Intraductal US in evaluation of biliary strictures without a mass lesion on CT scan or magnetic...

ORIGINAL ARTICLE: Clinical Endoscopy

Intraductal US in evaluation of biliary strictures without a mass lesionon CTscan or magnetic resonance imaging: significance of focal wallthickening and extrinsic compression at the stricture site

Naveen B. Krishna, MBBS, Saradhi Saripalli, Rizwan Safdar, MD, Banke Agarwal, MD

St. Louis, Missouri, USA

Background and Objective: The clinical utility of intraductal US (IDUS) for evaluating biliary strictures hasbeen limited because of a lack of easily recognized morphologic criteria to distinguish benign and malignantstrictures. We studied the clinical value of 2 easily assessed IDUS findings: wall thickness and extrinsic compres-sion at the stricture site.

Design and Setting: A retrospective, single-center study.

Patients and Methods: Forty-five patients without an identifiable mass on CT/magnetic resonance imaging,who underwent ERCP/IDUS for evaluation of biliary strictures were studied. IDUS pictures were reviewed spe-cifically to measure wall thickness and to look for extrinsic compression at the stricture site.

Main Outcome Measurements and Results: The mean age of the patients was 64.2 � 13.3 years. Thirty pa-tients had jaundice at presentation, and in 15 patients a stricture was suspected on imaging. The mean length ofbiliary strictures was 15.1 � 7.8 mm. Strictures were distal (distal common bile duct) in 25 patients and proximal(mid/proximal common bile duct or common hepatic duct) in 20 patients. Fourteen strictures were finally di-agnosed to be malignant. Strictures in 20 patients were caused by extrinsic compression, and tissue diagnosiswas readily obtained by EUS-FNA in all these patients. Of 25 strictures without extrinsic compression, 6 weremalignant (wall thickness 9-16 mm) and 19 were benign (wall thickness %9 mm). Bile duct wall thickness%7 mm at the stricture site, in the absence of extrinsic compression, had a negative predictive value of 100%for excluding malignancy in this cohort.

Limitations: Retrospective study and relatively small number of patients.

Conclusions: Evaluation of wall thickness and the presence of extrinsic compression at the site of biliary stric-tures by IDUS can help in further management of these patients. (Gastrointest Endosc 2007;66:90-6.)

Biliary intraductal US (IDUS) is easy to perform and canpotentially be used as an adjunct to biliary brushing andcholangiographic appearance for evaluation of biliary stric-tures encountered during ERCP.1-5 However, it has not be-come popular in clinical practice. One of the main reasonsfor this is that the EUS criteria used for assessing biliarystrictures are complex.1-6 Second, published studies haveexamined the use of IDUS to diagnose malignancy.1-5

But morphologic findings by IDUS, howsoever suggestive,cannot be diagnostic of cancer. It is possible to get a defin-itive cytologic diagnosis of malignancy by EUS-FNA in

See CME section; p. 126.

Copyright ª 2007 by the American Society for Gastrointestinal Endoscopy

0016-5107/$32.00

doi:10.1016/j.gie.2006.10.020

90 GASTROINTESTINAL ENDOSCOPY Volume 66, No. 1 : 2007

most of the patients with biliary strictures with about100% specificity,7 obviating the need for IDUS, whose re-ported specificity ranges from 86% to 95%.1-4

The majority of patients with suspected biliary obstruc-tion are imaged with CT, magnetic resonance imaging(MRI), or US. If no mass lesion is discernible, EUS is recom-mended as the next diagnostic test. But for various reasons,including limited availability of EUS, ERCP is frequently per-formed for diagnosis and therapy. The usual causes of bili-ary obstruction, common bile duct (CBD) stones andstrictures, are easily identified during ERCP. However, de-termination of the cause of the stricture on the basis ofthe morphologic features and brush cytologic studies ofthe stricture is not reliable.8 Although EUS-FNA can helpby providing a cytologic diagnosis of malignancy in a major-ity of these patients,7 it is still impossible to conclusively

www.giejournal.org

Krishna et al IDUS for biliary strictures

rule out malignancy, particularly in patients with obstruc-tive jaundice.

There are no definite guidelines for the management ofbiliary strictures of indeterminate etiology. A proactive ap-proach with surgical exploration is advocated in some cen-ters on the basis of earlier observation that the majority ofbiliary strictures are malignant. This is definitely the case ifall patients with jaundice and biliary stricture are included.However, if patients with an identifiable mass lesion onCT/MRI scan are excluded, the risk of malignancy is about51% to 70%.1,3,5 Surgical exploration in all these patientsmay not be justifiable. There is a need for a test that couldpotentially help identify patients with a low risk of malig-nancy in whom conservative management might be ap-propriate and those with a high risk for malignancy whowould most benefit from surgical exploration.

On the basis of our clinical observations, we reasonedthat IDUS could potentially help in the management ofthese patients on the basis of 2 easily assessed character-istics: presence of an extrinsic compression and wall thick-ness of bile duct at the stricture site. If a mass lesion isnoted causing extrinsic compression, biopsy by EUS-FNAfor a definitive diagnosis can easily be performed. In theabsence of extrinsic compression, it might be possible toidentify a cutoff value for wall thickness that can separatepatients with high and low probabilities of malignancy.These IDUS criteria might allow more judicious use of ex-ploratory surgery in patients with indeterminate biliarystrictures.

PATIENTS AND METHODS

This retrospective analysis was performed on patients(n Z 58) who underwent ERCP and IDUS at Saint LouisUniversity Hospital from June 2002 to July 2005 for evalu-ation of biliary strictures and who did not have an identifi-able mass lesion causing the stricture on CT or MRI.Patients with strictures above the bifurcation of the com-mon hepatic duct (CHD) (n Z 6) and patients with am-pullary tumors (n Z 4) were excluded. IDUS pictureswere reviewed specifically to measure wall thickness andto look for extrinsic compression at the stricture site.

This study was approved by the Institutional ReviewBoard at Saint Louis University School of Medicine. All pa-tients provided consent for the procedure performed.

IDUS examinationERCP and IDUS were performed with the patient under

conscious sedation and performed by one operator (B.A.).The procedure was performed in a standard fashion witha side-viewing therapeutic duodenoscope (Pentax, Or-angeburg, NY). IDUS was performed with a 20-MHz UScatheter probe (UMG20-29R, Olympus, Melville, NY) con-nected to a standard EUS processor (EU-M30, Olympus).This probe provides radial scanning perpendicular to its

www.giejournal.org

Capsule Summary

What is already known on this topic

d The use of intraductal ultrasonography to distinguishbenign and malignant biliary strictures has been limitedby the lack of easily recognizable morphologic criteria.

What this study adds to our knowledge

d In a retrospective single-center study of 45 patients whounderwent intraductal ultrasonography, biliary strictureswere caused by extrinsic compression in 20 patients, anda tissue diagnosis was made by EUS-FNA in all.

d In 25 patients with no extrinsic compression, bile ductwall thickness %7 mm at the stricture site was highlypredictive of a benign stricture.

axis. The probe has a monorail design for passage overa guidewire and has a radiopaque tip for fluoroscopic visu-alization.

Cholangiography was performed with a standard iodin-ated contrast medium. When a stricture was noted,a guidewire was placed in the bile duct across the stric-ture. The IDUS probe was then passed over the guidewirebeyond the stricture, and IDUS imaging was obtainedwhile the probe was withdrawn. Biliary sphincterotomyand stricture dilation, if needed, were deferred until afterthe IDUS examination was completed. Stricture dilationwas not performed for purposes of IDUS examination.Brushings of the biliary stricture were obtained after theIDUS examination. The remainder of the ERCP procedurewas then performed according to the individual needs ofeach patient.

For the purposes of the study, the ERCP and IDUS imageswere reviewed (without knowledge of the final diagnosis)specifically to evaluate (1) presence of extrinsic compres-sion at the site of the stricture and (2) thickness of thebile duct wall at the site of the stricture. Extrinsic compres-sion was considered present if a focal lesion was seen press-ing on the bile duct at the stricture site. To distinguishbetween extrinsic tumor infiltrating the bile duct and eccen-tric thickening of the bile duct growing outside the bileduct, we arbitrarily set a size limit: eccentric wall thickeningof more than 2 cm in thickness was considered to be due toextrinsic compression. This cutoff value was chosen be-cause the outer margins of lesions bigger than 2 cm cannotbe visualized adequately by IDUS, and for purposes of diag-nosis biopsy is easily performed by EUS-FNA for eccentricwall thickening or extrinsic lesions R2 cm.

Final diagnosis and follow-upThe final diagnosis was based on definitive cytologic

studies diagnostic for malignancy, surgical pathology stud-ies, or clinical follow-up of at least 6 months. Repeat ERCPand biliary brushing was done as clinically indicated.

Volume 66, No. 1 : 2007 GASTROINTESTINAL ENDOSCOPY 91

IDUS for biliary strictures Krishna et al

Clinical follow-up included imaging with abdominal CTwith or without EUS-FNA. Follow-up information wasalso obtained from hospital medical records, by contactingprimary care physicians/referring physicians, and withfollow-up phone calls to patients that we routinely makeas a part of our clinical practice.

RESULTS

Patient characteristicsForty-eight patients met the inclusion criteria. Two pa-

tients had biliary strictures caused by stones that were im-pacted into the wall and could not be identified bycholangiogram. One patient died of unrelated causes 1month after ERCP/IDUS. These patients were excludedfrom subsequent analysis. The mean age of the 45 patientsincluded in the study was 64.2 � 13.3 years (range 35-91years) (Table 1). Thirty patients had jaundice at presenta-tion. In 15 patients without jaundice, the stricture wasidentified by imaging (MRCP, ERCP, or intraoperative chol-angiogram during laparoscopic cholecystectomy). Five pa-tients had a history of orthotopic liver transplantation.The mean length of biliary strictures was 15.1 � 7.8 mm(range 5-40 mm). Twenty-five strictures were located inthe distal CBD (distal strictures), and 20 were located inthe mid or proximal CBD or CHD (proximal strictures).Fourteen patients were finally diagnosed to have malig-nant strictures.

Biliary strictures resulting from extrinsiccompression

Twenty of the 45 patients had extrinsic compression atthe stricture site (Table 2). Extrinsic compression inproximal strictures (n Z 9) was due to malignancy in 4patients and was benign in 5 patients as a result of vascular

TABLE 1. Patient characteristics (N Z 45)

Age (y) 64.2 � 13.3 (range 35-91)

Sex (male/female) 27:18

Length of stricture (mm) 15.1 � 7.8 (range 5-40)

Location of stricture

Proximal duct 20

Distal CBD 25

Presenting symptoms

Jaundice 30

Abnormal imaging 15

Final diagnosis

Benign 31

Malignant 14


compression (n Z 2) (Fig. 1), enlarged gallbladder withcholecystitis (n Z 1) (Fig. 2), periductal abscess (n Z 1)(Fig. 3), and scar tissue (n Z 1). The patients with peri-ductal abscess and scar tissue had a history of liver trans-plantation. Of the 11 patients with extrinsic compressionin the distal CBD, 4 patients had pancreatic cancer(Fig. 4A) and 3 patients had chronic pancreatitis(Fig. 4B). The remaining 4 patients had a hypoechoiclesion in the pancreas adjacent to the stricture but hadno cytologic evidence of malignancy (brush cytology dur-ing ERCP and EUS-FNA), did not meet the morphologicEUS criteria for chronic pancreatitis, and did not displayclinical evidence of malignancy during follow-up.

It was possible to correctly diagnose malignancy in allthe patients (n Z 8) who had extrinsic compression asa result of malignant lesions in our cohort by EUS-FNA.In the patient with periductal abscess (Fig. 3), IDUSshowed a discreet cystic lesion pressing on the CBD,and pus was aspirated from the lesion with EUS-FNAwith subsequent relief of jaundice and normalization ofliver function test results. Two patients with vascular com-pression had a stricture with proximal dilation of the ductsseen on MRCP and confirmed by ERCP but had no historyof jaundice. They did not display obstructive jaundice dur-ing follow-up and had a stable radiologic appearance. Thepatient with acute cholecystitis underwent emergency sur-gery and was found to have gangrenous cholecystitis. Thejaundice and abnormal liver function test results resolvedafter cholecystectomy in this patient. The diagnosis of scartissue was based on eccentric wall thickening (more than

TABLE 2. IDUS characteristics of malignant and benign

biliary strictures (N Z 45)

Wall thickness (mm)

(mean ± SD) (range)

Extrinsic compression present

Proximal duct

Malignant 4 d

Benign 5 d

Distal duct

Malignant 4 d

Benign 7 d

Extrinsic compression absent

Proximal duct

Malignant 3 12.7 � 3.5 (9-16)

Benign 8 5.5 � 2.0 (3-9)

Distal duct

Malignant 3 9.6 � 1.1 (9-11)

Benign 11 4.5 � 2.8 (1-9)

www.giejournal.org


2 cm) at the site of anastamosis of donor and recipient bileducts, and EUS-FNA specimens revealed fibrous tissuewithout any malignant cells and no clinical evidence ofmalignancy on follow-up. Chronic pancreatitis was diag-nosed on the basis of suggestive EUS morphologic charac-teristics (R5 criteria),9 suggestive cytologic findings inEUS-FNA specimens (presence of fibrous tissue andabsence of malignant cells), and clinical follow-up of atleast 6 months.

Biliary strictures without extrinsiccompression

Of the 25 patients without extrinsic compression (Table2), 6 patients with a malignant stricture had wall thicknessranging from 9 to 16 mm, and 19 patients with a benignstricture had wall thickness %9 mm. There was a signifi-cant difference (P Z .0015) in thickness of the bile ductwall at the site of the stricture between benign strictures(4.9 � 2.4 mm) and malignant strictures (11.2 � 2.8mm). Fig. 5 shows representative IDUS images illustratingwall thickness of a benign and a malignant stricture. Biliarywall thickness %7 mm in the absence of extrinsic com-pression had a negative predictive value of 100% forexcluding malignancy in this cohort. Of the 8 patientswith a wall thickness of R9 mm, 6 patients had cancer(positive predictive value 75%). Brush cytologic findingswere positive for malignancy in 3 of these 6 patients,and in 3 patients additional tissue was procured by EUS-FNA to obtain a definitive cytologic diagnosis ofmalignancy.

Figure 1. IDUS image demonstrating extrinsic compression by a promi-

nent vessel. There was discreet narrowing at the junction of left hepatic

and common hepatic ducts with proximal dilation of the left hepatic duct

that was detected by MRCP in a patient with abnormal liver function test

results but no history of jaundice.

www.giejournal.org

DISCUSSION

In this retrospective analysis of patients with biliarystrictures, we evaluated 2 easily identifiable morphologiccharacteristics by IDUS examination, presence of extrinsiccompression and wall thickness at the site of the biliarystricture, as potential aids in the further management ofthese patients. In patients with extrinsic compression atthe stricture site, definitive cytologic diagnosis of malig-nancy was obtained by EUS-FNA. In the absence of extrin-sic wall compression, wall thickness %7 mm at the siteof the stricture was highly suggestive of a benign biliarystricture.

Currently, there is no consensus on management of bil-iary strictures of indeterminate etiology. If all patients whoundergo ERCP for obstructive jaundice are included, thenthe majority of biliary strictures are malignant. Therefore,in several centers in the country these patients are rou-tinely taken for surgery unless there is a contraindicationor the patient refuses surgery. However, if patients withan identifiable mass on lesion on CT scan or MRI are ex-cluded from this group, then 30% to 49% of patientshave benign strictures.1,3,5 Furthermore, biliary stricturesare often identified by MRCP or ERCP on intraoperativecholangiograms in patients with abnormal liver functiontest results but without obstructive jaundice. The likeli-hood of malignancy in patients with biliary strictures with-out obstructive jaundice is probably not as high as in thosewith obstructive jaundice. Recommending exploratorysurgery in all these patients may not be justifiable becausemany of these strictures can be managed by endoscopic

Figure 2. IDUS image showing extrinsic compression of the common

hepatic duct by enlarged gallbladder. This elderly diabetic patient had

painless jaundice with leukocytosis and abnormal transaminase values.

ERCP showed a stricture in the CHD.




therapies. But, because of fear of missing early cancer,a policy of watchful waiting is distressing both for the phy-sicians and the patients.

There is, therefore, a need for a test that could helpidentify patients with a low risk of malignancy who couldbe followed up clinically without exploratory surgery. Wefound that a wall thickness %7 mm in he absence of ex-trinsic compression suggests benign stricture. This is con-sistent with another previously published study suggestingthat strictures more than 8 mm in thickness were likely tobe malignant.10,11 IDUS findings, howsoever suggestive ofa benign stricture, should be used in conjunction with bil-iary brush cytologic studies. If any enlarged periportallymph nodes are identified by IDUS in these patients,they should also be sampled by EUS-FNA to rule out ma-lignancy. In addition, IDUS can readily identify benign eti-ologies such as vascular compression, Mirizzi’s syndrome,and enlarged gallbladder pressing on the bile duct. In pa-tients without evidence of malignancy on brush cytologicexamination, the above findings can potentially help selectpatients in whom exploratory surgery is likely to havea low yield for malignancy.

In our study cohort, 6 of the 8 patients with wall thick-ness R8 mm had a malignant stricture justifying a moreproactive approach in these patients, including explor-atory surgery. The patients with wall thickening R8 mmmay also be taken for EUS-FNA in an attempt to obtaina definitive preoperative diagnosis of malignancy,7 orthey may directly be taken for surgery, depending on pref-erence of the treating physicians. In patients with extrinsiccompression of the bile duct, EUS-FNA is helpful in iden-tifying malignant lesions. Pancreatic cancer and chronicpancreatitis were the 2 most common causes of extrinsiccompression of the distal CBD in our study patients. Itis difficult to distinguish them morphologically by IDUS(Fig. 4), but EUS-FNA can usually provide a definitive diag-nosis in these patients.12

The question then arises: ‘‘Where does IDUS figure inthe diagnostic evaluation of patients with a biliary stric-ture?’’ If a biliary stricture is encountered in a patient un-dergoing ERCP, it might be reasonable to perform IDUSduring the same examination to provide information ad-junctive to the endoscopic and cholangiographic andbrush cytologic findings. That would require interven-tional endoscopists trained in IDUS interpretation andavailability of appropriate IDUS equipment in ERCP suites.Identification of simple IDUS criteria such as wall thick-ness and presence or absence of extrinsic compressionat the site of the stricture could easily be learned bymost interventional endoscopists with limited additional

Figure 3. US images showing extrinsic compression of the CBD by

a periductal abscess. A and B, IDUS images just above and at the level

of the biliary stricture. C, Image obtained with radial echoendoscope

(EUM-130, Olympus) of the abscess bulging into the CBD.

=

www.giejournal.org


training. They could also use videos or pictures to confirmthe findings with their endosonographer colleagues. EUS-FNA is a more appropriate diagnostic test if the patient hasobstructive jaundice without a mass lesion on CT/MRI or ifa biliary stricture is noted during intraoperative cholagio-gram or by MRCP. However, if the etiology of the biliarystricture cannot be ascertained by EUS-FNA, then ERCPwith IDUS might provide additive information to help de-cide on the need for exploratory surgery. If a potential butnot a definitive cause of biliary stricture is noted duringEUS (such as enlarged periportal lymph node or an aber-rant vessel), IDUS may help verify that the biliary stricture

Figure 4. IDUS image showing extrinsic compression of the distal CBD

by (A) pancreatic cancer and (B) a focal hypoechoic lesion in a patient

with chronic pancreatitis. Focal hypoechoic lesions are often noted in

the background of chronic pancreatitis, and they are not always the result

of pancreatic cancer.

www.giejournal.org

is indeed due to the lesion noted on EUS so that alternatesignificant pathologic conditions in the bile duct are notmissed.

Because of its retrospective design, the findings of thisstudy are not definitive. We made every effort to minimizethe limitations of a retrospective analysis. The number ofpatients in this study is also relatively small. The study isbased on objective, easily identifiable IDUS criteria. Ourfinding that wall thickness of %8 mm at the stricturesite is suggestive of a benign etiology is consistent withan earlier published study.10,11 The findings of this studyare important because they demonstrate that it might bepossible to stratify the patients with indeterminate biliarystrictures into those with low risk and high risk of malig-nancy on the basis of simple IDUS criteria that potentiallycan be learned by most interventional endoscopists. This

Figure 5. IDUS image demonstrating bile duct wall thickness at the site

of benign (A) and malignant (B) biliary strictures in absence of extrinsic

compression of the bile duct.



may provide useful adjunctive information in choosing ap-propriate strategies for further management of patientswith this difficult clinical scenario.

In conclusion, assessment by IDUS of focal wall thick-ening and extrinsic compression at the site of the biliarystricture can potentially help in further management of in-determinate biliary strictures.

DISCLOSURE

The authors have no disclosures.

REFERENCES

1. Farrell RJ, Agarwal B, Brandwein SL, et al. Intraductal US is a useful ad-

junct to ERCP for distinguishing malignant from benign biliary stric-

tures. Gastrointest. Endosc 2002;56:681-7.

2. Vasquez-Sequeiros E, Baron TH, Clain JE, et al. Evaluation of indetermi-

nate bile duct strictures by intraductal US. Gastrointest Endosc

2002;56:372-9.

3. Damagk D, Poremba C, Dietl K-H, et al. Endoscopic transpapillary biop-

sies and intraductal ultrasonography in the diagnosis of bile duct stric-

tures: a prospective study. Gut 2002;51:240-4.

4. Tamada K, Tomiyama T, Wada S, et al. Endoscopic transpapillary bile

duct biopsy with combination of intraductal ultrasonography in the

diagnosis of biliary strictures. Gut 2002;50:326-31.

5. Stavropoulos S, Larghi A, Verna E, et al. Intraductal ultrasound for the

evaluation of patients with biliary strictures and no abdominal mass

on computed tomography. Endoscopy 2005;37:715-21.


6. Tamada K, Kanai N, Wada S, et al. Utility and limitations of intraductal

ultrasonography in distinguishing longitudinal cancer extension along

the bile duct from inflammatory wall thickening. Abdom Imaging

2001;26:623-31.

7. Eloubeidi MA, Chen VK, Jhala NC, et al. Endoscopic ultrasound-guided

fine needle aspiration biopsy of suspected cholangiocarcinoma. Clin

Gastroenterol Hepatol 2004;2:207-8.

8. Fogel EL, deBellis M, McHenry L, et al. Effectiveness of a new long cy-

tology brush in the evaluation of malignant biliary obstruction: a pro-

spective study. Gastrointest Endosc 2006;63:71-7.

9. Sahai AV, Zimmerman M, Aabakken L, et al. Prospective assessment of

the ability of endoscopic ultrasound to diagnose, exclude, or establish

the severity of chronic pancreatitis found by endoscopic retrograde

cholangiopancreatography. Gastrointest Endosc 1998;48:18-25.

10. Tamada K, Sugano K. Diagnosis and non-surgical treatment of bile

duct carcinoma: developments in the past decade. J Gastroenterol

2000;35:319-25.

11. Tamada K, Tomiyama T, Wada S, et al. Detection of bile duct cancer in

an early stage by intraductal ultrasonography [in Japanese with En-

glish abstract]. Endosc Dig 1999;11:1164-72.

12. Varadarajulu S, Tamhane A, Eloubeidi MA. Yield of EUS-guided FNA of

pancreatic masses in the presence or the absence of chronic pancre-

atitis. Gastrointest Endosc 2005;62:737-41.

Received March 27, 2006. Accepted October 9, 2006.

Current affiliations: Division of Gastroenterology and Hepatology, Saint

Louis University School of Medicine, St. Louis, Missouri, USA.

Reprint requests: Banke Agarwal, MD, Division of Gastroenterology and

Hepatology, Saint Louis University School of Medicine, 3635 Vista Ave,

St. Louis, MO 63110.

www.giejournal.org

Intraductal US in evaluation of biliary strictures without a mass lesion on CT scan or magnetic...

Documents

Transcript of Intraductal US in evaluation of biliary strictures without a mass lesion on CT scan or magnetic...