| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Gastrointestinal Imaging |
1 Department of Radiological Diagnostics and Therapy, German Cancer Research Center, Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany (H.H., S.O.S., M.V.K., M.E., G.v.K.)
2 Department of Radiology, St Josefs Hospital, Heidelberg, Germany (P.M.).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Breath-hold fast spoiled gradient-echo 3D gadolinium-enhanced MR angiography was performed in 18 patients with 35 hepatic lesions. Measurements of signal intensity were obtained for 27 seconds in each phase, with 23-second delays between the three phases. Lesion-liver visibilities at each phase on the MR angiographic, precontrast T1-weighted, T2-weighted, and postcontrast T1-weighted images were compared. The MR angiographic functional lesion characterization was based on the combined assessment of spatial variations and the evolution of contrast material enhancement in all three phases.
RESULTS: All 35 lesions were correctly characterized on the MR angiographic images, which is significantly (P < .01) better than the precontrast T1-weighted (n = 14 [40%]), T2-weighted (n = 23 [66%]), and postcontrast T1-weighted (n = 25 [71%]) imaging results. Analysis of the spatial variations and the evolution of contrast material enhancement significantly (P < .01) improved lesion characterization in 66% (23 of 35) of all lesions.
CONCLUSION: Multiphase breath-hold 3D gadolinium-enhanced MR angiography is feasible and robust and significantly improves the morphologic detection of benign or malignant lesions during the early arterial phase. It further improves the functional characterization of hepatic lesions, combining an arterial, portal-venous, and late MR angiographic phase of contrast enhancement.
Index terms: Liver neoplasms, MR, 761.12142, 761.30 • Magnetic resonance (MR), contrast enhancement, 761.12143 • Magnetic resonance (MR), three-dimensional, 761.12117 • Magnetic resonance (MR), vascular studies, 761.12142
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Until recently, multiphase acquisition for the characterization of hepatic lesions was limited to spiral CT (2) or dynamic two-dimensional magnetic resonance (MR) imaging (4) studies. However, the recent introduction of fast three-dimensional (3D) MR angiographic techniques allows the acquisition of a complete 3D data set within a breath hold (5–7). Therefore, detailed information on various hepatic lesions can be acquired by assessing the spatial variations and evolution of contrast enhancement that occur during the arterial, portal venous, and late venous phases in fast hepatic imaging.
The aim of this preliminary study was to investigate the value of breath-hold, 3D, gadolinium-enhanced MR angiography performed in the arterial, portal venous, and late venous phases of contrast enhancement for the detection and morphologic and functional characterization of various hepatic lesions.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Twenty-one malignant lesions in 10 patients and 14 benign lesions in eight patients were recorded. A detailed analysis with histopathologic confirmation proved malignancy (hepatocellular carcinoma in three patients, including one patient with three coexisting lesions and the same radiologic features as those of lesions proved at histologic examination; metastasis of a different origin in seven patients; and thereof three patients with three, four, and five additional lesions and the same radiologic findings) and benignity (focal nodular hyperplasia in two patients; hemangioma in six patients; and thereof three patients with two and three with four coexisting lesions and the same radiologic findings).
MR Imaging
MR imaging was performed with a 1.5-T system (Magnetom Vision; Siemens, Erlangen, Germany) by using a four-element body phased-array coil for signal detection. All patients underwent axial and, optionally, coronal, sagittal, or both coronal and sagittal MR examinations with the following sequences: (a) breath-hold T1-weighted fast low-angle-shot (FLASH) MR imaging (166/5 [repetition time msec/echo time msec], 256 x 192 matrix, 260–400-mm field of view, one signal acquired, 22-second acquisition time); (b) T2-weighted fast spin-echo MR imaging (4,200/120; echo train length, 15; 256 x 192 matrix; 260–400-mm field of view; three to five signals acquired; 3-minute, 42-second acquisition time); (c) breath-hold multiphase 3D gadolinium-enhanced MR angiography (see next paragraph); and (d) breath-hold, contrast-enhanced, T1-weighted FLASH MR imaging (with parameters identical to those for precontrast T1-weighted imaging).
Multiphase 3D Gadolinium-enhanced MR Angiography
We adhered to a strict 3D gadolinium-enhanced MR angiography protocol to minimize errors derived from contrast material administration and to optimize the data acquisition. Therefore, the following considerations were taken into account.
1. To examine systematically nearly the entire liver in a 3D data set, 64 partitions were acquired in the coronal plane with an effective section thickness of only 2.5 mm.
2. To obtain both high-spatial-resolution and high-temporal-resolution 3D gadolinium-enhanced MR angiographic images (5/2, 300–450-mm field of view, 64 repetitions, 27-second acquisition time), a fast T1-weighted spoiled FLASH sequence in which the data are acquired in central k space was used.
3. To control for adequate timing of injection of the bolus in the arterial phase, the transport time from the site of injection to the celiac trunk was measured in all patients by using a single-section turbo FLASH sequence with 60 consecutive repetitions (one image per second) during fast administration of 1 mL of gadopentetate dimeglumine (Magnevist; Schering, Berlin, Germany). Thereafter, a signal intensity–time curve was evaluated to account for the mean arrival of the bolus in each patient (Fig 1). The time delay for the start of the 3D gadolinium-enhanced MR angiographic sequence was calculated in accordance with the delay in a previous report (8) as the time from the beginning of the contrast material injection to the time of 50% of peak enhancement.
|
5. To obtain functional lesion information, three repetitive image acquisitions were performed during a breath hold at the arterial, portal venous, and late venous phases. Figure 2 gives an overview of the sequence timing.
|
For quantitative assessment, the percentages of enhancement of the lesion and of the hepatic parenchyma were calculated for all three phases as follows: Enhancement = (Spost - Spre)/Spre x 100, where S is signal intensity and post and pre indicate postcontrast and precontrast. The contrast-to-noise ratio on the contrast-enhanced images was calculated as follows: Contrast-to-noise ratio = (Slesion - Sliver)/noise, where Slesion is the signal intensity of the lesion and Sliver is the signal intensity of the normal hepatic parenchyma. Noise (mean value) was measured by placing the region of interest outside the patient in the phase-encoding direction. The signal intensity and contrast-to-noise ratio were calculated separately in the center and periphery of the hepatic lesion.
Semiquantitative comparison of the lesion-liver visibility at all three phases was based on a five-point scale: "1" indicated that the lesion was strongly hypointense relative to normal hepatic tissue; "2," that the lesion was moderately hypointense; "3," that the lesion was isointense; "4," that the lesion was moderately hyperintense; and "5," that the lesion was strongly hyperintense.
Moreover, lesion localization, delineation, arterial feeding, and venous drainage were recorded.
Comparative Evaluation and Functional Characterization of 3D MR Angiographic and Conventional MR Imaging Sequences
Conventional MR imaging and 3D gadolinium-enhanced MR angiographic sequences were evaluated prospectively and independently by three of the authors (H.H., S.O.S., M.E.) without knowledge of the US, CT, or histopathologic findings. In 32 of the 35 suspected lesions, the three authors concurred on the lesion classification. In three cases, agreement was reached by means of a consensus of these authors.
Lesion-liver visibility in each phase was compared between the 3D gadolinium-enhanced MR angiographic images and the T2- and T1-weighted precontrast and the postcontrast images. Comparative evaluation of the images was based on a five-point scale: "1" indicated that the lesion was not visible on 3D gadolinium-enhanced MR angiographic images; "2," that the lesion was less visible; "3," that the lesion was equally visible; "4," that the lesion was more visible; and "5," visible only on 3D gadolinium-enhanced MR angiographic images.
Comprehensive Lesion Assessment
The functional lesion characterization on the 3D gadolinium-enhanced MR angiographic images was based on the spatial variations and the evolution of contrast enhancement during all three phases. This characterization was compared with that in the conventional MR imaging sequences. The final diagnoses made on the basis of the conventional and 3D gadolinium-enhanced MR angiographic images were compared to the diagnoses made on the basis of the bioptic, surgical, or both bioptic and surgical specimen findings. There were four final histopathologic diagnoses: hemangioma, focal nodular hyperplasia, metastasis, and hepatocellular carcinoma.
Statistics
The statistical analysis was performed by using SAS software (SAS for Windows, version 6.10, Cary, NC). The paired Student t test was used to determine whether differences between different MR imaging sequences were significant. A P value of less than .05 was considered to indicate a significant difference.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Signal-to-Noise Ratio and Contrast-to-Noise Ratio
The mean signal-to-noise ratios and contrast-to-noise ratios calculated for normal hepatic parenchyma and hepatic lesions at all three phases are presented in Table 1.
|
|
The results of hepatic lesion characterization on 3D gadolinium-enhanced MR angiographic images and conventional MR images are compared with the bioptic and surgical specimen findings in Table 3. All 35 hepatic lesions were correctly characterized on 3D gadolinium-enhanced MR angiographic images, which was significantly (P < .01) better than the 15 (43%) characterized on precontrast T1-weighted images, the 24 (69%) characterized on T2-weighted images, and the 26 (74%) characterized on postcontrast T1-weighted images.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The use of a short repetition time in a spoiled gradient-echo sequence helps to suppress unenhanced surrounding tissue. This is a well-known effect that has been described for 3D MR angiography as improving the contrast between vascular structures and background tissue during the first pass of gadolinium (5–7). In our present study, we found this effect particularly helpful in improving the contrast between early enhancing lesions and the surrounding hepatic tissue in the arterial phase. Each hepatic lesion type could be identified and characterized on the 3D gadolinium-enhanced MR angiographic images by analyzing the spatial variations and evolution of contrast enhancement. For example, the very early and strong contrast enhancement of hepatocellular carcinomas or the ringlike pattern of enhancement in metastatic lesions would have been obscured from liver parenchyma in the portal venous or late venous phase. Metastatic lesions showed an early rim of contrast enhancement and, at a later stage, central filling with a characteristically late peripheral washout. This early rim of enhancement is different from the peripheral nodular cotton-wool–like appearance of hemangiomas in the arterial phase.
Due to the short imaging times, acquisition of a complete 3D data set with high spatial and high temporal resolution depicting spatial variations in hepatic lesion contrast enhancement could be combined with the evolution of contrast enhancement provided by multiphase acquisition. Our introduced protocol, with a 27-second breath hold and a 23-second interval of free breathing, was feasible for all patients and ensured reproducible acquisition of images in the arterial, portal venous, and late venous phases. This combined information significantly improved lesion detection and lesion characterization as compared with information provided with conventional T1- and T2-weighted MR sequences.
Recent articles (2,11) have described the enhancement characteristics of hemangiomas at dynamic gadolinium-enhanced MR imaging as similar to those at dynamic CT, with early peripheral enhancement and complete hyperintense or isointense fill-in. We found that all hemangiomas could be characterized by analyzing the spatial distribution of contrast enhancement with a hyperintense nodular cotton-wool–like rim or reticular pattern in the arterial phase and a progressive, hyperintense, and centripetal enhancement in the portal venous or late venous phase.
These study findings demonstrate that metastatic lesions may initially enhance with peripheral hyperintensity and fill in over time with progressive centripetal hyperintensity, which potentially simulates the enhancement of hemangiomas. However, this potential pitfall is avoided by assessing the evolution of contrast enhancement during three phases when the initial hyperintense peripheral ring becomes isointense during the portal venous and strongly hypointense during the late venous phase.
One of the principle values of 3D gadolinium-enhanced MR angiographic imaging is that significantly more malignant hepatic lesions not visible at conventional MR imaging could be detected. In particular, more hepatocellular carcinomas and metastases were detected and characterized on the 3D gadolinium-enhanced MR images. Hepatocellular carcinoma is considered the prototypic hypervascular hepatic neoplasm (11). Enhancement in hepatocellular carcinomas may be accentuated as a result of hypertrophied hepatic arterial flow in patients with both cirrhosis and portal hypertension.
MR study findings have shown that dynamic two-dimensional gadolinium-enhanced multisection imaging of various hepatic lesions is a feasible technique that significantly improves lesion detection and characterization as compared with conventional T2- and postcontrast T1-weighted images (2,11,12). However, the results of this preliminary study show that hepatic lesion detection and characterization can even be further optimized by using a multiphase breath-hold 3D gadolinium-enhanced MR angiographic approach.
The following advantages could be identified with this technique: The 3D gadolinium-enhanced MR angiographic approach allowed us to obtain a complete 3D data set with 64 contiguous sections and an effective, nearly isotropic voxel size of only 1.8 x 2.3 x 2.5 mm, which resulted in a sufficiently large liver volume coverage that is substantially better than the eight to 19 sections of 8–10-mm section thickness reported in the former two-dimensional dynamic MR imaging studies (4,11). It is well conceivable that the detection of small hepatic lesions (<8–10 mm) will therefore be markedly improved with the 3D gadolinium-enhanced MR angiographic approach.
In addition, maximum intensity projections and multiplanar reconstructions can be obtained in multiple planes, which improves lesion delineation. This also potentially enables the detection of the feeding and draining vessels that were observed in a substantial number of hepatic lesions. This additional information may help to guide surgical resection or embolization of hepatic lesions, especially in cases of hepatocellular carcinoma. Another strength was the optimized timing of contrast material administration in combination with fast acquisition of the 3D data sets. Individualized bolus timing accounted for optimal lesion contrast during the arterial, portal, and late venous phases.
Despite the long breath holds of 27 seconds, we did not find image degradation due to breathing artifacts, which may occur during non-breath-hold two-dimensional dynamic MR imaging, in any of the patients. With centrically reordered 3D sequences as used here, possible image degradation due to incomplete breath holds affects only peripheral k-space data, which produces even less breathing artifact. By synchronizing imaging with the arterial phase of bolus injection, it is possible to begin the imaging with the collection of central k-space (centric-encoding) data (8).
Although our preliminary results indicate that 3D gadolinium-enhanced MR angiographic imaging has a potential role in hepatic lesion detection and characterization, several limitations of this study and our approach must be mentioned. First, histologic proof could not be obtained on a rigorous lesion-to-lesion basis, because the decision to perform biopsy was made by the referring clinician and we had no influence on his or her decision. However, multiple coexistent lesions occurred in a minority of patients. In addition, we assumed the same histology only if all hepatic lesions in a patient exhibited identical signal intensity characteristics in the arterial, portal venous, and late venous phases. Furthermore, to gain histologic specimens from multiple hepatic lesions is not easy and not practical.
Second, our results are certainly biased, because all malignant hepatic lesions were at least in part well hypervascularized. Therefore, lesion characteristics of hypovascular malignant lesions cannot be ascertained. However, it is well conceivable that hypovascular lesions can also be detected and characterized by analyzing the spatial variations and the evolution of contrast enhancement. Short, high-frequency pulses result in inferior section profiles, with the exclusion of the outermost sections from a diagnostic image analysis. Furthermore, the study group examined is small, therefore excluding this 3D gadolinium-enhanced MR angiographic method presently as a general, valid approach in hepatic lesion imaging. However, the current findings strongly indicate that the combined morphologic and functional 3D gadolinium-enhanced MR angiographic approach is a technically feasible approach that significantly improves hepatic lesion detection and characterization and thus justifies publication at this early stage. Furthermore, recent technical improvements with even shorter repetition times and echo times and k-space interpolation probably result in either images with higher temporal resolution or higher spatial resolution.
Multiphase breath-hold gadolinium-enhanced 3D imaging is a feasible and robust technique that significantly improves the morphologic detection of benign or malignant lesions during the early arterial phase and that further improves the functional characterization of hepatic lesions during the arterial, portal venous, and late venous phases, without substantial prolongation of a standard liver examination.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: FLASH = fast low-angle shot 3D = three-dimensional
Author contributions: Guarantors of integrity of entire study, G.v.K., H.H., S.O.S.; study concepts and design, S.O.S., H.H.; definition of intellectual content, H.H., P.M.; data acquisition, H.H., P.M.; data analysis, H.H., M.E., S.O.S.; statistical analysis, H.H., M.E., S.O.S.; manuscript preparation, H.H., S.O.S., M.V.K.; manuscript editing, H.H., S.O.S.; manuscript review, G.v.K.
Received February 16, 1998;
revision requested April 6, 1998; revision received May 13, 1998;
accepted July 20, 1998.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Ward, P. J. Robinson, J. A. Guthrie, S. Downing, D. Wilson, J. P. A. Lodge, K. R. Prasad, G. J. Toogood, and J. I. Wyatt Liver Metastases in Candidates for Hepatic Resection: Comparison of Helical CT and Gadolinium- and SPIO-enhanced MR Imaging Radiology, October 1, 2005; 237(1): 170 - 180. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Goehde, P. Hunold, F. M. Vogt, W. Ajaj, M. Goyen, C. U. Herborn, M. Forsting, J. F. Debatin, and S. G. Ruehm Full-Body Cardiovascular and Tumor MRI for Early Detection of Disease: Feasibility and Initial Experience in 298 Subjects Am. J. Roentgenol., February 1, 2005; 184(2): 598 - 611. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. C. Lauenstein, S. C. Goehde, C. U. Herborn, M. Goyen, C. Oberhoff, J. F. Debatin, S. G. Ruehm, and J. Barkhausen Whole-Body MR Imaging: Evaluation of Patients for Metastases Radiology, October 1, 2004; 233(1): 139 - 148. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W Ajaj, S C Goehde, N Papanikolaou, G Holtmann, S G Ruehm, J F Debatin, and T C Lauenstein Real time high resolution magnetic resonance imaging for the assessment of gastric motility disorders Gut, September 1, 2004; 53(9): 1256 - 1261. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Sahani, A. Mehta, M. Blake, S. Prasad, G. Harris, and S. Saini Preoperative Hepatic Vascular Evaluation with CT and MR Angiography: Implications for Surgery RadioGraphics, September 1, 2004; 24(5): 1367 - 1380. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. McKenzie, D. Lim, B. J. Ransil, M. Morrin, I. Pedrosa, E. N. Yeh, D. K. Sodickson, and N. M. Rofsky Shortening MR Image Acquisition Time for Volumetric Interpolated Breath-hold Examination with a Recently Developed Parallel Imaging Reconstruction Technique: Clinical Feasibility Radiology, February 1, 2004; 230(2): 589 - 594. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Huppertz, T. Balzer, A. Blakeborough, J. Breuer, A. Giovagnoni, G. Heinz-Peer, M. Laniado, R. M. Manfredi, D. G. Mathieu, D. Mueller, et al. Improved Detection of Focal Liver Lesions at MR Imaging: Multicenter Comparison of Gadoxetic Acid-enhanced MR Images with Intraoperative Findings Radiology, January 1, 2004; 230(1): 266 - 275. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Erden, I. Erden, S. Karayalcin, and C. Yurdaydin Budd-Chiari Syndrome: Evaluation with Multiphase Contrast-Enhanced Three-Dimensional MR Angiography Am. J. Roentgenol., November 1, 2002; 179(5): 1287 - 1292. [Full Text] [PDF]
|
|
|
|
|
|
T. C. Lauenstein, S. C. Goehde, C. U. Herborn, W. Treder, S. G. Ruehm, J. F. Debatin, and J. Barkhausen Three-Dimensional Volumetric Interpolated Breath-Hold MR Imaging for Whole-Body Tumor Staging in Less Than 15 Minutes: A Feasibility Study Am. J. Roentgenol., August 1, 2002; 179(2): 445 - 449. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. H. Coulam, F. P. Chan, and K. C. P. Li Can a Multiphasic Contrast-Enhanced Three-Dimensional Fast Spoiled Gradient-Recalled Echo Sequence Be Sufficient for Liver MR Imaging? Am. J. Roentgenol., February 1, 2002; 178(2): 335 - 341. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Glockner Three-dimensional Gadolinium-enhanced MR Angiography: Applications for Abdominal Imaging RadioGraphics, March 1, 2001; 21(2): 357 - 370. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. T. Lavelle, V. S. Lee, N. M. Rofsky, G. A. Krinsky, and J. C. Weinreb Dynamic Contrast-enhanced Three-dimensional MR Imaging of Liver Parenchyma: Source Images and Angiographic Reconstructions to Define Hepatic Arterial Anatomy Radiology, February 1, 2001; 218(2): 389 - 394. [Abstract] [Full Text]
|
|
|
|
|
|
V. S. Lee, M. T. Lavelle, N. M. Rofsky, G. Laub, D. M. Thomasson, G. A. Krinsky, and J. C. Weinreb Hepatic MR Imaging with a Dynamic Contrast-enhanced Isotropic Volumetric Interpolated Breath-hold Examination: Feasibility, Reproducibility, and Technical Quality Radiology, May 1, 2000; 215(2): 365 - 372. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||
