| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology and Nuclear Medicine, Hospital Brothers of St. John of
God, Grosse Mohrengasse 9, Vienna, Austria A-1020.
2 Department of Diagnostic Radiology, Eberhardt Karls University, Tuebingen,
Germany.
3 Department of Diagnostic Radiology, Homburg University Hospital, Homburg/Saar,
Germany.
4 Department of Radiology, Hospital Valduce, Como, Italy.
5 Institute for Diagnostic Radiology, University Hospital, Basel,
Switzerland.
6 Medical Prevention Center, University Mxedical Center, Hamburg-Eppendorf,
Germany.
7 Institute of Clinical Radiology, Ludwig Maximilians University, Munich,
Germany.
8 U.C.O. di Radiologia, Ospedale di Cattinara, Trieste, Italy.
9 Department of Radiology, Ospedale Cà Foncello, Treviso, Italy.
10 Crawford Long Hospital, Atlanta, GA.
11 Department of Radiology, Università G. D'Annunzio, Chieti, Italy.
12 Institut für Radiologie, Charité-Universitäts medizin,
Berlin, Germany.
13 Department of Radiology, Johns Hopkins Medical Center, Baltimore, MD.
14 Klinikum der Johannes, Gutenberg Universität, Mainz, Germany.
15 Worldwide Medical and Regulatory Affairs, Bracco Imaging SpA, Milan,
Italy.
16 Worldwide Medical and Regulatory Affairs, Bracco Diagnostics Inc., Princeton,
NJ.
Received March 5, 2007;
accepted after revision May 19, 2007.
G. Schneider is a speaker for Bracco Imaging, D. A. Bluemke is a consultant
for Bracco, and M. A. Kirchin and G. Pirovano are salaried employees of
Bracco.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Two hundred seventy-two patients underwent MR
angiography and digital subtraction angiography of the iliofemoral arteries.
MR angiography was performed before (2D time-of-flight acquisitions) and after
(spoiled gradient-echo acquisitions) the administration of 0.1 mmol/kg of
gadobenate dimeglumine at 1–2 mL/s. Contrast-enhanced MR angiography and
digital subtraction angiography of the calf arteries were performed in 241 of
272 participants. Images were evaluated on-site and by four blinded reviewers
(three for MR angiography, one for digital subtraction angiography).
Comparative diagnostic performance for the detection of significant (
51%
vessel lumen narrowing) disease was evaluated using the McNemar test and
generalized estimating equations. Interobserver agreement was assessed with
generalized kappa statistics. The chi-square test was used to compare
technical failure rates.
RESULTS. Digital subtraction angiography confirmed significant
disease (597 stenoses, 386 occlusions) in 983 iliofemoral segments. The
sensitivity (54–80.9%), specificity (89.7–95.3%), and accuracy
(85–87.5%) of contrast-enhanced MR angiography for the detection of
significant iliofemoral disease were significantly (p < 0.001, all
reviewers) better than those of time-of-flight MR angiography
(33.2–62.8%, 74.3–88.9%, and 68–77.3%, respectively).
Similar diagnostic performance was obtained for the calf arteries. The
technical failure rate with contrast-enhanced MR angiography (2.5–3.4%)
was similar to that of digital subtraction angiography (1.4%) and
significantly (p < 0.001) lower than that of time-of-flight MR
angiography (6.2–18.0%). Significantly better reproducibility
(p < 0.001) was obtained with contrast-enhanced MR angiography
(82% vs 65.2% agreement; 
= 0.66 vs 0.45).
CONCLUSION. Improved diagnostic performance and reproducibility are achievable with gadobenate dimeglumine–enhanced MR angiography in patients with peripheral arterial occlusive disease.
Keywords: contrast agents • diagnostic performance • gadobenate dimeglumine • MR angiography • peripheral arterial occlusive disease
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In patients with symptoms suggestive of peripheral arterial occlusive disease, information about the number and severity of vascular lesions is essential for planning appropriate revascularization therapy. Traditionally, assessment of peripheral arterial occlusive disease before treatment has been performed with conventional catheter angiography. However, conventional angiography is a highly invasive procedure that carries substantial risk to the patient [2]. The advent of alternative minimally invasive procedures such as MDCT angiography (MDCTA) [3–6]and contrast-enhanced MR angiography (CE-MRA) [7–16] has markedly reduced the need for preoperative diagnostic catheter angiography, effectively limiting its use to patients undergoing interventional treatment.
Of the minimally invasive techniques available, CE-MRA has the advantage of not requiring ionizing radiation or large volumes of iodinated contrast material. Moreover, with improvements in MRI hardware and sequence design [17–20]that permit greater spatial resolution and faster image acquisition, CE-MRA is increasingly considered the method of choice for imaging large vascular territories such as the peripheral runoff vessels. The advent of MR contrast agents with beneficial properties for vascular imaging may further improve the diagnostic impact of CE-MRA.
Gadobenate dimeglumine (MultiHance, Bracco Imaging) is a gadolinium contrast agent whose r1 relaxivity in blood is roughly two times higher than the r1 values of conventional gadolinium contrast agents at available magnetic field strengths up to 3 T [21, 22]. The increased r1 relaxivity derives from weak and transient interaction of the Gd-BOPTA contrast-effective chelate of gadobenate dimeglumine with serum albumin [23] and results in significantly better vascular contrast enhancement and better vessel delineation than that achieved with conventional gadolinium agents at equivalent or higher doses [24–28]. In the peripheral vasculature, a 0.1 mmol/kg dose of gadobenate dimeglumine has been shown to be superior to an equivalent dose of gadopentetate dimeglumine in terms of diagnostic image quality [27] and to permit better visualization of the arterial vasculature, particularly in the lower runoff territory [27, 28].
Previously, sensitivity and specificity values of 94% and 89–93% have
been reported for the diagnostic accuracy of gadobenate
dimeglumine–enhanced MR angiography in patients with peripheral arterial
occlusive disease [28].
However, these values were obtained in a relatively small single-center
population of just 28 patients using a standard contrast agent volume of 34 mL
per patient. Our study was performed in a much larger multinational,
multicenter patient population using a standard gadobenate dimeglumine dose
per patient of 0.1 mmol/kg. Values for sensitivity, specificity, and overall
diagnostic accuracy for the detection of significant ( 51%) stenoocclusive
disease were determined using digital subtraction angiography as the reference
standard and were compared with values obtained using unenhanced 2D
time-of-flight MR angiography (TOF MRA).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Study Population
All patients were enrolled between May 2003 and November 2004. Men and
women were eligible for enrollment if they were 18 years old or older and had
known or suspected peripheral arterial occlusive disease in the iliofemoral
arteries based on clinical examination or sonographic findings. All patients
were required to undergo a conventional digital subtraction angiography
examination within 1–30 days before or after the CE-MRA examination and
to exhibit no change in clinical symptoms related to peripheral arterial
occlusive disease between the two procedures. Patients were not permitted to
undergo any therapeutic intervention for vascular disease between the MR
angiography and the digital subtraction angiography procedures or any other
surgical procedure within 24 hours after the administration of gadobenate
dimeglumine.
Patients with known allergies to one or more of the study agent ingredients or a known history of hypersensitivity to metals, including gadolinium or iodinated contrast media, were ineligible for inclusion, as were patients who received any other investigational agent within 30 days before the study or any other contrast agent within 24 hours before or after gadobenate dimeglumine administration. Similarly, patients who suffered severe claustrophobia, had class III or IV congestive heart failure according to the American Heart Association classification [31], or had a pacemaker, metallic cardiac valve, or metallic vascular stent in one or more of the vessels of interest were also ineligible for inclusion. Finally, pregnant or lactating women were ineligible for inclusion, as were patients with any medical condition or other circumstances that would significantly decrease the chances of obtaining reliable data or of achieving study objectives. A total of 294 patients with known or suspected peripheral arterial occlusive disease based on clinical or sonographic findings were enrolled. Of these 294 patients, 287 (207 men, 80 women; mean age, 65.7 ± 9.95 years; range, 40–93 years) underwent TOF MRA and CE-MRA.
MR Angiography
The 287 participants underwent MR angiography at 1.5 T on commercially
available MR scanners equipped with a gradient of 20 mT/m. The MR
scanners used for the study were from Siemens Medical Solutions (Symphony,
n = 69 [24.0%]; Sonata, n = 75 [26.1%]; Avanto, n =
12 [4.2%]), Philips Medical Systems (Gyroscan Intera, n = 69
[24.0%]), or GE Healthcare (Genesis Signa, n = 40 [13.9%]; Excite,
n = 22 [7.7%]).
MR angiography was performed using a 2D TOF MRA sequence before contrast agent administration and a 3D spoiled gradient-echo MR angiography sequence immediately after administration of gadobenate dimeglumine. The large number of investigating centers involved in the study and the wide variety of imaging systems used resulted in necessarily slightly different parameters among centers for the TOF MRA and CE-MRA sequences. Nevertheless, each sequence at each center was selected to meet minimal requirements for image acquisition and interpretability.
The parameters for the TOF MRA sequence varied among centers as follows:
axial orientation; TR/TE range, 
60/4.2–7.2; flip angle,
30–70°; excitations, 1–2; slice thickness, < 4 mm; matrix,
256 x 160; overall acquisition time, 5–12 minutes. ECG gating
was performed for 75% of the patients. CE-MRA of the peripheral arteries was
performed with a dedicated phased-array peripheral coil (225 patients) or with
a body coil (62 patients) and a bolus chase technique using the following
sequence parameters: coronal orientation; TR range/TE range,
2.3–6/0.78–2.15; flip angle, 25–45°; excitations,
0.5–1; slice thickness, 1–3.5 mm; matrix, 256 x 224;
true in-plane spatial resolution, 0.68 x 0.68–1.3 x 1.3 mm;
overall acquisition time, 50 seconds. The iliofemoral field of view for
both the TOF MRA and CE-MRA sequences was tailored for each patient to include
arterial vasculature from 2 cm above the aortic bifurcation to a point on the
popliteal artery at the level of the knee joint line. MR angiography of the
calf arteries was optional at all investigational centers as a secondary
acquisition after full CE-MRA of the iliofemoral arteries.
The CE-MRA sequence was acquired after the administration of gadobenate dimeglumine at a dose of 0.1 mmol/kg of body weight. Contrast agent administration was performed using a power injector at a rate of 1–2 mL/s, followed by a 20-mL saline flush at the same rate. Timing for the CE-MRA sequence was achieved by means of a bolus timing acquisition (n = 173 participants) or through the use of an automatic or MR fluoroscopic bolus detection technique (SmartPrep [GE Healthcare], BolusTrak [Philips Medical Systems], or CARE Bolus [Siemens Medical Solutions], depending on the scanner manufacturer; n = 114 subjects). The test bolus timing approach involved acquisition of 45–60 dynamic single-slice T1-weighted fast gradient-echo images of the common femoral artery at a frequency of one image per second after the administration of a 1- to 2-mL bolus of gadobenate dimeglumine.
Digital Subtraction Angiography
Conventional digital subtraction angiography was performed by injecting an
iodinated contrast medium through a pigtail or straight 4- to 5-French
catheter inserted via a femoral artery puncture using the Seldinger technique.
The catheter tip was positioned in the abdominal aorta 5–10 cm above the
aortic bifurcation. Anteroposterior, right anterior oblique, and left anterior
oblique projections at angulations of 15–30° were obtained of the
aortoiliac station as appropriate according to each center's standard
operating procedure. Anteroposterior projections were obtained for the upper
and lower leg stations. Most digital subtraction angiography examinations were
performed using iodinated contrast media having iodine concentrations of >
200 mg I/mL (200–300 mg I/mL in 58% of the subjects; > 300 mg I/mL in
42% of the subjects). The total volume of contrast medium administered was
50–200 mL. Individual injections of 15–40 mL were administered at
rates of 4–12 mL/s depending on the vessel of interest.
Image Evaluation
Images were evaluated by on-site investigators and by four off-site
independent, experienced (at least 10 years of experience in vascular imaging)
board-certified radiologists (three for MR angiography, one for digital
subtraction angiography) who were not affiliated with any of the study sites
and who were fully blinded to all patient information and to the results of
other diagnostic procedures.
Off-site evaluation of digital MR angiography and digital subtraction angiography images was performed at an independent core imaging laboratory equipped with two separate Windows (Microsoft)-based workstations (AquariusNet Viewer, Tera-Recon) for evaluation of images (two monitors) and for recording of assessment findings using an electronic Case Report Form (e-CRF) system. All TOF MRA and CE-MRA images were combined into a single randomization pool, and each image set for each patient was reviewed separately, one at a time and in random order. In each case, axial source images and volumetric maximum-intensity-projection (MIP) reconstructions were always displayed on the two monitors set up for image evaluation. All routine image review tools (window and level, zoom, pan, and so forth) were available to the reviewers.
The three off-site reviewers of MR angiography images performed their evaluations independently in a fully blinded fashion. Evaluation of the iliofemoral arterial anatomy from 2 cm above the aortic bifurcation to the popliteal arteries at the level of the knee joint was performed on a segmental basis in which standard segments comprised the left and right common, internal, and external iliac arteries; the left and right common, superficial, and deep femoral arteries; and the left and right popliteal arteries.
Initial off-site evaluation was performed to determine the technical
adequacy (quality of visualization) of the TOF MRA and CE-MRA image sets. If
any segment was not entirely in the field of view or was considered
technically inadequate for any reason, no further assessment was performed for
that segment. Assessment of all segments considered technically adequate was
then performed using a 3-point scale in which 1 = stenosis of 50% (vessel
with no clinically significant disease), 2 = stenosis of 51–99% (vessel
with clinically significant disease), and 3 = occlusion (vessel with 100%
blockage of the vessel lumen). Evaluation of calf vessels was performed using
similar assessment methodology with the calf vasculature divided into segments
comprising the left and right tibiofibular trunk, the left and right anterior
and posterior tibial arteries, and the left and right peroneal arteries.
The presence and location of collateral circulation was assessed in a yes-or-no manner. Collateral circulation was assigned to one of the following locations per side (right and left): station 1 (from the abdominal aorta to the external iliac artery), between stations 1 and 2 (from the abdominal aorta to the popliteal artery), station 2 (from the common femoral artery to the popliteal artery), between stations 2 and 3 (from the common femoral artery to the tibial arteries), and station 3 and below (from the tibiofibular trunk to downstream). Finally, the presence and type of associated disease in each iliofemoral segment was recorded as none, aneurysm, dissection, or other.
Off-site evaluation of digital subtraction angiography images was performed by a fourth independent radiologist who had 12 years of experience in angiographic procedures, was not affiliated with the study centers, and was blinded to all clinical and radiologic information. Digital subtraction angiography images were combined into a second pool, different from the MR angiography pool, for blinded reviewing purposes, but were evaluated using similar assessment methodology and criteria.
On-site evaluation of MR angiography images was performed using similar criteria to those of the off-site evaluation. Evaluation of MR angiography and digital subtraction angiography images was performed independently by two experienced radiologists (one for each technique) at each investigational site. Each reviewer was fully blinded to the results of the other imaging technique.
Safety Evaluations
Physical examination was performed within 24 hours before gadobenate
dimeglumine administration and at 24 hours after administration. Measurement
of vital signs (blood pressure, heart rate) was performed within 24 hours
before dosing and before the participant entered the magnet, and at 30
minutes, 1 hour, and 24 hours after gadobenate dimeglumine administration.
Recording of ECGs was similarly performed before the patient entered the bore
of the magnet and at 1 hour and 24 hours after administration of gadobenate
dimeglumine.
In addition, blood and urine samples were collected within 24 hours before gadobenate dimeglumine administration and at 24 hours after administration. Laboratory evaluation of collected samples was performed for hematology (hematocrit, hemoglobin, and counts of RBCs, WBCs, and platelets), blood chemistry (glucose, creatinine, total bilirubin, total protein, albumin, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, gamma glutamyl transpeptidase, sodium, potassium, and chloride), and urinalysis (protein, glucose, ketones, blood, and pH).
Finally, the safety of gadobenate dimeglumine was assessed in terms of the incidence of adverse clinical events from the time of signed informed consent until 24 hours after gadobenate dimeglumine administration. Adverse events were classified as serious (i.e., death, life-threatening, or requiring or prolonging hospitalization) or not serious (rated as mild, moderate, or severe). The relationship of each adverse event to the study agent was classified as probable, possible, not related, unknown, or missing.
Statistical Analysis
The primary objectives were to determine the diagnostic accuracy of CE-MRA
with gadobenate dimeglumine for detection of significant stenoocclusive
disease (defined as stenosis of 51% or occlusion) of the iliofemoral
arteries using digital subtraction angiography as a reference standard, and to
compare the diagnostic performance of CE-MRA with that of unenhanced TOF MRA.
Data from each of the three off-site MR angiography reviewers and from the
on-site investigators were analyzed and presented separately.
Sensitivity for detection of significant stenoocclusive disease was defined
as the number of correctly identified significantly ( 51%) diseased
segments on TOF MRA or CE-MRA divided by the total number of significantly
( 51%) diseased segments on digital subtraction angiography. Specificity
was defined as the number of correctly identified segments in TOF MRA or
CE-MRA that were not diseased or not significantly (< 51%) diseased divided
by the total number of segments on digital subtraction angiography that were
not diseased or not significantly (< 51%) diseased. Accuracy was defined as
the number of correctly identified segments (either diseased or nondiseased)
on MR angiography divided by the total number of segments evaluated on digital
subtraction angiography. All uninterpretable MR angiography images were
considered inaccurate for all determinations of diagnostic performance. If a
segment was technically inadequate on MR angiography, this segment was
considered false-positive if the corresponding digital subtraction angiography
revealed a stenosis of 50%; however, this segment was considered
false-negative if the corresponding digital subtraction angiography revealed a
stenosis of 51% or occlusion.
The sensitivity, specificity, and accuracy of TOF MRA and CE-MRA were compared using the McNemar test. In addition, supplemental supporting analyses of sensitivity and specificity were performed using generalized estimating equations (GEEs) [32] to eliminate potential correlation-related bias caused by evaluation of multiple segments for each patient. "Reviewer" was considered a fixed effect in the GEE model.
Determinations of positive predictive value (PPV), negative predictive
value (NPV), positive likelihood ratio, and negative likelihood ratio were
performed and compared descriptively for the two MR angiography sequences.
Determination of interobserver agreement was evaluated by means of the
generalized kappa () coefficient and by the percentage of concordance
among the three MR angiography reviewers.
The technical failure rate for each MR angiography sequence was defined as the total number of technically inadequate segments divided by the total number of segments included in the field of view. Comparison of the technical failure rate for TOF MRA with that for CE-MRA for the iliofemoral arteries was performed using the chi-square test.
A power calculation was performed on the basis of the assumption that the
expected difference in sensitivity for detecting significant ( 51%)
stenoocclusive disease between TOF MRA and CE-MRA was 10%. Assuming that the
proportion of discordant pairs was 0.26 and that the sensitivity of TOF MRA
was 0.70, then for a two-sided 0.05 alpha level McNemar test of equality of
paired proportion, 225 participants with at least one significant stenosis
were required for statistical power of 85%. Assuming a 10% dropout rate, at
least 265 subjects were required to enter the study. The total number of
subjects assessed for sensitivity was also sufficient for specificity because
each subject could have many negative segments contributing to
specificity.
All statistical analyses were performed using the statistical software package SAS version 8.2 (SAS Institute).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Most patients (171/287, 59.6%) were 65 years old or older; the remainder were 40–64 years old. Most patients who received gadobenate dimeglumine (165/287, 57.5%) presented with moderate to severe claudication (stage IIb according to the classification of Fontaine et al. [33]). The remaining patients had mild claudication (stage IIa; 53/287, 18.5%), ischemic pain at rest (stage III; 38/287, 13.2%), ulceration or gangrene (stage IV; 29/287, 10.1%), or were asymptomatic (2/287, 0.7%) at presentation. Most participants underwent MR angiography to confirm or evaluate a previously detected stenosis (131/287, 45.6%) or because of clinical symptoms suggestive of stenosis (117/287, 40.8%). A smaller proportion of participants (39/287, 13.6%) underwent MR angiography to guide revascularization or as follow-up to a previous treatment.
Of the 287 participants to undergo MR angiography, only 272 (94.8%) also underwent the required digital subtraction angiography examination. Consequently, assessment of diagnostic performance was performed for 272 participants overall.
Technical Adequacy and Quality of Segment Visualization
The technical adequacy of TOF MRA and CE-MRA for evaluation of the
iliofemoral arteries in all 287 participants who received gadobenate
dimeglumine is shown in Table
1. The technical failure rate of CE-MRA for reviewers 1, 2, and 3
(2.8%, 2.5%, 3.4%, respectively) was in all cases significantly (p
< 0.001) lower than the failure rate of TOF MRA (18.0%, 11.3%, 6.2%,
respectively). Similar findings were noted by the on-site investigators (6.6%
of iliofemoral segments were considered inadequate on CE-MRA compared with
32.2% on TOF MRA; p < 0.001). Overall, the technical failure rate
of CE-MRA for the iliofemoral arteries was low in absolute terms for each
reviewer and comparable to the failure rate of digital subtraction angiography
(1.4%).
|
Off-site assessment of the calf arteries revealed slightly greater numbers of technically inadequate segments for both TOF MRA and CE-MRA when compared with the iliofemoral arteries (technical failure rate, 45.0%, 26.6%, and 25.3% for TOF MRA compared with 12.0%, 8.2%, and 14.2% for CE-MRA rated by reviewers 1, 2, and 3, respectively). However, a similar trend for significantly (p < 0.001, all reviewers) lower numbers of technically inadequate images on CE-MRA was apparent. Moreover, the number of technically inadequate calf segments on CE-MRA was comparable to the failure rate of digital subtraction angiography (10.2%).
Diagnostic Performance
Iliofemoral arteries—Of the 272 participants to undergo both
MR angiography and digital subtraction angiography, only eight (2.9%) did not
have any iliofemoral segment with significant ( 51%) stenoocclusive
disease; the remaining 264 (97.1%) participants had at least one iliofemoral
segment with significant disease. These 264 participants comprised 14 (5.1%)
who had one segment with clinically significant disease, 29 (10.7%) who had
two segments with clinically significant disease, 43 (15.8%) who had three
segments with clinically significant disease, 41 (15.1%) who had four segments
with clinically significant disease, and 137 (50.4%) who had five or more
segments with clinically significant disease.
Overall, 4,003 iliofemoral segments were considered in the field of view on
digital subtraction angiography. Of these, 2,962 (74.0%) segments were
considered to be without significant disease, whereas 983 (24.6% segments were
considered to have significant disease (597 [14.9%] segments with significant
( 51%) stenosis, 386 [9.6%] segments with occlusions). The remaining 58
(1.4%) segments were considered to be technically inadequate.
The diagnostic performance of MR angiography for the detection of significant stenoocclusive disease of the iliofemoral arteries using digital subtraction angiography as the reference standard is summarized in Table 2. All off-site reviewers and on-site investigators obtained significantly (p < 0.001) higher sensitivity, specificity, and overall accuracy for the detection of significant stenoocclusive disease for CE-MRA compared with TOF MRA. After eliminating the possibility of within-cluster correlation effect, the results of the GEE model analysis confirmed the significantly better sensitivity (odds ratio [OR], 4.3 [95% CI, 3.4–5.5]) and specificity (OR, 3.6 [2.7–4.8]) of CE-MRA compared with TOF MRA. Overall increases in accuracy of 17.0% (15.5–18.5%), 15.9% (14.2–17.5%), and 9.9% (8.5–11.3%) were determined for off-site reviewers 1, 2, and 3, respectively, whereas a greater increase in accuracy of 26.1% (24.4–27.8%) was determined for the on-site investigators. Examples of the improved image quality achievable on CE-MRA with 0.1 mmol/kg of gadobenate dimeglumine compared with TOF MRA and of the excellent correlation of gadobenate dimeglumine–enhanced MR angiography and digital subtraction angiography are shown in Figures 1A, 1B, 1C, 1D, 2A, 2B, 2C, 2D, 2E, 2F, 2G, 3A, 3B, 3C, 3D, 3E, 3F, 3G, 4A, 4B.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
A significant (p < 0.00001) increase in agreement among the
three MR angiography reviewers was noted for the CE-MRA image sets (82%
agreement; , 0.66) compared with the TOF MRA image sets (65.2%
agreement; , 0.45) (Table
3).
|
Determinations of PPV, NPV, positive likelihood ratio, and negative likelihood ratio confirmed the better performance for CE-MRA compared with TOF MRA for both off-site reviewers and on-site investigators (Table 4).
|
Calf arteries—CE-MRA of the calf arteries was performed for 263 (91.6%) of the 287 participants who underwent CE-MRA of the iliofemoral arteries. Conversely, only 141 (49.1%) participants underwent TOF MRA of the calf arteries. Digital subtraction angiography correlation was available for 241 (91.6%) of the 263 participants who underwent CE-MRA of the calf arteries.
A total of 1,507 calf artery segments were considered to be in the field of
view on digital subtraction angiography. Significant stenoocclusive disease
was noted in 465 (30.9%) of these 1,507 segments (150 [10.0%] segments with
significant ( 51%) stenosis, 315 [20.9%] segments with occlusions),
whereas 889 (59.0%) of 1,507 segments were considered to be without
significant disease. The remaining 153 (10.2%) segments were considered
technically inadequate.
The diagnostic performance of MR angiography for the detection of significant stenoocclusive disease of the calf arteries relative to digital subtraction angiography is summarized in Table 5. Although slightly lower values for sensitivity, specificity, and accuracy were obtained compared with values obtained for the iliofemoral arteries, significantly better performance was noted in all cases by each MR angiography reviewer for CE-MRA compared with TOF MRA. Similar trends to those observed in the iliofemoral arteries were also noted for PPV, NPV, positive likelihood ratio, and negative likelihood ratio.
|
Collateral Circulation and Associated Disease
Collateral circulation was detected in a greater percentage of stations on
CE-MRA (8.5–10.3% of the stations examined at MR angiography and digital
subtraction angiography) compared with TOF MRA (3.7–4.0% of the
stations) across the three blinded MR angiography reviewers. A similar trend
was noted by on-site investigators (26.8% on CE-MRA compared with 10.1% on TOF
MRA). Accuracy for the detection of collateral circulation ranged from 82.0%
to 89.0% for CE-MRA and from 71.5% to 88.3% for TOF MRA when compared with the
collateral circulation detected on digital subtraction angiography. The
difference in accuracy between TOF MRA and CE-MRA for the detection of
collateral circulation was significant for MR angiography reviewer 3
(p = 0.027) and for the on-site investigators (p < 0.001)
but not for MR angiography reviewers 1 and 2.
Blinded evaluation of digital subtraction angiography images revealed associated disease in 172 (4.4%) of 3,945 iliofemoral segments (aneurysm in 45 [1.1%] segments and associated disease classified as "other" in 127 [3.2%] segments). In 3,773 (95.6%) segments, no associated disease was present. The three blinded MR angiography reviewers detected associated aneurysms in 43 (1.1%), 22 (0.6%), and 34 (0.9%) segments, respectively, on CE-MRA and in three (0.1%), 18 (0.5%), and eight (0.2%) segments, respectively, on TOF MRA. No associated disease was recorded for 98.3–98.8% of segments on CE-MRA and 99.2–99.7% of segments on TOF MRA.
Safety
A total of 43 nonserious adverse events were reported by 30 (10.5%) of 287
participants overall, of which 32 events in 22 (7.7%) participants were
considered of probable, possible, unknown, or missing relationship to the
administration of gadobenate dimeglumine. The remaining events were considered
unrelated. All events were either mild (87%) or moderate (13%) in intensity,
and no serious adverse events were reported. The most commonly reported events
that were considered of potential relationship to the administration of
gadobenate dimeglumine were injection site warmth (1.7%, 5/287), increased
systolic blood pressure (1.4%, 4/287), increased diastolic blood pressure
(1.0%, 3/287), and nausea (0.7%, 2/287). All increases in blood pressure
reported as adverse events were recorded in one participant at one site. No
other event was reported by more than one participant. No clinically
meaningful time-related changes were noted for vital signs or clinical
laboratory investigations, and no significant effects were noted for any
cardiac electrophysiology parameter.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
51%
is considered the threshold for defining patients for whom therapeutic
intervention should be considered. Accordingly, our study was performed
predominantly in patients with moderate to severe peripheral arterial
occlusive disease (232/287 [80.8%] patients with stage IIb–IV disease
according to the staging of disease by Fontaine et al.
[33]) because these patients
are the population most likely to undergo routine CE-MRA of the peripheral
vasculature. Although our primary objective was to assess the diagnostic performance of CE-MRA with 0.1 mmol/kg of gadobenate dimeglumine relative to conventional digital subtraction angiography, additional comparison with unenhanced TOF MRA was performed to meet guidelines issued by the U.S. Food and Drug Administration (FDA) [34] and the Committee for Proprietary Medicinal Products (CPMP) [35] concerning the development of diagnostic imaging agents. However, because the unacceptably long acquisition times of TOF MRA sequences precluded successful unenhanced TOF imaging of the entire runoff vasculature including the calf station in a large number of patients (only approximately half the patients underwent TOF MRA of the calf arteries, and many could not lie still in the bore of the magnet for the entire acquisition time of the sequence), the primary focus of our study was the iliofemoral vasculature, comprising the aortoiliac and femoropopliteal stations.
As was to be expected on the basis of findings from previous comparisons of
CE-MRA with unenhanced MR angiography
[36–38],
CE-MRA of the iliofemoral arteries with gadobenate dimeglumine was
significantly (p < 0.001) better than TOF MRA for all main
measures of diagnostic performance (i.e., sensitivity, specificity, and
accuracy for detection of significant stenoocclusive disease). Each of the
three off-site blinded reviewers reported values of 85% for the
diagnostic accuracy of CE-MRA for the detection of significant stenoocclusive
disease of the iliofemoral arteries compared with digital subtraction
angiography. These values compare favorably with values reported elsewhere for
CE-MRA with other gadolinium-based MR contrast agents
[8,
9,
39] and imply that correct
classification of patients in terms of the need for interventional treatment
or conservative follow-up is achievable in at least 85% of cases. Although the
on-site investigators reported a slightly lower combined value for overall
accuracy (81.3%), these investigators, like the off-site reviewers, were fully
blinded to the results of the digital subtraction angiography procedure.
Because of the artificial full-blinding conditions of the off-site reviewers to all patient clinical and radiologic information in this study, a reduction of the diagnostic accuracy by 10–15% might not be unexpected [40]. Notably, a recent study comparing the conventional extracellular gadolinium contrast agents gadodiamide and gadopentetate dimeglumine for the detection of hemodynamically relevant stenosis reported accuracy values of 71–100% for the common iliac arteries, 57–71% for the external iliac arteries, and 12–50% for the internal iliac arteries [41]. Moreover, two recent phase III clinical trials performed to obtain regulatory approval for gadofosvesetfor peripheral CE-MRA determined diagnostic accuracy values of 83.8–90.3% [42] and 80.3–87.6% [43] for the detection of significant aortoiliac occlusive disease. Each of these studies was performed using an MR angiography protocol design similar to that used in our study, using similar CE-MRA sequences and similar assessment methodology involving fully blinded evaluation of all MR angiography images by three highly experienced independent reviewers.
Regarding other parameters of diagnostic performance, whereas the values
for specificity ( 89.7% for all three off-site reviewers) were in line
with values reported elsewhere
[8–12,
37,
39,
41–45],
slightly lower values for sensitivity (33.2–62.8% for TOF MRA,
54.0–80.9% for CE-MRA) were obtained. In particular, the sensitivity for
off-site reviewer 1 was lower than the values determined for the other two
reviewers. On the other hand, strong interreviewer agreement for the diagnosis
of clinically significant stenoocclusive disease was shown both for agreement
between two reviewers and for agreement among all three reviewers.
Specifically, all three blinded reviewers agreed in 82% ( = 0.66) of
segment evaluations on CE-MRA compared with just 65.2% ( = 0.45) on TOF
MRA. On the basis of the guidelines provided by Landis and Koch
[46] for describing the
clinical value of degree of concordance, the kappa value obtained in this
study leads to the conclusion that gadobenate dimeglumine–enhanced MR
angiography is a diagnostic test with substantial reproducibility. Similar
consensus among reviewers was noted regarding the technical adequacy of CE-MRA
for visualization of the iliofemoral arteries. For each off-site blinded
reviewer, the technical failure rate reported for CE-MRA (2.5–3.4%) was
significantly (p < 0.001) lower than that reported for TOF MRA
(6.2–18.0%) and comparable to that reported for digital subtraction
angiography (1.4%).
Because sensitivity and specificity may provide an incomplete picture of the clinical usefulness of MR angiography, additional assessment was made of the PPV and NPV values. Although the PPV indicates the actual likelihood of disease in instances of a positive examination, the NPV indicates the likelihood of no disease in instances of a negative examination. The PPV determinations for the blinded reviewers in this study indicate that a positive iliofemoral segment on CE-MRA with gadobenate dimeglumine is likely to be significant stenoocclusive disease in approximately 80% of cases. These results can be judged positively, especially considering the artificial environment in which they were obtained. The NPV results (86.2–93.4%) indicate that the risk of overlooking stenoocclusive disease on CE-MRA with gadobenate dimeglumine is low. Therefore, a normal study at CE-MRA with gadobenate dimeglumine should obviate further potentially hazardous conventional angiographic or surgical procedures.
Differently from predictive values and values for sensitivity and
specificity, the values for positive likelihood ratio and negative likelihood
ratio are not affected by the prevalence of disease
[47]. Thus, determination of
these values offers an approach to assessing diagnostic performance that is
unaffected by the condition being evaluated in the population. The positive
likelihood ratio indicates the effect of a positive examination finding on the
probability that the condition in question exists, and the negative likelihood
ratio addresses the effect of a negative examination on the probability that
the condition in question is present. Likelihood ratio values therefore
provide quantification of the effect of MR angiography on diagnostic
thinking—that is, the impact of the MR angiography test result on the a
priori probability of clinically significant stenoocclusive disease versus the
a posteriori probability of such disease
[48]. A positive likelihood
ratio of 7.9 for each blinded reviewer in this study suggests that a
positive finding on CE-MRA of the iliofemoral arteries would in each case lead
to a moderate to large and often conclusive shift in the probability of
51% stenoocclusive disease being present. The PPV, NPV, and likelihood ratios
determined for CE-MRA were consistently superior to those determined for TOF
MRA, thereby providing supplemental confirmation of the superiority of CE-MRA
for diagnostic imaging of the iliofemoral arteries.
Concerning the calf arteries, excellent visualization has previously been shown with both unenhanced MR angiography [49–51] and CE-MRA [7, 11, 44, 52], although venous contamination and reduced signal-to-noise ratio (SNR) have sometimes been reported for the latter approach [53, 54]. In this study, markedly better diagnostic performance (sensitivity, specificity, and accuracy) and favorable predictive values and likelihood ratios were noted by all three off-site blinded reviewers for CE-MRA of the calf arteries compared with TOF MRA. As in the iliofemoral arteries, the technical failure rate determined by the off-site reviewers for CE-MRA of the calf arteries (8.2–14.2%) was comparable to that of digital subtraction angiography (10.2%). Although the technical failure rate and overall accuracy values (74.5–77.5%) for detection of significant stenoocclusive disease of the calf arteries were slightly inferior to those of the iliofemoral arteries, the caliber of the calf vessels is much smaller than that of the iliofemoral arteries and the acquisition of good-quality images is more technically challenging.
Although comparison of the diagnostic performance of gadobenate dimeglumine with that of other gadolinium agents was not performed in this study, the calf arteries are one vascular territory in which the greater relaxivity of gadobenate dimeglumine compared with that of other agents [21–23] is likely to prove beneficial in terms of vessel visualization and diagnostic performance. In this regard, previous studies [27, 28] have shown that the contrast enhancement (SNR and contrast-to-noise ratio) and visualization of below-the-knee segments is significantly better after the administration of 0.1 mmol/kg of gadobenate dimeglumine compared with an equivalent dose of gadopentetate dimeglumine [27], and that gadobenate dimeglumine may have a significant beneficial effect on the ability to assess below-the-knee segments [28].
Specifically, a study by Wyttenbach et al. [28] not only showed better diagnostic performance for gadobenate dimeglumine compared with gadoterate meglumine at an equivalent dose but also noted significantly fewer nonassessable below-the-knee segments after gadobenate dimeglumine. Wyttenbach et al. administered a standard volume of 34 mL of either gadobenate dimeglumine or gadoterate meglumine to all patients. Given that both agents are formulated at concentrations of 0.5 mol/L [22], this volume equates to an administered dose of almost 0.25 mmol/kg of body weight for an average 70-kg person. Although this dose may be appropriate in the case of gadoterate meglumine and other conventional gadolinium agents [12, 45, 55] with standard r1 relaxivity [21, 22], previous studies have shown that doses of gadobenate dimeglumine of 0.2 mmol/kg of body weight have at best minimal and at worst slightly deleterious effects on overall CE-MRA image quality [56, 57]. Studies performed in vitro have supported these clinical observations in showing that the r1 relaxivity of gadobenate dimeglumine is concentration-dependent, with higher relaxivity values, and hence greater signal intensity enhancement, at lower concentrations [58]. Possibly Wyttenbach et al. might have obtained even better results for peripheral CE-MRA with gadobenate dimeglumine had just a single 0.1 mmol/kg dose been used.
A single 0.1 mmol/kg dose of gadobenate dimeglumine has previously been shown to be equivalent to a double dose of a conventional gadolinium agent (gadopentetate dimeglumine) for CE-MRA of the renal arteries [26] and superior to a double dose of this agent for CE-MRA of the carotid arteries [25], with particular benefits noted for the visualization of small or narrow vessels. Given the current widespread concern among the radiology community about the use of double and triple doses of gadolinium contrast agents, particularly in patients with renal insufficiency who may be at risk of nephrogenic systemic fibrosis [59, 60], our results with just a single 0.1 mmol/kg dose of gadobenate dimeglumine might be of considerable additional interest, especially given the relatively favorable physicochemical properties of this agent compared with other agents [58, 61]. Furthermore, part (4–5%) of the injected dose of gadobenate dimeglumine is eliminated by the hepatobiliary system [62], and a reduced 0.1 mmol/kg dose may result in potential cost savings.
A principal limitation of our study inherent to its multicenter design is the range of MRI systems and sequence parameters used. Moreover, because innovative technology such as parallel imaging [63] and time-resolved MR angiography [38] were in their infancy and not widely available at the time the study was planned and conducted, and because most of the 26 investigational centers did not have access to state-of-the-art MRI systems, the study was performed using conventional MR angiography technology appropriate to the respective imaging capabilities of the individual centers. Both image quality and diagnostic performance might have improved had this more advanced technology been available. Because gadobenate dimeglumine has a higher r1 relaxivity and boosts intravascular signal more than other available extracellular gadolinium agents [24–28], it may prove useful in conjunction with parallel imaging, which penalizes SNR, especially in vascular territories such as the peripheral arteries, for which increased spatial resolution or speed is beneficial. Further investigation of this effect may be of interest in the future.
In conclusion, our study shows that CE-MRA of the lower extremities with gadobenate dimeglumine is significantly more efficacious than TOF MRA, and that CE-MRA is an appropriate alternative to invasive digital subtraction angiography for the diagnostic evaluation of the pelvic and lower leg vasculature in patients with known or suspected peripheral arterial occlusive disease. Moreover, the administration of gadobenate dimeglumine was safe, and no clinically meaningful effects on vital signs, clinical laboratory investigations, or cardiac electrophysiology parameters were observed.
Acknowledgments
The authors would like to thank Ningyan Shen and Usha Halemane (Bracco
Diagnostics Inc., Princeton, NJ) for the statistical analysis performed for
the study and the following investigators for enrolling patients: In-golf
Arlart, Institut fur Radiologie, Katharinenhospital, Stuttgart, Germany; James
Meaney, MRI Department, St. James's Hospital, Dublin, Ireland; Gerardo
Cardenas, Hospital y Clinica OCA, Monterrey, N.L., Mexico; Dan Reimer, Sunbelt
Research Group, LLC, Mobile, AL; Lorenzo Bonomo, Department of Radiology,
Policlinico Agostino Gemelli Roma, Italy; Michael Lee, Department of
Radiology, Beaumont Hospital & Royal College of Surgeons, Dublin, Ireland;
Matthijs Oudkerk, Department of Radiology, Academisch Ziekenhuis Groningen,
Groningen, The Netherlands; Kevin DeMarco, Laurie Imaging Center, UMDNJ, New
Brunswick, NJ; Val Runge, Radiology Department, Scott and White Clinic and
Hospital, Texas A&M University Health Science Center, Temple, TX; Orjan
Smedby, Department of Radiology, Universititetssjukhuset, Linkoeping, Sweden;
Pablo Niedmann, Department of Radiology, Hospital Clinico de la Universidad de
Chile, Santiago, Chile; Luigi Grazioli, Department of Radiology, University of
Brescia, Italy; Jelle Barentsz, Department of Radiology, Universitair Medisch
Centrum St. Radboud, Nijmegen, The Netherlands; Robert Lookstein, Department
of Radiology, Mount Sinai Hospital, New York, NY; W. Ross Stevens, Department
of Radiology MRI, St. John's Hospital, Southern Illinois University School of
Medicine, Springfield, IL; and William Romano, Department of Radiology,
William Beaumont Hospital, Royal Oak, MI.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
