Professor of Radiology Gerald D. Dodd, Jr. Distinguished Chair
Diagnostic Imaging Director of Academic DevelopmentUniversity of
Texas M.D. Anderson Cancer Center Houston, TX
Janio Szklaruk, MD, PhD University of Texas M.D. Anderson Cancer
Center, Houston, TX
Chaan S. Ng, MD University of Texas M.D. Anderson Cancer Center,
Eric Tamm, MD University of Texas M.D. Anderson Cancer Center,
Terry J. Cooper, MBA, MT (ASCP)
University of Texas M.D. Anderson Cancer Center, Houston, TX
Imaging and Clinical Staging of Biliary Tract Tumors with
Janio Szklaruk, MD, PhD and Paul M. Silverman, MD
Cholangiocarcinomas are malignant tumors of the biliary tract
that represent <2% of all neoplasms.
Biliary tract cancer is, however, the second most common primary
hepatobiliary malignancy, following hepatocellular carcinoma. In
the United States, an estimated 2000 to 3000 bile duct cancers are
diagnosed annually with a 1.2/100,000 rate of occurrence.
Cholangiocarcinoma is more common in males (1.5:1) with an average
onset in the 5th decade. These tumors can be divided into
intrahepatic (8% to 13%) and extrahepatic location (87% to 92%).
Extrahepatic tumors are divided into proximal, middle, or distal
ductal tumors. Tumors located at the confluence of the right and
left hepatic ducts with the proximal common hepatic duct are termed
Klatskin tumors (Table1). Increased incidence of cholangiocarcinoma
occurs in patients with choledochal cysts, liver fluke (
) infestation, thorotrast exposure, and ulcerative colitis with and
without primary sclerosing cholangitis. The latter is the most
common predisposing factor although in most cases the etiology is
indeterminate. The clinical manifestations of cholangiocarcinoma
are secondary to bile duct obstruction and include jaundice,
prutitus, clay-colored stools, weight loss, cola-colored urine, and
pain. The staging of the disease uses TNM classification. The
primary tumor is defined by confinement to the bile duct and by
local invasion. The thin wall of the bile ducts predisposes to
periductal invasion. Extension of tumor to the liver, pancreas,
ampulla of Vater, duodenum, colon, omemtum, stomach, or gallbladder
are common. Late in the disease, distant metastases in the liver,
lung, and peritoneum may occur. Surgery is the only curative
treatment modality. Tumors are considered unresectable if there are
distant metastases, bilateral involvement of the secondary radicles
of the biliary tree, main portal vein involvement, or portal vein
involvement with lobar atrophy. The application of chemotherapy and
radiotherapy are currently under clinical trials.
Imaging plays an essential role in the clinical management of
bile duct cancers. The goal of imaging is to identify the cause of
biliary obstruction. If it is secondary to a mass lesion, the mass
must be localized and its local extent determined, including
invasion of adjacent vascular structures and hepatic parenchyma.
Imaging must also identify the presence of distant metastases.
CT is the most common imaging modality for staging bile duct
tumors. Helical CT or multislice CT (MSCT) with dynamic contrast
enhancement provides high-detail images for the diagnosis and
staging of biliary tract cancer. The normal intrahepatic bile ducts
are barely visible with a diameter <2 mm. The normal
extrahepatic bile duct and common bile duct are visible with a
paper-thin wall. The CT findings of extrahepatic cholangiocarcinoma
are: (1) an infiltrating stenotic lesion with focal or eccentric
wall thickening (Figure 1); (2) a polypoid mass; or (3) a
mucin-producing intraductal tumor. Dilatation of bile duct, an
abrupt change in duct caliber, non-union of bile duct radicles, and
lobar atrophy are indirect signs of bile duct cancer.
With conventional and single-slice helical CT, the detection
rate for hilar cholangiocarcinoma is poor, with reported rates
varying from 40% to 68%.
Current biphasic and thin-section scanning techniques using MSCT
has improved lesion detection to 82% to 100%.
Using the biphasic helical scanning, the density of these tumors
during the early phase, as compared with the surrounding liver, is
quite variable; the tumors are reportedly hyperdense in 30% to 100%
of cases. Tumors were identified to be hyperdense in either or both
phases in 81%-100% of cases. Multiphasic CT for the staging of
hilar cholangiocarcinoma results in a high rate of accuracy in the
diagnosis of advanced disease. However, the technique still needs
improvement in predicting resectability, as its rate of accuracy
varies from 50% to 60%.
CT tends to underestimates the extent of bile duct involvement,
portal vein or hepatic artery involvement, and nodal metastasis and
Currently at our institution, we use a biphasic scanning
technique (the late arterial phase and the portal venous phase)
following an intravenous (IV) injection of nonionic contrast
material (150 to 180 mL) delivered by a power injector at a rate of
5 mL/sec (Szklaruk J, Silverman PM, Charnsangavej C. Imaging in the
diagnosis, staging, treatment, and surveillance of biliary tumors.
Unpublished data, 2003). The initial phase is scanned from the dome
of the diaphragm inferiorly to the distal bile duct 30 seconds
after the start of contrast injection. A breath hold of only 15 to
20 seconds is necessary to acquire images of this region at a slice
thickness of 1 to 3 mm. This late arterial phase of imaging is
utilized to maximize the enhancement of the tumor in the hilar
region and to enhance visualization of its relationship to the
hepatic artery. The second phase, or portal venous phase, is
obtained after a pause of 7 to 10 seconds for breathing following
the completion of the first phase. This phase emphasizes the
relationship between the tumor and the portal vein and adjacent
Applications of MSCT
The gold standard for biliary tree evaluation is direct but
invasive contrast opacification of the biliary system using
percutaneous cholangiography (PTC) or endoscopic retrograde
cholangiopancreatography (ERCP). The complication rates for these
procedures are 3.4% and 0.5% to 5%, respectively.
Magnetic resonance cholangiopancreatography (MRCP) has gained
popularity as a noninvasive technique.
CT technology with three-dimensional (3D) reconstruction is
evolving in its application in the evaluation of the biliary tree.
CT cholangiography can provide images of the biliary tree using two
techniques. In the first technique, the biliary tree is opacified
by the oral or IV administration of a radiodense biliary contrast
agent. Thin-section axial images are then reconstructed in the
desired plane, typically a coronal or coronally obliqued plane,
using a maximum intensity projection (MIP) algorithm. More
recently, a technique has been developed that avoids the
administration of a cholangiographic agent. In this technique,
thin-section axial images acquired after the administration of
conventional IV contrast are reconstructed in the desired plane
using a minimum intensity projection (Min, Figures 2 and 3), or a
volume-rendering (VR, Figure 4) technique. There is a good
correlation between the images produced using these techniques and
those generated by ERCP or PTC. An additional advantage of this
second technique is that the axial source images can also be
postprocessed to display arterial and venous anatomy in any desired
plane (CT angiography/CT portography, Figure 5). Thus, the
information that can be provided by a single multiphase, multislice
CT examination can be used to provide a variety of perspectives
(angiographic, cholangiographic, and morphologic) for aiding
For preoperative MSCT assessment of patients with bile duct
tumors, we currently use a combination of thin axial-source images,
acquired in multiple phases as outlined above, with several
postprocessing techniques. Coronally oriented MIP images are
generated of the arterial and portal venous structures (CT
arteriography and CT portography, Figure 5). Oblique coronally
oriented MIP (Figure 6) and Min images (Figures 3 and 4) are
generated of the biliary tree and the pancreatic ducts. We
currently also use a 3D volume postprocessing technique to estimate
segmental volume in the assessment of the preoperative patient. Our
experience with evaluating biliary tract cancer using this
combination of techniques has had promising results.
In conclusion, the combination of axial and multiplanar
reconstructed images from MSCT allows an improved sensitivity of
detection, staging, and surgical planning of bile duct tumors.
1. Blumgart LH, Fong Y, Jarnagin WR.
. Lewiston, NY: B.C. Decker; 2001.
2. Choi B, Lee JH, Han MC, et al. Hilar cholangiocarcinoma:
Comparative study with sonography and CT.
3. Tillich M, Milschinger HJ, Preisegger KH, et al. Multiphasic
helical CT in diagnosis and staging of hilar cholangiocarcinoma.
AJR Am J Roentgenol
4. Loperfido S, Angelini G, Benedetti G, et al. Major early
complications from diagnostic and therapeutic ERCP: A prospective
Multislice CT in 3D Imaging for Partial Nephrectomy
Chaan S. Ng, MD, Marc J. Fenstermacher, MD, and Paul M.
Refinements in surgical technique have allowed for increased
application by urologists of partial nephrectomy for small renal
tumors. For the radiologist who has multislice CT (MSCT), it
represents a challenge and opportunity to demonstrate its
outstanding imaging capabilities in helping urologists assess and
plan surgery. The rationale for pursuing partial, rather than total
or radical, nephrectomies is the motivation to sacrifice as little
normal renal tissue as possible in the process of removing the
tumor, hence nephron-sparing surgery. The approach is further
strengthened by emerging data that indicates that nephron-sparing
surgery is as efficacious as total nephrectomy for renal tumors,
with reported local recurrence rates of <2% and 5-year survivals
of 87% to 90%, which are comparable with those from radical
The indications for nephron-sparing partial nephrectomy are
evolving. At our institution, the current requirements are that the
tumor is <4 cm in size and that it is located peripherally.
Specific indications are absence of a contralateral kidney, renal
insufficiency, and conditions in which patients are at an increased
risk for bilateral renal tumors, such as in patients with von
Surgical approaches to partial nephrectomy include open surgery,
laparoscopic surgery, and hand-assisted laparoscopic surgery. The
advantages of the laparoscopic approaches include a faster
postoperative recovery for patients, with all the associated
Important determinants for the feasibility of and surgical
approach for partial nephrectomy include the location of the tumor,
in particular its relationship to the renal pelvic structures,
specifically the renal artery(s) and vein(s), and the collecting
system. Identification of accessory renal arteries is particularly
important in assisting planning. Other aspects of preoperative
staging also need to be addressed; for example, exclusion of renal
vein thrombosis, and detection of adrenal masses, contralateral
renal masses, and abnormally enlarged lymph nodes.
At our institution, MSCT evaluation of renal tumors with a view
to partial nephrectomy involves a multiphasic protocol encompassing
from above the kidneys superiorly to the proximal common iliac
arteries inferiorly. Scans are undertaken using 5-mm collimation
and 15-mm table speed (pitch 1.5, 120 kV, 370 mA, HS mode,
LightSpeed scanner, GE Medical Systems, Milwaukee, WI). Data is
reconstructed prospectively at 2.5-mm slice thickness, 1.5-mm
intervals. The raw data prior to reconstruction is not saved due to
archival constraints. Patients are given oral contrast 45 minutes
prior to their scan. Precontrast images are obtained of the
abdomen. We administer 150 mL of 300 mg/mL nonionic contrast at 5
mL/sec through a peripheral line, and scan at delays of 25 sec
(arterial/early cortico-medullary phase), 100 sec (nephrographic
phase), and 180 sec (excretory phase), utilizing breath-holds of
approximately 12 to 15sec for each phase. If required, an
evaluation of the chest and/or abdomen (and pelvis) can be
undertaken in the interscan periods, with the abdomen (and pelvis)
scan typically being interposed between the arterial and
nephrographic phases of the renal scan. The latter are undertaken
using the HQ mode of the GE LightSpeed scanner utilizing 7.5-mm
collimation, table speed of 15 mm, 120 kV, and 220 mA.
The reconstructed 2.5-mm thick, 1.5-mm interval source images
from each of the renal phases are transferred to a workstation
(Vitrea2, Vital Images Inc., Minneapolis, MN, or General Electric
Advantage Windows) for image processing. Image manipulation
includes multiplanar orthogonal, oblique and/or curved
reformations, MIP reconstruction, and 3D volume rendering. Although
all phases of the study are potentially useful, the arterial phase
images are particularly useful for determining the number and
origin of the renal arteries, and evaluating the tumor in relation
to the arteries, veins, and collecting system.
Variant or aberrant anatomy and relationship of the tumor to
structures in the renal pelvis are vital in the evaluation for
surgical and preoperative planning, particularly when considering
laparoscopic approaches (Figure 1). Urologists find the
presentation of 3D data, particularly surface-rendered
presentations, extremely helpful in allowing them to plan their
surgery, as it provides them with a realistic sense of what they
are likely to see at surgery.
It should noted however, that 3D manipulations can create
artifacts, and careful review of the source images is advised in
With the increased incidence of renal tumor detection,
particularly of small tumors detected incidentally in the course of
other radiologic studies, including CT body screening, there is
likely to be a concomitant increase in the demand for
nephron-sparing surgery. With the addition of 3D techniques, MSCT
is an important tool to assist urologists in their planning for
1. Novick AC. Advances in the management of localized renal cell
Can J Urol.
2. Sheth S, Scatarige JC, Horton KM, et al. Current concepts in
the diagnosis and management of renal cell carcinoma: Role of
multidetector CT and three-dimensional CT.
3. Urban BA, Ratner LE, Fishman EK. Three-dimensional
volume-rendered CT angiography of the renal arteries and veins:
Normal anatomy, variants, and clinical applications.
Pitfalls in the Diagnosis of Pancreatic Cancer by CT
Eric Tamm, MD and Paul M. Silverman, MD
Because of its poor prognosis, pancreatic cancer remains the
fifth leading cause of cancer death in the Western hemisphere.
Surgical resection remains the only option for cure. However,
nearly 60% of patients have advanced disease at the time of
diagnosis. Accurate staging by CT may mean the difference between
unnecessary surgery and the possibility for cure.
Optimal technique is essential. When multislice CT (MSCT) is
used at our institution, dual-phase imaging technique is utilized.
Patients are injected with 150 mL of nonionic intravenous (IV)
contrast at a rate of 5 mL/sec. Imaging during the first phase, the
pancreatic parenchymal phase, begins 25 seconds after the start of
IV contrast injection. Imaging begins at the level of the diaphragm
and extends through the pancreas to the horizontal portion of the
duodenum at 2.5-mm slice thickness at a table speed of 7.5 mm/sec,
reconstructed to 1.25-mm slice thickness. Imaging of the pancreas
therefore occurs at 35 to 45 seconds after the start of injection
of IV contrast. This first phase is utilized to evaluate for the
primary tumor, detection of vascular variants, and for involvement
During the second phase (the portal venous phase [PVP]) imaging
begins 55 seconds after beginning IV contrast injection at the
level of the diaphragm through to the horizontal portion of the
duodenum at 5-mm slice thickness, table speed of 15-mm/sec,
reconstructed to 2.5-mm slice thickness. This second phase is
utilized to evaluate for possible liver metastases and to offer a
second opportunity to evaluate the portal venous system for local
The first possible pitfall is the result of poor technique. Low
injection rates of contrast or poor timing of image acquisition can
result in poor differentiation between the tumor and normal
pancreatic parenchyma, which results in failure to diagnose
Another potential pitfall occurs when CT is performed too soon
after other diagnostic procedures (Figures 1 and 2). Often, CT
examinations are performed following ERCP. In patients who complain
of upper abdominal pain, the first diagnostic procedure is
typically an abdominal ultrasound. When such a study reveals a
dilated common bile duct and no evidence of cholecystitis, ERCP is
often utilized next. When ERCP shows possible malignant narrowing
of the common bile duct, brushings are obtained and, typically, a
stent is placed. CT may follow shortly. However, postprocedure
inflammatory changes from pancreatitis and artifacts directly
related to a stent can result in either a misdiagnosis of cancer or
a failure to detect malignancy. These inflammatory changes can also
infiltrate the fat surrounding the superior mesenteric artery as
well as the common hepatic artery resulting in a misdiagnosis of
tumor involvement. Therefore, caution should be exercised when
interpreting such CT studies. Follow-up examinations, allowing for
resolution of possible inflammatory changes, would be advised.
Similar difficulties are encountered when CT follows immediately
after an endoscopic ultrasound (EUS)-guided, fine-needle aspiration
Signs that are useful for diagnosing pancreatic cancer include
dilatation of the pancreatic duct with abrupt change in caliber,
the presence of a hypodense mass, focal enlargement of the
pancreas, soft-tissue changes surrounding vasculature, and mass
effect upon vasculature, as well as signs of spread of disease,
including adenopathy, peritoneal disease, and metastatic disease to
the liver and lungs. The presence of multiple signs increases the
probability of diagnosis of pancreatic cancer.
Hypodense changes from focal chronic pancreatitis can mimic
pancreatic cancer closely
; chronic pancreatitis can also be associated with stricture of the
pancreatic duct, distortion of the contour of the pancreas, and
thrombosis of the splenic vein. Therefore, the presence of
pancreatic cancer must be determined by biopsy, which is typically
obtained with endoscopic ultrasound guidance. The greater
sensitivity of EUS and the high specificity of EUS-guided FNA,
also allows for the evaluation of cases that are negative on
In summary, MSCT can provide a comprehensive evaluation for
pancreatic cancer. Furthermore, optimal technique and knowledge of
the patient's clinical history, including the nature of prior
work-up, is essential in avoiding diagnostic pitfalls. Biopsy is
required to confirm the diagnosis of pancreatic cancer. In the case
of negative EUS, negative biopsy, and negative MSCT, close
follow-up by CT and EUS may be necessary to prove the absence of
1. Jemal A, Thomas A, Murray T, Thun M. Cancer statistics, 2002.
CA Cancer J Clin.
2. Kim T, Murakami T, Takamura M, et al. Pancreatic mass due to
chronic pancreatitis: Correlation of CT and MR imaging features
with pathologic findings.
AJR Am J Roentgenol.
3. Harewood GC, Wiersema MJ. Endosonography-guided fine needle
aspiration biopsy in the evaluation of pancreatic masses.
Am J Gastroenterol
4. Shin HJ, Lahoti S, Sneige N. Endoscopic ultrasound-guided
fine-needle aspiration in 179 cases: The M.D. Anderson Cancer
The Privacy Act and Its Impact on Patient Care
Terry J. Cooper, MBA, MT (ASCP)
Well, it seems that this HIPAA thing is not going away anytime
soon. So, just like everything else in life, those of us in health
care will have to adapt to the changes required to be in compliance
with the HIPAA Privacy Rule by April 14, 2003.
The American public continues to express their concerns about
protecting their identifiable health information, also known as
protected health information (PHI). The Health Privacy Project, at
Georgetown University's Institute for Health Care Research and
Policy, cited the following data
* 1 of 5 American adults believes that a healthcare provider,
insurance plan, government agency, or employer has improperly
disclosed personal medical information.
* 74 % of the respondents said the privacy of their personal
health information is very important.
* 75% of people are concerned about health Web sites sharing
information without their permission.
The purpose and intent of the HIPAA is to protect the right of
individuals to keep information about themselves from being
disclosed. The HIPAA Privacy Rule is designed to safeguard records
that hold personal health information. Paper records, electronic
records, and oral communication are covered. Oral communication
runs the gamut from paging patients, to whispering in corridors and
elevators, and talking on cell phones. Confidential information is
not necessarily relegated to a patient's name.
In fact, it includes information related to a person's past,
present, or future physical or mental health condition and anything
associated with healthcare services or treatment.
The Privacy Rule can be summed up as:
* Creating limits on the uses and disclosure of PHI;
* Giving patients new rights to access their medical records and
to know who else has accessed them;
* Restricting most disclosure of PHI to the minimum needed for
the intended purpose (minimum necessary guidelines);
* Creating new criminal and civil sanctions for improper use or
disclosure of PHI; and
* Establishing new requirements for access to records by
researchers and others.
While these rights are available to all patients, healthcare
providers should review any requests carefully. The provider must
weigh a patient's request against prudent patient care to prevent
disruption of that care. Providers should explain to patients that
excessive restrictions could impede timely care.
So, in general, what does the HIPAA Privacy Rule require the
average healthcare provider to do? For the average provider, the
Privacy Rule requires activities, such as:
* Notifying patients about their privacy rights and how their
information can be used;
* Adopting and implementing privacy procedures for its practice,
hospital, or plan;
* Training employees so that they understand the privacy
* Designating an individual to be responsible for seeing that
the privacy procedures are adopted and followed; and
* Securing patient records containing individually identifiable
health information so that they are not readily available to those
who do not need them.
Responsible healthcare providers and businesses already take
many of the kinds of steps required by the Privacy Rule to protect
patients' privacy. To ease the burden of complying with the new
requirements, the Privacy Rule gives needed flexibility for
providers to create their own privacy procedures, tailored to fit
their size and needs. The scalability of the Privacy Rule provides
a more efficient and appropriate means of safeguarding protected
health information than would any single standard.
For example, the training requirement may be satisfied by a
small physician's practice when each new member of the workforce is
provided with a copy of its privacy policies and documentation is
obtained that new members have reviewed the policies. In contrast,
a large health organization may provide training through live
instruction, video presentations, or interactive software
Under HIPAA, a provider must issue a written Notice of Privacy
Practices (NPP) to patients outlining its privacy practices and
patients' rights. Anyone can request the NPP, although first-time
patients must receive theirs at the time of initial contact where
PHI is exchanged. In cases of phone communications, the NPP should
be sent out via mail within 24 hours (preferably) or given to
patients upon arrival for their first visit or treatment. If the
provider has a Web site, the NPP should be posted and be
downloadable. Regardless of whether the Notice is mailed or
electronically downloaded, a good faith effort must be made to
acquire a written acknowledgement from the patient.
HIPAA is not another Y2K event, but an evolving piece of
legislation. The Department of Health and Human Services has the
authority to modify the privacy standards, as the Secretary may
deem appropriate. However, a standard can be modified only once in
a 12-month period. For those in healthcare, the HIPAA Privacy Rule
uses the term reasonable several times throughout the regulation.
Our approach to the Privacy Rule must be reasonable, not to over
self-regulate your practices to the point of impeding quality