Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print October 10, 2006, 10.1148/radiol.2413060053
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Table E1
Right arrow Appendix E1
Right arrow All Versions of this Article:
2413060053v1
241/3/861    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Nikolaou, K.
Right arrow Articles by Schoenberg, S. O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Nikolaou, K.
Right arrow Articles by Schoenberg, S. O.
(Radiology 2006;241:861-872.)
© RSNA, 2006

Technical Developments

High-Spatial-Resolution Multistation MR Angiography with Parallel Imaging and Blood Pool Contrast Agent: Initial Experience1

Konstantin Nikolaou, MD, Harald Kramer, MD, Christina Grosse, MD, Dirk Clevert, MD, Olaf Dietrich, PhD, Mike Hartmann, Paul Chamberlin, MD, Stefan Assmann, PhD, Maximilian F. Reiser, MD and Stefan O. Schoenberg, MD

1 From the Department of Clinical Radiology, Ludwig-Maximilians University of Munich, Grosshadern Campus, Marchioninistr 15, 81377 Munich, Germany (K.N., H.K., C.G., D.C., O.D., M.F.R., S.O.S.); EPIX Pharmaceuticals, Cambridge, Mass (M.H., P.C.); and Siemens Medical Solutions, Erlangen, Germany (S.A.). Received January 11, 2006; revision requested March 9; revision received April 13; accepted May 17; final version accepted June 16. Supported by EPIX Pharmaceuticals. Address correspondence to K.N. (e-mail: Konstantin.Nikolaou{at}med.uni-muenchen.de).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
The purpose of this study was to prospectively evaluate the diagnostic accuracy of reader detection of 75% or greater stenosis at high-spatial-resolution multistation magnetic resonance (MR) angiography performed with matrix coils and a blood pool contrast agent. Ten healthy volunteers and 10 patients were examined. All participants provided informed consent to participate in this institutional review board–approved study. For contrast agent–enhanced multistation MR angiography, an albumin-binding gadolinium chelate, gadofosveset trisodium, was used. Imaging was performed during the first-pass and steady-state phases of the contrast agent. Vessel conspicuity on the first-pass MR angiograms obtained in both volunteers and patients was rated as excellent for 93% of vessels. At steady-state imaging, vessel conspicuity was rated as excellent or good for 89% of vessels. Gadofosveset trisodium–enhanced MR angiography yielded sensitivities of 100% and 97% and specificities of 96% and 97% for detection of significant disease in the carotid and lower extremity arteries, respectively.

Supplemental material:
radiology.rsnajnls.org/cgi/content/full/2413060053/DC1
radiology.rsnajnls.org/cgi/content/full/2413060053/DC2

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Magnetic resonance (MR) angiography has already become the diagnostic method of choice for a variety of angiographic applications, such as MR angiography of the carotid arteries (1), renal arteries (2), and peripheral vessels (3), and has replaced invasive catheter-based angiography in many cases. However, a number of abnormalities—for example, atherosclerosis, diabetic vasculopathy, and inflammatory vessel changes like Takayashu arteritis—must be interpreted as systemic diseases that simultaneously affect several body regions. Investigators in a number of studies have evaluated the feasibility of whole-body MR angiography (4,5), which involves anatomic coverage of most vascular territories.

To assess vascular abnormalities in more than one vascular region, the investigators in several of these previous studies have used dedicated rolling platforms to which the receiver coils are fixed and can be moved along the patient's body. These devices were combined with fast imaging modalities to image the entire body within a short examination time (6). Initial results have been promising but accompanied by reduced image quality and spatial resolution. In consecutive studies of multistation MR angiography performed with matrix coils and parallel imaging techniques, depiction of the entire vasculature without patient repositioning has been demonstrated (7). Still, limitations and compromises in spatial resolution or anatomic coverage due to the limited time frame of the first pass of the contrast agent were described in all of these studies.

To extend the duration of arterial phase imaging, several approaches, such as venous compression for imaging peripheral runoff vessels (8,9), have been tested. Although initial results showing the value of MR angiography—for diagnostic assessment of renal artery stenosis, for example—have been promising (10), subsequent large-scale multicenter trials have revealed that even with use of state-of-the-art techniques, MR angiography of the renal arteries has limited accuracy (11). Investigators in a recent study of renal MR angiography described the need for submillimeter high-spatial-resolution MR angiography with parallel imaging for reliable evaluation of renal artery stenosis on cross-sectional images reconstructed from MR angiography data sets (2).

With use of intravascular paramagnetic blood pool contrast agents for enhancement, limitations might be overcome and imaging strategies for multistation MR angiography could be further developed and improved (12). One such intravascular contrast agent is gadofosveset trisodium (Vasovist; EPIX Pharmaceuticals, Cambridge, Mass; and Schering, Berlin, Germany). This is a small-molecule contrast agent that binds noncovalently to serum albumin. This reversible binding to albumin enhances the paramagnetic effectiveness of gadolinium and allows the administration of lower contrast agent doses compared with the doses of conventional MR contrast agents needed (13). Most important, the albumin-binding characteristic extends the vascular lifetime of the agent and thus allows longer vascular imaging time, potentially higher spatial resolution, and larger anatomic coverage.

The purpose of our study was to prospectively evaluate the diagnostic accuracy of reader detection of 75% or greater stenosis at high-spatial-resolution multistation MR angiography performed with matrix coils and a blood pool contrast agent, with standard imaging techniques as the reference standards.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Informed consent was obtained from all participants in this institutional review board–approved, clinical phase II study. EPIX Pharmaceuticals provided the blood pool contrast agent used (gadofosveset trisodium) and financial support for each patient included. Two authors (M.H., P.C.) are employees of and shareholders of stock in EPIX Pharmaceuticals. Only those authors not employed by this company had control of the data and information submitted for publication.

Healthy Volunteers
For the initial setup and selection of the imaging parameters to be used with the intravascular contrast agent in the patients, 10 healthy male volunteers (mean age, 28 years ± 3 [standard deviation]; range, 25–35 years) were enrolled during a 3-month period (May to July 2004) (Fig 1). The volunteers had to be aged 18–45 years and have a low likelihood of atherosclerotic disease based on medical history and epidemiologic criteria—that is, they had to be nondiabetic and have less than a 10% risk of coronary artery disease, as determined according to Framingham criteria.


Figure 1
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Outline description of the study subjects.

 
Patients and Reference-Standard Modalities
After sequence adaptation with the healthy volunteers, patients were enrolled in the study. For 10 months (August 2004 to May 2005), we prospectively examined 10 patients (two female, eight male; mean age, 62 years ± 11; range, 50–81 years) with known atherosclerotic disease of the carotid, renal, or peripheral arteries (Fig 1). Patients participated voluntarily after we explained the nature and all details of the study. Atherosclerotic disease in the above-mentioned vascular territories was confirmed with state-of-the art peripheral MR angiography (seven cases) or MR angiography of the renal arteries (two cases) enhanced with conventional gadolinium-based extracellular (ie, non–protein-binding) MR contrast agents; Doppler ultrasonography (US) of the carotid arteries (five cases); or digital subtraction angiography of the abdominal aorta, renal arteries, and peripheral arteries (three cases). All of these reference-standard examinations had to be performed at least 72 hours before or 24 hours after administration of gadofosveset trisodium. The mean time between gadofosveset trisodium administration and any reference-standard examination was 25 days ± 17 (range, 3–55 days).

MR Protocols and Intravascular Contrast Agent
All gadofosveset trisodium–enhanced examinations were performed with a 1.5-T MR imager (Magnetom Avanto; Siemens Medical Solutions, Erlangen, Germany) that provided a gradient strength of 40 mT/m and a maximum slew rate of 200 (mT · m–1)/msec. Surface coils were used for signal reception in all body regions. The surface coils, which covered the entire body, consisted of up to 76 coil elements that could be assigned to 32 independent receiver channels—that is, up to 32 coil elements could be used simultaneously in the field of view. No repositioning was required during the examination. The multistation MR angiography protocol used in our study was adapted from earlier studies involving the use of a 32-channel whole-body MR imager (7); specific information on the exact number of coil elements used over a given field of view can be found in that published work. Before the examination, a 20-gauge intravenous catheter was placed in a right arm antecubital vein for contrast agent injection.

For contrast agent–enhanced MR angiography, gadofosveset trisodium, which binds strongly to albumin and has a half-life of about 15 hours, was injected at a flow rate of 1 mL/sec. In plasma at 1.5-T imaging, the relaxivity of gadofosveset trisodium is approximately five to seven times higher than that of gadopentetate dimeglumine (13,14). The administered volume of gadofosveset trisodium was calculated according to the formula [W · 0.03 mmol/kg]/0.25 mmol/mL, where W is the subject's weight in kilograms. With mean body weights of 77 kg for the volunteers and 74 kg for the patients, the mean injected volume was 9 mL for both groups. At an injection flow rate of 1 mL/sec, the mean injection duration was 9 seconds.

After gadofosveset trisodium injection, there were two acquisition phases: During the arterial first-pass (ie, dynamic) phase, arterial angiograms similar to any arterial angiogram acquired with a conventional extracellular MR contrast agent were obtained. During this dynamic phase after contrast agent injection, angiograms of the carotid or renal arteries and the lower leg (ie, calf) vessels were obtained during the first pass of gadofosveset trisodium. The carotid or renal arteries and the lower leg arteries can be imaged after a single injection because with the 32-channel MR imager, the table movement is faster than the passage of the contrast agent bolus. The time required for the table to move from the height of the carotid or renal arteries to the isocenter of the lower leg arteries for MR angiogram acquisition is about 6 seconds (table speed, 20 cm/sec).

After the dynamic phase, gadofosveset trisodium is distributed throughout the entire body vasculature, including the arterial and venous systems, and reaches an equilibrium phase within about 5 minutes. During this steady-state (ie, equilibrium) phase, high spatial resolution of the vascular beds not affected by cardiac or respiratory motion, like the lower extremity vessels, can be achieved during several minutes of imaging. In the volunteers, the primary vascular bed was the lower leg arteries and the secondary target was the carotid arteries. In the patients, the primary target was again the lower leg arteries and the secondary targets were the carotid (five patients) and renal (five patients) arteries.

Imaging Protocol for Volunteers
During all first-pass and steady-state acquisitions, a parallel (ie, generalized autocalibrating partially parallel acquisition [15]) imaging factor of two was applied in the phase-encoding direction. The k-space trajectory for all MR angiography protocols was a centrically reordered readout scheme. The mean total acquisition time for first-pass and steady-state acquisitions in the volunteers was 65 minutes ± 8.

First-pass imaging.—In the 10 healthy volunteers, after injection of gadofosveset trisodium, the carotid and lower leg arteries were imaged during the first pass of the contrast agent (Table E1 [radiology.rsnajnls.org/cgi/content/full/2413060053/DC1]). In all volunteer and patient examinations, the bolus timing for first-pass imaging was performed by using a test bolus of 1 mL of gadofosveset trisodium injected at 1 mL/sec and followed by a 20-mL saline bolus injected at the same rate. The time for passage of the test bolus was assessed by means of repeated two-dimensional acquisition of one transverse section at the height of the common carotid arteries. For first-pass carotid MR angiography, imaging started after a delay of the test bolus time plus 3 seconds. In the lower leg, four acquisitions were performed during the first pass of the contrast agent to ensure the acquisition of at least one truly arterial phase image.

Steady-state imaging.—During the steady-state phase (Table E1 [radiology.rsnajnls.org/cgi/content/full/2413060053/DC1]), the carotid, thoracic, abdominal (including common iliac arteries), upper leg (including external iliac arteries), knee, and lower leg arterial territories were imaged. To optimize image quality, several acquisitions in the lower leg and carotid arteries were performed at increasing spatial resolutions. Neither electrocardiographic gating nor respiratory triggering was used to image the thoracic and abdominal vessels.

Imaging Protocol for Patients
The mean total acquisition time for first-pass and steady-state acquisitions was 48 minutes ± 6. The parameters used for first-pass imaging of the lower leg, carotid, and renal arteries are described in Table E1 (radiology.rsnajnls.org/cgi/content/full/2413060053/DC1).

To keep the total acquisition time within an acceptable range in the patients, steady-state examinations of only the following vessels were performed: Carotid arteries at an isotropic voxel length of 0.80 mm, upper leg and knee arteries at a voxel length of 0.50 mm, and lower leg arteries at a voxel length of 0.50 mm (Table E1 [radiology.rsnajnls.org/cgi/content/full/2413060053/DC1]). For further optimization of image quality, the thoracic and abdominal examinations were modified (Table E1 [radiology.rsnajnls.org/cgi/content/full/2413060053/DC1]). For five patients, two sagittal slabs of the thorax were acquired for separate imaging of the right and left pulmonary arterial systems. In the abdomen, renal arteries were imaged separately by using oblique coronal slabs oriented along the course of the renal arteries. However, a slab thickness and field of view sufficient to depict the entire abdominal aorta and the common iliac arteries were chosen as well. These modified thoracic and abdominal MR angiography examinations were performed during single breath holds.

Reference-Standard and Intravascular Contrast-enhanced Imaging
Details of the peripheral MR angiogram, renal MR angiogram, conventional angiogram, and carotid Doppler US image acquisitions (reference standards) and interpretations are given in Appendix E1 (radiology.rsnajnls.org/cgi/content/full/2413060053/DC2).

For all image quality assessments, the following vessels were examined separately and bilaterally on the first-pass and steady-state images: the common carotid, internal carotid, and vertebral (carotid) arteries; in the thorax, the ascending aorta, descending thoracic aorta, pulmonary main stem, and left and right pulmonary arteries; in the abdomen, the suprarenal aorta, infrarenal aorta, left and right renal arteries, and common iliac artery; in the thigh, the external iliac, common femoral, and superficial femoral arteries; in the knee, the distal superficial femoral, popliteal, and proximal lower leg arteries; and in the calf, the anterior tibial, posterior tibial, and peroneal (fibular) arteries.

All images were transferred to a three-dimensional postprocessing workstation (Leonardo; Siemens Medical Solutions) for qualitative analysis with use of the original source images, maximum intensity projections, and multiplanar reformations (MPRs). The quality of all (volunteer and patient) first-pass and steady-state images was evaluated by two fellows in cardiovascular imaging (H.K. and K.N., 3 and 6 years experience, respectively) independently and then in a consensus reading. The fellows were supervised by an attending radiologist (S.O.S., 10 years experience). A four-point scale was used to assess image quality in terms of vessel conspicuity, which was rated as excellent (optimal vascular signal intensity [SI]), good (diagnostically sufficient SI), moderate (somewhat decreased SI that possibly impairs diagnostic assessment), or poor (insufficient SI such that vessel is not completely identifiable and image is not diagnostic). Presence of artifacts was rated by using a three-point scale: none, mild (artifacts present but not impairing diagnostic information), or major (extensive artifacts impair diagnosis). Venous overlap was rated by using a three-point scale for the first-pass images: none, mild (venous overlap present but not impairing diagnostic information), or major (extensive venous overlap impairs diagnosis). A two-point scale was used to rate the venous overlap on steady-state images: Venous enhancement causes a diagnostic problem or does not cause a diagnostic problem.

Quantitative Analysis of Arterial and Venous Vessel Discrimination on Steady-State Images
Quantitative analysis was performed by one of the fellows who assessed image quality and diagnostic accuracy (K.N.), with the same supervising attending radiologist (S.O.S.). To test for improved discrimination of arterial and venous vessels at increasing spatial resolutions on the steady-state images of the lower leg in the volunteers, the SI profiles were measured along a line perpendicular to the vascular structures by using thin (2-mm) MPRs to display the adjacent vessels. With use of a simple self-programmed software tool, the absolute SI values (expressed as arbitrary units) were read along this line for 40 lower leg data sets (four data sets of increasing spatial resolution per volunteer) and transferred to spreadsheet software (Windows Office Excel 2003; Microsoft, Redmond, Wash) for further evaluation of the data. With increasing spatial resolutions (1.000, 0.512, 0.125, and 0.074 mm3), the profiles of the central arterial vessels and accompanying venous structures were evaluated. In all 10 volunteers, these evaluations were performed for the anterior tibial, posterior tibial, and/or fibular artery, depending on the alignment of the vascular structures. Ideally, the MPR would display all three vessels (one artery, two accompanying veins).

Assessment of Vascular Abnormalities
The MR images of all patients were independently evaluated by the same two reviewers for detection of vascular abnormalities in the carotid (including vertebral) arteries (five patients, 30 vessel segments [six vessels per patient]), renal arteries (five patients, 10 vessel segments [two vessels per patient]), and peripheral runoff vessels (10 patients, 140 vessels). These evaluations were performed by using the first-pass and steady-state data in a combined fashion—first independently by each reviewer and then in a consensus reading. For the renal and carotid arteries, both first-pass and steady-state data were available. For the peripheral runoff vessels, both first-pass and steady-state images of the lower leg vessels only (80 vessels) were available. For the iliac and upper leg arteries (60 vessels), however, only steady-state images were available for diagnostic assessment. Thus, a total of 140 peripheral runoff vessel segments were analyzed. The readers were blinded to all reference-standard imaging results. At review of the gadofosveset trisodium–enhanced MR angiography data, vascular abnormalities were rated as follows: no (0%) stenosis, moderate (<75%) stenosis, significant (≥75%) stenosis, or total vessel occlusion (100% stenosis).

Statistical Analyses
For statistical analysis, two Windows-based software products were used (MedCalc, version 7.0.0.2, 2002, MedCalc Software, Mariakerke, Belgium; SPSS 12.0.1, 2003, LEAD Technologies, Chicago, Ill). For all statistical tests, a significance level of P < .05 was set. Continuous variable data are presented as means ± standard deviations. The sensitivity and specificity, and corresponding 95% confidence intervals, for reader detection of occlusion or significant vascular disease (stenosis ≥ 75%) were calculated.

Intermodality agreement was quantified in terms of the percentage of diagnoses made by using gadofosveset trisodium–enhanced MR angiography that were consistent with the diagnoses made by using the reference-standard imaging modalities, on a per-vessel basis. Interobserver agreement for the two reviewers in terms of assessment of image quality and detection or exclusion of substantial vascular abnormalities on a per-vessel basis was quantified by using {kappa} values: A {kappa} value of 0.20 or less indicated poor agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate agreement; 0.61–0.80, good agreement, and 0.81–1.00, very good agreement. To assess the significance of differences in improved arterial versus venous vessel delineation on the lower leg steady-state images at increasing spatial resolutions, a Student t test was performed to determine the absolute difference in SI between the maximum SI in the central lower leg artery and the minimum SI in the intravascular space (between vein and artery), according to increasing spatial resolution. Receiver operating characteristic curve analysis was performed to assess reader detection of significant (≥75%) stenosis in the carotid, renal, and peripheral arteries.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Image Quality: Volunteers
The two readers rated vessel conspicuity on the first-pass images as excellent for 85% and 98% of carotid and lower leg vessels, respectively, with only minor artifacts and no or mild venous overlap in most instances (Figs 2, 3; Table 1). On the steady-state images, 98% of the lower leg vessels were found to have excellent conspicuity at spatial resolutions of 0.125 and 0.074 mm3 voxel size. For the carotid arteries, image quality was rated best at a spatial resolution of 0.512 mm3 voxel size (excellent and good conspicuity in 53% and 40% of vessels, respectively). Conspicuity was rated as good for 53% and 76% of thoracic and abdominal vessels, respectively, with no excellent results. For upper leg vessels examined at a 0.125-mm3 spatial resolution, conspicuity was excellent (82%) and good (18%) (Table 2).


Figure 2
View larger version (85K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: First-pass and steady-state MR angiograms in 28-year-old healthy man. A1, A2, First-pass coronal maximum intensity projections. During arterial first pass of gadofosveset trisodium, pure arterial phase images of, A1, carotid arteries (3.4/1.3 [repetition time msec/echo time msec], 30° flip angle), and, A2, calf vessels (4.3/1.4, 25° flip angle) were obtained after a single injection of contrast agent, with an isotropic spatial resolution of 1.000 mm3. B1–B3, Steady-state coronal MPRs. During steady-state imaging of the calf vessels, increasing spatial resolutions of, B1, 1.00-mm (1.000-mm3) (4.3/1.4, 25° flip angle); B2, 0.80-mm (0.512-mm3) (4.3/1.4, 25° flip angle); and, B3, 0.42-mm (0.074-mm3) (7.5/2.5, 25° flip angle) voxel lengths were acquired. The data set with highest spatial resolution (B3) depicts the smallest calf vessels in great detail while maintaining a very high signal-to-noise ratio. C, D, Steady-state 20-mm-thick coronal maximum intensity projections of the thoracic and abdominal vessels (4.3/1.4, 25° flip angle).

 

Figure 3
View larger version (97K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Discrimination of arterial and venous vessels in the lower leg on paracoronal MPRs obtained along the course of the vessels in 30-year-old healthy man. Magnified views of anterotibial artery (vessel in the center) accompanied by two veins (vessels to the left and right) are shown at increasing spatial resolutions—specifically, isotropic voxel lengths of, A, 1.0 mm; B, 0.8 mm; C, 0.5 mm; and, D, 0.4 mm. SI profiles at the bottom of each image were measured along the horizontal line perpendicular to the vascular structures. With increasing spatial resolution, the profile of the central arterial vessel is delineated more clearly, proving that the level of differentiation between arterial and venous vessels on steady-state images is dependent on the spatial resolution. a.u. = arbitrary units.

 

View this table:
[in this window]
[in a new window]

  		Table 1. Image Quality at First-Pass Imaging in 10 Healthy Volunteers

 

View this table:
[in this window]
[in a new window]

  		Table 2. Image Quality at Steady-state Imaging in 10 Healthy Volunteers

 
SI measurements along a perpendicular line across selected lower leg arteries (accompanied by two veins each) revealed mean SI differences between the artery and the intervascular tissue of 71 ± 35 at a voxel length of 1.00 mm, 109 ± 44 at 0.80 mm, 196 ± 51 at 0.50 mm, and 220 ± 65 at 0.42 mm. Student t tests revealed significant differences in SI between the 1.00- or 0.80-mm voxel length and the 0.50- or 0.42-mm voxel length (P = .015), with no significant differences between the 1.00- and 0.80-mm voxel lengths (P = .13) or between the 0.50- and 0.42-mm voxel lengths (P = .38). {kappa} Values of interobserver agreement on image quality ratings in the volunteers ranged between 0.71 and 0.89 for the first-pass data and between 0.52 and 1.00 for the steady-state data.

Image Quality: Patients
On the first-pass images, the conspicuity of the carotid, renal, and lower leg arterial segments was rated as excellent in 90%, 83%, and 100% of cases, respectively (Table 3). On the steady-state images, excellent and good conspicuity was observed in 95% and 5% of the lower leg (ie, calf) artery segments, respectively, with 0.074-mm3 spatial resolution; in 20% and 63% of the carotid artery segments, respectively, with 0.512-mm3 spatial resolution; in 71% and 25% of the thigh artery segments, respectively, with 0.125-mm3 spatial resolution; and in 69% and 31% of the knee artery segments, respectively, with 0.125-mm3 spatial resolution. Owing to adaptation of the imaging protocol used to assess the thoracic and abdominal vessels in the volunteers (Table E1 [radiology.rsnajnls.org/cgi/content/full/2413060053/DC1]), the quality of the thoracic and abdominal images obtained in the patients was improved, with excellent and good conspicuity in 20% and 60% of thoracic artery segments, respectively, and in 3% and 83% of abdominal artery segments, respectively (Table 4). In the patients, {kappa} values ranged between 0.69 and 1.00 for the first-pass data and between 0.60 and 1.00 for the steady-state data.


View this table:
[in this window]
[in a new window]

  		Table 3. Image Quality at First-Pass Imaging in 10 Patients

 

View this table:
[in this window]
[in a new window]

  		Table 4. Image Quality at Steady-state Imaging of Arterial Segments in 10 Patients

 
Diagnostic Accuracy: Patients
In the 10 patients, the carotid arteries in two patients (two vessels) and the peripheral arteries in eight (35 vessels) (Table 5) were significantly diseased—that is, 75% or greater stenosis or total occlusion was present according to the reference-standard imaging results (Figs 4–6). The gadofosveset trisodium–enhanced images had sensitivities of 97%–100% and specificities of 96%–100%. Intermodality agreement between the gadofosveset trisodium–enhanced imaging and reference-standard imaging data was 90%–93%, and {kappa} values ranged between 0.69 and 1.00 (Table 6). The area under the receiver operating characteristic curve (Fig 3) was 0.96 (95% confidence interval: 0.91, 1.00).


View this table:
[in this window]
[in a new window]

  		Table 5. Diagnostic Evaluation of Vascular Abnormalities in 10 Patients

 

Figure 4
View larger version (9K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Receiver operating characteristic analysis curve illustrates the accuracy of gadofosveset trisodium–enhanced MR angiography, relative to the reference-standard imaging modalities, in diagnosing the disease state of 180 vessels (30 supraaortic, 10 renal, 140 peripheral runoff), including 37 cases of significant (≥75%) stenosis.

 

Figure 5
View larger version (175K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5a: Images in 54-year-old man with long history of diabetes and increasing claudication of right leg. (a, b) Coronal steady-state gadofosveset trisodium–enhanced MR image (a) and magnified view (b) of thigh and knee (4.3/1.4, 25° flip angle) at 0.125-mm3 spatial resolution show large aneurysm of right distal superficial femoral artery, with complete occlusion of vessel, which is substituted by collateral vessels (arrow). A small aneurysm (arrowhead) of the distal superficial femoral artery without significant stenosis is also seen in left leg. (c) Curved MPRs show course of distal superficial femoral artery and popliteal artery in right and left legs, sites of vessel occlusion (arrows) in right leg, and vessel wall irregularities in left leg. Contrast-enhanced venous structures like the right popliteal vein (arrowhead) do not interfere in making the clinical diagnosis. (d) Findings on state-of-the-art MR angiogram of lower leg vessels enhanced with a conventional MR contrast agent (reference standard) and obtained at 1-mm3 spatial resolution confirm superficial femoral artery occlusion on right side, but image fails to depict details of aneurysms on both sides, as in c.

 

Figure 5
View larger version (177K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5b: Images in 54-year-old man with long history of diabetes and increasing claudication of right leg. (a, b) Coronal steady-state gadofosveset trisodium–enhanced MR image (a) and magnified view (b) of thigh and knee (4.3/1.4, 25° flip angle) at 0.125-mm3 spatial resolution show large aneurysm of right distal superficial femoral artery, with complete occlusion of vessel, which is substituted by collateral vessels (arrow). A small aneurysm (arrowhead) of the distal superficial femoral artery without significant stenosis is also seen in left leg. (c) Curved MPRs show course of distal superficial femoral artery and popliteal artery in right and left legs, sites of vessel occlusion (arrows) in right leg, and vessel wall irregularities in left leg. Contrast-enhanced venous structures like the right popliteal vein (arrowhead) do not interfere in making the clinical diagnosis. (d) Findings on state-of-the-art MR angiogram of lower leg vessels enhanced with a conventional MR contrast agent (reference standard) and obtained at 1-mm3 spatial resolution confirm superficial femoral artery occlusion on right side, but image fails to depict details of aneurysms on both sides, as in c.

 

Figure 5
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5c: Images in 54-year-old man with long history of diabetes and increasing claudication of right leg. (a, b) Coronal steady-state gadofosveset trisodium–enhanced MR image (a) and magnified view (b) of thigh and knee (4.3/1.4, 25° flip angle) at 0.125-mm3 spatial resolution show large aneurysm of right distal superficial femoral artery, with complete occlusion of vessel, which is substituted by collateral vessels (arrow). A small aneurysm (arrowhead) of the distal superficial femoral artery without significant stenosis is also seen in left leg. (c) Curved MPRs show course of distal superficial femoral artery and popliteal artery in right and left legs, sites of vessel occlusion (arrows) in right leg, and vessel wall irregularities in left leg. Contrast-enhanced venous structures like the right popliteal vein (arrowhead) do not interfere in making the clinical diagnosis. (d) Findings on state-of-the-art MR angiogram of lower leg vessels enhanced with a conventional MR contrast agent (reference standard) and obtained at 1-mm3 spatial resolution confirm superficial femoral artery occlusion on right side, but image fails to depict details of aneurysms on both sides, as in c.

 

Figure 5
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5d: Images in 54-year-old man with long history of diabetes and increasing claudication of right leg. (a, b) Coronal steady-state gadofosveset trisodium–enhanced MR image (a) and magnified view (b) of thigh and knee (4.3/1.4, 25° flip angle) at 0.125-mm3 spatial resolution show large aneurysm of right distal superficial femoral artery, with complete occlusion of vessel, which is substituted by collateral vessels (arrow). A small aneurysm (arrowhead) of the distal superficial femoral artery without significant stenosis is also seen in left leg. (c) Curved MPRs show course of distal superficial femoral artery and popliteal artery in right and left legs, sites of vessel occlusion (arrows) in right leg, and vessel wall irregularities in left leg. Contrast-enhanced venous structures like the right popliteal vein (arrowhead) do not interfere in making the clinical diagnosis. (d) Findings on state-of-the-art MR angiogram of lower leg vessels enhanced with a conventional MR contrast agent (reference standard) and obtained at 1-mm3 spatial resolution confirm superficial femoral artery occlusion on right side, but image fails to depict details of aneurysms on both sides, as in c.

 

Figure 6
View larger version (147K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6a: Images in 65-year-old man with transient ischemic attacks. (a) Coronal maximum intensity projection of first-pass gadofosveset trisodium–enhanced MR data at spatial resolution of 1.0 mm3 shows significant stenosis (arrow) in proximal right internal carotid artery. (b) Findings on color-coded Doppler US image (reference standard), showing 75% or greater stenosis with a maximal flow velocity of 1.7 m/sec, confirm the findings in a. (c) Findings on MPRs, on which first-pass data (3.4/1.3, 30° flip angle) are compared with steady-state data (6.5/2.2, 25° flip angle), also confirm the internal carotid artery stenosis (arrows) seen in a.

 

Figure 6
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6b: Images in 65-year-old man with transient ischemic attacks. (a) Coronal maximum intensity projection of first-pass gadofosveset trisodium–enhanced MR data at spatial resolution of 1.0 mm3 shows significant stenosis (arrow) in proximal right internal carotid artery. (b) Findings on color-coded Doppler US image (reference standard), showing 75% or greater stenosis with a maximal flow velocity of 1.7 m/sec, confirm the findings in a. (c) Findings on MPRs, on which first-pass data (3.4/1.3, 30° flip angle) are compared with steady-state data (6.5/2.2, 25° flip angle), also confirm the internal carotid artery stenosis (arrows) seen in a.

 

Figure 6
View larger version (75K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6c: Images in 65-year-old man with transient ischemic attacks. (a) Coronal maximum intensity projection of first-pass gadofosveset trisodium–enhanced MR data at spatial resolution of 1.0 mm3 shows significant stenosis (arrow) in proximal right internal carotid artery. (b) Findings on color-coded Doppler US image (reference standard), showing 75% or greater stenosis with a maximal flow velocity of 1.7 m/sec, confirm the findings in a. (c) Findings on MPRs, on which first-pass data (3.4/1.3, 30° flip angle) are compared with steady-state data (6.5/2.2, 25° flip angle), also confirm the internal carotid artery stenosis (arrows) seen in a.

 

View this table:
[in this window]
[in a new window]

  		Table 6. Performance Values for Contrast-enhanced MR Angiography Diagnosis

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 References
 
The results of our study, in which gadofosveset trisodium enhancement and a 32-channel MR imager were combined with parallel imaging techniques to perform high-spatial-resolution multistation MR angiography, are encouraging. The acquisition strategies used during the first pass of the contrast agent were very similar to conventional arterial MR angiography, and consistently high image quality was achieved during first-pass imaging. Also, because the relaxivity of gadofosveset trisodium is five to seven times higher than that of conventional 0.5 mol/L extracellular contrast agents (used at 1.5 T), the injection of gadofosveset trisodium at 1 mL/sec translates into a conventional contrast agent injection rate of about 5–7 mL/sec. This might be one reason for the very high vascular SI and the high image quality of the gadofosveset trisodium–enhanced first-pass imaging data. This benefit of high signal-to-noise ratio during the first pass of gadofosveset trisodium is also potentially advantageous for time-resolved imaging.

During steady-state imaging, the acquisition strategies had to be adapted. Owing to the several-hour half-life of gadofosveset trisodium, steady-state imaging can be performed for up to several minutes in one vascular bed. Thus, high spatial resolutions of up to 0.074 mm3 could be achieved, especially in vascular territories not affected by cardiac or respiratory motion. If we compare voxel sizes, this translates into an increase in spatial resolution by a factor of 13.5 compared with the spatial resolution achieved at conventional MR angiography—about 1.0 mm3. However, because the matrix size used was as high as 896 x 896 in certain vascular beds, some acquisition times were longer than 6 minutes. Also, the image reconstruction time increases with these high matrices: up to a maximum of about 7–10 minutes for 128 images at a matrix of 896 x 896, which has to be noted as a drawback. Still, despite the long acquisition times at 0.125- and 0.074-mm3 spatial resolutions in the volunteers, image quality was rated as excellent in almost all cases.

In the patients, however, the quality of the high-spatial-resolution lower leg images was diminished owing to motion artifacts in a few instances. The risk of motion artifacts impairing image quality increases with increasing acquisition time. On the other hand, the small-caliber arterial and venous structures in close vicinity—especially those in the lower leg—could be clearly differentiated only at a high spatial resolution of 0.50- or 0.42-mm voxel length. At voxel lengths of 0.80 and 1.00 mm, venous overlap was present and was considered a potential diagnostic problem in 38% and 80% of the lower leg steady-state data, respectively. As a compromise, a voxel length of 0.50 mm seems to be acceptable for imaging the peripheral runoff vessels: It enables clear differentiation of the arterial and venous structures and an acceptable acquisition time.

In the other vascular territories imaged during the steady state of gadofosveset trisodium, cardiac and breathing motion artifacts posed considerable problems in acquisition optimization. For example, in the thorax and abdomen, the acquisition time is limited to the time frame of a single breath hold; thus, the potential of the intravascular contrast agent to facilitate high spatial resolution cannot be fully exploited. To optimize image quality in the patients, the protocol for imaging the thoracic and abdominal vessels in the volunteers was adapted for the final set of five patients. In the thorax, the spatial resolution was improved owing to the reduced acquisition time facilitated by acquiring two sagittal slabs covering the pulmonary arteries and both lungs. Similarly, in the abdomen, by acquiring two oblique coronal slabs along the course of the renal arteries, we increased the parallel imaging factor to three, increased the spatial resolution, and improved the overall image quality of abdominal MR angiograms compared with that of the angiograms obtained in the volunteers. In addition, refined acquisition techniques involving free-breathing navigator approaches—possibly combined with electrocardiographic gating—could be effective for both the thoracic and the abdominal vasculature.

The steady-state images of the carotid arteries acquired with a 0.80-mm voxel length (acquisition time, 40 seconds) had superior image quality compared with the image quality achieved with a 1.00-mm voxel length (acquisition time, 23 seconds). However, at a 0.50-mm voxel length, image quality decreased—probably owing to cardiac or pulsation artifacts and/or patient movement and to the reduced signal-to-noise ratio of the small voxels acquired (acquisition time, 1 minute 41 seconds). Therefore, for the carotid arteries, a voxel length of 0.80 mm appeared to be the ideal compromise for steady-state imaging, as venous structures are more easily differentiated from arteries in this vascular territory than in the lower leg.

Regarding diagnostic accuracy, gadofosveset trisodium–enhanced MR angiography data sets had high sensitivity (100% and 97%, respectively) for detection of significant stenosis in the carotid (two significantly stenosed vessels) and lower extremity (35 significantly stenosed vessels) arteries. In the present study, first-pass and steady-state images were read in a combined fashion, as this would be the typical approach for the use of this recently available contrast agent in a clinical setting. We acknowledge that special reconstruction techniques, such as three-dimensional volume rendering, which is described as helpful in the comprehensive assessment of MR angiography data sets (16), cannot be used with steady-state data owing to venous overlap. Other studies (17,18) of gadofosveset trisodium–enhanced MR angiography for detection of peripheral arterial disease have revealed significantly improved effectiveness at a dose of 0.03 mmol/kg, as compared with the effectiveness of nonenhanced MR angiography, and the safe and effective MR examination of patients with aortoiliac occlusive disease. However, these studies were limited to the examination of a single-vessel territory and comparison of diagnostic accuracy between gadofosveset trisodium–enhanced MR angiography and nonenhanced MR angiographic techniques only.

Our study had limitations: First, only 10 volunteers and 10 patients were included. However, the receiver operating characteristic analysis performed in the patients revealed combined first-pass and steady-state gadofosveset trisodium–enhanced MR angiography to have high potential clinical value. Furthermore, the combined reading of first-pass and steady-state gadofosveset trisodium–enhanced images might be considered a limitation, as separate results would have made the increase in diagnostic accuracy potentially achieved with the additional steady-state data more obvious. However, owing to the limited number of patients and the significantly stenosed vessels in this initial study, a power analysis to compare the diagnostic value of first-pass images, steady-state images, and the combined data was not possible. Also, as mentioned earlier, in a clinical setting, first-pass and steady-state images would most likely always be used in a combined fashion, as was done in our study.

We must admit also that for the carotid, renal, and lower leg arteries, both first-pass and steady-state data were available for assessment of diagnostic accuracy. However, for the iliac and upper leg arteries (60 vessel segments), only steady-state data were available for diagnostic assessment. Furthermore, conventional gadolinium-enhanced MR data might not have been the optimal reference standard for comparison with our gadofosveset trisodium–enhanced MR data. Finally, the image quality at steady-state angiography was optimized primarily in the carotid and lower leg arteries, as these were the target vessels for sequence optimization. The image quality at steady-state angiography of the thoracic and abdominal vessels was not yet optimized and cannot be considered state-of-the-art compared with the image quality of current first-pass MR angiography techniques enhanced with extracellular contrast agents. Acquisition techniques will have to be further adapted in future studies. Perhaps the use of higher contrast agent doses or newer electrocardiographic or respiratory gating techniques can further improve the image quality that is achievable in these vascular territories.

In conclusion, multistation MR angiography enhanced with an intravascular contrast agent can be performed with high spatial resolution and thus maintain high diagnostic accuracy. Future studies will have to address the optimization of imaging time, reconstruction time, and interpretation strategies—as well as the further optimization of image quality—for steady-state imaging data, especially in the thorax and abdomen.


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


   ACKNOWLEDGMENTS
 
The authors thank Denise Friedrich, RT, Anja Struwe, RT, and Frank Stadie, RT, for assistance in setting up the gadofosveset trisodium–enhanced MR angiography protocol and for performing the MR examinations. The authors also thank Bozena Brzuska, BA, for assistance in organizing and conducting this study.


   FOOTNOTES
 

Abbreviations: MPR = multiplanar reformation • SI = signal intensity

See Materials and Methods for pertinent disclosures.

Author contributions: Guarantors of integrity of entire study, K.N., S.O.S.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, K.N., O.D., S.O.S.; clinical studies, K.N., H.K., C.G., D.C., M.H., M.F.R., S.O.S.; statistical analysis, K.N., H.K., S.O.S.; and manuscript editing, K.N., H.K., C.G., O.D., P.C., M.F.R., S.O.S.


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

  1. Buskens E, Nederkoorn PJ, Buijs-Van Der Woude T, et al. Imaging of carotid arteries in symptomatic patients: cost-effectiveness of diagnostic strategies. Radiology 2004;233:101–112.[Abstract/Free Full Text]
  2. Schoenberg SO, Rieger J, Weber CH, et al. High-spatial-resolution MR angiography of renal arteries with integrated parallel acquisitions: comparison with digital subtraction angiography and US. Radiology 2005;235:687–698.[Abstract/Free Full Text]
  3. Meissner OA, Rieger J, Weber C, et al. Critical limb ischemia: hybrid MR angiography compared with DSA. Radiology 2005;235:308–318.[Abstract/Free Full Text]
  4. Ruehm SG, Goyen M, Barkhausen J, et al. Rapid magnetic resonance angiography for detection of atherosclerosis. Lancet 2001;357:1086–1091.[CrossRef][Medline]
  5. Herborn CU, Goyen M, Quick HH, et al. Whole-body 3D MR angiography of patients with peripheral arterial occlusive disease. AJR Am J Roentgenol 2004;182:1427–1434.[Abstract/Free Full Text]
  6. Ruehm SG, Goyen M, Quick HH, et al. Whole-body MR angiography on a rolling table platform (AngioSURF) [in German]. Rofo 2000;172:670–674.[Medline]
  7. Kramer H, Schoenberg SO, Nikolaou K, et al. Cardiovascular screening with parallel imaging techniques and a whole-body MR imager. Radiology 2005;236:300–310.[Abstract/Free Full Text]
  8. Vogt FM, Ajaj W, Hunold P, et al. Venous compression at high-spatial-resolution three-dimensional MR angiography of peripheral arteries. Radiology 2004;233:913–920.[Abstract/Free Full Text]
  9. Bilecen D, Jager KA, Aschwanden M, Heidecker HG, Schulte AC, Bongartz G. Cuff-compression of the proximal calf to reduce venous contamination in contrast-enhanced stepping-table magnetic resonance angiography. Acta Radiol 2004;45:510–515.[CrossRef][Medline]
  10. Vasbinder GB, Nelemans PJ, Kessels AG, Kroon AA, de Leeuw PW, van Engelshoven JM. Diagnostic tests for renal artery stenosis in patients suspected of having renovascular hypertension: a meta-analysis. Ann Intern Med 2001;135:401–411.[Abstract/Free Full Text]
  11. Vasbinder GB, Nelemans PJ, Kessels AG, et al. Accuracy of computed tomographic angiography and magnetic resonance angiography for diagnosing renal artery stenosis. Ann Intern Med 2004;141:674–682.[Abstract/Free Full Text]
  12. Perreault P, Edelman MA, Baum RA, et al. MR angiography with gadofosveset trisodium for peripheral vascular disease: phase II trial. Radiology 2003;229:811–820.[Abstract/Free Full Text]
  13. Lauffer RB, Parmelee DJ, Dunham SU, et al. MS-325: albumin-targeted contrast agent for MR angiography. Radiology 1998;207:529–538.[Abstract/Free Full Text]
  14. Rohrer M, Bauer H, Mintorovitch J, Requardt M, Weinmann HJ. Comparison of magnetic properties of MRI contrast media solutions at different magnetic field strengths. Invest Radiol 2005;40:715–724.[CrossRef][Medline]
  15. Griswold MA, Jakob PM, Heidemann RM, et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 2002;47:1202–1210.[CrossRef][Medline]
  16. Mallouhi A, Schocke M, Judmaier W, et al. 3D MR angiography of renal arteries: comparison of volume rendering and maximum intensity projection algorithms. Radiology 2002;223:509–516.[Abstract/Free Full Text]
  17. Goyen M, Edelman M, Perreault P, et al. MR angiography of aortoiliac occlusive disease: a phase III study of the safety and effectiveness of the blood-pool contrast agent MS-325. Radiology 2005;236:825–833.[Abstract/Free Full Text]
  18. Rapp JH, Wolff SD, Quinn SF, et al. Aortoiliac occlusive disease in patients with known or suspected peripheral vascular disease: safety and efficacy of gadofosveset-enhanced MR angiography—multicenter comparative phase III study. Radiology 2005;236:71–78.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 RadiologyHome page
D. R. Hadizadeh, J. Gieseke, S. H. Lohmaier, K. Wilhelm, J. Boschewitz, F. Verrel, H. H. Schild, and W. A. Willinek
Peripheral MR Angiography with Blood Pool Contrast Agent: Prospective Intraindividual Comparative Study of High-Spatial-Resolution Steady-State MR Angiography versus Standard-Resolution First-Pass MR Angiography and DSA
Radiology, September 3, 2008; (2008) 2492072033.
[Abstract] [Full Text]

Home page
 Am. J. Roentgenol.Home page
H. Ersoy and F. J. Rybicki
MR Angiography of the Lower Extremities
Am. J. Roentgenol., June 1, 2008; 190(6): 1675 - 1684.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Table E1
Right arrow Appendix E1
Right arrow All Versions of this Article:
2413060053v1
241/3/861    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Nikolaou, K.
Right arrow Articles by Schoenberg, S. O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Nikolaou, K.
Right arrow Articles by Schoenberg, S. O.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH