| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Medical Imaging, University of Toronto, Toronto, ON,
Canada.
2 The Toronto General Hospital, University Health Network, 585 University Ave.,
Toronto, ON M5G 2N2, Canada.
3 Department of Medical Biophysics, University of Toronto Imaging Research, and
Sunnybrook and Women's College Health Sciences Centre, Toronto, ON,
Canada.
Received January 4, 2005;
accepted after revision March 23, 2005.
Supported by a phase 2 clinical trial funded by Bristol-Myers Squibb
Medical Imaging, by the Canadian Institutes of Health Research, and by the
Terry Fox Programme of the National Cancer Institute of Canada.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Ninety-six lesions in 92 patients were evaluated with DMP 115 (Definity)-enhanced pulse-inversion sonography, comprising 44 malignancies (29 hepatocellular carcinomas, 12 metastases, two peripheral cholangiocarcinomas, and one hepatic lymphoma) and 52 benign lesions (26 hemangiomas, 20 focal nodular hyperplasias, and six others). All had continuous low-mechanical-index imaging through the arterial and portal venous phase. A three-person blind review evaluated single images at baseline, early and peak arterial phases, and through the extended portal phases with a movie showing arterial phase wash-in. Reviewers assessed lesional vascularity and enhancement blindly but did not make a diagnosis. Combinations of answers were compared with independently determined final diagnoses to develop an algorithm for liver mass diagnosis.
RESULTS. Portal phase enhancement comprises the first step of the algorithm, with positive or sustained enhancement identifying 48 (92%) of 52 benign lesions and negative enhancement or washout present in 41 (93%) of 44 malignancies. Sustained portal phase enhancement with arterial phase peripheral nodularity and centripetal progression predicted 24 (92%) of 26 of the hemangiomas; diffuse arterial phase enhancement greater than the liver identified 19 (95%) of 20 of the focal nodular hyperplasias. With negative portal phase enhancement, arterial phase information was less effective at differentiating hepatocellular carcinoma (25 [86%] of 29 cases) from another hepatic malignancy (11 [73%] of 15 cases).
CONCLUSION. A simple diagnostic algorithm for interpretation of microbubble-enhanced sonography provides sensitive and accurate diagnosis of commonly encountered liver masses.
Keywords: contract media • dynamic sonography • liver • oncologic imaging • sonography
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Microbubble contrast agents for sonography were introduced in the 1990s and have steadily gained popularity for the investigation of liver masses [9]. A unique component of a contrast-enhanced sonography study is the real-time evaluation of lesional and hepatic vasculature when performed with a low-mechanical-index, bubble-specific nonlinear imaging technique. Pulse-inversion imaging, used here, is one such technique that suppresses echoes from tissue in favor of those from bubbles [10]. When used at a low mechanical index, pulse-inversion imaging does not cause disruption of the microbubbles and therefore preserves their integrity, allowing continuous assessment of the vessels as the contrast agent traverses the imaging field. The detailed depiction of vessels often shows their distribution and morphology and is thus in some ways comparable with angiographic assessment of tumor vessels, providing information that is only occasionally seen on CT or MRI. In addition, contrast-enhanced sonography also shows the lesional and parenchymal enhancement analogous to that seen on CT and MR images, but with higher temporal resolution (typically 10-15 frames per second). Microbubble enhancement has added significantly to the ability of gray-scale sonography in the evaluation of focal liver disease, with a number of early reports suggesting promise for the method in the differential diagnosis of liver masses [11-17]. In this study, we used blinded interpretations of 100 liver contrast studies combined with a knowledge of the vascular features of hepatic lesions seen on angiography, CT, and MRI to develop and assess a simple algorithm for the differential diagnosis of focal liver masses using microbubble-enhanced pulse-inversion sonography.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Patients
Patients were accepted in the study if they were referred for sonography at
our institution for evaluation of an unknown focal liver mass or if a
previously undiagnosed focal liver mass was detected during sonography
performed for another reason. Patients were recruited nonconsecutively and
accepted if their lesion was clearly appreciated on a baseline sonography scan
and the likelihood of a confirmatory diagnosis was high. One hundred such
patients were recruited for the contrast sonography study. After examination,
eight subjects were eliminated (five technically inadequate examinations, one
with no confirmed lesion, one intraductal Klatskin cholangiocarcinoma with no
liver involvement, and one protocol deviation with radiofrequency ablation
before the sonography), leaving a study population of 92 patients. There were
46 men and 46 women, with a mean age of 51 years (range, 19-85 years). When
patients had more than one lesion with the same diagnosis within a single
field of view, this was counted as a single lesion. Four patients had two
separate lesions in different lobes of the liver, not visible in a single
view. Of these, three had the same diagnosis for each lesion and one patient
had two diagnoses, FNH and hemangioma, yielding a total of 96 lesions studied
in 92 patients. Lesions varied in maximum diameter from 1 to 13 cm (mean, 4.5
cm). Seventeen lesions were less than 2 cm in diameter and another 17 lesions
were larger than 8 cm. Final diagnoses for the 96 lesions included HCC
(n = 29; mean diameter, 5.3 cm); non-HCC malignancies (n =
15; mean diameter, 4.2 cm), comprising 12 metastases, two peripheral
cholangiocarcinomas, and a single hepatic lymphoma; FNH (n = 20; mean
diameter, 3.8 cm); hemangioma (n = 26; mean diameter, 4.9 cm); and
six other benign lesions (mean size, 2.6 cm) including three regenerative
nodules, two adenomas, and one lipoma.
Final diagnosis was made by histopathology in 34 of 44 malignant lesions,
including one lymphoma, nine of 12 metastases, 22 of 29 HCCs, and two
peripheral cholangiocarcinomas. In the remaining 10 patients with malignant
disease, seven HCCs and three metastases, diagnoses were established with CT
(n = 7), MRI (n = 2), and CT and MRI (n = 1) using
standard diagnostic criteria. These include a hypervascular mass in the
arterial phase with negative enhancement in the portal venous phase for
hepatomas [18] and a variably
enhancing mass often with rim enhancement or hypovascularity in the arterial
phase with negative enhancement in the portal venous phase for metastases.
Clinical evidence for the seven HCCs without biopsy included
-fetoprotein level over 400 (n = 4), positive viral serology
for hepatitis B or C (n = 6), and proven alcoholic cirrhosis
(n = 1). The clinical evidence for the diagnosis of the three
metastases without biopsy included overwhelming evidence of malignancy with
bone metastases (n = 1) and new liver lesions in patients with a
known primary malignancy (n = 3). Final diagnosis was made by
histopathology in seven of 52 benign lesions, including one of two adenomas,
two of 20 FNHs, two of three regenerative nodules, one lipoma, and one of 26
hemangiomas. In the 45 benign lesions without biopsy, 28 had diagnoses
confirmed by CT and 13 by MRI using standard diagnostic criteria. Hemangioma
is characterized by peripheral nodular enhancement with centripetal
progression of enhancement often to complete fill
[19] and FNH by homogeneous
hypervascularity in the arterial phase with sustained enhancement in the
portal venous phase and delayed enhancement of a central scar
[20]. Further, nine FNHs had a
sulfur colloid liver-spleen scan showing a positive uptake of radiotracer in
the location of the liver mass, and three hemangiomas had a positive labeled
RBC scan.
Sonography Technique and Interpretation
Sonography scans were performed by a single physician. The majority were
performed on an ATL HDI 5000 (Philips Ultrasound). Scans were also performed
on an Acuson Sequoia (Siemens Medical Solutions). A standard convex linear
array transducer was used, operated in pulse-inversion mode at a mechanical
index of less than 0.15. In this mode, real-time imaging of microbubble
contrast is seen throughout the arterial and portal phases in both vessels and
liver parenchyma [10]. In all
scans, the transmit focal zone was placed deep in relation to the lesion. The
contrast agent was administered to a maximum total dose of 50 µL/kg. At the
initiation of this trial, DMP 115 was given by IV infusion at 0.5-10 mL/min.
It was also administered by multiple small boluses of 0.1-0.2 mL followed by a
5-mL saline flush, to the same maximum total dose. By the completion of the
19th patient, the infusion technique was abandoned and subsequent patients
were examined using the multiple small bolus technique. A single 1.3-mL vial
of contrast agent sufficed for all patients; in most cases, less than half of
the vial was used.
Sonographic technique comprised continuous observation of the lesion from the time of injection for about 4 min, using the low-mechanical-index pulse-inversion technique. With the arrival of the first microbubbles in the field of view, the arterial phase was timed for 45-60 sec. A 90-frame digital cine loop was stored for later review, starting at the first arrival of the bubbles in the field of view and including the interval to peak enhancement. The beginning of the portal venous phase was timed from about 60 sec after completion of injection. As these microbubbles are purely intravascular and show no late or "Kupffer" phase enhancement [21], there was no interstitial or equilibrium phase. For the remainder of the observation period, therefore, the enhancement showed progressive decline until the baseline appearance was again observed (usually about 4-5 min after completion of the saline flush). We refer to this period of observation from 60 sec to the end of the observation period as the "extended" portal venous phase. From these scanning sequences, images were selected for interpretation as follows: to show lesional vascularity, arterial phase enhancement, and extended portal venous phase enhancement.
Lesional vascularity—Lesional vascularity refers to the presence, distribution, and morphology of lesional vessels. Specific morphologic features of interest include linear vessels, stellate vascular morphology, vascular dysmorphology, marginal vascularity, and the presence of peripheral puddles and pools of enhancement without linear vessel morphology.
Lesional enhancement relative to the liver in the arterial phase—As the entire image appears dark, or hypoechoic, at the initiation of a low-mechanical-index sequence, any echogenicity that subsequently appears in the liver or the lesion is attributed to the presence of the contrast agent. Therefore, arterial phase enhancement was assessed by comparing the echogenicity of the lesion relative to the echogenicity of the liver at the peak of contrast enhancement.
Lesional enhancement relative to the liver in the extended portal venous phase—As the liver derives most of its blood supply from the portal veins, normal liver enhancement progressively increases from the time of injection through the arterial phase to its peak in the portal phase, at which time the enhancement of the lesion is again compared with the enhancement of the adjacent liver parenchyma. The portal venous phase imaging was timed beginning at 60 sec after the completion of the flush. Images were recorded until completion of the scanning sequence, usually at about 4 min. If liver enhancement overtook that of the lesion (often referred to as portal phase washout), images were recorded at the peak of the contrast difference between the liver and lesion.
A three-person blind review was conducted to determine which combination of the resulting observations, if any, provided the ability to diagnose focal liver masses with microbubble-enhanced pulse-inversion sonography. The reviewers were all radiologists trained in the use and interpretation of contrast agents in the liver. A clinical fellow in body imaging trained in contrast sonography who neither scanned the patients nor was a reader selected representative images and videotape files to show the baseline appearance, lesional vascularity, and arterial phase and portal venous phase enhancement of the lesion relative to the enhancement of the liver. Image sets were randomized and shown to the reviewers, who remained blinded to all clinical information and patient details. The reviewers were not asked to provide a diagnosis. The questions were designed to determine the salient features of the lesional vascularity and enhancement as follows: First, in the arterial phase, are there discrete lesional vessels? If yes, how many? Is there a stellate pattern? Second, is there peripheral puddling or pooling of contrast? Third, is the predominant degree of arterial phase enhancement greater than, equal to, or less than the adjacent liver? Fourth, is the pattern of arterial phase enhancement diffuse, rim, peripheral nodular, or sparse? Fifth, if diffuse, is there a nonenhancing portion? If yes, is it a scar or necrosis? Sixth, is there centripetal progression of the enhancement? Seventh, in the extended portal venous phase, is there sustained enhancement of the lesion equal to or greater than the liver?
Before interpretation, the reviewers were shown examples of contrast-enhanced sonograms to establish a standardized approach to the interpretation of the information provided on the imaging sequences. For the arterial phase, reviewers were asked to describe the enhancement of the lesion relative to the liver on the basis of its relative echogenicity. Lesional enhancement greater than the adjacent liver parenchyma was termed "positive" enhancement; less than the adjacent liver was termed "negative" enhancement. Lesions with the same enhancement as the liver were called "isoechoic." To ensure consistency, reviewers were trained with examples of positive, negative, and isoechoic enhancement, and diffuse, sparse, rimlike, and peripheral nodular enhancement.
Analysis
In analyzing the results of the interpretation, consensus agreement was
defined as the answer agreed upon by at least two of the three reviewers in
their response to a question. Consensus answers to the questions in the blind
review were then analyzed in conjunction with the final diagnosis of each mass
to optimize an algorithm for the differential diagnosis of the masses with
contrast-enhanced sonography.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
If sustained enhancement in the portal venous phase indicated a benign lesion, the additional presence of peripheral nodular enhancement in the arterial phase with centripetal progression correctly predicted the diagnosis of hemangioma in 24 (92%) of 26 cases (Figs. 2A, 2B, 2C, 2D, and S2). Diffuse arterial phase enhancement greater than the adjacent liver correctly identified 19 (95%) of 20 FNHs (Figs. 3A, 3B, 3C, and S3). Although negative portal venous phase enhancement of the lesion relative to the liver was predictive of a malignant lesion, the arterial phase enhancement was less successful at differentiating the type of malignant lesion. Arterial phase diffuse positive enhancement identified 23 (85%) of 29 HCCs (Figs. 4A, 4B, 4C, 4D, and S4), and negative enhancement identified 10 (71%) of 14 metastases (Figs. 5A, 5B, 5C, and S5). The algorithmic diagnoses are shown in comparison with the final diagnoses in Table 2. Two incorrect algorithmic diagnoses are shown in Figures 6A, 6B, 6C, 7A, 7B, and 7C, both related to the appearance of the lesion in the portal venous phase that led to its mis-classification as benign or malignant. Figures S2-S5 and S8 can be seen in the AJR electronic supplement to this article, available at www.ajronline.org, and present more detail than the figures printed here.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Initial publications on the use of microbubble-enhanced sonography for improved characterization of focal liver masses used the first-generation air-based contrast agent SH U 508 (Levovist, Schering) and earlier versions of nonlinear imaging technology. Although reports were optimistic [12-14], they also reflected the limited potential of air-based contrast agents that require a high mechanical index for their visualization. Bubble destruction at a high mechanical index eradicates the bubbles as they appear in the field of view. Consequently, only vessels with rapidly moving bubbles, which replenish those destroyed within the interframe interval, are seen in real time. Although parenchymal perfusion can be seen by a single interval delay image in which the bubbles are disrupted, continuous perfusion imaging is not possible.
The more recent introduction of secondgeneration perfluorocarbon agents, together with newer imaging techniques, has provided much better imaging capabilities. These agents may be imaged at a low mechanical index, preserving the microbubble population and allowing continuous imaging of both the lesion and its vessels. Current investigators report improved ability to characterize liver lesions [16, 27, 28] and improved ability to distinguish benign from malignant lesions on the basis of their portal venous phase enhancement [15].
In this study, the development of an algorithm for the diagnosis of focal liver masses with contrast-enhanced sonography was based on responses to a questionnaire in a blind review format. The reviewers were not asked to make a diagnosis of the lesion, only to answer questions regarding their observations of the vascularity of the lesion and its enhancement over time. The diagnostic algorithm is based on enhancement appearances that have similarities to those used for interpretation of contrast-enhanced CT and MRI scans and will therefore be familiar to body imagers. Thus, we anticipate that such an algorithm could be easily incorporated into clinical practice.
One of the strongest outcomes of the analysis (thus chosen as the first step in the algorithmic approach to an unknown mass) quantifies the ability of microbubble sonography to differentiate malignant and benign lesions on the basis of their portal phase enhancement. The real-time nature of the sonography scan allows for detection of washout when it occurs, and our study design included continuous observation of the lesion for several minutes. Therefore, we did not define a precise moment when the enhancement of the lesion relative to the liver would be evaluated in either phase. Rather, in the arterial phase, we chose an image that showed the peak of enhancement of the lesion; and in the portal venous phase, we observed the lesion until it was obvious whether the lesion showed greater, equal, or lower enhancement than the adjacent liver. This evaluation generally differentiates benign from malignant lesions, which then allows for more specific diagnosis on the basis of the arterial phase enhancement pattern, including diffuse, sparse, rimlike, or peripheral nodular, and enhancement intensity relative to the adjacent liver and on the arterial phase vascular morphology.
Benign lesions, hemangioma, and FNH were correctly predicted in more than 90% of cases. Both lesions consistently showed sustained enhancement such that the mass appeared equal to or greater in echogenicity than the adjacent liver for the duration of the scanning interval, often as long as 4 min. Peripheral nodular enhancement in the arterial phase, an essential component for the diagnosis of hemangioma, is distinctly different from the appearance of the vascularity seen with all other lesions. The microbubbles collecting in the vascular spaces of the hemangioma show pooling of the contrast agent without the linear appearance of tumor vessels seen in other lesions. Centripetal progression of this enhancement from the lesion periphery toward the center is seen on the real-time sonography examination regardless of the rapidity with which it occurs (Fig. S8).
Infrequently, a metastatic tumor might show peripheral enhancement in the arterial phase, which, on the basis of a single frame, may be similar to the peripheral nodular enhancement of a hemangioma. However, subsequent imaging of the metastasis reveals washout or negative enhancement of the lesion relative to the liver in the portal venous phase, which is easily distinguished from the centripetal progression with sustained portal venous enhancement that characterizes the hemangioma.
The concept of washout of a lesion in the portal venous phase is recognized to be a reflection of not only the blood supply of the lesion but also the hemodynamics of the liver itself. The normal liver, with its dual blood supply, shows first enhancement in the arterial phase as the contrast agent fills the hepatic artery, with progressively greater enhancement as the contrast agent arrives in the portal vein. A severely arterialized cirrhotic liver will show different hemodynamics, including greater relative enhancement of the liver in the arterial phase with less difference between the liver enhancement in the arterial and the portal venous phases. It is our belief, therefore, that HCC in a severely cirrhotic liver may not show this washout phenomenon until much later, if at all. A study using CT found similarly that enhancement of the cirrhotic liver was significantly lower during the portal venous and delayed phases, presumably because of decreased peripheral portal perfusion [29]. The addition of a delayed scanning sequence at 180 sec has improved both detection and characterization of HCC over a standard dual-phase CT technique [30, 31]. Sonography contrast studies in the liver have also created our impression that small and well-differentiated HCC may not show this washout as reliably. These facts may have influenced our results, as we did not recruit small nodules in cirrhotic livers in this study. Rather, the HCCs in this study were relatively large (mean diameter, 5.3 cm). Also, although many of our patients with HCCs had significant risk factors for the development of HCC, they did not have end-stage cirrhotic livers. Consequently, it is likely that a similar study of the difficult noninvasive diagnosis of small nodules in severely cirrhotic livers would produce less impressive results.
Our algorithm is highly simplified and only uses answers to seven questions. For example, questions about stellate vascularity and regions of nonenhancement in the arterial phase were asked but not used in the final algorithm. Nonetheless, this should not be taken to indicate that these features are irrelevant to clinical interpretation. For example, in lesions with diffuse enhancement in the arterial phase, both tumor necrosis and a central scar may contribute to areas of nonenhancement. As necrosis is very common in HCC, especially with increasing tumor size, and scars are a regular feature of FNH, their identification and differentiation are important features for correct diagnosis. Although, on a case-by-case basis, we know that this is very important, the questions about inhomogeneity within a diffusely enhancing tumor did not capture this distinction sufficiently in the blind review to make an impact on the developed algorithm. We expect that the failure of this question may have been influenced by the large number of hemangiomas that frequently show inhomogeneous enhancement in the arterial phase as only the peripheral nodules are enhanced. We had designed the question to differentiate FNH from HCC and had not anticipated this problem, and we suspect that had the hemangiomas been removed from the analysis, the responses regarding inhomogeneity of enhancement might have been more discriminating. Furthermore, experience suggests that identification of stellate vascularity and a highly tortuous feeding artery are highly suggestive of an FNH (Figs. 3A, 3B, and 3C). However, questions about vessel morphology did not differentiate FNH from other lesions in the blind review. Stellate vessels were specific but not sensitive for the diagnosis of FNH, identified in only nine of 20 lesions. To some extent, this may be because of technical limitations in the imaging method, which might be overcome using new techniques [32].
In this trial, the initial use of an infusion technique was discontinued in favor of bolus administration for two reasons. First, whereas bolus injection allowed for evaluation of arterial phase wash-in and arterial vascular morphology before the parenchymal image was dominated by the portal venous phase, drip infusion produced a gradual wash-in that concealed the onset of the portal phase. Second, although a single bolus injection of 0.2 mL was often sufficient to provide all of the answers to the questions in our blind review, the infusion technique produced a chronic but relatively weak enhancement of the liver parenchyma that required much higher doses of the contrast agent, sometimes requiring a second vial.
Benign lesions erroneously identified as malignant by the algorithm
included four benign lesions showing washout in the portal venous phase. These
were one regenerative nodule, two adenomas, and one FNH (Figs.
6A,
6B, and
6C). Diagnosis was confirmed
by biopsy in two lesions and by further imaging with 3-year follow-up in the
other two. Three malignant lesions, all HCC, were misclassified as benign
because they showed sustained enhancement in the extended portal venous phase
(Figs. 7A,
7B, and
7C). Diagnosis of malignancy
in these lesions was confirmed by biopsy in one case; the other two had
confirmatory imaging, known cirrhosis, and significant elevation of the
-fetoprotein level.
Limitations of our study include those inherent to all sonography studies, operator dependence and the negative influence of large patient habitus. Sonography is not as versatile for looking at the entire liver as CT or MRI and certainly not for analyzing a liver with many masses not necessarily having the same diagnosis. Further, a cirrhotic liver is even more difficult to scan than a normal liver on the basis of its small size and more limited acoustic access. Five of the 100 examinations failed technically for a variety of reasons, including equipment failure. Study entry criteria specified an identified lesion visible on sonography, which introduced some selection bias in our patient population. We also analyzed the results of only four pathologies, so that additional pathologies were, by definition, misclassified. This included the hepatic adenomas (n = 2), the solitary lipoma, and the regenerative nodules (n = 3). Furthermore, as the algorithm was developed on the basis of the findings from a blind interpretation, it is retrospective. A prospective study is now required to test it in a new population of patients.
The proposed algorithm relies heavily on the enhancement of the lesion in the extended portal venous phase. Although we recognize the strength of the data in this analysis to differentiate a malignant from a benign lesion based solely on this observation, we think that in clinical practice, it is unrealistic to rely on a single observation to diagnose a liver mass. We emphasize, instead, multiple observations made throughout the entire study and stress the value added by the arterial phase observations as well. It is particularly important to appreciate that not all masses show typical enhancement features, and that reliance on a single feature will increase the likelihood of diagnostic error. Furthermore, in clinical practice, the arterial phase enhancement and lesional vessels are the first things seen when performing a contrast-enhanced sonography scan. Their appearance should be integrated with the portal venous phase behavior to arrive at a final diagnosis. Furthermore, portal venous phase behavior alone may be insufficient information for patient management, particularly for malignant lesions where treatment options are quite different for HCC and metastasis. Precise differentiation of these two lesions requires attention to the arterial phase. As hypervascular metastases and HCC show identical arterial phase features on contrast sonography, as they do on contrast-enhanced CT or MRI, clinical history contributes crucially to their ultimate differentiation.
There are also unique strengths of microbubble-enhanced sonography. First, as the studies are performed in real time, rapidly changing enhancement (for example, that which occurs in type 1 hemangiomas or in most metastases), may be shown easily (Figs. 5A, 5B, 5C, and 8). In our study, the peak of arterial phase enhancement of a lesion was highly variable among individual patients. These differences were inconsequential as the peak was appropriately recorded regardless of the time of its occurrence. Washout also occurs at varying intervals from the injection time, and this is captured if the lesion is observed for several minutes from the completion of the contrast agent injection. This temporal resolution is a unique feature of sonography imaging. It should be noted that the microbubbles themselves are intravascular and do not pass into the interstitium of tumor or tissue, as they are approximately the same size as RBCs. The enhancement seen is therefore truly representative of the vascular distribution of the bubbles, removing the potential for erroneous interpretation to which CT and MRI are prone, where the contrast agents have a recognized ability to pass through the vascular endothelium into the tumor or tissue interstitium. The designation "equilibrium" phase as related to CT and MRI, therefore, does not apply to sonography. The term "extended" portal venous phase reflects the continuous imaging of the sonography contrast enhancement from the moment that the bubbles arrive in the portal vein until such time as the scan is discontinued. Finally, microbubble sonography shows lesional vessels much like an angiogram, providing some of the same diagnostic capability of angiographic techniques but without the need for arterial access.
Acknowledgments
The authors acknowledge the advice and guidance of Martin Rosenberg of
Bristol-Myers Squibb.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  T. Takayama, M. Makuuchi, M. Kojiro, G. Y. Lauwers, R. B. Adams, S. R. Wilson, H.-J. Jang, C. Charnsangavej, and B. Taouli Early Hepatocellular Carcinoma: Pathology, Imaging, and Therapy Ann. Surg. Oncol., April 1, 2008; 15(4): 972 - 978. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. R. Wilson, H.-J. Jang, T. K. Kim, H. Iijima, N. Kamiyama, and P. N. Burns Real-Time Temporal Maximum-Intensity-Projection Imaging of Hepatic Lesions with Contrast-Enhanced Sonography Am. J. Roentgenol., March 1, 2008; 190(3): 691 - 695. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. K. Kim, H.-J. Jang, P. N. Burns, J. Murphy-Lavallee, and S. R. Wilson Focal Nodular Hyperplasia and Hepatic Adenoma: Differentiation with Low-Mechanical-Index Contrast-Enhanced Sonography Am. J. Roentgenol., January 1, 2008; 190(1): 58 - 66. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Lanka, H.-J. Jang, T. K. Kim, P. N. Burns, and S. R. Wilson Impact of Contrast-Enhanced Ultrasonography in a Tertiary Clinical Practice J. Ultrasound Med., December 1, 2007; 26(12): 1703 - 1714. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Murphy-Lavallee, H.-J. Jang, T. K. Kim, P. N. Burns, and S. R. Wilson Are Metastases Really Hypovascular in the Arterial Phase?: The Perspective Based on Contrast-Enhanced Ultrasonography J. Ultrasound Med., November 1, 2007; 26(11): 1545 - 1556. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-J. Jang, T. K. Kim, P. N. Burns, and S. R. Wilson Enhancement Patterns of Hepatocellular Carcinoma at Contrast-enhanced US: Comparison with Histologic Differentiation Radiology, |
