Reference Section
a report by
William R Brugge, MD
Director, Gastrointestinal Endoscopy Unit, Massachusetts General Hospital and
Associate Professor of Medicine, Harvard Medical School
Endoscopic Ultrasound ý A Review
B USINESS BRIEFING: US GASTROENTEROLOGY REVIEW 2005
1
Introduction
Endoscopic ultrasound (EUS) represents the most
significant development in endoscopy over the past 20
years.The role of EUS has continued to expand with
the introduction of EUS-guided fine needle aspiration
(FNA) of mural, extra-mural, and pancreatic lesions.


This technique has also been used for several
therapeutic applications.


Instrumentation
Endosonoscopes may be linear or radial in design,
depending on the orientation of the ultrasound
transducer. Radial imaging provides easily
interpretable cross-sectional imaging with either a
probe or an endosonoscope. In contrast, linear arrayed
imaging uses a side-viewing probe to examine
sagitally. Although interpretation of linear arrayed
images is not as intuitive as that of radial images, linear
arrayed imaging offers advantages in the ability to
perform EUS-directed FNA.


Ultrasound transducers can be positioned in the upper
gastrointestinal (UGI) tract using dedicated
endosonoscopes, non-endoscopic probes, or small
probes that can be placed through the channel of an
endoscope. Most small probes can be placed through
the channel of an endoscope and directed to the area of
interest, such as a mural mass. Imaging via these probes
offers a relatively simple means of determining the type
of tissue in mural lesions (fatty, cystic, or vascular). Non-
endoscopic probes are used solely for esophageal or
rectal imaging because they cannot be guided into the
stomach or duodenum.


Radial endosonoscopes provide high- and mid-
resolution cross-sectional images of the UGI tract.The
transducers are located at the tip of a gastroscope and
are connected to a dedicated ultrasound console.The
transmission and reception of ultrasound waves from
the transducer are enhanced by the use of water
contained in a balloon covering the transducer.


Linear array transducers are becoming widespread
because of the ability to direct FNA. The laterally
oriented electronic transducers are placed at the tip of
an oblique-viewing gastroscope. In contrast to most
radial instruments, the ultrasound console for linear
endosonoscopes has a range of ultrasound features such
as color Doppler, pulse Doppler, and cine-loops and
provides higher quality images.The endoscopic biopsy
channel can accommodate a needle aspiration device
that can be directed into mural and extra-mural lesions.


Clinical Effectiveness
Staging of Gastroesophageal Masses
The imaging obtained from EUS provides a clear
differentiation between a mural lesion and extrinsic
compression. The echo patterns of intramural masses
such as cysts, lipomas, and solid masses are quite
characteristic and nearly diagnostic. Lipomas also arise
from the third layer, and are bright, echo-rich, and
smooth. In contrast, the solid intramural tumors of the
stomach and esophagus, such as fibromas, leiomyomas,
and carcinoids, are difficult to differentiate from each
other.1 Leiomyomas or GI stromal cell tumors (GIST),
the most common intramural lesion, may arise from
either the inner or the outer muscularis layer and
appear as a non-homogenous, echo poor round lesion.


FNA of stromal cell tumors will provide diagnostic
cytology in 75% of cases.


In addition to evaluating intramural lesions of the
esophagus and stomach, EUS can evaluate the cause of
thickened gastric folds.After identifying and excluding
isolated gastric varices, EUS can define the layers of the
stomach involved in the fold thickening and suggest the
pathologic process.Thickening of the deeper layers is
usually associated with malignancy such as lymphoma
or linitus plastica.


Tumor staging is another common indication for EUS.


A large number of studies have documented the
accuracy (70% to 90%) of radial EUS in the tumor (T)
William R Brugge, MD, is the
Director of the Gastrointestinal
Endoscopy Unit at Massachusetts
General Hospital, and Associate
Professor of Medicine at Harvard
Medical School. He is Co-author of
the recent New England Journal of
Medicine reviews of upper
gastrointestinal endoscopy and
pancreatic-biliary endoscopy. Dr
Brugge has pioneered the use of
endoscopic ultrasound (EUS),
particularly in pancreatic diseases.


In addition to his work on the
staging and diagnosis of pancreatic
cancer, he is directing a large
multicenter investigation on the use
of EUS for the diagnosis of
pancreatic cystic neoplasms. His
interest in ultrasound as an
imaging modality in pancreatic
biliary diseases originated in his
work on biliary lithotripsy at the
Hospital of the University of
Pennsylvania where he was the
Clinical Director of Gastroenterology.


Using ultrasound, he described the
effects of lithotripsy and gallstone
dissolution on gall bladder motility.


With the demise of biliary
lithotripsy in the US, he sought
training in endoscopic ultrasound by
traveling to Europe and Japan. He
quickly became enthralled with the
ability of EUS to image the
pancreas and direct biopsies. Dr
Brugge received his MD from the
Baylor College of Medicine in Texas
in 1974 He is a prolific author,
and has written over 50 articles.


1. Arantes V, Logrono R, Faruqi S,Ahmed I,Waxman I, Bhutani M S,ýEndoscopic sonographically guided fine-needle aspiration
yield in submucosal tumors of the gastrointestinal tractý, J. Ultrasound Med. (2004), 23: pp. 1,141ý1,150.


2
B USINESS BRIEFING: US GASTROENTEROLOGY REVIEW 2005
Reference Section
and nodal (N) staging of gastroesophageal malignancies.


EUS can determine whether esophageal tumors are
localized to the mucosa-submucosa (T1-T2 lesions), or
have invaded through the muscularis propria (T3).2 The
invasion of extra-esophageal organs such as the aorta,
pericardium, and bronchi, corresponding to stage T4,
can be demonstrated with EUS.


Tumor staging accuracy by EUS (80% to 90%) is
significantly higher than that reported for computer
tomography (CT) scanning (50% to 60%). EUS staging
is predictive of patient survival and is cost-effective.


Patients undergoing EUS staging and radiation-
chemotherapy have a survival advantage over patients
not staged pre-operatively.


Nodal staging with EUS is in a state of evolution with
the increasing use of FNA. Previous studies of EUS
have demonstrated accuracies that surpass CT. Not
only the stage grouping of tumour-node-metastasis
(TNM) classification, but also the number of lymph
node metastases determined by ultrasound and
endoscopic ultrasound before surgery may be useful
for predicting prognosis in patients with esophageal
carcinoma. Patients with malignant celiac nodal
disease as identified by EUS have a significantly worse
post-operative survival than patients with malignant
celiac adenopathy. Celiac nodes can be better
evaluated with EUS, compared with CT imaging.


EUS-FNA provides more accurate assessment of
nodes than EUS imaging alone.


Using the same principles of EUS-FNA for esophageal
malignancy and adenopathy, many centers have started
to use EUS as the primary method for the evaluation of
mediastinal adenopathy and lung cancer staging. This
approach has proven effective in patients with post-
surgical abnormalities seen on CT scans, including
mediastinal cystic lesions. Molecular diagnostic testing
has demonstrated evidence of malignancy in ýbenigný
nodes aspirated with EUS.


Tumor staging of gastric cancer uses the same principles
as for esophageal tumors. Lesions contained within the
gastric wall (T1ýT2) can be identified with accuracy
rates of nearly 90% and may be amenable to curative
resection.3 The differentiation between lesions within
the mucosa/deep mucosa and those invading the
muscularis is increasingly important as endoscopic
mucosectomy gains more acceptance. Endoscopic
mucosal resection of T1 lesions has provided high rates
of successful resection and long-term disease-free
survival. EUS is particularly sensitive for the finding of
ascites and small metastases in the left lobe of the liver.4,5
The staging of mucosa-associated lymphoid tissue
(MALT) lymphoma has a direct impact on patient
management. Lymphoma staging by EUS is similar to
the staging of gastric cancer and the accuracy rates are
comparable. If MALT lymphomas are localized to the
mucosa and submucosa, they can be successfully treated
with antibiotics.


Pancreatic-biliary Diseases
The role of endosonography in pancreatic diseases is
considerably different to its role in gastroesophageal
diseases. EUS may be used as a primary diagnostic
procedure in a wide range of benign, malignant, and
inflammatory disorders. Endoscopic ultrasound is often
used to complement the findings of endoscopic
retrograde cholangio pancreatography (ERCP) and/or
CT scanning such as a focal stricture or a suspected
mass. The sensitivity (98%) of EUS for diagnosing
pancreatic cancer is superior to all other imaging
modalities, including US (75%), CT (80%), and
angiography (89%).


In addition to the use of EUS for diagnosing pancreatic
cancer, EUS has been used to stage pancreatic masses.6
Using the TNM staging classification, EUS can be used
to determine what structures have been invaded by a
pancreatic mass. Small intra-pancreatic masses (<2cm)
that do not involve the bile duct or duodenum are staged
as T1.More commonly,masses are seen involving the bile
duct and/or duodenal wall and are staged as T2. Larger
lesions extending out of the pancreas, often invading the
arteries and veins adjacent to the pancreas, are classified
as T3.6 In a compilation of a large number of studies
involving 350 patients, EUS had an accuracy rate of 80%
for predicting the T-stage and 72% for detecting nodal
metastases. These rates for predicting T-stage were
2. Vazquez-Sequeiros E,Wiersema M J, Clain J E, et al., ýImpact of lymph node staging on therapy of esophageal carcinomaý,
Gastroenterology (2003), 125: pp. 1,626ý1,635.


3. Shimoyama S,Yasuda H, Hashimoto M, et al.,ýAccuracy of linear-array EUS for preoperative staging of gastric cardia cancerý,
Gastrointest. Endosc. (2004), 60: pp. 50ý55.


4. Prasad P, Schmulewitz N, Patel A, et al., ýDetection of occult liver metastases during EUS for staging of malignanciesý,
Gastrointest. Endosc. (2004), 59: pp. 49ý53.


5. Chu K M, Kwok K F, Law S,Wong K H,ýA prospective evaluation of catheter probe EUS for the detection of ascites in patients
with gastric carcinomaý, Gastrointest. Endosc. (2004), 59: pp. 471ý474.


6. DeWitt J, Devereaux B, Chriswell M, et al.,ýComparison of endoscopic ultrasonography and multidetector computed tomography
for detecting and staging pancreatic cancerý, Ann. Intern. Med. (2004), 141: pp. 753ý763.


superior to those obtained with CT scanning (44%) and
US (35%). EUS has also proven to be a highly diagnostic
test for the diagnosis of portal vein thrombosis in the
setting of pancreatic malignancy.7
A variety of other pancreatic disorders can be diagnosed
with EUS. Chronic pancreatitis can be diagnosed with
EUS with greater sensitivity than US, CT, or ERCP by
finding parenchymal changes, small pseudocysts, or
microcalcifications.8 Cystic lesions, such as
cystadenomas, are easily seen with EUS.9 However,
EUS imaging alone cannot differentiate between a
benign and malignant cyst, and fine needle aspiration is
required. Aspirated cyst fluid CEA is useful for the
diagnosis of mucinous cystic lesions.10 Pseudocysts may
be diagnosed and drained with primary introduction of
EUS-guided stents.


Islet cell tumors, including gastrinomas, can be
diagnosed in more than 80% of patients, and EUS is
clearly more sensitive than CT scanning. EUS may be
the first test used for acute relapsing pancreatitis. EUS,
a less invasive test than ERCP,demonstrated an etiology
in two-thirds of patients with idiopathic acute
pancreatitis. Recently, EUS has been used to diagnose
autoimmune pancreatitis.11
The staging of ampullary masses with EUS is based on
similar principles as used with pancreatic cancer. Large
lesions are easily localized and the accuracy of staging is
80% to 90%. Smaller lesions, <2cm, can also be staged
but with a lower accuracy rate, approximately 70%.


Superficial ampullary malignancies may be managed
with endoscopic resection.12
EUS is an ideal tool for diagnosis bile duct stones.


Several prospective, blinded trials have demonstrated
superior sensitivity of EUS over CT and US. EUS is
capable of detecting 90% to 96% of bile duct stones,
only missing an occasional stone located high in the
intrahepatic ducts.The enhanced ability of EUS was
particularly notable in patients with a normal
diameter bile duct and with stones <1cm. In these
cases, EUS was clearly superior to US and CT
imaging and comparable to ERCP for detecting bile
duct stones. EUS appears useful for the diagnosis and
staging of gallbladder and cholangiocarcinoma.13
EUS-guided FNA may be safely performed in
gallbladder and bile duct masses.


Rectal Diseases
Malignancies contained within the rectal wall are
staged as T1ý2, and EUS has an accuracy rate of 73%
to 94%. Masses extending through the muscularis are
staged as T3 and metastatic lesions are staged as T4.


Masses that are localized to the rectum can be treated
with local excision and more advanced lesions can be
treated with radiation.14 EUS can also localize an
intramural rectal mass such as a carcinoid tumor.


Examination of the anal sphincter can detect
traumatic tears in the external sphincter and a loss of
continuity in the internal sphincter. These findings
can be useful in the evaluation of the patients with
fecal incontinence. EUS can also be used to evaluate
patients with perianal Crohnýs, looking for abscesses
and fistulas.


EUS-guided Fine Needle Aspiration
The introduction of linear array imaging has made EUS-
directed FNA biopsies possible. Under direct and
continuous ultrasound guidance, a 22-gauge needle can
be directed into submucosal lesions, lymph nodes, liver
masses, and pancreatic masses.15 The needle can be
extended as far as 5cm, providing access to tissue
3
B USINESS BRIEFING: US GASTROENTEROLOGY REVIEW 2005
Reference Section
7. Lai L, Brugge W R,ýEndoscopic ultrasound is a sensitive and specific test to diagnose portal venous system thrombosis (PVST)ý,
Am. J. Gastroenterol. (2004), 99: pp. 40ý44.


8. Yusoff I F, Raymond G, Sahai A V, ýA prospective comparison of the yield of EUS in primary vs. recurrent idiopathic acute
pancreatitisý, Gastrointest. Endosc. (2004), 60: pp. 673ý678.


9. Brugge W R, Lauwers G Y, Sahani D, Fernandez-del Castillo C,Warshaw A L,ýCystic neoplasms of the pancreasý, N. Engl.


J. Med. (2004), 351: pp. 1,218ý1,226.


10. Brugge W R, Lewandrowski K, Lee-Lewandrowski E, et al.,ýDiagnosis of pancreatic cystic neoplasms: a report of the cooperative
pancreatic cyst studyý, Gastroenterology (2004), 126: pp. 1,330ý1,336.


11. Farrell J J, Garber J, Sahani D, Brugge W R,ýEUS findings in patients with autoimmune pancreatitisý, Gastrointest. Endosc.


(2004), 60: pp. 927ý936.


12. Kahaleh M, Shami V M, Brock A, et al.,ýFactors predictive of malignancy and endoscopic resectability in ampullary neoplasiaý,
Am. J. Gastroenterol. (2004), 99: pp. 2,335ý2,339.


13. Fritscher-Ravens A, Broering D C, Knoefel W T, et al.,ýEUS-guided fine-needle aspiration of suspected hilar cholangiocarcinoma
in potentially operable patients with negative brush cytologyý, Am. J. Gastroenterol. (2004), 99: pp. 45ý51.


14. Harewood G C,ýAssessment of clinical impact of endoscopic ultrasound on rectal cancerý, Am. J. Gastroenterol. (2004), 99: pp.


623ý627.


15. Vander Noot M R, 3rd, Eloubeidi M A, Chen V K, et al.,ýDiagnosis of gastrointestinal tract lesions by endoscopic ultrasound-
guided fine-needle aspiration biopsyý, Cancer (2004), 102: pp. 157ý163.


surrounding the esophagus, stomach and the entire
pancreas.Pulse and color Doppler imaging aid in avoiding
vessels, contributing to the low complication rates of
needle aspiration, <1%. Mediastinal masses and lymph
nodes should be evaluated with FNA since cytology has
proven to be more accurate than imaging alone.Recently,
a Tru-cut needle device has been introduced.16
Aspirated pancreatic tissue may be diagnostic of
adenocarcinoma, islet cell tumors, or cystic neoplasms.


EUS-guided FNA of pancreatic cancer may have
some advantages over CT-guided biopsies, including
shorter needle track, smaller diameter needles, and
better access to small intra-pancreatic masses.17
Peritoneal seeding of cells should be even less of a
concern than with percutaneous biopsies. Liver
metastases, malignant lymph nodes, and ascites can also
be aspirated. Cystic neoplasms of the pancreas can be
aspirated under EUS guidance and the fluid analyzed
not only for cytology but also for tumor markers
(CEA) associated with cystic neoplasms.


Therapeutic EUS
The needle aspiration device that is used for biopsies can
also be used to inject therapeutic agents under
ultrasound guidance. Celiac neurolysis has been
performed by injecting local anesthetics or neurolytic
agents into the celiac ganglia. In patients with chronic
abdominal pain secondary to pancreatic cancer, an
ethanol-bupivacaine injection into the celiac ganglia can
significantly reduce the severity of pain and the need for
chronic narcotic use in more than 80% of patients. Pain
scores were lower two weeks after EUS celiac plexus
neurolysis, an effect that was sustained for 24 weeks
when adjusted for morphine use and adjuvant therapy.


Summary
EUS has proven itself as an important new diagnostic
tool in endoscopy.Therapeutic applications will continue
to evolve based primarily on the ability to direct
injections into small malignancies of the pancreas. a73
Endoscopic Ultrasound
B USINESS BRIEFING: US GASTROENTEROLOGY REVIEW 2005
4
16. Varadarajulu S, Fraig M, Schmulewitz N, et al.,ýComparison of EUS-guided 19-gauge Trucut needle biopsy with EUS-guided
fine-needle aspirationý, Endoscopy (2004), 36: pp. 397ý401.


17. Takahashi K,Yamao K, Okubo K, et al.,ýDifferential diagnosis of pancreatic cancer and focal pancreatitis by using EUS-guided
FNAý, Gastrointest. Endosc. (2005), 61: pp. 76-9.