| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vascular and Interventional Radiology |
1 From the Department of Radiology-MRI, New York University Medical Center, 530 First Ave, New York, NY 10016. Received June 9, 2002; revision requested August 14; revision received August 29; accepted October 24. Address correspondence to V.S.L. (e-mail: vivian.lee@med.nyu.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: A series of 100 patients underwent routine MR imaging of the upper abdomen at 1.5 T; imaging included multiple dynamic contrast-enhanced fat-suppressed transverse three-dimensional spoiled gradient-echo acquisitions (3.6–4.5/1.5–1.9 [repetition time msec/echo time msec], 12° flip angle). Arterial phase acquisitions were obtained during suspended respiration at end expiration (n = 50) or at end inspiration (n = 50), and venous phase acquisitions were obtained at the opposite respiratory phase. Two radiologists, blinded to patient identity and the phase of respiration, independently assessed the degree of narrowing on reconstructed oblique sagittal images. Radiologists reached consensus in 97 patients, who formed the cohort for this study. The percentage of stenosis of the celiac artery relative to its origin and the angle formed by the proximal celiac artery and the aorta were also measured in all patients. This angle and the arcsine transformation of the percentage of stenosis were compared for the two respiratory phases by using a paired Student t test. 
2 analysis was used to evaluate whether the degree of narrowing was independent of the breath-hold protocol that was used.
RESULTS: In total, 55 (57%) of 97 patients had at least mild artery narrowing at end expiration, of whom 40 (73%) had less narrowing at end inspiration and 11 (20%) had no change. The average percentage of stenosis at end expiration (21% ± 16) was significantly higher than that at end inspiration (11% ± 11; P < .001). At end expiration, the average celiac artery angle was significantly lower in patients with mild to severe narrowing (41° ± 19) than in those without narrowing (50° ± 19; P < .03).
CONCLUSION: Accentuation of celiac artery compression at end expiration can give rise to a potential pitfall of breath-hold abdominal imaging. When compression is suspected, imaging should be performed during inspiration.
© RSNA, 2003
Index terms: Arteries, celiac, 951.92 • Arteries, stenosis or obstruction, 951.721 • Magnetic resonance (MR), vascular studies
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The character and position of the median arcuate ligament are highly variable. In an autopsy series of 75 cases, Lindner and Kemprud (6) observed that the celiac artery origin was at or above the median arcuate ligament in 33% (25 of 75) of cases. Results of conventional angiographic studies dating to the early 1970s showed that the positions of the median arcuate ligament, celiac artery, and aorta vary considerably during respiration and that median arcuate ligament compression is often accentuated during expiration (7,8) (Fig 1).
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MR Imaging
All patients underwent imaging at 1.5 T (Quantum or Vision; Siemens Medical Systems, Erlangen, Germany) with a torso phased-array coil. Prior to imaging, a 22-gauge intravenous catheter was placed in an arm vein and was attached to an MR-compatible power injector (Spectris; Medrad, Pittsburgh, Pa). In addition to routine breath-hold transverse T1-weighted gradient-echo imaging and short inversion time inversion-recovery fast spin-echo imaging, a series of 3D spoiled gradient-echo acquisitions was obtained in the liver and upper abdomen both before and after intravenous administration of contrast material. Evaluation of the 3D data sets formed the basis of this study.
All 3D MR imaging was performed with a transverse 3D radiofrequency-spoiled gradient-echo sequence (3.6–4.5/1.5–1.9 [repetition time msec/echo time msec], 12° flip angle) with an intermittent (once per partition loop) fat-saturation pulse, which was modified from a sequence described in the literature (11,12). By using a field of view of 300–425 mm and an image matrix of 256 x 80–135, with a rectangular field of view optimized to patient body habitus, in-plane spatial resolution of 2.0–3.3 x 1.2–1.7 mm was attained. Asymmetric echo sampling in the section-select direction was used to obtain 40–64 data points, which were then interpolated with zero filling to produce 80–128 partitions. We used slab thicknesses ranging from 160 to 260 mm to ensure full coverage of the liver (including the celiac artery and superior mesenteric artery) in all cases, thereby yielding a partition thickness of 1.5–2.5 mm. Acquisition times were no more than 30 seconds (average, 21.5 seconds ± 2.1; range, 16–30 seconds).
In all patients, an unenhanced 3D acquisition was performed first and was followed by a timing examination performed according to a modification of a method described in the literature (13). In brief, a 1-mL test dose of gadopentetate dimeglumine (Magnevist; Berlex, Wayne, NJ) was administered intravenously and was followed by a 20-mL saline flush; both were injected at a rate of 2 mL/sec with a power injector. Repeated transverse two-dimensional magnetization-prepared spoiled gradient-echo (turbo fast low-angle shot, 5.2/3, 15° flip angle, 300-msec inversion time) imaging, which was synchronized with the injection, was performed through the aorta at a rate of one image every 2 seconds. The time to peak aortic contrast enhancement was used to determine circulation time (transit time from arm vein to aorta).
For contrast-enhanced imaging, all patients were administered 19 mL of gadopentetate dimeglumine, which resulted in an average dose of 0.12 mmol/kg (range, 0.07–0.18 mmol/kg). Four 3D acquisitions were subsequently obtained in succession; the first was timed for arterial enhancement on the basis of the test-bolus timing examination (13). Two additional 3D acquisitions were obtained, after a 45-second and an additional 90-second delay. The first three contrast-enhanced imaging sequences were performed during suspended respiration either at end expiration (n = 50) or at end inspiration (n = 50). After an additional 20-second delay, a fourth 3D imaging sequence was performed, in which the patient was instructed to suspend respiration at the opposite phase of respiration. The delays between each 3D acquisition allowed patients to breathe normally and to prepare for the next breath hold. Breath-holding instructions and techniques were reviewed with each patient before the start of the examination and again during the examination.
Image Analysis
In all patients, the arterial phase and delayed phase volumetric 3D acquisitions were reformatted in the sagittal or oblique sagittal plane in order to view the origin and first several centimeters of the celiac artery through the midline of the vessel. This was performed by two investigators (J.N.M., A.G.S.T.) who worked together at a satellite workstation and used commercially available software (Siemens). The typically acute angle formed between the aorta and the proximal celiac artery at its origin was also measured at the workstation. Reformatted images were printed on film. The craniocaudal diameter of the celiac artery was measured on the film images by using a vernier caliper (Fowler, Newton, Mass) at several locations: the origin, the narrowest point, and just beyond the narrowest point. These points were measured in all patients and at both respiratory phases. The diameter was measured orthogonal to the axis of the vessel. The window and level of film images were assigned subjectively by the two investigators. The percentage of stenosis was calculated by using the diameter at the point of greatest narrowing divided by the diameter at the celiac artery origin. If ostial stenosis was present, the percentage of stenosis was determined by using the diameter of the vessel distal to the stenosis as reference. Each of two radiologists (G.A.K., J.A.B.), who were blinded to patient identity and to the phase of respiration, graded the degree of narrowing on the following scale: none, minimal, mild, mild to moderate, moderate, and severe. Graders were instructed that the five abnormal grades should reflect quintiles of stenosis (1%–20%, 20%–40%, 40%–60%, 60%–80%, >80%). Prior to grading, the two radiologists also reviewed a range of clinical studies together to reach agreement on definitions of each grading category. Once independent grading was performed, discrepancies were resolved by consensus, and consensus readings were used for data analysis.
The following additional findings were also considered by the two graders, and consensus observations were used: poststenotic dilatation in the celiac artery, presence of collateral vessels, presence of atheromatous plaque of the aorta, and stenosis of the superior mesenteric artery.
Data Analysis
Of the 100 patients who underwent imaging, three were excluded from analysis for reasons that included inability of the patient to suspend respiration (n = 1) and inability of the graders to reach consensus in interpretation of the images (n = 2). Results from the remaining 97 patients (51 men, 46 women; average age, 56 years ± 13) were used for analysis. The prevalence of celiac artery narrowing at end-inspiratory and end-expiratory imaging was calculated. The celiac artery-aorta angles for the two phases of respiration were compared by using a paired Student t test. The distribution of data on the percentage of stenosis was skewed and therefore subject to arcsine transformation in order to approximate a normal distribution before application of a paired Student t test to compare the stenosis for the two phases of respiration. Patients were also assigned to one of two subgroups: those with at least mild celiac artery narrowing during end expiration and those with minimal or no narrowing. Patient age and weight and the celiac artery-aorta angle were compared between the two groups by using a Student t test. In 49 of the 97 patients, end-expiratory imaging was performed during arterial phase enhancement; this was followed by end-inspiratory imaging performed during delayed venous phase enhancement. In the remaining 48 patients, this protocol was reversed. To evaluate whether observations of celiac artery narrowing (subjective categories) were independent of the protocol used, a 2 analysis was performed.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
2 analysis, the presence of celiac artery narrowing during end-expiratory imaging was determined to be independent of the phase of vascular enhancement (P > .2). The percentage of celiac artery stenosis, which was measured relative to the diameter of the celiac artery origin, averaged 21% ± 16 during end expiration. During end inspiration, the percentage of celiac artery stenosis was significantly lower and averaged 11% ± 11 (P < .001). Images in a total of 22 patients (23%) demonstrated a change in the percentage of stenosis between inspiratory and expiratory imaging that measured greater than or equal to 20%; in only three of these patients was the degree of stenosis greater at end inspiration than at end expiration (Fig 4).
|
The data were further analyzed by assigning patients to one of two groups: those with at least mild celiac artery narrowing during end expiration (n = 55, 27 men and 28 women) and those with minimal or no narrowing (n = 42, 24 men and 18 women). In those with at least mild narrowing, the average percentage of celiac artery stenosis was significantly higher at end expiration than at end inspiration (P < .01) (Table 2). There was no significant difference between average age of the two groups: 56.2 years ± 12.0 for those with narrowing versus 55.4 years ± 14.6 for those without narrowing (P = .8). Those patients with at least mild narrowing weighed on average slightly less (73.9 kg ± 17.3, versus 75.9 kg ± 14.7 for those with minimal or no narrowing), although this difference did not achieve statistical significance (P > .1).
|
Additional findings were uncommon. Poststenotic dilatation was recorded in three patients: two with severe narrowing at end expiratory imaging and one with moderate narrowing. In all, the dilatation persisted at end-inspiratory imaging. In one patient, the severe narrowing remained unchanged at inspiratory imaging. In the remaining two patients, the narrowing decreased to mild narrowing at inspiration. Collateral vessels were not identified in any patients. Atheromatous plaque in the aorta was noted in five patients; in one, the plaque extended into the orifice of the celiac artery and resulted in fixed severe narrowing at end-inspiratory and end-expiratory imaging. The superior mesenteric artery was without stenosis in all patients.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The accentuation of celiac artery compression during deep expiration has been identified at lateral angiography performed in patients with abdominal bruit or abdominal symptoms (7). In one case, Erden et al (15) used Doppler sonography to document increases in flow velocities during end expiration. In 1973, Reuter and Bernstein (8) studied eight patients in whom surgical clips had been placed along the free edge of the median arcuate ligament during abdominal operations. By observing the respiratory motion of the clips at fluoroscopy, the investigators found that the ligament, aorta, and celiac artery moved caudad during inspiration and cephalad during expiration in all patients (Fig 1). The relative motion of each structure caused the celiac artery and median arcuate ligament to move toward each other during expiration, which accentuated celiac artery compression in most cases.
To achieve improved reproducibility, MR imaging of the chest and abdomen and MR angiography of the torso are frequently performed during suspended respiration at end expiration (9,10). Reproducibility may be important in several clinical scenarios. In the evaluation of renal masses, for example, the detection of enhancing components may depend on accurate subtraction if the lesions are small or already hyperintense on unenhanced images (16). MR angiographic images are typically reconstructed by using maximum intensity projection reconstructions of subtracted data sets; subtraction improves image quality by minimizing signal contributions from background tissues (17). Although patients are typically able to suspend respiration for a shorter duration at end expiration, the improved consistency of end-expiratory imaging has been thought to outweigh this disadvantage.
The results of our study suggest an additional disadvantage to end-expiratory imaging of the abdomen. In 97 patients who were referred to undergo abdominal MR imaging for reasons other than suspected mesenteric ischemia, 46 (46%) were found to have mild to moderate celiac artery narrowing and nine (9%) were found to have severe narrowing at MR imaging during suspended respiration at end expiration; the degree of narrowing lessened considerably in the majority of patients (40 of 55, 73%) when imaging was performed at end inspiration. It is possible that some of these patients with substantial compression had unreported clinical symptoms attributable to celiac insufficiency, but for the majority of patients, the celiac artery compression should be considered a false-positive finding. The paucity of additional findings such as poststenotic dilatation (n = 3, 3%) supports this assessment.
There are recognized limitations to our study. First, our study was performed retrospectively, which introduces possible bias in the interpretation. To minimize this effect, the two independent readers were blinded to the phase of respiration and to patient identity. Second, to assess the caliber of the celiac artery, we performed sagittal or oblique sagittal reconstructions of transverse 3D source data. The near isotropic pixels achieved with interpolation allowed for reconstructed images with comparable apparent in-plane resolution (11,12). Third, to compare different phases of respiration, we used the first and fourth contrast-enhanced acquisitions, which resulted in a decline in contrast enhancement on the latter acquisitions. However, because the order of breath holding was reversed for the second group of 50 patients in this study, the effects of reduced image quality did not bias our results. Fourth, we were unable to verify the degree of cooperation with breath-holding instructions that patients were able to achieve. Our technologists are skilled in instructing and practicing breath-holding techniques with patients, and in all cases, breath-holding instructions were reviewed both before and during the examination. As discussed previously, although we excluded patients known to have or suspected of having mesenteric ischemia, patients with occult cases may have inadvertently been included in this series. If this occurred, the number of patients would likely be small, and the effects on our results would be negligible. Last, conventional contrast-enhanced angiography, which is considered a reference standard for vascular disease, was not available in this population.
We conclude that when celiac artery compression syndrome is suspected on the basis of end-expiratory imaging, findings should be correlated with clinical history and results of physical examination and, if clinically indicated, confirmed with imaging at end inspiration. In patients suspected of having mesenteric ischemia, imaging of the abdominal aorta should be performed exclusively at end inspiration. Techniques such as MR image measurements of blood-oxygen saturation in mesenteric veins may be useful as a means of determining the physiologic significance of celiac artery narrowing, and these techniques deserve further investigation (18).
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, V.S.L.; study concepts and design, V.S.L., J.N.M., A.G.S.T.; literature research, J.N.M., V.S.L.; clinical studies, J.N.M., A.G.S.T., P.V.P., J.A.B.; statistical analysis, V.S.L.; manuscript definition of intellectual content, V.S.L., J.C.W.; data acquisition and analysis/interpretation, all authors; manuscript preparation, editing, revision/review, and final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E Rosenthal, S R Sangle, P Taylor, M A Khamashta, G R V Hughes, and D P D'Cruz Treatment of mesenteric angina with prolonged anticoagulation in a patient with antiphospholipid (Hughes) syndrome and coeliac artery stenosis. Ann Rheum Dis, October 1, 2006; 65(10): 1398 - 1399. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
), the degree of stenosis increased at end inspiration, while the remaining 19 patients (
) had markedly less stenosis at end inspiration.