HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2421051006v1 242/1/162    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Burns, P. N. Articles by Wilson, S. R. Search for Related Content PubMed PubMed Citation Articles by Burns, P. N. Articles by Wilson, S. R.

Focal Liver Masses: Enhancement Patterns on Contrast-enhanced Images—Concordance of US Scans with CT Scans and MR Images¹

¹ From the Departments of Medical Biophysics (P.N.B.) and Medical Imaging (P.N.B., S.R.W.), University of Toronto, Toronto, Ontario, Canada; Department of Imaging Research, Sunnybrook Health Sciences Centre, 2075 Bayview Ave, S660, Toronto, ON, Canada M4N 3M5 (P.N.B.); and Toronto General Hospital, University Health Network, Toronto, Ontario, Canada (S.R.W.). Received June 15, 2005; revision requested August 18; revision received November 2; accepted December 8; final version accepted April 7, 2006. Supported by the Canadian Institutes of Health Research and the Terry Fox Programme of the National Cancer Institute of Canada and in part by Bristol-Myers Squibb Medical Imaging. Address correspondence to P.N.B. (e-mail: Burns{at}swri.ca).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Purpose: To assess prospectively the concordance of enhancement patterns of focal liver masses on contrast material–enhanced ultrasonographic (US) scans with patterns on contrast-enhanced computed tomographic (CT) scans or magnetic resonance (MR) images.

Materials and Methods: This study was approved by the institutional review board; patients gave informed consent. Contrast-enhanced US and contrast-enhanced CT or MR imaging were performed in 135 patients (62 men, 73 women; mean age, 51 years) with 144 confirmed liver masses. Masses included 49 hepatocellular carcinomas, 13 metastases, 30 hemangiomas, 41 lesions of focal nodular hyperplasia, and 11 others. Randomized image sets from each modality were shown independently to three blinded readers, who answered identical questions about enhancement of the lesion and liver in the arterial and portal venous phases and changes with time. Concordance for modalities was calculated from answers of readers and consensus answers between readers, with 95% confidence intervals (CIs). The values were calculated for interreader agreement.

Results: Features of arterial phase enhancement showed concordance of more than 76% for modalities. The highest concordance of 92% (132 of 144), with 95% CI of 86% and 95% ( > 0.84), was for the presence of peripheral pools and centripetal progression. Concordance in the portal venous phase was lower, with agreement for predominant enhancement of the lesion in 61% (86 of 142), with 95% CI of 52% and 68% ( > 0.83). Portal venous phase washout occurred in 75% (106 of 142), with 95% CI of 67% and 81% ( > 0.81). The majority of discordances were for malignancies for which only US depicted no sustained enhancement in the portal venous phase.

Conclusion: US shows high concordance with CT or MR imaging, especially for the arterial phase. Discordance in the portal venous phase may reflect the tendency of CT and MR contrast agents, unlike microbubbles, to diffuse into interstitium.

© RSNA, 2006

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Historically, diagnosis of liver masses has been accomplished at histopathologic analysis following excisional or percutaneous biopsy. During the last two decades, the necessity for biopsy of liver lesions has diminished. Most benign and many malignant lesions now are correctly diagnosed noninvasively with either computed tomography (CT) or magnetic resonance (MR) imaging, and the diagnosis is determined principally from vascular information obtained from contrast enhancement in the arterial and portal venous phases (1–7).

Nonenhanced CT is relatively limited in its capability to distinguish hepatic masses, for isoattenuating masses suggest a hepatocellular origin and the findings of gross fat, calcification, or acute hemorrhage within a lesion contribute to the limitation of the differential diagnosis (8). Current protocols for nonenhanced MR imaging (8,9) exploit the greater capability of this modality to differentiate tissue components. Furthermore, T2-weighted sequences aid in the characterization of cysts and hemangiomas (10); chemical shift helps in the identification of fat within a lesion; and opposed-phase and chemical fat-saturation techniques assist in the detection of small and large amounts of lipids, respectively (11). Notwithstanding these capabilities, it is widely held that both CT and MR imaging ultimately are dependent on intravascular contrast enhancement for the definitive diagnosis of focal liver masses.

Nonenhanced ultrasonography (US) of the liver provides good spatial resolution and inherent soft-tissue contrast, which alone allow the characterization of many liver lesions. Though color Doppler US and power Doppler US provide good depiction of large-vessel flow, they fail to yield the information that is provided with contrast material–enhanced CT and MR imaging about vascularity of the parenchyma or the lesion. This is particularly true for lesions that are small, deep, or subject to artifact from cardiac or respiratory motion. The recent introduction of microbubble contrast agents for use with clinical US has offered the potential for US to show enhancement of liver lesions comparable to that shown at contrast-enhanced CT and MR imaging (12–15). In combination with nonlinear imaging techniques, such as pulse inversion (16), real-time assessment of liver vascularity also is possible without the blooming and motion artifacts that limit color Doppler US (16,17).

Promising preliminary experience has been reported of the use of first- and second-generation contrast agents for diagnosis of focal liver masses (18–23). The exact role that US can play in the treatment of patients known to have focal liver disease, however, remains unclear. Thus, the purpose of this study was to assess prospectively the concordance of enhancement patterns of focal liver masses on contrast-enhanced US scans with patterns on contrast-enhanced CT scans or MR images.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Partial financial support for this study was provided by Bristol-Myers Squibb Medical Imaging, Billerica, Mass. The authors had exclusive control of data and information presented in the manuscript.

Study Population
This prospective study had approval of the institutional ethics review board of Toronto General Hospital, Toronto, Ontario, Canada; all patients gave informed consent. Nonconsecutive patients with a solid liver mass visible at routine US were recruited for contrast-enhanced US at the time of their US examination. The inclusion criterion was that the patient had undergone contrast-enhanced CT or MR imaging: For malignant lesions, the requirement was that imaging was performed within an interval of 65 days from contrast-enhanced US. For benign lesions, there was no time limit. Pregnant or lactating patients, those undergoing active therapy, and those in whom a satisfactory baseline liver US examination could not be performed were excluded.

Between August 1999 and February 2003, 135 patients were recruited (62 men and 73 women; mean age, 51 years; range, 19–86 years) in whom 144 solid focal liver masses were depicted. These liver masses comprised 49 hepatocellular carcinomas (HCCs), 13 metastases, 30 hemangiomas, 41 lesions of focal nodular hyperplasia, three cholangiocarcinomas, three adenomas, three regenerative nodules, one lymphoma, and one lipoma. Pathologic confirmation of the diagnosis was obtained in 59 of 66 malignant lesions: 45 HCCs, 10 metastases, three cholangiocarcinomas, and one lymphoma. At imaging, three patients with a known extrahepatic primary tumor had widespread metastases in the liver, lung, and bone, with disease progression at 3-year follow-up. Three patients with HCCs had an -fetoprotein level of more than 4000 µg/L; in one patient with HCC, the diagnosis was determined only at imaging. Pathologic confirmation was obtained in 10 of 78 benign lesions, which included one lipoma, two of three regenerative nodules, two of three adenomas, and five of 41 lesions of focal nodular hyperplasia. All other benign lesions were confirmed by using supportive CT or MR imaging, with continued imaging and clinical observation for a mean period of 49 months (range, 31–70 months). Lesions ranged from 1 to 11 cm in maximal diameter (mean, 4.8 cm).

Imaging Technique
Masses were evaluated with contrast-enhanced US (n = 144) and contrast-enhanced CT (n = 116), MR imaging (n = 33), or both (n = 5). The mean time between US and CT or MR imaging was 19 days for malignant lesions (in 34, CT or MR imaging was performed before US; in 20, CT or MR imaging was performed after US) and was 85 days for benign lesions (in 35, CT or MR imaging was performed before US; in 30, CT or MR imaging was performed after US). In 25 lesions (12 malignant, 13 benign), all imaging was performed on the same day. US examinations were performed on a solitary mass or on a dominant mass in patients with multiple lesions; however, both lesions were examined in nine patients who had two masses, for which the diagnoses were believed to be potentially different, in separate lobes of the liver.

US was performed by one physician (S.R.W.) with more than 20 years of experience in liver US. All examinations were performed by using microbubble-specific imaging. One hundred twenty-eight examinations were performed with one machine (ATL HDI 5000; Philips Medical Systems, Bothell, Wash) by using pulse-inversion imaging; 16 were performed with another unit (Acuson Sequoia; Siemens Medical Solutions, Santa Clara, Calif) by using contrast pulse sequence imaging. Curvilinear transducers were employed with a center frequency of 1.1–2.2 MHz. The US contrast agent, gas-filled perflutren lipid microspheres (Definity; Bristol-Myers Squibb Medical Imaging), was administered intravenously as three to six (median, four) small boluses of 0.1–0.4 mL (median, 0.2 mL) each, followed by a saline flush, to a maximum total dose of 10 µL per kilogram of body weight. Injections were administered more than 5 minutes apart to minimize any cumulative effect of contrast agent.

With the scanning technique, real-time imaging with a low mechanical index of less than 0.1, which preserves the microbubble population and allows evaluation of vessels of the lesion and enhancement of the lesion and liver, was used. The transmit focal zone was positioned distal to the lesion of interest. Image frame rate was typically 10–15 frames per second. The region of interest was observed continuously from the time of injection for about 5 minutes. The arterial phase was timed for 45 seconds after the completion of the flush, after which we defined an extended portal venous phase, and this phase encompassed the commonly described interval from 45 to 70 seconds after injection, as well as the remainder of the observation period. Because the microbubbles of the contrast agent are purely intravascular, there is no interstitial or equilibrium phase: Enhancement in the extended portal venous phase shows progressive decay during about 3 minutes until the baseline appearance is again observed.

CT was performed by using four- or eight-detector scanners (LightSpeed QX/i or LightSpeed Ultra; GE Medical Systems, Milwaukee, Wis). A triphasic liver protocol with standard delays was used with nonenhanced imaging during the arterial (30 seconds after injection) and portal venous (60 seconds after injection) phases. Iohexol (Omnipaque; Amersham Health, Princeton, NJ) at 300 mg of iodine per milliliter or iodixanol (Visipaque; Amersham Health) at 270 mg of iodine per milliliter was used as an intravenous contrast agent, with 2 mL/kg injected at 4–5 mL/sec up to 200 mL. Reconstruction was performed with a standard algorithm, with section thickness of 5 mm and 50% overlap.

MR imaging was performed by using 1.5-T systems (EchoSpeed LX; GE Medical Systems) with a phased-array torso coil. The standard protocol included the following sequences: coronal single-shot fast spin-echo T2-weighted MR imaging (effective echo time, 60 msec; section thickness, 6 mm; acquisition time, 23 seconds), transverse dual-echo in- and out-of-phase spoiled gradient-echo T1-weighted MR imaging (repetition time msec/echo time msec, 150/2.1, 4.2; flip angle, 70°; section thickness, 5–8 mm; acquisition time, 23 seconds), transverse breath-hold fast-recovery fast spin-echo T2-weighted fat-saturated MR imaging (2600/90; echo train length, 17; section thickness, 5–7 mm; spacing, 1 mm; acquisition time, 26 seconds), transverse single-shot fast spin-echo T2-weighted fat-saturated MR imaging (echo time, 160–180 msec; section thickness, 5–7 mm; acquisition time, 2 seconds per section), and dynamic two-dimensional fast spoiled gradient-echo MR imaging (150–200/minimum; flip angle, 80°; section thickness, 5–8 mm; acquisition time, 24 seconds). The sequences were performed at nonenhanced imaging during the arterial (15 seconds after injection), portal venous (50 seconds after injection), and delayed (85 seconds after injection) phases. Gadodiamide (Omniscan; Amersham Health), 0.1 mmol/kg, was manually injected intravenously and was followed by a saline flush.

Blinded Reading
To determine the concordance of the contrast-enhanced US examinations with contrast-enhanced CT or MR imaging, a three-person blinded reading was conducted for evaluation of specific features of enhancement. All three readers were trained in body imaging and had been working continuously for 3, 6, and 10 years, respectively, in the reading of US, CT, and MR images. They were proficient in the interpretation of contrast-enhanced US images but had no involvement with the patients in this study. A clinical fellow in body imaging, who was neither an author nor a blinded reader, selected optimal images to show the lesion at baseline and in the arterial and the portal venous phases of enhancement for each modality. At US, therefore, hundreds of frames were available for selection for both phases, whereas at CT and MR imaging, one image at the appropriate location was selected for each phase that showed the optimal contrast between enhancement of the lesion and of the parenchyma.

For US, a brief real-time movie also was included, and the movie showed the wash-in of the contrast agent to the liver in real time. On all images, the lesion was marked with arrows. Window level, gray-scale map, and dynamic range were optimized as the images were selected so that no changes could be made during the reading. CT scans, MR images, and US scans from the same patient were separated, and then all 288 images were randomized and assigned an identification number. Image sets were shown separately in a single sitting to the three readers, who were blinded to the clinical information and the name of the patient and could not confer. Randomization helped to ensure that no US scan was shown consecutively with its corresponding CT scan or MR image. Images were displayed on a workstation in the same sequence to each of the readers. Nine identical questions (Table 1) were then asked for each lesion for each modality.

An interpretive component of the reading described the change in enhancement with time. Centripetal progression indicated progressive enhancement from the periphery toward the center of the lesion. Portal venous phase enhancement less than that of the liver, frequently referred to as washout, was determined if the lesion initially showed partial or complete enhancement greater than that of the liver in the arterial phase and then showed less enhancement in the portal venous phase. Sustained enhancement occurred if the lesion showed enhancement in the arterial phase and greater or equal enhancement in the portal venous phase relative to that of the liver. Readers were not asked to provide a diagnosis of the liver mass from the images they were shown either before or after contrast enhancement.

Data and Statistical Analysis
Data analysis was based on individual concordance, the proportion of exact agreement between answers of an individual reader to a specific question with regard to US and CT or MR imaging. From individual concordance, the mean individual concordance was derived with averaging of the concordance of individual readers. We also defined consensus concordance, for which the consensus answer (ie, agreement of at least two of three readers about the answer to a specific question) was determined from all three answers to a single question for each modality. Consensus concordance was then calculated as the proportion of cases in which this answer was the same for US and CT or MR imaging. Concordance is expressed as a percentage of cases analyzed per question in which there was exact agreement between modalities. The 95% confidence intervals (CIs) were derived from the concordance values by using the method described by Agresti and Coull (24).

Cases in which a reader responded "can't assess" to a question for both modalities were not considered concordant and were excluded from analysis of individual concordance. In the five cases with both MR and CT studies, the MR images were analyzed. Cases in which no consensus was reached in regard to the answer to a question were excluded from analysis of consensus concordance. The value, an estimate of the proportion of agreement among modalities that was not due to chance, was used to estimate the degree of agreement among the readers for each question and each modality. The value also was calculated for the agreement of the consensus of the readers' responses to each question about the US scans and the CT scans or MR images.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Technical Performance
A total of 144 examination pairs were analyzed. All examinations were technically satisfactory for US and for CT or MR imaging. In two instances, questions could not be answered concerning portal venous phase behavior at US because of a technical error while the study was recorded.

Blinded Reading
The concordance was calculated for the consensus among the three readers (Table 2) (144 evaluated pairs) and for each individual reader (Table 3) (432 evaluated pairs). Consensus agreement among the readers who assessed the images was high. Consensus was not reached (ie, three readers gave three different answers) on only six (0.5%) occasions of 1296 examination questions for CT or MR imaging and on nine (0.7%) occasions of 1288 examination questions for US. The difference between consensus concordance and the mean concordance of the individual readers was small (mean difference, 1%; mean absolute difference, 1.4%), and this finding indicated a high degree of agreement among readers. The values for agreement between readers were between 0.51 and 0.94 and, except for one instance, were all greater than 0.7. For the two questions in regard to portal venous phase washout, the values (Table 2) indicated that there was a greater degree of agreement between the readers in regard to US (0.94 and 0.90) than there was in regard to CT or MR imaging (0.81 and 0.80).

View larger version (176K): [in this window] [in a new window] [Download PPT slide]   Figure 1a: Hemangioma in 46-year-old woman, with concordance of arterial phase imaging. Transverse images of right lobe of liver at arterial phase (a) microbubble-enhanced US and (b) iodine-enhanced CT and (c) portal venous phase gadolinium-enhanced T1-weighted fast spoiled gradient-echo MR imaging (170/1.6; flip angle, 80°). All images show a similar pattern of bright peripheral nodular enhancement (arrow). This pattern progressed to complete fill-in on images for all three modalities (not shown). Readers agreed completely; all reported that enhancement of peripheral puddles was greater than that of adjacent liver.

View larger version (168K): [in this window] [in a new window] [Download PPT slide]   Figure 1b: Hemangioma in 46-year-old woman, with concordance of arterial phase imaging. Transverse images of right lobe of liver at arterial phase (a) microbubble-enhanced US and (b) iodine-enhanced CT and (c) portal venous phase gadolinium-enhanced T1-weighted fast spoiled gradient-echo MR imaging (170/1.6; flip angle, 80°). All images show a similar pattern of bright peripheral nodular enhancement (arrow). This pattern progressed to complete fill-in on images for all three modalities (not shown). Readers agreed completely; all reported that enhancement of peripheral puddles was greater than that of adjacent liver.

View larger version (173K): [in this window] [in a new window] [Download PPT slide]   Figure 1c: Hemangioma in 46-year-old woman, with concordance of arterial phase imaging. Transverse images of right lobe of liver at arterial phase (a) microbubble-enhanced US and (b) iodine-enhanced CT and (c) portal venous phase gadolinium-enhanced T1-weighted fast spoiled gradient-echo MR imaging (170/1.6; flip angle, 80°). All images show a similar pattern of bright peripheral nodular enhancement (arrow). This pattern progressed to complete fill-in on images for all three modalities (not shown). Readers agreed completely; all reported that enhancement of peripheral puddles was greater than that of adjacent liver.

View larger version (159K): [in this window] [in a new window] [Download PPT slide]   Figure 2a: Hemangioma in 71-year-old woman, with concordance of arterial phase enhancement at US, CT, and MR imaging. Transverse contrast-enhanced images of left lobe of liver at arterial phase (a) US and (b) CT and (c) portal venous phase gadolinium-enhanced T1-weighted fast spoiled gradient-echo MR imaging (170/1.6; flip angle, 80°) are concordant. All images show brightly enhanced mass (arrow on a and d) with small central area of nonenhancement. Readers agreed completely that there was an enhancing component in arterial phase, with sustained enhancement in portal venous phase (not shown). (d–f) Sequential low-mechanical-index real-time sagittal US images of the mass in left lobe obtained earlier than a. (d) Focal lobulated solid hypoechoic mass (arrow) prior to arrival of contrast agent. (e) Enhancement of liver, with a faint rim of enhancement around the lesion. (f) Peripheral nodules with evidence of centripetal progression of enhancement. Because of more rapid image acquisition at US, the filling of this lesion, with early peripheral nodules, and centripetal progression are not appreciated on CT or MR images.

View larger version (189K): [in this window] [in a new window] [Download PPT slide]   Figure 2b: Hemangioma in 71-year-old woman, with concordance of arterial phase enhancement at US, CT, and MR imaging. Transverse contrast-enhanced images of left lobe of liver at arterial phase (a) US and (b) CT and (c) portal venous phase gadolinium-enhanced T1-weighted fast spoiled gradient-echo MR imaging (170/1.6; flip angle, 80°) are concordant. All images show brightly enhanced mass (arrow on a and d) with small central area of nonenhancement. Readers agreed completely that there was an enhancing component in arterial phase, with sustained enhancement in portal venous phase (not shown). (d–f) Sequential low-mechanical-index real-time sagittal US images of the mass in left lobe obtained earlier than a. (d) Focal lobulated solid hypoechoic mass (arrow) prior to arrival of contrast agent. (e) Enhancement of liver, with a faint rim of enhancement around the lesion. (f) Peripheral nodules with evidence of centripetal progression of enhancement. Because of more rapid image acquisition at US, the filling of this lesion, with early peripheral nodules, and centripetal progression are not appreciated on CT or MR images.

View larger version (150K): [in this window] [in a new window] [Download PPT slide]   Figure 2c: Hemangioma in 71-year-old woman, with concordance of arterial phase enhancement at US, CT, and MR imaging. Transverse contrast-enhanced images of left lobe of liver at arterial phase (a) US and (b) CT and (c) portal venous phase gadolinium-enhanced T1-weighted fast spoiled gradient-echo MR imaging (170/1.6; flip angle, 80°) are concordant. All images show brightly enhanced mass (arrow on a and d) with small central area of nonenhancement. Readers agreed completely that there was an enhancing component in arterial phase, with sustained enhancement in portal venous phase (not shown). (d–f) Sequential low-mechanical-index real-time sagittal US images of the mass in left lobe obtained earlier than a. (d) Focal lobulated solid hypoechoic mass (arrow) prior to arrival of contrast agent. (e) Enhancement of liver, with a faint rim of enhancement around the lesion. (f) Peripheral nodules with evidence of centripetal progression of enhancement. Because of more rapid image acquisition at US, the filling of this lesion, with early peripheral nodules, and centripetal progression are not appreciated on CT or MR images.

View larger version (140K): [in this window] [in a new window] [Download PPT slide]   Figure 2d: Hemangioma in 71-year-old woman, with concordance of arterial phase enhancement at US, CT, and MR imaging. Transverse contrast-enhanced images of left lobe of liver at arterial phase (a) US and (b) CT and (c) portal venous phase gadolinium-enhanced T1-weighted fast spoiled gradient-echo MR imaging (170/1.6; flip angle, 80°) are concordant. All images show brightly enhanced mass (arrow on a and d) with small central area of nonenhancement. Readers agreed completely that there was an enhancing component in arterial phase, with sustained enhancement in portal venous phase (not shown). (d–f) Sequential low-mechanical-index real-time sagittal US images of the mass in left lobe obtained earlier than a. (d) Focal lobulated solid hypoechoic mass (arrow) prior to arrival of contrast agent. (e) Enhancement of liver, with a faint rim of enhancement around the lesion. (f) Peripheral nodules with evidence of centripetal progression of enhancement. Because of more rapid image acquisition at US, the filling of this lesion, with early peripheral nodules, and centripetal progression are not appreciated on CT or MR images.

View larger version (143K): [in this window] [in a new window] [Download PPT slide]   Figure 2e: Hemangioma in 71-year-old woman, with concordance of arterial phase enhancement at US, CT, and MR imaging. Transverse contrast-enhanced images of left lobe of liver at arterial phase (a) US and (b) CT and (c) portal venous phase gadolinium-enhanced T1-weighted fast spoiled gradient-echo MR imaging (170/1.6; flip angle, 80°) are concordant. All images show brightly enhanced mass (arrow on a and d) with small central area of nonenhancement. Readers agreed completely that there was an enhancing component in arterial phase, with sustained enhancement in portal venous phase (not shown). (d–f) Sequential low-mechanical-index real-time sagittal US images of the mass in left lobe obtained earlier than a. (d) Focal lobulated solid hypoechoic mass (arrow) prior to arrival of contrast agent. (e) Enhancement of liver, with a faint rim of enhancement around the lesion. (f) Peripheral nodules with evidence of centripetal progression of enhancement. Because of more rapid image acquisition at US, the filling of this lesion, with early peripheral nodules, and centripetal progression are not appreciated on CT or MR images.

View larger version (142K): [in this window] [in a new window] [Download PPT slide]   Figure 2f: Hemangioma in 71-year-old woman, with concordance of arterial phase enhancement at US, CT, and MR imaging. Transverse contrast-enhanced images of left lobe of liver at arterial phase (a) US and (b) CT and (c) portal venous phase gadolinium-enhanced T1-weighted fast spoiled gradient-echo MR imaging (170/1.6; flip angle, 80°) are concordant. All images show brightly enhanced mass (arrow on a and d) with small central area of nonenhancement. Readers agreed completely that there was an enhancing component in arterial phase, with sustained enhancement in portal venous phase (not shown). (d–f) Sequential low-mechanical-index real-time sagittal US images of the mass in left lobe obtained earlier than a. (d) Focal lobulated solid hypoechoic mass (arrow) prior to arrival of contrast agent. (e) Enhancement of liver, with a faint rim of enhancement around the lesion. (f) Peripheral nodules with evidence of centripetal progression of enhancement. Because of more rapid image acquisition at US, the filling of this lesion, with early peripheral nodules, and centripetal progression are not appreciated on CT or MR images.

View larger version (156K): [in this window] [in a new window] [Download PPT slide]   Figure 3a: HCC in 67-year-old man, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass in posterior right lobe. Arterial phase images b and e show nonuniformly enhancing hypervascular mass (M). Portal venous phase images c and f show lesion as less echogenic at US and less attenuating at CT than adjacent liver. Readers agreed completely that both modalities showed diffuse arterial phase enhancement with a nonenhancing region and negative enhancement in the portal venous phase with washout.

View larger version (151K): [in this window] [in a new window] [Download PPT slide]   Figure 3b: HCC in 67-year-old man, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass in posterior right lobe. Arterial phase images b and e show nonuniformly enhancing hypervascular mass (M). Portal venous phase images c and f show lesion as less echogenic at US and less attenuating at CT than adjacent liver. Readers agreed completely that both modalities showed diffuse arterial phase enhancement with a nonenhancing region and negative enhancement in the portal venous phase with washout.

View larger version (161K): [in this window] [in a new window] [Download PPT slide]   Figure 3c: HCC in 67-year-old man, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass in posterior right lobe. Arterial phase images b and e show nonuniformly enhancing hypervascular mass (M). Portal venous phase images c and f show lesion as less echogenic at US and less attenuating at CT than adjacent liver. Readers agreed completely that both modalities showed diffuse arterial phase enhancement with a nonenhancing region and negative enhancement in the portal venous phase with washout.

View larger version (133K): [in this window] [in a new window] [Download PPT slide]   Figure 3d: HCC in 67-year-old man, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass in posterior right lobe. Arterial phase images b and e show nonuniformly enhancing hypervascular mass (M). Portal venous phase images c and f show lesion as less echogenic at US and less attenuating at CT than adjacent liver. Readers agreed completely that both modalities showed diffuse arterial phase enhancement with a nonenhancing region and negative enhancement in the portal venous phase with washout.

View larger version (155K): [in this window] [in a new window] [Download PPT slide]   Figure 3e: HCC in 67-year-old man, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass in posterior right lobe. Arterial phase images b and e show nonuniformly enhancing hypervascular mass (M). Portal venous phase images c and f show lesion as less echogenic at US and less attenuating at CT than adjacent liver. Readers agreed completely that both modalities showed diffuse arterial phase enhancement with a nonenhancing region and negative enhancement in the portal venous phase with washout.

View larger version (154K): [in this window] [in a new window] [Download PPT slide]   Figure 3f: HCC in 67-year-old man, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass in posterior right lobe. Arterial phase images b and e show nonuniformly enhancing hypervascular mass (M). Portal venous phase images c and f show lesion as less echogenic at US and less attenuating at CT than adjacent liver. Readers agreed completely that both modalities showed diffuse arterial phase enhancement with a nonenhancing region and negative enhancement in the portal venous phase with washout.

View larger version (171K): [in this window] [in a new window] [Download PPT slide]   Figure 4a: Focal nodular hyperplasia in 38-year-old asymptomatic woman, with concordance of arterial and portal venous phase enhancement. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass (arrows) in right lobe of liver. Arterial phase images b and e show enhancement consistent with hypervascular mass (arrows). Portal venous phase images c and f show sustained enhancement of mass (arrows), which continues to show more enhancement than adjacent liver. With both modalities, a benign lesion is correctly suggested. All readers agreed that there was arterial phase enhancement at US and CT. One reader interpreted the portal venous phase CT image as showing washout.

View larger version (127K): [in this window] [in a new window] [Download PPT slide]   Figure 4b: Focal nodular hyperplasia in 38-year-old asymptomatic woman, with concordance of arterial and portal venous phase enhancement. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass (arrows) in right lobe of liver. Arterial phase images b and e show enhancement consistent with hypervascular mass (arrows). Portal venous phase images c and f show sustained enhancement of mass (arrows), which continues to show more enhancement than adjacent liver. With both modalities, a benign lesion is correctly suggested. All readers agreed that there was arterial phase enhancement at US and CT. One reader interpreted the portal venous phase CT image as showing washout.

View larger version (142K): [in this window] [in a new window] [Download PPT slide]   Figure 4c: Focal nodular hyperplasia in 38-year-old asymptomatic woman, with concordance of arterial and portal venous phase enhancement. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass (arrows) in right lobe of liver. Arterial phase images b and e show enhancement consistent with hypervascular mass (arrows). Portal venous phase images c and f show sustained enhancement of mass (arrows), which continues to show more enhancement than adjacent liver. With both modalities, a benign lesion is correctly suggested. All readers agreed that there was arterial phase enhancement at US and CT. One reader interpreted the portal venous phase CT image as showing washout.

View larger version (136K): [in this window] [in a new window] [Download PPT slide]   Figure 4d: Focal nodular hyperplasia in 38-year-old asymptomatic woman, with concordance of arterial and portal venous phase enhancement. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass (arrows) in right lobe of liver. Arterial phase images b and e show enhancement consistent with hypervascular mass (arrows). Portal venous phase images c and f show sustained enhancement of mass (arrows), which continues to show more enhancement than adjacent liver. With both modalities, a benign lesion is correctly suggested. All readers agreed that there was arterial phase enhancement at US and CT. One reader interpreted the portal venous phase CT image as showing washout.

View larger version (158K): [in this window] [in a new window] [Download PPT slide]   Figure 4e: Focal nodular hyperplasia in 38-year-old asymptomatic woman, with concordance of arterial and portal venous phase enhancement. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass (arrows) in right lobe of liver. Arterial phase images b and e show enhancement consistent with hypervascular mass (arrows). Portal venous phase images c and f show sustained enhancement of mass (arrows), which continues to show more enhancement than adjacent liver. With both modalities, a benign lesion is correctly suggested. All readers agreed that there was arterial phase enhancement at US and CT. One reader interpreted the portal venous phase CT image as showing washout.

View larger version (171K): [in this window] [in a new window] [Download PPT slide]   Figure 4f: Focal nodular hyperplasia in 38-year-old asymptomatic woman, with concordance of arterial and portal venous phase enhancement. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of right lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Baseline images a and d show subtle mass (arrows) in right lobe of liver. Arterial phase images b and e show enhancement consistent with hypervascular mass (arrows). Portal venous phase images c and f show sustained enhancement of mass (arrows), which continues to show more enhancement than adjacent liver. With both modalities, a benign lesion is correctly suggested. All readers agreed that there was arterial phase enhancement at US and CT. One reader interpreted the portal venous phase CT image as showing washout.

View larger version (157K): [in this window] [in a new window] [Download PPT slide]   Figure 5a: Small HCC in 53-year-old woman, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show small mass (arrow) in liver segment 7. Arterial phase images b and e show uniform enhancement of mass (arrow). Portal venous phase images c and f show less enhancement of mass (arrow) than of liver. Readers agreed completely about interpretation among modalities.

View larger version (192K): [in this window] [in a new window] [Download PPT slide]   Figure 5b: Small HCC in 53-year-old woman, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show small mass (arrow) in liver segment 7. Arterial phase images b and e show uniform enhancement of mass (arrow). Portal venous phase images c and f show less enhancement of mass (arrow) than of liver. Readers agreed completely about interpretation among modalities.

View larger version (172K): [in this window] [in a new window] [Download PPT slide]   Figure 5c: Small HCC in 53-year-old woman, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show small mass (arrow) in liver segment 7. Arterial phase images b and e show uniform enhancement of mass (arrow). Portal venous phase images c and f show less enhancement of mass (arrow) than of liver. Readers agreed completely about interpretation among modalities.

View larger version (149K): [in this window] [in a new window] [Download PPT slide]   Figure 5d: Small HCC in 53-year-old woman, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show small mass (arrow) in liver segment 7. Arterial phase images b and e show uniform enhancement of mass (arrow). Portal venous phase images c and f show less enhancement of mass (arrow) than of liver. Readers agreed completely about interpretation among modalities.

View larger version (154K): [in this window] [in a new window] [Download PPT slide]   Figure 5e: Small HCC in 53-year-old woman, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show small mass (arrow) in liver segment 7. Arterial phase images b and e show uniform enhancement of mass (arrow). Portal venous phase images c and f show less enhancement of mass (arrow) than of liver. Readers agreed completely about interpretation among modalities.

View larger version (162K): [in this window] [in a new window] [Download PPT slide]   Figure 5f: Small HCC in 53-year-old woman, with concordance of enhancement in arterial and portal venous phases. (a–c) Transverse US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show small mass (arrow) in liver segment 7. Arterial phase images b and e show uniform enhancement of mass (arrow). Portal venous phase images c and f show less enhancement of mass (arrow) than of liver. Readers agreed completely about interpretation among modalities.

View larger version (182K): [in this window] [in a new window] [Download PPT slide]   Figure 6a: Cholangiocarcinoma of liver in 64-year-old woman, with discordance in portal venous phase. (a–c) Oblique US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show focal mass (arrow) in liver segment 7 adjacent to inferior vena cava. Arterial phase US image b shows enhancement. Arterial phase MR image e shows enhancing nodule within mass (arrow). Portal venous phase US image c shows hypoechoic mass (arrow) relative to enhanced liver: Washout suggests malignancy. Portal venous phase MR image f shows that enhancing nodule within the mass (arrow) continues to show slightly greater signal intensity than adjacent liver and has increased in size with a nonenhancing rim. This finding was interpreted clinically as an indeterminate MR result. With blinded reading, there was lack of agreement for MR images; two readers reported peripheral puddles with centripetal progression and sustained enhancement. At US, readers agreed completely about enhancement in arterial phase with portal venous phase washout.

View larger version (154K): [in this window] [in a new window] [Download PPT slide]   Figure 6b: Cholangiocarcinoma of liver in 64-year-old woman, with discordance in portal venous phase. (a–c) Oblique US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show focal mass (arrow) in liver segment 7 adjacent to inferior vena cava. Arterial phase US image b shows enhancement. Arterial phase MR image e shows enhancing nodule within mass (arrow). Portal venous phase US image c shows hypoechoic mass (arrow) relative to enhanced liver: Washout suggests malignancy. Portal venous phase MR image f shows that enhancing nodule within the mass (arrow) continues to show slightly greater signal intensity than adjacent liver and has increased in size with a nonenhancing rim. This finding was interpreted clinically as an indeterminate MR result. With blinded reading, there was lack of agreement for MR images; two readers reported peripheral puddles with centripetal progression and sustained enhancement. At US, readers agreed completely about enhancement in arterial phase with portal venous phase washout.

View larger version (156K): [in this window] [in a new window] [Download PPT slide]   Figure 6c: Cholangiocarcinoma of liver in 64-year-old woman, with discordance in portal venous phase. (a–c) Oblique US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show focal mass (arrow) in liver segment 7 adjacent to inferior vena cava. Arterial phase US image b shows enhancement. Arterial phase MR image e shows enhancing nodule within mass (arrow). Portal venous phase US image c shows hypoechoic mass (arrow) relative to enhanced liver: Washout suggests malignancy. Portal venous phase MR image f shows that enhancing nodule within the mass (arrow) continues to show slightly greater signal intensity than adjacent liver and has increased in size with a nonenhancing rim. This finding was interpreted clinically as an indeterminate MR result. With blinded reading, there was lack of agreement for MR images; two readers reported peripheral puddles with centripetal progression and sustained enhancement. At US, readers agreed completely about enhancement in arterial phase with portal venous phase washout.

View larger version (114K): [in this window] [in a new window] [Download PPT slide]   Figure 6d: Cholangiocarcinoma of liver in 64-year-old woman, with discordance in portal venous phase. (a–c) Oblique US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show focal mass (arrow) in liver segment 7 adjacent to inferior vena cava. Arterial phase US image b shows enhancement. Arterial phase MR image e shows enhancing nodule within mass (arrow). Portal venous phase US image c shows hypoechoic mass (arrow) relative to enhanced liver: Washout suggests malignancy. Portal venous phase MR image f shows that enhancing nodule within the mass (arrow) continues to show slightly greater signal intensity than adjacent liver and has increased in size with a nonenhancing rim. This finding was interpreted clinically as an indeterminate MR result. With blinded reading, there was lack of agreement for MR images; two readers reported peripheral puddles with centripetal progression and sustained enhancement. At US, readers agreed completely about enhancement in arterial phase with portal venous phase washout.

View larger version (122K): [in this window] [in a new window] [Download PPT slide]   Figure 6e: Cholangiocarcinoma of liver in 64-year-old woman, with discordance in portal venous phase. (a–c) Oblique US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show focal mass (arrow) in liver segment 7 adjacent to inferior vena cava. Arterial phase US image b shows enhancement. Arterial phase MR image e shows enhancing nodule within mass (arrow). Portal venous phase US image c shows hypoechoic mass (arrow) relative to enhanced liver: Washout suggests malignancy. Portal venous phase MR image f shows that enhancing nodule within the mass (arrow) continues to show slightly greater signal intensity than adjacent liver and has increased in size with a nonenhancing rim. This finding was interpreted clinically as an indeterminate MR result. With blinded reading, there was lack of agreement for MR images; two readers reported peripheral puddles with centripetal progression and sustained enhancement. At US, readers agreed completely about enhancement in arterial phase with portal venous phase washout.

View larger version (125K): [in this window] [in a new window] [Download PPT slide]   Figure 6f: Cholangiocarcinoma of liver in 64-year-old woman, with discordance in portal venous phase. (a–c) Oblique US images of right lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse T1-weighted fast spoiled gradient-echo MR images (185/1.4; flip angle, 80°) of right lobe. (d) Nonenhanced and gadolinium-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced images a and d show focal mass (arrow) in liver segment 7 adjacent to inferior vena cava. Arterial phase US image b shows enhancement. Arterial phase MR image e shows enhancing nodule within mass (arrow). Portal venous phase US image c shows hypoechoic mass (arrow) relative to enhanced liver: Washout suggests malignancy. Portal venous phase MR image f shows that enhancing nodule within the mass (arrow) continues to show slightly greater signal intensity than adjacent liver and has increased in size with a nonenhancing rim. This finding was interpreted clinically as an indeterminate MR result. With blinded reading, there was lack of agreement for MR images; two readers reported peripheral puddles with centripetal progression and sustained enhancement. At US, readers agreed completely about enhancement in arterial phase with portal venous phase washout.

View larger version (162K): [in this window] [in a new window] [Download PPT slide]   Figure 7a: Colorectal metastasis in 63-year-old woman, with discordance in portal venous phase. (a–c) Transverse US images of left lobe of liver. (a) Nonenhanced and contrast-enhanced images in (b) arterial and (c) portal venous phases. (d–f) Transverse CT scans of left lobe. (d) Nonenhanced and contrast-enhanced images in (e) arterial and (f) portal venous phases. Nonenhanced US image a shows focal echogenic mass (M) in left lobe. Nonenhanced CT scan d shows low-attenuation mass (M) in same location. Arterial phase images b and e show nonuniform enhancement of a portion of the mass (arrows). Portal venous phase US image c shows washout of enhancement: Lesion is uniformly hypoechoic relative to enhanced liver except for a faint rim (arrowheads) of remaining enhancement. Portal venous phase CT image f shows enhancing component of mass (arrows), with slight attenuation relative to adjacent liver. This persistent enhancement may have been caused by contrast agent within the tumor interstitium.

View larger version (156K): [in this window]