| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Gastrointestinal Imaging |
1 From the Department of Radiology, Thomas Jefferson University Hospital, Philadelphia, Pa. Received March 22, 1999; revision requested May 10; revision received July 26; accepted August 12. Address correspondence to K.I., Department of Radiology, Yamaguchi University School of Medicine, 1-1-1 Minami-kogushi, Ube, Yamaguchi 755-8505, Japan (e-mail: itokatsu@po.cc.yamaguchi-u.ac.jp).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Seventy-two patients (37 control subjects, 35 patients with cirrhosis) underwent imaging with a high-performance–gradient (25 mT/m) system.
RESULTS: In the 37 subjects in the control group, the MPV was visualized in 37; splenic vein (SV), in 37; superior mesenteric vein (SMV), in 37; inferior mesenteric vein (IMV), in 35; posterior superior pancreaticoduodenal vein (PSPDV), in 35; gastrocolic trunk (GT), in 34; right gastroepiploic vein (RGEV), in 31; right colic vein, in 30; anterior superior pancreaticoduodenal vein, in 22; middle colic vein (MCV), in 29; and first jejunal vein (FJV), in 36. Satisfactory visualization (mean ratings of 2 or higher) was achieved in the MPV, SV, SMV, IMV, PSPDV, GT, RGEV, and FJV in the control group. Mean diameters of the SV, SMV, MCV, and FJV were significantly larger in the cirrhosis group than in the control group (P < .001, P < .001, P =.048, and P = .002, respectively).
CONCLUSION: Thin-section 3D contrast-enhanced dynamic fat-suppressed MR imaging can facilitate precise visualization of the MPV and its tributaries. Dilatation of the tributaries may be a nonspecific secondary finding that is suggestive of cirrhosis.
Index terms: Liver, cirrhosis, 761.794 • Liver, MR, 761.121412, 761.121415, 761.12143, 761.794 • Magnetic resonance (MR), contrast enhancement, 761.12143, 957.12943 • Magnetic resonance (MR), fat suppression, 761.121415, 957.129415 • Magnetic resonance (MR), thin-section, 957.12918 • Magnetic resonance (MR), three-dimensional, 957.12917 • Portal vein, MR, 957.12917, 957.12918, 957.129412, 957.129415, 951.12943
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Recent improvements in gradient technology and software design have allowed implementation of thin-section three-dimensional (3D) contrast-enhanced dynamic fat-suppressed MR imaging in abdominal studies. This technique has the potential to improve the visualization of the main portal vein (MPV) and its tributaries, including the small peripancreatic veins, at routine contrast-enhanced abdominal MR imaging because of the good continuity between individual sections, suppression of high signal intensity from intraabdominal fat, and wide coverage during imaging.
Previous findings suggest that the size of the MPV is altered in patients with cirrhosis because of the change in the hemodynamics of the portal venous system (5,6). Improved visualization of the vessels may facilitate detection of dilated portal venous branches in cirrhosis and, thus, contribute to the discrimination between control subjects and patients with cirrhosis. This may lead to a more precise assessment of the severity of portal hypertension in patients with cirrhosis. The purpose of this study was to evaluate the visibility of the MPV and its tributaries in control subjects at thin-section 3D contrast-enhanced dynamic fat-suppressed MR imaging and to determine whether this technique can provide useful additional information in the evaluation of patients with cirrhosis.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Patients were excluded from this study if they had (a) a history of acute or chronic pancreatitis (n = 12); (b) malignant tumors in the liver or pancreas (n = 18); (c) thrombus in the MPV, splenic vein (SV), and/or superior mesenteric vein (SMV) (n = 4); or (d) MR images that did not sufficiently cover the vessels to be analyzed (n = 8). Consequently, this study population included 72 patients (42 men, 30 women; age range, 22–87 years; mean age, 53.9 years), 37 in the control group and 35 in the cirrhosis group.
Patients in the control group had undergone MR imaging evaluation of suspected benign hepatic lesions (eg, hepatic cyst, hemangioma or focal fatty sparing), for benign diseases of the biliary tract (eg, stones or strictures), or to exclude metastatic hepatic lesions from gastrointestinal malignancies. Patients in the cirrhosis group had been referred for MR imaging for evaluation of the severity of cirrhosis and portal hypertension, for preoperative evaluation for liver transplantation, and/or for screening or further examination of hepatic lesions that were suggested at imaging with use of other modalities. The diagnosis of cirrhosis was made on the basis of findings at percutaneous liver biopsy (n = 26) or clinical evaluation (n = 9). Cirrhosis was caused by viral infection (hepatitis B [n = 12] or C [n = 23]).
MR Imaging Technique
All examinations were performed with a 1.5-T high-performance–gradient MR imaging system (Signa; GE Medical Systems, Milwaukee, Wis) and a phased-array torso coil. The following pulse sequences were used during routine abdominal MR imaging: (a) T2-weighted fast spin-echo fat-suppressed sequence with a respiratory trigger (4,000/100 [repetition time msec/echo time msec]), (b) single-shot fast spin-echo sequence (
/102), (c) T1-weighted in-phase gradient-echo sequence (120/4.2, single breath hold), (d) T1-weighted opposed-phase gradient-echo (120/2.1, single breath hold), (e) 3D enhanced fast spoiled gradient-echo dynamic fat-suppressed sequence before and after the administration of contrast material. Contrast-enhanced dynamic imaging was performed as a part of the routine evaluation of abdominal disease.
For the purpose of data analysis, only contrast-enhanced images were reviewed. In all patients, a series of transverse 3D contrast-enhanced fast spoiled gradient-echo dynamic fat-suppressed images (spectral inversion at lipids) were obtained. Fat suppression was segmented so that 16–32 excitation pulses followed each spectral inversion pulse. The imaging parameters were as follows: 7–9/2–3/31 (repetition time msec/echo time msec/spectral inversion time msec); flip angle, 20°; matrix, 256 x 128–160; section thickness, 5 mm; receiver bandwidth, ±32 kHz; and number of signals acquired, one-half. A rectangular field of view was used to reduce the number of phase-encoding views. The k space was zero-filled in the section-encoding direction to decrease the increment between sections to 2.5 mm and to double the number of reconstructed imaging sections (7). As a result, a total of 65–75 sections were acquired during a 20–24-second breath hold.
A sagittal test-bolus gradient-echo timing sequence (6–9/1.5–2.0; flip angle, 90°; section thickness, 20 mm) was used to optimize imaging during the arterial phase (8,9). An initial test bolus of 2 mL of gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) was intravenously injected by hand; this injection was followed by a 20-mL normal saline flush. Sequential MR images were obtained at the level of the abdominal aorta. After the circulation time was determined, multiphasic contrast-enhanced dynamic MR imaging was performed before and after the intravenous administration of 0.1 mmol of gadopentetate dimeglumine per kilogram of body weight followed by a 20-mL bolus of normal saline. Dynamic images were acquired during the arterial, portal, and delayed phases. Postprocessing techniques such as maximum intensity projection were not used in this study.
Image Interpretation
MR images were reviewed retrospectively and independently by three radiologists (K.I., R.B., S.M.H.) who were experienced in abdominal MR imaging and who were blinded to the final diagnosis with regard to the presence of cirrhosis. Images from the two groups were randomly mixed. Images were evaluated for the visibility of the MPV and its tributaries, including the SV, SMV, inferior mesenteric vein (IMV), posterior superior pancreaticoduodenal vein (PSPDV), gastrocolic trunk (GT), right gastroepiploic vein (RGEV), right colic vein (RCV), anterior superior pancreaticoduodenal vein (ASPDV), middle colic vein (MCV), and first jejunal vein (FJV).
The visibility of each vessel was graded and recorded by using the following four-point scale: 3 was excellent, 2 was good, 1 was poor, 0 was no visibility. In the MPV and SMV, the ratings were defined as follows: 3 was excellent (vessel was clearly visible as a round or oval structure throughout the entire MPV and SMV), 2 was good (vessel was moderately visible), 1 was poor (vessel was visible, but the image of the vessel was blurred), and 0 was no visibility (vessel was not visible). In the SV, GT, RGEV, RCV, ASPDV, MCV, and FJV, the ratings was defined as follows: 3 was excellent (confluence to a larger vessel [eg, SMV] and distribution to the affected areas [eg, spleen, intestine or pancreas] were clearly visible), 2 was good (vessel was moderately visible), 1 was poor (vessel was visible, but the image of the vessel was blurred), and 0 was no visibility (vessel was not visible). In the IMV and PSPDV, the following rating criteria were used: 3 was excellent (confluence to a larger vessel [eg, SMV or MPV] and the normal course of the vessel were clearly visible), 2 was good (vessel was moderately visible), 1 was poor (vessel was visible, but the image of the vessel was blurred), and 0 was no visibility (vessel was not visible). For data analyses, portal phase (second-pass) images were reviewed.
When the radiologists disagreed about whether the vessel was visible or not, a majority opinion was used as the final decision. For instance, a vessel was determined to be visible when the ratings of at least two readers were 1 or more. For the comparison of the visibility of each vessel between the groups, the mean rating was used. The venous anatomy described in previous studies (10–13) was used as the reference for the evaluation of the MR images (Fig 1). Since a true GT is present in only 72% of cases (14), the segment of the RGEV distal to the ASPDV was defined as the GT (as described by Crabo et al [11]) when an RCV joining these vessels was not identified or when an RCV directly joined the SMV.
|
Statistical Analysis
The assessment of interobserver variability in the interpretation of MR images was performed by means of weighted 
statistics (15). The level of agreement was defined as follows: values of less than 0 were considered to indicate no agreement; values of 0.00–0.40, poor agreement; values of 0.41–0.75, good agreement; and values of 0.76–1.00, excellent agreement. The 
2 test was used to compare the difference of the frequency of each vessel visualization between the groups. The mean scores of visibility of the vessels were compared by using an unpaired two-tailed t test. Comparisons of the diameter of the vessels in both groups were also made by using an unpaired two-tailed t test. A P value of less than .05 was considered to indicate a statistically significant difference.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
values indicated good or excellent agreement (0.61–0.79) in the rating of visibility of the MPV and its tributaries. The frequency of visualization of each vessel in the control and cirrhosis groups is given in Table 1 (Figs 2, 3). There was no statistically significant difference in the frequencies of vessel visualization between the control and cirrhosis groups.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Several technical advantages are attributable to the high frequencies of visualization of the vessels in this study; these advantages include the optimal timing of data acquisition after injection of contrast materials, thickness of the sections, 3D data acquisition, availability of fat-suppression technique, and use of high-performance–gradient system. In our study, 3D volumetric data were acquired during the portal phase in which excellent vascular enhancement was seen, and thin-section (5-mm) images with a 2.5-mm section overlap were reconstructed in the transverse plane with good continuity between individual sections. The small vessel (PSPDV) with the cephalocaudal course was easy to trace continuously (16). The small veins (RGEV, RCV, ASPDV, and MCV) that coursed in the transverse plane were readily recognized because of fewer partial volume effects and/or less section misregistration. Thin-section imaging without an intersection gap is crucial for the visualization of small vessels (17).
Fat suppression can improve the visibility of small veins by eliminating chemical shift artifacts and high signal intensity from intraabdominal fat and by producing greater contrast between the enhancing vessels and the low-signal-intensity fat.
Regarding the small peripancreatic veins (PSPDV and ASPDV), the timing of data acquisition after enhancement with gadolinium-based contrast materials may further improve the visibility of these veins. The frequency of visualization of the small peripancreatic veins correlates approximately with the degree of contrast between the MPV and the pancreas (11). On the portal phase MR images, small peripancreatic veins on the surface of the pancreas were easy to visualize because of their peak enhancement, since they were not obscured by pancreatic tissue. Excellent intravascular contrast enhancement is important in the discrimination of these veins from the adjacent pancreatic parenchyma. Three-dimensional portal phase contrast-enhanced dynamic fat-suppressed MR imaging is ideally suited for the evaluation of small peripancreatic veins.
Previous investigators have indicated that analysis of small peripancreatic veins is valuable in the precise staging of pancreatic carcinoma at CT (10,13,18,19). The present findings indicate that visualization of the small peripancreatic veins (such as the GT, ASPDV, and PSPDV) at MR imaging may be sufficient to allow evaluation and staging of pancreatic carcinoma. Thin-section 3D contrast-enhanced dynamic fat-suppressed MR imaging may be useful in the early detection of vascular invasion by tumors. However, a further prospective study is necessary to determine whether this technique can improve the accuracy of staging of pancreatic carcinoma.
In patients with cirrhosis, several hemodynamic changes are induced in the liver. Portal venous flow into the liver is reduced probably because of increased intrahepatic resistance caused by hepatic fibrosis. In addition, cirrhosis is usually accompanied by intestinal vascular vasodilatation and increased splanchnic blood flow (20). In the current study, mean diameters of the SV, SMV, MCV, and FJV were significantly larger in the cirrhosis group than in the control group. In addition, the mean rating of the visibility of the GT and its branches (RGEV, RCV, ASPDV) tended to be greater in the cirrhosis group. These observations suggest that tributaries of the MPV serve as collateral vessels in the setting of high portal pressure in patients with cirrhosis.
Prior investigators (5) showed that there was a tendency (although not a significant one) toward increased SV and SMV diameters, as measured at angiography, in cirrhotic patients compared with control subjects. Thickening of the colonic wall in cirrhosis, especially in the ascending and/or transverse colon, has been reported at CT evaluation and is probably related to increased colonic venous flow secondary to portal hypertension (21). Therefore, recognition of the dilated tributaries of the MPV may be an additional secondary sign of cirrhosis, although the distribution of the diameters in the two groups overlapped considerably. This sign should be used in conjunction with other intra- and extrahepatic findings of cirrhosis described in prior reports (22,23).
A potential limitation of our study is that conventional angiography was not performed to provide an anatomic reference standard. Since there is no clinical indication for conventional angiography in control patients, its use was not justified. In addition, precise identification of control small peripancreatic veins may be difficult at conventional angiography. Another limitation is that the related arteries were not separately studied to enable a comparison with the venous anatomy. However, even in the confusing cases, it was easy to differentiate the veins from their accompanying arteries in the comparison of the arterial phase and portal phase images. These limitations do not reduce the validity of the basic conclusions from this study.
In summary, small tributaries of the MPV can be delineated on thin-section 3D contrast-enhanced dynamic fat-suppressed MR images by using the optimized techniques we have described. These images can be obtained at routine abdominal MR imaging examinations without the need to perform dedicated contrast-enhanced MR angiography; the images can be used to evaluate pancreatic and hepatic parenchymal and/or focal disease. This technique can facilitate precise visualization of the anatomy of the MPV and its tributaries. Dilatation of the tributaries of the MPV depicted with this technique may be a nonspecific but additional secondary finding suggestive of cirrhosis. Further studies are necessary to determine the actual clinical implications and value of visualization of small portal tributaries with improved MR imaging techniques.
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, K.I., D.G.M.; study concepts, K.I., R.B.; study design, K.I.; definition of intellectual content, K.I., D.G.M.; literature research, K.I.; clinical studies, K.I., D.G.M.; data acquisition, all authors; data and statistical analyses, K.I.; manuscript preparation, K.I.; manuscript editing and review, D.G.M.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Shimada, H. Isoda, T. Okada, T. Kamae, S. Arizono, Y. Hirokawa, and K. Togashi Unenhanced MR Portography With a Half-Fourier Fast Spin-Echo Sequence and Time-Space Labeling Inversion Pulses: Preliminary Results Am. J. Roentgenol., July 1, 2009; 193(1): 106 - 112. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Ito, S. Koike, A. Shimizu, M. Tanabe, C. Jo, M. Miyazaki, and N. Matsunaga Portal Venous System: Evaluation with Unenhanced MR Angiography with a Single-Breath-Hold ECG-Synchronized 3D Half-Fourier Fast Spin-Echo Sequence Am. J. Roentgenol., August 1, 2008; 191(2): 550 - 554. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. H. Wachsberg Inferior Mesenteric Vein: Gray-Scale and Doppler Sonographic Findings in Normal Subjects and in Patients with Portal Hypertension Am. J. Roentgenol., February 1, 2005; 184(2): 481 - 486. [Full Text] [PDF]
|
|
|
|
 |
|
 J K Willmann, D Weishaupt, T Bohm, T Pfammatter, B Seifert, B Marincek, and P Bauerfeind Detection of submucosal gastric fundal varices with multi-detector row CT angiography Gut, June 1, 2003; 52(6): 886 - 892. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
