty-six patients (22 men, mean age of 48.3 years [range 27-68 years]) with indeterminate stricture or mass or filling defect underwent SpyGlass cholangioscopy (Table 1). Indeterminate lesions were located mostly in the hilum and/or mid common bile duct (CBD). Indications for cholangioscopy were based on previous ERCP findings (Table 2).
Cholangioscopic findings In 3 cases, SpyGlass cholangioscopy revealed CBD calculi without a stricture. In the previous ERCP, these cases had been reported as indeterminate filling defects, and the 514 GASTROINTESTINAL ENDOSCOPY Volume 74, No. 3 : 2011
Ramchandani et al
diagnosis of CBD calculi was missed. They were included for assessment of diagnostic accuracy of cholangioscopic visual impression and of complications, but were excluded from assessment of diagnostic accuracy of cholangioscopically guided biopsy. Cholangioscopic findings or SpyGlass visual impressions are described in Table 2.
Histopathology of cholangioscopy-guided biopsies and procedural success Excluding the 3 patients with CBD calculi and no stricture, biopsy samples were obtained by using SpyBite forceps under direct visualization in 33 patients even if no suspicious lesions were identified on cholangioscopy. SpyBite biopsy specimens were deemed adequate for histology in 27 of 33 patients (Table 2), thus yielding a malignant or benign histological assessment in 27 of 33 patients and an inconclusive result in 6 of 33 patients, and cholangioscopy alone successfully determined the presence of CBD stones without a stricture in 3 patients. Hence, procedure success was attained in 30 of 36 (83%) patients.
Final diagnosis The final diagnosis of a malignant lesion was established by surgical resection (8 patients), diagnostic laparoscopy and biopsy (5 patients), clinical follow-up (3 patients), EUS-FNA cytology (2 patients), ascitic fluid cytology (2 patients), and CT-guided FNA cytology (2 patients). The final diagnosis of benign biliary stricture was made by surgical resection (6 patients) and clinical follow-up (5 patients). In 3 patients, cholangioscopy demonstrated CBD calculi without a stricture. In 11 patients who had benign SpyBite biopsy samples, the mean follow-up was 12.2 months (range 6-18 months). Final diagnoses (Table 2) were malignant in 22 patients (cholangiocarcinoma in 21 patients [Fig. 3] and gallbladder carcinoma in 1 patient) and corresponded to benign histology in 11 patients (oriental cholangiohepatitis [4 patients], autoimmune pancreatitis with biliary stricture [2 patients] (Fig. 4), primary sclerosing cholangitis with benign stricture [2 patients], idiopathic stricture [2 patients], and tubercular stricture [1 patient]).
Diagnostic accuracy of cholangioscopic impression in 36 patients The diagnostic accuracy of the SpyGlass visual impression as a predictor of a malignant or benign final diagnosis was assessed in all 36 enrolled patients. Definite malignant impression and suspected malignant impression were lumped together and termed malignant impression. Of 22 patients with a final malignant diagnosis, cholangioscopic impression was malignant in 21 patients (95%) and benign in 1 patient (5%). Of 14 patients with a final benign diagnosis including the 3 patients with CBD stones www.giejournal.org
Ramchandani et al
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
TABLE 2. ERCP findings, SpyGlass visual impressions, SpyBite biopsy results, final diagnosis ERCP findings (n ⴝ 36)
SpyGlass visual impression (n ⴝ 36)
SpyBite biopsy results (n ⴝ 33)
Final diagnosis (n ⴝ 36)
Indeterminate biliary stricture (n ⫽ 23)
Definite malignant (n ⫽ 7), suspected malignant (n ⫽ 12), benign (n ⫽ 4)
Malignant (n ⫽ 14), benign (n ⫽ 3), inadequate sample (n ⫽ 6)
Cholangiocarcinoma (n ⫽ 18), autoimmune pancreatitis with biliary stricture (n ⫽ 2), tubercular stricture (n ⫽ 1), idiopathic stricture (n ⫽ 2)
Indeterminate filling defects (n ⫽ 7)
Definite malignant (n ⫽ 4), CBD calculi without stricture* (n ⫽ 3)
Malignant (n ⫽ 4)
Carcinoma gallbladder (n ⫽ 1), cholangiocarcinoma (n ⫽ 3), choledocholithiasis* (n ⫽ 3)
Intrahepatic strictures and filling defects (n ⫽ 4)
Intrahepatic calculi with stricture: suspected malignant (n ⫽ 1), benign (n ⫽ 3)
Benign (n ⫽ 4)
Oriental cholangiohepatitis with benign strictures and intrahepatic calculi (n ⫽ 4)
PSC with dominant stricture (n ⫽ 2)
Benign (n ⫽ 2)
Benign (n ⫽ 2)
PSC with benign stricture (n ⫽ 2)
CBD, Common bile duct; PSC, primary sclerosing cholangitis. *No biopsy samples taken from 3 patients with CBD calculi and no strictures.
and no stricture, cholangioscopic impression was malignant in 3 patients (21%) and benign in 11 patients (79%). Eleven patients with cholangioscopic findings of a mass and dilated tortuous vessels were classified as definite malignant. All of these patients had a final diagnosis of malignant disease. In 13 patients with suspected malignant disease on visual impressions, the findings were villous or papillary projections and nodularity. Of these 13 patients, 3 had a final diagnosis of benign disease. Thus, the overall accuracy of SpyGlass visual impression for differentiating malignant from benign ductal lesions that were inconclusive on previous ERCP-guided brushing or biopsy was 89% (32/36) with follow-up until a definitive malignant diagnosis or at least 6 months, whichever came first. Table 3 provides detailed diagnostic measures.
Diagnostic accuracy of cholangioscopically guided biopsies in 33 patients The diagnostic accuracy of the SpyBite biopsies to determine the malignant or benign nature of intraductal lesions was assessed in all 33 patients in whom SpyBite biopsy samples were taken. The 6 patients with inconclusive histology were retained for the assessment of diagnostic accuracy of the SpyGlass-guided biopsies, thus yielding an intent-to-treat analysis. Of 22 patients with a final malignant diagnosis, SpyBite specimens yielded malignant histology in 18 patients (82%), benign histology in none, and inconclusive results in 4 patients (18%). Of 11 patients with a final benign diagnosis, SpyBite biopsies yielded malignant histology in none, benign histology in 9 patients (82%), and inconclusive results in 2 patients (18%). Thus, the overall accuracy of SpyBite biopsies for differentiating malignant from benign ductal lesions that www.giejournal.org
were inconclusive on previous ERCP-guided brushing or biopsy was 82% (27/33) in an intent-to-treat analysis with follow-up until definitive malignant diagnosis or at least 6 months, whichever came first. Table 4 provides detailed diagnostic measures.
Concordance of diagnostic measures Consideration of concordance or discordance of SpyGlass cholangioscopic impressions, histology of the SpyBite biopsies, and final diagnoses is of particular interest in cases of false-positive or false-negative diagnostic results. Three benign lesions at final diagnosis were false positives during cholangioscopic visual examination. These were wrongly characterized as malignant on the basis of nodularity. One of these patients had benign disease based on a SpyBite biopsy. Two patients with possible malignant bile duct strictures by cholangioscopy and inadequate biopsy subsequently underwent surgery. The diagnosis of benign stricture was confirmed on histology of the surgical specimen for both patients. One malignant lesion on a final diagnosis was diagnosed as benign on visual impression; however, the SpyBite biopsy sample was suggestive of malignant disease.
Complications In 36 enrolled patients, cholangitis that resolved with antibiotic therapy developed in 2 patients (5.6%), and mild pancreatitis developed in 1 patient (2.8%). No other complications were observed.
DISCUSSION Cholangioscopy has distinct advantages over other diagnostic modalities such as ERCP and EUS in allowing the Volume 74, No. 3 : 2011 GASTROINTESTINAL ENDOSCOPY 515
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
Ramchandani et al
Figure 3. A, Cholangiographic view of a indeterminate mass involving the left hepatic duct. B, C, Cholangiographic image showing SpyScope, SpyGlass, and SpyBite forceps inside the biliary tree. D, Endoscopic view of a left hepatic duct mass during SpyGlass examination. E, SpyGlass-directed forceps biopsy revealed cholangiocarcinoma. Histology demonstrated dysplastic ductal epithelial cells in glandular pattern (H&E, orig. mag. ⫻100). Inset, Histological appearance of glands lined by columnar cells with pleomorphic nuclei (H&E, orig. mag. ⫻400).
direct examination of the bile duct and in obtaining targeted biopsy samples. This is especially important in developing countries such as India where bile duct strictures have varied etiologies such as tuberculosis, parasitic cholangiopathy, and AIDS cholangiopathy, apart from bile duct malignancy. Methods of peroral cholangioscopy currently in clinical practice are conventional mother-baby cholangioscopy, SpyGlass cholangioscopy, and cholangioscopy by using ultraslim upper endoscopes. The major disadvantage of conventional POC is that it requires 2 operators, has poor 516 GASTROINTESTINAL ENDOSCOPY Volume 74, No. 3 : 2011
irrigation capability, is expensive, is easily damaged, and is difficult to use. SpyGlass cholangioscopy is designed to overcome these disadvantages. We wanted to assess the efficacy of SpyGlass cholangioscopy with cholangioscopically guided biopsies by using SpyBite forceps in the setting of our practice. In this prospective, single-center study in 36 patients, SpyGlass cholangioscopy without biopsy in 3 patients and with SpyBite biopsy in 33 patients was performed to evaluate filling defects and strictures that had inconclusive evaluations in previous ERCPguided brushing or biopsy. Six of 36 evaluations (17%) www.giejournal.org
Ramchandani et al
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
Figure 4. A, Cholangiographic view of hilar stricture in a patient with autoimmune pancreatitis. B, Cholangioscopic image revealing benign smooth stricture. C, Histology demonstrating stricture lined by bland columnar epithelium with moderate chronic inflammation in subepithelial tissue (H&E, orig. mag. ⫻100 and ⫻400 [inset]). D, Cholangiographic picture showing resolution of hilar stricture after steroid therapy.
remained inconclusive after SpyGlass cholangioscopy and SpyBite biopsy. The fewer biopsy samples (3-4 per case) and small size of the biopsy sample acquired with SpyBite forceps may be the reason for these inconclusive results. Further studies are needed to study the optimum number of biopsy samples to be taken during each cholangioscopic procedure when assessing a stricture.13 www.giejournal.org
Chen et al5 reported their data on the application of the SpyGlass Direct Visualization System in biliary diseases. Of 35 patients, 22 had indeterminate strictures and 5 had indeterminate filling defects. The procedure was successful in 91% of patients. SpyBite biopsy samples were taken from 20 patients in their study, with 71% sensitivity and 100% specificity in differentiating malignant versus benign pathologies. Volume 74, No. 3 : 2011 GASTROINTESTINAL ENDOSCOPY 517
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
TABLE 3. Diagnostic accuracy of SpyGlass cholangioscopic impression in 36 patients Final diagnosis Malignant
Benign
Malignant
21
3
Benign
1
11
Visual impression
Sensitivity
95% (21/22)
Specificity
79% (11/14)
Positive predictive value
88% (21/24), 95% CI, 68.9%-95.6%
Negative predictive value
92% (11/12), 95% CI, 64.6%-98.5%
Overall accuracy
89% (32/36)
CI, Confidence interval.
TABLE 4. Diagnostic accuracy of SpyBite cholangioscopic biopsy in 33 patients Final diagnosis Malignant
Benign
Malignant
18
0
Benign
0
9
Inconclusive
4
2
Biopsy
Sensitivity
82% (18/22)
Specificity
82% (9/11)
Positive predictive value
100% (18/18); 95% CI, 82.4%-100%
Negative predictive value
100% (9/9); 95% CI, 70%-100%
Overall accuracy
82% (27/33)
CI, Confidence interval.
One of the major advantages of cholangioscopy is to accurately localize the lesion and take targeted biopsy samples. Tissue sampling14-16 at ERCP is dependent on localization of pathology based on the fluoroscopic appearance of a stricture or an intraductal lesion. This might be particularly difficult in patients with primary sclerosing cholangitis17-19 in which there may be diffuse involvement, and targeted biopsy samples are difficult to obtain. Overall, the sensitivity of ERCP for detecting malignancy varies from 30% to 70% in various series. Shah et al20 reported that 92% of the cholangioscopic biopsy samples were adequate for diagnosis, but in their 518 GASTROINTESTINAL ENDOSCOPY Volume 74, No. 3 : 2011
Ramchandani et al
study, biopsy samples were obtained under direct visualization (cholangioscopy directed) or through the duodenoscope (cholangioscopy assisted). We did not perform any cholangioscopy-assisted biopsies in our series because the patients in this series had inconclusive pathology and/or cytology in previous biopsy and/or brushing tissue acquisition. In our study, SpyBite biopsy specimens were deemed adequate for histology in 81% of patients. The visual impression seems to be most sensitive to diagnose malignancy on cholangioscopy.21 In our series, the sensitivity of the SpyGlass visual impression was 95%, with a specificity of 79%. There are no reported trials comparing conventional POC and SpyGlass cholangioscopy. Fukuda et al9 reported an accuracy of 93% in distinguishing malignant from benign biliary strictures by using conventional POC, which is comparable to that of SpyGlass cholangioscopy (89%) in our study. In this series, we classified the lesions as definite and suspected malignant lesions. A lesion was diagnosed as definite malignant if a mass with dilated tortuous vessels was seen. The rest of the lesions, such as papillary and villous projections and intraductal nodularity, were defined as suspected malignant lesions. Benign lesions were characterized by smooth surface mucosa, without definite neovascularization, and homogeneous granular mucosa without a primary mass. Nimura22 and Nimura and Kamiya23 reported that irregularly dilated and tortuous vessels were not found in benign bile duct disease. Previous studies found that irregularly dilated and tortuous vessels were only seen in malignant biliary disorders, with both specificity and a positive predictive value of 100%.24 In our study, tortuous dilated vessels were seen in malignant disease only, whereas in benign inflammatory disease, none had such a finding. Another cholangioscopic feature that is highly suspicious of malignancy was granular or papillary projection. Wakai et al25 also demonstrated that a fine granular or fine papillary mucosal appearance on cholangioscopy is pathognomonic of superficial ductal spread of papillary cholangiocarcinoma. The limitations of this study are that it was not randomized or multicentric. In addition, definitions used for malignant and benign cholangioscopic visual impressions corresponded to previous published criteria for malignancy, but full validation of proposed classification criteria may require a future randomized study. In summary, we report a series of SpyGlass cholangioscopies and biopsies for the evaluation of patients from India with indeterminate biliary lesions. The results indicate that SpyGlass cholangioscopy is safe, allows adequate directed tissue sampling, and has an excellent predictive value for confirming or excluding malignancy in patients with indeterminate strictures or equivocal ERCP findings. Cholangioscopic confirmation of CBD stones without a stricture or confirmed malignant or benign histology of SpyBite biopsy specimens provide an accurate definitive www.giejournal.org
Ramchandani et al
Single-operator peroral cholangioscopy in the diagnosis of indeterminate biliary lesions
diagnosis in the majority (83%; 30/36) of patients in whom previous ERCP-guided brushing and biopsies had yielded inconclusive diagnoses. A randomized trial comparing the diagnostic accuracy and complication profile of SpyGlass cholangioscopy with SpyBite biopsy with that of ERCPguided brushing and biopsy in patients with indeterminate strictures or indeterminate filling defects on previous imaging studies is warranted. REFERENCES 1. Röesch W, Koch H, Demling L. Peroral cholangioscopy. Endoscopy 1976; 8:172-5. 2. Siddique I, Galati J, Ankoma-Sey V, et al. The role of choledochoscopy in the diagnosis and management of biliary tract disease. Gastrointest Endosc 1999;50:67-73. 3. Urakami Y, Seifert E, Butke H. Peroral direct cholangioscopy (PDCS) using routine straight-view endoscope: first report. Endoscopy 1977;9: 27-30. 4. Chen YK, Parsi MA, Binmoeller KF, et al. Peroral cholangioscopy (PO) using a disposable steerable single operator catheter for biliary stone therapy and assessment of indeterminate strictures: a multicenter experience using Spyglass [abstract]. Gastrointest Endosc 2008;67:AB103. 5. Chen YK, Pleskow DK. SpyGlass single-operator peroral cholangiopancreatoscopy system for the diagnosis and therapy of bile-duct disorders: a clinical feasibility study (with video). Gastrointest Endosc 2007; 65:832-41. 6. Bogardus ST, Hanan I, Ruchim M, et al. “Mother-baby” biliary endoscopy: the University of Chicago experience. Am J Gastroenterol 1996;91: 105-10. 7. Soda K, Shitou K, Yoshida Y, et al. Peroral cholangioscopy using new fine-caliber flexible scope for detailed examination without papillotomy. Gastrointest Endosc 1996;43:233-8. 8. Sander R, Poesl H. Initial experience with a new babyscope for endoscopic retrograde cholangiopancreaticoscopy. Gastrointest Endosc 1996;44:191-4. 9. Fukuda Y, Tsuyuguchi T, Sakai Y, et al. Diagnostic utility of peroral cholangioscopy for various bile-duct lesions. Gastrointest Endosc 2005; 62:374-82. 10. Itoi T, Sofuni A, Itokawa F, et al. Peroral cholangioscopic diagnosis of biliary-tract diseases by using narrow-band imaging (with videos). Gastrointest Endosc 2007;66:730-6.
www.giejournal.org
11. Yeh TS, Chen NH, Jan YY, et al. Obstructive jaundice caused by biliary tuberculosis: spectrum of the diagnosis and management. Gastrointest Endosc 1999;50:105-8. 12. Behera A, Kochhar R, Dhavan S, et al. Isolated common bile duct tuberculosis mimicking malignant obstruction. Am J Gastroenterol 1997;92: 2122-3. 13. Tamada K, Kurihara K, Tomiyama T, et al. How many biopsies should be performed during percutaneous transhepatic cholangioscopy to diagnose biliary tract cancer? Gastrointest Endosc 1999;50:653-8. 14. Ferrari Júnior AP, Lichtenstein DR, et al. Brush cytology during ERCP for the diagnosis of biliary and pancreatic malignancies. Gastrointest Endosc 1994;40:140-5. 15. Howell DA, Parsons WG, Jones MA, et al. Complete tissue sampling of biliary strictures at ERCP using a new device. Gastrointest Endosc 1996; 43:498-502. 16. Farrell RJ, Jain AK, Brandwein SL, et al. The combination of stricture dilation, endoscopic needle aspiration, and biliary brushings significantly improves diagnostic yield from malignant bile duct strictures. Gastrointest Endosc 2001;54:587-94. 17. Siqueira E, Schoen RE, Silverman W, et al. Detecting cholangiocarcinoma in patients with primary sclerosing cholangitis. Gastrointest Endosc 2002;56:40-7. 18. Awadallah NS, Chen YK, Piraka C, et al. Is there a role for cholangioscopy in patients with primary sclerosing cholangitis? Am J Gastroenterol 2006;101:284-91. 19. Tischendorf JJ, Krüger M, Trautwein C, et al. Cholangioscopic characterization of dominant bile duct stenoses in patients with primary sclerosing cholangitis. Endoscopy 2006;38:665-9. Erratum in: Endoscopy 2006; 38:852. 20. Shah RJ, Langer DA, Antillon MR, et al. Cholangioscopy and cholangioscopic forceps biopsy in patients with indeterminate pancreaticobiliary pathology. Clin Gastroenterol Hepatol 2006;4:219-25. 21. Pleskow D, Parsi MA, Chen YK, et al. Biopsy of indeterminate biliary strictures-does direct visualization help? A multicenter experience [abstract]. Gastrointest Endosc 2008;67:AB103. 22. Nimura Y. Staging of biliary carcinoma: cholangiography and cholangioscopy. Endoscopy 1993;25:76-80. 23. Nimura Y, Kamiya J. Cholangioscopy. Endoscopy 1998;30:182-8. 24. Kim HJ, Kim MH, Lee SK, et al. Tumor vessel: a valuable cholangioscopic clue of malignant biliary stricture. Gastrointest Endosc 2000;52:635-8. 25. Wakai T, Shirai Y, Hatakeyama K. Peroral cholangioscopy for noninvasive papillary cholangiocarcinoma with extensive superficial ductal spread. World J Gastroenterol 2005;11:6554-6.
Volume 74, No. 3 : 2011 GASTROINTESTINAL ENDOSCOPY 519