| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Institute of Diagnostic Radiology, University Hospital Zurich, Ramistr. 100,
CH8091 Zurich, Switzerland.
2
Department of Diagnostic Radiology, University Hospital Essen, Hufelandstr.
55, D-45122 Essen, Germany.
3
Nycomed Amersham, Nycomed Arzneimittel, COMET, Fraunhoferstr. 7, 0-85737
Ismaning, Germany.
4
Nycomed Amersham Imaging, Sandakerveien 100, 1-0401 Oslo, Norway.
Received August 2, 1999;
accepted after revision December 7, 1999.
Address correspondence to J. F. Debatin.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Twenty patients with hemodynamically
significant stenosis (
50% of luminal diameter) of the iliac or renal
arteries or an aortic aneurysm documented by digital subtraction angiography
underwent MR angiography at 1.5 T after administration of NC100150.
Three-dimensional MR angiographic data sets were collected in the equilibrium
phase. In a prospective analysis, each vascular segment (16 segments per
arterial tree) was evaluated.
RESULTS. All patients tolerated the NC100150 administration well. Mean contrast-to-noise ratios of the vascular data collected in the equilibrium phase of NC100150 was 30.3 ± 15.9. Compared with digital subtraction angiography, the sensitivity and specificity of MR angiography for the renal arteries were 82% and 98%, respectively; for the common iliac arteries, 86% and 97%, respectively; for the external iliac arteries, 80% and 100%, respectively; and for the internal iliac arteries, 71% and 97%, respectively. All 83 aneurysmal changes revealed by digital subtraction angiography of the aortoiliac arteries were well displayed on the MR angiographic data sets.
CONCLUSION. Equilibrium-phase NC100150-enhanced three-dimensional MR angiography shows high specificity when evaluating the abdominal and pelvic vascular systems, but the attendant venous overlap can limit the assessment of stenosis in renal and pelvic arterial segments.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The extracellular nature of the paramagnetic gadolinium chelates induces rapid distribution of the agent into the extracellular spaces, resulting in rapidly decreasing arterial signal intensity. The short intravascular half-life mandates optimal timing of the contrast-material bolus and fast data collection [7]. Limitations regarding the maximal allowable contrast dose restrict 3D MR angiography in the abdomen and chest to the assessment of one or two vascular territories. Repeated imaging of questionable regions is feasible, but requires complex timing and time-consuming image subtraction strategies [10]. Blood pool contrast agents with a long intravascular half-life can overcome these limitations because the data collection time is not limited [11, 12]. After administration of a single dose of contrast agent, multiple body regions may be imaged.
NC100150 (Clariscan; Nycomed Amersham, Wayne, PA) is a new colloidal preparation of ultrasmall superparamagnetic iron oxide particles. This MR contrast agent was designed as a blood pool agent for MR imaging.
This study evaluates the diagnostic usefulness of NC100150 for assessing the aortoiliac and renal arteries. Three-dimensional MR angiography data sets were collected in the equilibrium phase after the administration of NC100150, using digital subtraction angiography as the standard of reference.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
50% of luminal diameter) of the iliac or renal
arteries or presence of an aortic aneurysm, as documented by digital
subtraction angiography; lack of preexisting liver diseases, iron storage
diseases, or renal insufficiency, as evidenced by physical examination and
blood chemistry testing; negative pregnancy testing for women not surgically
sterile; lack of contraindications to MR imaging (pacemaker, claustrophobia,
and so forth); and completion of both digital subtraction angiography and MR
angiography examinations within a 6-week interval. Patients were excluded on the basis of a history of allergic reaction to iron supplements, dextran, starch, or polyethylene glycol or its derivatives.
MR Contrast Agent
NC100150 is a colloidal preparation of ultrasmall superparamagnetic iron
oxide particles with an oxidized starch coating. Each particle has an
approximate diameter of 20 nm. The preparation exerts a relatively high T1 (r1
= 21.8 mmol-1 sec-1) with a low r2:r1 ratio of 1.62 at
20 MHz and 40° C. NC100150 contains 30 mg of iron [Fe]/ml. The vascular
half-life is dose dependent and ranges between 3 and 4 hr
[13]. NC100150 was
administered to each patient by hand in three bolus steps up to a cumulative
dose of 5.0 mg Fe/kg of body weight (first dose, 0.75 mg Fe/kg of body weight;
second dose, 1.25 mg Fe/kg of body weight; third dose, 3 mg Fe/kg of body
weight). The injection rate was 2 ml/sec in all patients. MR imaging was
limited to the equilibrium phase commencing no sooner than 2 min after
completion of the cumulative NC100150 administration. Although MR imaging was
performed after application of each dose of the three bolus steps, only the MR
data sets collected with a cumulative dose of 5 mg Fe/kg of body weight were
used for image assessment in this study.
As part of the phase II safety assessment, blood pressure, heart rate, and body temperature were monitored over the first 2 hr at set intervals and at 24 and 72 hr after NC100150 injection. A 12-lead EKG and laboratory testing, including serum chemistry, hematology, and urine analysis, were performed at 2, 24, and 72 hr after injection of NC100150. All data were compared with baseline values determined 1-2 hr before contrast injection.
MR Imaging
All MR imaging was performed on a 1.5-T scanner (Signa Horizon LX; General
Electrics Medical Systems, Milwaukee, WI) equipped with a fast three-axis
gradient system characterized by a maximum amplitude of 40 mT/m and slew rate
of 150 mT/m/msec. An anteroposterior phased array surface coil was used for
signal reception. The coil was placed to cover the expected volume that
contained the lower portion of the abdominal aorta, including the renal and
pelvic arteries, to the level of the inguinal ligaments.
On the basis of an axial two-dimensional, gradient-recalled echo localizing sequence (TR/TE, 5.6/1.7; flip angle, 30°) performed within a 24-sec breath-hold, 3D MR angiography image sets were prescribed in the coronal plane using a 3D Fourier transformed gradient-recalled echo sequence with spoiling gradients [9]. The sequence used the following parameters: TR/TE, 5.5/1.6; flip angle, 30°; sampling bandwidth, ±62.5 MHz; slice thickness, 2.2-2.8 mm. Combined with a 256 x 192 matrix, a 32-38 x 34 cm field of view resulted in an in-plane resolution of 1.2-1.5 x 1.8 mm. Zero interpolation in all three planes improved the latter to 0.6-0.7 x 0.9 mm. Each image set, consisting of 64-100 sections was collected with a breath-hold of 24-32 sec.
Digital Subtraction Angiography
Digital subtraction angiography was performed using a transfemoral approach
with a 5-French pigtail catheter (AngiOptic; Angiodynamics, Queensbury, NY) on
one of two units (Digitron 3, Siemens, Erlangen, Germany; or Integris V3000
Version 13, Philips Medical Systems, Best, the Netherlands). For evaluation of
the abdominal aorta and renal arteries, the catheter tip was positioned
between the 12th thoracic and first lumbar vertebral body and 20 ml of
iodinated contrast material ([320 mg ioxaglate/ml] Hexabrix 320; Labaratoire
Guerbet, Aulnay-sous-Bois, France) were injected. The catheter tip was
subsequently positioned above the aortic bifurcation for digital subtraction
angiography of the pelvic arteries with 20 ml of contrast material. As
required, the examination was supplemented with oblique views. Digital
subtraction angiography preceded MR angiography in all patients.
Image Analysis
For quantitative analysis of the MR angiographic data sets, signal
intensities were measured in regions of interest positioned in the artery
lumen, in the stationary tissues adjacent to the arteries, and outside of the
body. Vessel signal-to-noise ratios and contrast-to-noise ratios were
calculated. Signal-to-noise ratios were determined by measuring the mean
signal intensity in a region of interest and dividing it by the standard
deviation of the signal intensity in a region of interest outside of the body.
Contrast-to-noise ratios were determined by measuring the mean signal
intensity in the region of interest and subtracting the mean signal intensity
in adjacent arterial tissue before dividing by the standard deviation of the
signal intensity in a region of interest outside of the body.
Digital subtraction angiography and 3D MR angiography image sets were interpreted in a prospective and blinded fashion. An experienced vascular radiologist unaware of the results of MR angiography assessed all digital subtraction angiography examinations. The MR angiography image sets were read by an experienced MR radiologist. For analysis of the MR angiography examinations, entire-volume maximum intensity projections, targeted maximum intensity projections, and source images were made available on radiographs. For patients with an abdominal aortic aneurysm, surface shaded-display images of the entire aneurysm were also available. All the postprocessing was performed by a third radiologist unaware of the patients' data and the digital subtraction angiography results. In addition, multiplanar reformations were rendered on an interactive workstation by the interpreting radiologist. Other postprocessing techniques were not used for image evaluation. For analysis the arterial vascular system was divided into the following 16 segments: the aorta, divided into a suprarenal and infrarenal segment; the renal arteries, divided into proximal, middle, and distal thirds as seen from the aortic origin to the renal hilum; the common iliac artery; the external iliac artery divided into a proximal and distal half; and the internal iliac artery.
Each segment was analyzed regarding image quality (diagnostic versus nondiagnostic) and assessed for the presence of occlusive disease on the basis of the five-point scale: 0 = normal vessel; 1 = vessel irregularity with less than 10% luminal reduction; 2 = mild stenosis with less than 50% luminal narrowing; 3 = severe stenosis exceeding 50% luminal narrowing, and 4 = occlusion.
Grading of the stenosis was performed at the interactive workstation with enlarged maximum-intensity-projection images using an electronic caliper with 0.1-mm accuracy.
Evidence of fibromuscular dysplasia was noted separately. Changes consistent with fibromuscular dysplasia in the renal arteries were graded as severe on the basis of the assumption that they were hemodynamically significant. Similarly, presence and location of aneurysmal changes were noted separately. A short-axis measurement greater than 3.5 cm was used as the diagnostic criterion for abdominal aneurysms. Aortic aneurysms were classified as suprarenal when the distance of the proximal extent, or "neck," of the aneurysm extended over the renal artery's origins by more than 1 cm. A juxtarenal aneurysm was defined as one with a neck of less than 1 cm, whereas an aneurysm was considered infrarenal if the infrarenal neck was 1 cm or greater. The relationship of the aneurysm to the proximal portions of the renal artery and superior mesenteric artery was assessed. The celiac trunk was not evaluated because it frequently was not included in the 3D MR angiographic image volume. An iliac aneurysm was diagnosed in the presence of a focal increase in arterial diameter exceeding the diameter of the adjacent vessel by more than 50%.
Sensitivity, specificity, positive predictive values, and negative
predictive values for determination of hemodynamically significant stenotic
lesions (50% of the luminal diameter) or occlusions were calculated for
all segments combined and for each vessel separately, and 95% confidence
intervals were calculated. When two or more pathologic changes were detected
in the same vessel, the most severe change was used for subsequent grading and
analysis.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
All 20 patients tolerated the NC100150 administration well. No serious adverse events or significant trends regarding any of the safety parameters were detected. Six patients reported adverse events graded as mild. These included diarrhea and mild fever in two patients, back pain in two patients, and orthostatic hypotonia (approximately 2 hr after drug administration) in two patients. None required treatment and all adverse events resolved before the patients were discharged from the study. Laboratory testing revealed a temporally limited increase in serum iron and transferrin (Fig. 1). Within 72 hr after injection of NC100150, the serum iron levels returned to normal in 15 patients. In the remaining five patients serum iron levels were minimally elevated.
|
All digital subtraction angiography and MR angiography examinations were technically adequate. All image sets contained the abdominal aorta, the renal arteries, and the pelvic arterial system. None of the data sets had to be repeated because of technical reasons. Quantitative analysis of the MR angiography data sets revealed a mean overall signal-to-noise ratio of 41.4 ± 19.2, and contrast-to-noise ratio of 30.3 ± 15.9. Signal-to-noise and contrast-to-noise ratios for individual vessels are summarized in Table 1.
|
Digital subtraction angiography revealed two accessory renal arteries in one patient, bringing the total number of segments that could be assessed to 326. Both accessory arteries in this patient were correctly identified on MR angiography. Venous overlap prohibited diagnostic assessment of six (1.8%) of 326 arterial segments in three patients (two segments in each patient) after administration of NC100150. Therefore, a total of 320 arterial segments were evaluated.
All nondiagnostic segments involved the middle and distal segments of the right renal artery. Analysis of the MR angiography data sets required multiplanar reformations in all 20 patients. Interpretation of NC100150 data sets required, on average, 6.2 min.
Table 2 summarizes the
location, number, and degree of occlusive disease of the 280 arterial segments
in digital subtraction angiography and MR angiography. Digital subtraction
angiography revealed seven occluded, 23 severely stenosed (>50% luminal
narrowing), 33 mildly stenosed (11-50% luminal narrowing) and 62 irregular
(<10% luminal narrowing) segments. Intravascular contrast-enhanced MR
angiography characterized 27 segments as irregular (<10% luminal
narrowing), 32 segments as mildly stenosed (<50% luminal narrowing), 21 as
severely stenosed, and eight as occluded. Regarding detection of
hemodynamically significant arterial occlusive disease with luminal narrowing
at least 50% or occlusions, MR angiography overestimated four severely
stenosed segments as mildly stenosed and one severely stenosed segment was
falsely interpreted as occluded. NC100150-enhanced MR angiography falsely
depicted one mildly stenosed segment as severely stenosed and five severely
stenosed segments as mildly stenosed or irregular. Sensitivity and specificity
for detection of hemodynamically significant stenosed segments (50%) or
occlusions for all 320 artery segments were 81% (95% confidence interval,
0.76-0.84) and 98% (95% confidence interval, 0.97-0.99), respectively.
Positive predictive value was 83% (95% confidence interval, 0.79-0.87), and
negative predictive value 98% (95% confidence interval, 0.97-0.99).
|
Sensitivity, specificity, positive predictive value, and negative predictive value for detection and grading of hemodynamically significant stenosis (>50%) and occlusions for the different artery segments are listed in Table 3 (Fig. 2A,2B).
|
|
|
Digital subtraction angiography showed evidence of fibromuscular dysplasia in a total of 12 renal segments in four patients. MR angiography revealed eight lesions in two patients. Fibromuscular dysplasia changes affecting two middle and two distal segments in two patients were missed using NC100150 because arterial delineation included venous overlap.
Digital subtraction angiography revealed aneurysmal changes in 83 segments. On digital subtraction angiography, suprarenal aortic aneurysms were revealed in two patients (four aneurysmatic aortic segments), infrarenal aneurysms in four patients (four aneurysmatic aortic segments), and juxtarenal aortic aneurysms in one patient (one aneurysmatic segment). NC100150-enhanced MR angiography revealed indicated location and extent of all aortic aneurysms (Fig. 3A,3B,3C). Focal aneurysms of the common iliac artery and external iliac artery were present in 22 and 30 arterial segments on both digital subtraction angiography and MR angiography. The 22 focal aneurysms of the internal iliac revealed on digital subtraction angiography were also visible on MR angriography.
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
NC100150 is one of two blood pool contrast agents currently under clinical evaluation. In comparison with the gadolinium-based albumin binding agent MS 325 (EPIX Medical, Cambridge, MA), NC100150 is an iron oxide-based contrast agent. By combining a small particle size (<20 nm) with a relatively high r1 and low r2, NC100150 meets the demands of an effective preparation for 3D MR angiography. Contrary to protein-bound agents, NC100150 remains totally insulated from saturation problems at higher doses, which can result in extravasation of unbound agent into extravascular spaces. However, the iron-particle nature of the agent requires use of short TEs (<2 msec) at high field strengths to prevent inherent T2 and T2* shortening. The use of partial echo-acquisition techniques is thus recommended, particularly if imaging in the early arterial phase is to be attempted [13].
The safety profile of NC100150 is favorable as documented in phase I [13] and ongoing phase II studies. There were no serious adverse events in this study. None of the mild events required any treatment. Clearly, more data will need to be provided before final conclusions regarding the safety of NC100150 can be drawn.
IV administration of NC100150 in a dose of 5 mg Fe/kg of body weight adds 350 mg or 9% of the body's iron burden in a 70-kg man. This increase will not significantly alter either iron status or metabolism [14]. Accordingly, in our patients laboratory analysis merely revealed a transient increase in total iron; the rapid decrease to normal values in most patients within 72 hr and simultaneous increase in ferritin reflects the incorporation of iron particles into the normal iron metabolism.
The equilibrium-phase MR images showed good image quality in all 3D data sets with high spatial resolution. This is mirrored by overall signal-to-noise and contrast-to-noise ratios generally exceeding 40 and 30, respectively. The values measured reflect the high T1 shortening of NC10050 and indicate that NC100150 is largely retained within the vasculature and the capillary bed remains intact. Data from the previous phase I study have shown a relative stability of contrast-to-noise ratio over an imaging period of 90 min [13]. Thus, the data acquisition is independent of issues related to timing of the IV contrast material administration and manual and automated timing protocols [7, 10, 15,16,17] become superfluous. Furthermore, the long imaging window provided by NC100150 permits repetition of the acquisition without sacrifice in image quality. Failure of the patient to comply with breathing commands are without consequence because the data collection may simply be repeated. By the same token, imaging of multiple vascular territories can be accomplished or imaging of equivocal findings may be repeated. Finally, the homogeneous contrast concentration during the equilibrium phase insulates the 3D acquisitions from K-space modulation effects, which are capable of introducing blurring and other artifacts [18]. This aspect was particularly evident when analyzing renal arteries for the presence of arterial changes in fibromuscular dysplasia (Fig. 4A,4B). Although assessment of renal arteries may be hampered by venous overlap, specifically in the middle and distal thirds of the renal arteries, the use of blood pool contrast agents may enhance the ability of 3D MR angiography to detect fibromuscular dysplastic changes, which may be difficult to detect using conventional extracellular paramagnetic agents [19].
|
|
Equilibrium-phase NC100150-enhanced 3D MR data sets reliably revealed aneurysmal changes. The frequency of external iliac aneurysms in our patients was higher than that previously reported [20]. All abdominal aortic aneurysms were also correctly identified with regard to location and extent of the aneurysm. Obtaining a second transaxial gradient-recalled echo or spin-echo data set for assessment of the arterial wall in aortic aneurysms should be considered mandatory because NC100150-enhanced MR angiography data sets merely provide endoluminal information. The presence and extent of thrombosed regions cannot be ascertained. The use of surface shaded-display images can provide a more 3D rendition of the underlying morphology. Although these displays may aid in surgical planning, they can be considered of limited diagnostic value.
Sensitivity and specificity for the detection of luminal narrowing of at least 50% in the aortoiliac and renal arteries were somewhat inferior compared with those reported in most studies evaluating first-pass 3D MR angiography on the basis of extracellular contrast agents [7, 21]. The somewhat limited diagnostic performance of NC100150-enhanced 3D MR angiography reflects the presence of venous overlap, particularly of the renal and pelvic vessels (Fig. 5A,5B). In addition, despite careful multiplanar reformation of all NC100150-enhanced data sets, resulting in a significant prolongation of data interpretation times, several arterial segments simply could not be assessed. Because these segments were not compared with digital subtraction angiographic images, the overall performance was limited. Segmentation approaches, possibly based on dynamic early arterial enhancement profiles, may solve this problem in the future.
|
|
An optimal approach to imaging the arterial system in the abdomen with a blood pool agent may combine arterial phase first-pass imaging with subsequent targeted equilibrium-phase imaging. Bolus application of NC100150 up to a dose of 5 mg Fe/kg of body weight has been shown to be safe in phase I and some of the ongoing phase II trials.
We acknowledge several limitations of this study. Inclusion of patients primarily based on the presence of a hemodynamically significant arterial stenosis or an arterial aneurysm as revealed by digital subtraction angiography introduced a selection bias. Although each interpreter knew that disease was present in at least one anatomic segment, the bias effect was mitigated by the large number of evaluated segments and the frequency of encounters with less severe concomitant vascular disease. A further limitation relates to the limited prevalence of hemodynamically significant arterial stenosis affecting 37 of the 346 segments. Clearly, our data need to be confirmed in larger series. Finally, the MR interpretations were performed by a single reviewer. Interobserver reliability needs to be assessed in further studies.
The availability of blood pool contrast agents such as NC100150 will open up diagnostic opportunities well beyond assessment of the arterial system in the abdomen and pelvis. Thus, portal and systemic venous imaging may be facilitated. Preliminary data also suggest the use of NC100150 for coronary artery and pulmonary imaging [22, 23]. In addition, intravascular agents are poised to play a dominant role in the detection and localization of intestinal and peritoneal hemorrhage [24] and MR-based guidance, monitoring of intravascular interventions [25], and performing MR guided biopsies. Finally, blood pool agents of different particle sizes may aid in characterizing perfusion properties of ischemic or even neoplastic tissues [26].
We conclude that equilibrium-phase NC100150-enhanced 3D MR angiography provides good-quality displays of the abdominal and pelvic arterial systems. Equilibrium-phase NC100150-enhanced 3D MR angiography shows high specificity when evaluating the abdominal and pelvic vascular systems, but the attendant venous overlap can limit the assessment of stenosis in renal and pelvic arterial segments.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Visser, M. C. J. M. Kock, K. M. Kuntz, M. C. Donaldson, G. S. Gazelle, and M. G. Myriam Hunink Cost-effectiveness Targets for Multi-Detector Row CT Angiography in the Work-up of Patients with Intermittent Claudication Radiology, June 1, 2003; 227(3): 647 - 656. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. J. Kroencke, M. N. Wasser, P. M. T. Pattynama, J. O. Barentsz, E. Grabbe, G. Marchal, M. V. Knopp, G. Schneider, L. Bonomo, D. J. Pennell, et al. Gadobenate Dimeglumine--Enhanced MR Angiography of the Abdominal Aorta and Renal Arteries Am. J. Roentgenol., December 1, 2002; 179(6): 1573 - 1582. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
