| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Clinical Observations |
1 Department of Radiology, Interventional Radiology Unit, Rabin Medical Center,
Beilinson Campus, Petah-Tiqva 49100, Israel.
2 Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel.
3 Department of Vascular Surgery, Rabin Medical Center, Petah-Tiqva 49100,
Israel.
4 Department of Thoracic Surgery, Rabin Medical Center, Petah-Tiqva 49100,
Israel.
Received March 19, 2004;
accepted after revision January 28, 2005.
Address correspondence to G. N. Bachar.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. MR angiography appears to be effective and reliable for use as the sole imaging method before endovascular repair of aortic aneurysms in patients with renal impairment.
Keywords: abdominal aortic aneurysm • digital subtraction angiography • endovascular treatment • MR angiography • thoracic aortic aneurysm
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MR Angiographic Technique
MR angiograms were acquired with a 1.5-T imager (Gyroscan Intera, Philips
Medical Systems) and a whole-body phased-array coil. After the sequences had
been localized around the area of the aneurysms, axial-plane gradient-echo
T1-weighted images (slice thickness, 7 mm; TR/TE, 10/4.6; field of view, 375
mm; reformatted of field view (RFOV), 70%; flip angle, 15°; number of
signals averaged, 2) and axial-plane gradient-echo T2-weighted images (slice
thickness, 7 mm; 800/80; field of view, 375 mm; RFOV, 70%; flip angle,
15°; number of signals averaged, 2) were obtained.
A two-chamber power injector, with contrast medium in one chamber and saline in the other, connected to an IV line was used. We injected 30 mL of gadopentetate dimeglumine (Magnevist, Schering) at a flow rate of 1.5-2 mL/sec, regardless of patient weight, followed by 20 mL of normal saline at the same rate. The bolus-chase technique with the live-view option was applied for scan timing. The bolus was followed at a rate of 1 frame per second until it reached the aortic bifurcation. The patients began to hold their breath for 17-25 sec, and a 3D contrast-enhanced breath-hold sequence was activated immediately (4.1/1.2; field of view, 400 mm). In patients with AAAs, one breath-hold contrast-injection scan was obtained, starting from the distal thoracic aorta and ending at the level of the femoral necks. In patients with TAAs, we obtained two contrast-injection scans of the aneurysm and of the distal aorta and pelvis. In the second run, 10 mL of contrast medium and 10 mL of saline were injected at a rate of 1 mL/sec with the same sequence parameters. The scanned volume consisted of 75 interpolated continuous slices, with centric-elliptic k-space data acquisition in the phase-encoding direction through one breath-hold. We included the coronal balanced turbo field-echo sequence of the abdomen (slice width, 7 mm; 5.9/3; flip angle, 90°; field of view, 400 mm; RFOV, 100%) for added vascular and extravascular information.
The source images were evaluated on a computer workstation (Easy Vision, Philips Medical Systems). Subvolume multiplanar reconstruction was performed using maximum intensity projections in different planes. Standard measurements were taken as shown in Figure 1 using the measuring toolbar available in the workstation. Total imaging time was less than 10 min, and the reconstructions and measurements took about 20 min.
|
Although the automatic vessel-tracking software is an accurate sizing method and considered a good standard, technical problems precluded its use in all patients. Measurements were obtained manually using calipers and were compared with the sizing catheter located in the artery.
Because of the renal function impairment, we tried to limit contrast injections to difficult and tortuous aneurysms shown by MR angiography. We obtained these images according to the best angle obtained with MR angiography. The injected volume varied by site, vessel dimensions, and blood creatinine levels and ranged from 80 to 100 mL.
Data Analysis and Statistical Analysis
The MR angiographic measurements were obtained separately and independently
by two experienced radiologists on the basis of the raw data and the axial,
sagittal, and coronal multiplanar reconstruction images and maximum intensity
projections. The radiologists were unaware of the clinical data. Although the
planning of projections in digital subtraction angiography was based on the
angiograms to spare the patients injection of contrast medium, the digital
subtraction angiographic measurements were performed separately by two other
experienced interventional radiologists unaware of the MR angiographic
measurements. In tortuous vessels, the vessel dimensions were determined at
the widest and smallest diameters in the projection judged by the reviewer to
be optimal. In those cases, we tried to use the same projections as for MR
angiography. Whenever there was a discrepancy between the two reviewers, the
final measurement was decided by consensus.
For the purposes of the study, in patients with AAA the reviewers
documented nine orthonormal diameters and two lengths by both MR angiography
and digital subtraction angiography, as shown in
Figure 1. Measurements made by
MR angiography and digital subtraction angiography were plotted using the
Bland-Altman method for analysis of differences in orthonormal diameters and
lengths between the methods. The differences in measurements were plotted
against their means. This method was used to detect any discordant
observations and possible bias or systematic error (Figs.
2 and
3). The SD of the differences
represented the variability between the methods, with a bias of ± 1.96
SD denoting the limits of agreement. In patients with TAAs the reviewers
documented five orthonormal diameters and three lengths, as shown in
Figure 4. Values are expressed
as means and SDs. The mean difference between the MR angiographic and digital
subtraction angiographic measurements was calculated, and the statistical
significance was determined by a two-tailed unpaired Student's t
test. Type I error 
= 0.05 and p < 0.05 were considered
significant.
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
High-quality angiographic images were obtained in all cases. Table 1 compares the measurements obtained with MR angiography and digital subtraction angiography for nine orthonormal diameters and two lengths in patients with AAA. In none of the procedures planned by MR angiography was there a failure to obtain access or a need for surgery. The MR angiographic measurements were similar to the digital subtraction angiographic measurements, and the difference between the two techniques was not statistically significant (Tables 1 and 2). Mean AAA diameter was 53 ± 18 mm by MR angiography and 54 ± 17 mm by digital subtraction angiography (p = 0.43). Corresponding values for the mean length of the infrarenal aorta were 111.9 ± 14.0 mm and 111.4 ± 14.4 mm (mean difference, 0.5 ± 4.7 mm; range, 6.0-11.9 mm; p = 0.73). The abdominal aorta measured 21.5 ± 2.9 mm by MR angiography and 21.1 ± 2.9 mm by digital subtraction angiography at the level of the renal artery ostium and 20.9 ± 2.3 mm by MR angiography and 20.6 ± 2.0 mm by digital subtraction angiography at 15 mm below the renal artery ostium. Values for the small diameter at the right external iliac artery were 7.5 ± 1.0 mm by MR angiography and 7.5 ± 1.1 mm by digital subtraction angiography, respectively (Tables 1 and 2). The range of differences in the measured diameter of the aorta at the proximal extent of the aneurysm was 0.5 mm (2.4%) to 1.6 mm (6.4%). The range of differences in the measured diameter of the aorta at the renal ostium was 0.3 mm (1.1%) to 1.6 mm (6.3%). Regression analysis showed an excellent correlation (R = 0.968; R2 = 0.917) in the diameter of the aorta at the proximal extent of the aneurysm between measurements made with the vessel analysis program with MR angiography and measurements made with digital subtraction angiography (Fig. 5). The Bland-Altman scatter diagram of the difference against the average of the aneurysmal diameter between the two techniques shows an approximately normal distribution. Both MR angiography and digital subtraction angiography showed aneurysmal dilatation (> 15 mm) of the right common iliac artery in two patients and of the left common iliac artery in one patient. The dimensions of the iliac artery aneurysms were correctly shown with MR angiography, as compared with the dimensions shown by digital subtraction angiography. Examples of digital subtraction angiographic and MR angiographic findings are provided in Figures 6A, 6B, 6C, 7A, 7B, and 7C.
|
|
|
|
|
|
|
|
|
Table 2 compares the measurements obtained with MR angiography and digital subtraction angiography for six orthonormal diameters and two lengths in patients with TAAs. The diameter of the aorta at the proximal extent of the aneurysm was 26.9 mm with MR angiography, compared with 27.0 mm with digital subtraction angiography (p = 0.86). The range of measured differences was -1 mm (3.5%) to 1.1 mm (3.5%). The proximal neck of the aneurysm was 12 mm with MR angiography, compared with 12 mm with digital subtraction angiography (p = 0.43). The range of measured differences was -1 mm (3.5%) to 0.7 mm (1.9%).
In all patients, the measurements used for stent planning were based only on MR angiographic images routinely performed before all endovascular procedures. Talent stent grafts (World Medical) were implanted in 11 patients (69%), Exclude stent grafts (Gore) in four (25%), and Zenite stent grafts (Cook) in one (6%).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Digital subtraction angiography with a calibrated catheter provides a 2D display of the approximate length of the aortic aneurysm and the vessel lumen diameters (lumenography). However, on the basis of digital subtraction angiography, the radiologist cannot accurately determine the diameter of the entire aneurysm, including the thrombotic area. Furthermore, digital subtraction angiography is an invasive procedure necessitating hospitalization and is associated with a low, but not negligible, complication rate [13, 14]. Longitudinal measurements can be achieved also with intravascular sonography, but we did not use this technique in the present study. CTA is considered useful for stent-graft planning and is being used increasingly as the sole imaging method for this purpose [4, 5]. Its disadvantages include the need for large volumes of hyperosmolar iodinated contrast material, precluding use in patients with impaired renal function or allergic to iodinated contrast material. MR angiography with gadolinium allows visualization of the aorta and major branches in multiple projections, without risk to renal function or the need to hospitalize the patient. Calibrated catheters, though widely used for obtaining measurements for stent placement in conventional angiography, have a noncentral course in the vessel lumen, especially in large vessels, causing a minor decrease or increase in vessel dimensions. The automatic vessel-tracking software is an accurate sizing method and is considered a good standard. However, we could not use it in all cases for technical reasons. Thus, for clinical practice and stent planning, we compared the MR angiographic measures to the angiographic images with the calibrated catheters.
Our data showed complete matching between measurements obtained by MR angiography and measurements obtained by conventional digital subtraction angiography with a marker catheter. The patients in the present study had impaired renal function, precluding the use of iodinated contrast agents. For this reason, we could not compare MR angiographic and CTA measurements. However, other studies have shown these techniques to agree well when used to determine patient suitability for an endovascular approach [3-5]. Our findings are in line with those of Lutz et al. [4], who used an automated software tool to analyze preoperative CTA and MR angiographic data sets in 20 patients with AAAs. The largest differences between techniques (1.7 mm) were for infrarenal aortic length, entire aneurysm length, and distance between the most caudal renal artery and both iliac bifurcations; all were considered minor by the authors [4]. In present study, the largest difference in the measurements of aortic diameter between MR angiography and digital subtraction angiography was minimal (1.6 mm, or 6.3%) and not statistically significant. In one patient with a very tortuous abdominal aorta, the difference in the length of the infrarenal aorta was 12 mm (11.6%). This value, too, appeared to be negligible, considering the mean distance of 111.9 mm for the entire vessel. In a comparison of CTA and MR angiography in 61 patients with AAA, Thurnher et al. [5] found the two methods to be equivalent for evaluation of the proximal extent of the aneurysm and all aortic dimensions. MR angiography was superior for assessing iliac vessels because of its larger field of view. Ludman and colleagues [3] performed MR angiography with MRI and CTA on 16 potential candidates for endovascular repair; only six were found to be suitable by MR angiography and CTA. The authors concluded that the two imaging techniques agreed with regard to patient eligibility for an endovascular approach.
Our goal in this study was to minimize the use of nephrotoxic agents before endograft repair in patients with renal impairment, to find patients suitable for endovascular repair, and to reduce the contrast injections needed for obtaining the right projections.
The mean creatinine level in our patients, not including the dialysis patients, was 1.3 ± 1.4 mg/dL before the procedure and 1.3 ± 0.8 mg/dL afterward. Although not statistically significant, this difference indicates the importance of reducing the contrast load whenever possible. Overall, gadolinium has been proven nonnephrotoxic and quite safe. A retrospective study of 15,000 patients who underwent gadolinium-enhanced MRI yielded no significant change in serum creatinine levels after the procedure (2.5-2.3 mg/dL) [6]. Accordingly, Neschis et al. [2], in a controlled study of patients with AAA and impaired renal function, investigated the outcome of stent-graft design based solely on MR angiography; the rates of intraoperative failures and complications were similar in the two groups (16.7% and 18.3%, respectively).
MR angiography allows the acquisition of truly 3D data, in contrast to digital subtraction angiography, with which only 2D projection data can be obtained. Therefore, MR angiographic images may provide more information for the interventional radiologist or surgeon in cases of more complex tortuous arterial anatomy and may increase the accuracy of measurements for stent-graft repair. Previous studies in which MR angiography was performed without contrast enhancement showed limited success in the evaluation of iliac arteries. In our study, because the injection of gadolinium minimized the saturation effect, the dimensions and tortuosity of the iliac arteries were best shown by dynamic contrast-enhanced MR angiography. Moreover, because incomplete opacification of the aorta and iliac vessels may result from inaccuracies in the estimation of circulation time, we used a power injector, which has been shown to improve the aortic signal-to-noise ratio [5].
Even with these modifications, however, MRI still is contraindicated in several patient groups: patients with pacemakers, intracranial aneurysm clips, or eye fragments after trauma and patients allergic to paramagnetic contrast agents. In addition, some people are too claustrophobic to tolerate an MRI examination or cannot hold their breath for the required duration.
Additional limitations of MR angiography are its susceptibility to artifacts from cardiac or respiratory motion, metallic objects, and air-containing structures and its failure to show the extent of arterial calcifications— particularly important with respect to access to aortic aneurysms via the femoral route. A highly calcified iliac artery or aortic bifurcation may impede the passage of the delivery system. To overcome this problem, we used T1- and T2-weighted sequences before injecting contrast medium to assess aortic wall calcifications (void areas in both T1 and T2). Other authors have recommended unenhanced CT of the abdomen and pelvis in addition to MR angiography to assess iliac vessel calcifications [2, 3].
Several limitations of this study should be addressed. First, a relatively high dose of gadopentetate dimeglumine (30 mL) was administered for MR angiography because high signal intensity and homogeneous vessel opacification were necessary for good measurements. Second, although measurements were obtained prospectively and the reviewers were unaware of the results of MR angiography, the digital subtraction angiographic projections and the surgical plans were based on the projections obtained by MR angiography to spare patients the need for contrast medium and to compare the two techniques.
In conclusion, MR angiography is a noninvasive, rapid, and reliable technique. Despite our small sample size (especially the TAA group), our findings indicate that MR angiography is effective and reliable for use as the sole imaging method before endovascular repair of aortic aneurysms in patients with renal impairment. Further studies on larger populations are warranted to confirm these findings.
Acknowledgments
We are indebted to A. Herskoviz for his help in preparing this article.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. J. Francois, D. Tuite, V. Deshpande, R. Jerecic, P. Weale, and J. C. Carr Unenhanced MR Angiography of the Thoracic Aorta: Initial Clinical Evaluation Am. J. Roentgenol., April 1, 2008; 190(4): 902 - 906. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
