Downloaded from www.ajronline.org by NYU Langone Med Ctr-Sch of Med on 07/31/15 from IP address 128.122.253.228. Copyright ARRS. For personal use only; all rights reserved
981
1991
Executive
ARRS Council Award
Imaging Value
of the Renal
Arteries:
of MR Angiography
.
JOrg F. Debatin1 Charles E. Spritzer1 Thomas M. Grist1 Craig Beam1 Laura P. Svetkey2 Glenn E. Newman1 H. Dirk Sostman1
We compared
for visualizing angiographic
the efficacy
of MR angiography
the renal arteries studies, consisting
with that of conventional
angiography
and detecting renovascular disease. Thirty-three MR of axial two-dimensional (2-D) phase-contrast, coronal 2-D time-of-flight acquisitions, were performed within
2-D phase-contrast, and coronal 48 hr of conventional arteriography. The studies were done to evaluate possible renovascular hypertension (n = 25) or potential donor nephrectomy (n = 8). The three MR image sets were interpreted independently, in random order by three observers, with
regard
to the number
overlap,
and
presence
of renal
arteries,
of renovascular
degree
of vessel
disease.
A fourth
visualization, interpretation
artenovenous was
based
on
the combined axial and coronal phase-contrast image sets. Evaluation was limited to the proximal 35 mm of each renal artery. Renal artery visualization and detection of renovascular disease were more complete with coronal phase-contrast (80% sensitivity, 91% specificity) than with time-of-flight (53% sensitivity, 97% specificity) images. Combined
axial and coronal
phase-contrast
imal 35 mm of all dominant sensitivity,
97%
renal
images
arteries
permitted
and detection
visualization
of the prox-
of 13 of 15 stenoses
Our data suggest that biplanar MR angiography has considerable potential noninvasive screening technique for the evaluation of renovascular disease. AJR
157:981-990,
November
March 1991.
20, 1991 : accepted
after
revi-
Presented at the annual meeting of the American Roentgen Ray Society, Boston, May 1991. 1 Department of Radiology, Box 3808, Duke University Medical Center, Durham, NC 27710. Address reprint requests to C. E. Spritzer. 2 Department of Medicine, Duke University Medical Center,
Durham,
NC 27710.
0361-803x/91/1575-0981 0 American Roentgen Ray Society
as a
1991
Interest in the identification of renovascular disease in patients with hypertension has increased because of the availability and effectiveness of treatment with percutaneous transluminal angioplasty [1 2]. To date, conventional arteriography and intraarterial digital subtraction angiography are the most accurate means of assessing the presence and severity of renal artery stenosis [3, 4]. Because these techniques are invasive, require contrast material, and are expensive, they are not well suited for screening a population of patients with a low prevalence of renal artery stenosis. Many other diagnostic screening techniques have thus been evaluated [1-8]. Although MR angiography has been reasonably successful in the evaluation of circulation in the head and neck [9, 10], renal artery imaging poses a number of particular challenges. These include artifacts from cardiac, respiratory, and bowel motion; problems with stationary tissue suppression; the inherent complex flow patterns and directions of the renal arteries; and vessel overlap from renal veins, the inferior vena cava, and the adrenal and gonadal vasculature [1 1 1 2]. These difficulties have translated into variable success in the assessment of renovascular disease with MR angiography, ranging from not accurate (50% sensitivity) [1 3] to very accurate (1 00% sensitivity) [14]. We performed a prospective study to evaluate four MR angiographic image sets (axial two-dimensional [2-D] phase-contrast, coronal 2-D phase-contrast, coronal 2-D time-of-flight, and combined axial and coronal 2-D phase-contrast) with respect to image quality and the degree of visualization of dominant renal arteries, acces,
Received sionJuly5,
(87%
specificity).
,
.
..
982
Downloaded from www.ajronline.org by NYU Langone Med Ctr-Sch of Med on 07/31/15 from IP address 128.122.253.228. Copyright ARRS. For personal use only; all rights reserved
sory
DEBATIN
renal
arteries,
arteriography
was
and
renovascular
used
Subjects
and Methods
Population
of Patients
as the
disease.
reference
age,
arteriography =
who
had
undergone
of possible
for evaluation
(e.g.
routine
renovascular
donor nephrectomy
within
pacemaker,
,
bia, or recent
surgery),
(n
=
our
ability
hypertension
(n
8). One patient was
the MR examination
to perform
Conventional
conventional
ocular metallic foreign bodies, claustropho-
and
48 hr of conventional
Conventional pigtail
MR angiography of the renal arteries was performed with a 1 .5-T MR imager (Signa System, General Electric, Milwaukee, WI) with the patient in shallow respiration. The body coil was used for signal transmission and reception. To localize the aorta and renal arteries, spin-echo scout images, 500/1 4 (TRITE), were obtained in the cor-
onal projection through the kidneys by using a 256 x 128 matrix, a 40-cm2 field of view, and 1 0-mm-thick sections. The MR angiography techniques, 2-D phase-contrast and 2-D timeof-flight, are part of the Signa System Vascular Magnetic Resonance Package. 2-D phase-contrast images were obtained in both axial and coronal planes with the following parameters: 40/8.4 (TRITE), 60#{176} angle, 256 x 1 28 matrix, 28-cm2 field of view, two excitations, and first-order gradient-motion nulling. The velocity-encoding value (velocity-yielding r radians of phase shift) was 40 cm/sec. Sevenflip
acquired
images
to the more
sections
with
2-mm
overlap
between
consecutively
were obtained from the more cephalad superior pole
caudad
inferior
pole of either
kidney.
The
same
slice
parameters were used to obtain coronal images from 10 mm anterior to 25 mm posterior to the aorta. For each phase-contrast image, the system acquires a magnitude gradient-recalled echo image and three raw data sets for the three flow-encoding
directions:
tenor. The magnitudes
right
to left,
superoinfenor,
and
anteropos-
of these data sets are then combined
into a
total flow diagram for that particular section [1 5]. Depending on kidney size and location, between 1 1 and 23 image sections (average,
16) were obtained 1 6 mm. projection
axial projection,
in the
with acquisition
times of 8-
On average, only 1 1 images were obtained in the coronal (range, 9-1 4), with acquisition times between 7 and 12
mm. 2-D
time-of-flight
MR
imaging
was
performed
in the coronal
plane
parameters: 40/8.4 (TRITE), 60#{176} flip angle, 256 x 128 matrix, 32-cm2 field of view, an average oftwo excitations, and first-order gradient-motion nulling. Between 20 and 28 consecutively acquired 4-mm-thick sections with 2-mm intersection overlap were obtained from 1 0 mm anterior to the aorta to 25 mm posterior to the aorta. Acquisition times ranged from 4 to 6 mm. with the following
scan
was
Contrast
material
performed was
by using
injected
at
25
a 5-French mI/sec
for
2
the arteriogram was repeated in oblique projections. If selective renal artery injections were required to adequately define the renovascular anatomy, a preformed end-hole catheter was used, and the injection rate varied with the vessel size.
signal from the inferior vena cava and saturation pulses were placed inferior to the lower pole of the kidneys and over the renal cortex on either
phase-contrast,
and coronal
and time-
time-of-flight,
and
a combination
were rendered
of axial
in random
order.
individual
sections.
Each image set was assessed with regard to the number of renal arteries present, the degree of vessel visualization, the degree of venous overlap, and the presence of renovascular disease. Each renal artery was divided into three segments: the ostium, the proximal
15 mm (1-15 mm), and the distal 20 mm (15-35 mm). Each segment was characterized as either seen, not seen owing to stenosis, or not seen owing to artifact. It was noted separately if venous overlap impaired proper assessment of the renal artery. On the basis of previous in vivo MR angiographic evaluations of stenosis (Porges R et al. presented at the annual meeting of the Radiological Society of North America, November 1990; Grist TM et al. unpublished data), renovascular disease was defined as a seg,
,
ment of luminal narrowing and/or signal loss and graded on a fourpoint scale ranging from mild stenosis to occlusion. Mild disease (