Download High-resolution magnetic resonance angiography of digital arteries

RHEUMATOLOGY
Rheumatology 2011;50:1712–1719
doi:10.1093/rheumatology/keq453
Advance Access publication 21 June 2011
Original article
High-resolution magnetic resonance angiography of
digital arteries in SSc patients on 3 Tesla: preliminary
study
Wei Zhang1, Jian Rong Xu1, Qing Lu1, Shuang Ye2 and Xiao Sheng Liu1
Abstract
Objective. We sought to evaluate the use of high-resolution three-dimensional time of flight (3D TOF)
magnetic resonance angiography (MRA) at 3 Tesla in the visualization of digital arteries in SSc patients.
Methods. A total of 33 patients with SSc [32 females and 1 male; mean (S.D.) age 37.0 (10.2) years;
median number of years since onset of RP 3.4 (2.8) years] and 7 healthy controls [6 females and 1 male;
mean (S.D.) age 30.7 (3.5) years] were examined on a 3T MR system. A modified high spatial resolution
(voxel size = 0.35 0.35 0.5 mm3) 3D TOF MRA (repetition time/echo time = 19/4.4 ms, flip angle = 15 ,
slice thickness = 1 mm) was performed during a total scan time of 8 min 22 s. The source images and
maximum intensity projection reconstruction were studied; the digital arteries count and lumen area of the
selective section of the vessel were measured independently by two experienced radiologists and compared with that of the control ones; and a four-level grading system was made according to the severity.
Statistical analysis was performed with t-test and P < 0.05 was used as the criterion.
CLINICAL
SCIENCE
Results. We detected the eight digital arteries in the four fingers (without the thumb) of each case and got
a 47.58% presentation in general in the SSc group, and artery No. 5 had the highest presentation rate
(70.97%). Statistics showed that the digital arteries No. 1 (P = 0.058) and No. 3 (P = 0.093) had no difference in the lumen area (P > 0.05). We finally got 3 cases in Grade 1, 14 cases in Grade 2, 9 cases in
Grade 3 and 5 cases in Grade 4.
Conclusion. MRA of the digital arteries in the described technique is a promising method for us to judge
the severity of microvascular involvement in finger vessels of SSc patients.
Key words: 3 Tesla magnetic resonance imaging, Micro-magnetic resonance angiography, Maximum intensity
projection, Three-dimensional, Time of flight, Proper digital artery, Systemic sclerosis, Raynaud’s phenomenon,
High resolution, Vascular abnormalities.
Introduction
SSc is a generalized connective tissue disorder characterized by autoimmune abnormality, vascular involvement
and collagen accumulation in the connective tissue.
Vascular abnormalities such as fingertip ulcers and
1
Department of Radiology and 2Department of Rheumatism, Shanghai
Clinical Centre of Rheumatic Diseases and Institute of Rheumatology,
Renji Hospital, Shanghai Jiao Tong University School of Medicine,
Shanghai, China.
Submitted 13 April 2010; revised version accepted 14 December 2010.
Correspondence to: Jian Rong Xu, Department of Radiology, Renji
Hospital, Shanghai Jiao Tong University School of Medicine, No. 145
Middle Shan Dong Road, Shanghai 200001, China.
E-mail: [email protected]
Raynaud’s syndrome as well as involvement of organs
including the lungs, heart, kidney and gastrointestinal
tract are prominent features of the disease [1–3]. In pathobiology, SSc is characterized by injury to the vascular wall
and extensive damage to the microvessels, including
endothelial impairment, lumen narrowing and fibrosis of
the microvasculature [4].
Distal digital ulcers occur in up to 50% of patients with
limited or diffuse SSc. RP is an early manifestation representing vascular involvement and is seen in 90–95% of
SSc patients [5–7]. Microvascular involvement in fingers
may reflect the damage of the finger and hand vessel
system. Intra-arterial digital subtraction angiography
(DSA) is still considered as the reference standard for
the diagnosis of vascular pathologies of the fingers. As
! The Author 2011. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: [email protected]
High-resolution MRA of digital arteries
an invasive procedure, the inherent limitations have made
it a not so widely accepted technique in practice.
Recent years have seen the clinical implementation of
magnetic resonance angiography (MRA) with and without
the use of contrast material as a safe, reliable and accurate modality for evaluation of vascular pathologies of the
hand and fingers [8–10]. However, imaging the small vessels of the finger places special demands on this technique. High resolution is necessary to accurately depict
the digital vessels, which are often <1 mm in diameter.
The introduction of 3T whole-body imaging systems into
clinical practice, with the promise of improved signalto-noise ratios (SNRs) compared with those with 1.5T, is
advantageous when highly accelerated parallel imaging is
considered.
In this preliminary study, we used a three-dimensional
time of flight (3D-TOF) technique with improved spatial
and time resolution to perform imaging of digital arteries
at 3T. A 1 mm thickness was used with the voxels of
0.35 0.35 0.5 mm3. High-quality images were obtained
and revealed well-defined margin of the arteries.
Materials and methods
Patients
Thirty-three patients confirmed with SSc [32 females and
1 male; mean (S.D.): age 37.0 (10.2) years; disease duration
5.8 (4.3) years; median number of years since onset of
RP 3.4 (2.8) (range 0–10) years] and 7 healthy controls
[6 females and 1 male; mean (S.D.) age 30.7 (3.5) (range
24.8–34.1) years] were recruited into the study. All patients satisfied the preliminary criteria of the ACR for the
classification of SSc [11]. We collected the following data
for all subjects: age, sex, disease duration (beginning
at the date when the first non-RP symptom occurred),
duration of RP, laboratory findings (especially kidney function), pulmonary artery pressure (colour Doppler showed
roughly), drug history (using pimobendane or not), pulmonary changes on high-resolution CT (HRCT) and current digital situation. All the details were listed as shown in
Table 1. None of the patients had chronic arterial occlusive risk factors such as hyperlipidaemia, diabetes mellitus or obesity. Hypertension in two patients was well
controlled with anti-hypertensive drugs. There was no history of sex hormone administration in any of the patients.
All subjects had been asked to refrain from smoking and
caffeine consumption and to suspend all the drugs 3 days
before the MR examination. We had two other volunteers
who had the DSA examination done. The study was
approved by the local ethics committee (Medical Ethics
Committee of Renji Hospital, Shanghai Jiao Tong
University School of Medicine, Shanghai, China), and all
study participants signed informed consent before the
study.
3D imaging protocol
All the examinations were performed on a 3T MR scanner
(Signa Excite 3.0T; GE HealthCare, Milwaukee, WI, USA)
fitted with high-performance gradients (peak gradient
www.rheumatology.oxfordjournals.org
amplitude 50 mT/m; slew rate 150 mT/m/s). A dedicated
multichannel receive-only head coil was used.
Subjects were imaged lying prone, with the arm raised
above the head and the hand placed at the centre of the
head coil (preferably palm down). To reduce acquisition
time, the hand had to be positioned with no rotation so
that the minimum number of thin sections could be used
to cover the vascular anatomy of the fingers. In cases of
clinically suspected vasospasm, a warm water bath of the
hand was performed before imaging, and the examination
room temperature was controlled within 23 (±1 ).The
hand was wrapped in the coil and fixed with tape, with
fingers spread slightly to prevent overlapping of the digital
arteries.
A fat-saturated, high spatial resolution 3D-TOF MRA
was performed in the axial plane by using a fast-gradient
echo sequence (Table 2). Zero interpolation improved the
in-plane resolution to 0.35 0.35 mm2. A parallel imaging
technique such as array spatial sensitivity encoding
TABLE 1 Characteristics of the 33 patients with SSc
Women/men
Cutaneous subtype, diffuse/limited
Disease duration, mean (S.D.), years
RP
Duration of RP, mean (S.D.), years
Current digital situation
Colour changed in fingertips
(red/white/cyanosis/dander under the nail)
Skin hardened
Current active digital ulcers
Distal fingertip amputations
Finger shortening and atrophia
History of digital ulcers, including
pitting scars
PAP (colour Doppler showed roughly),
mean (S.D.), mmHg
Pulmonary fibrosis on CT scan
Urine protein (positive)
Vasodilators required
Hypertension
32/1
23/10
5.8 (4.3)
33
3.4 (2.8)
23
26
6
4
5
13
30.1 (7.6)
19
16 (33)
14
2
PAP: pulmonary artery pressure.
TABLE 2 Imaging parameters for high spatial resolution
3D MRA
Parameter
Value
Repetition time/echo time, ms
Flip angle, degrees
Bandwith, kHz/pixel
FOV, mm
Matrix size
In-plane spatial resolution, mm2
Section thickness, mm
Voxel dimension, mm3
Acquisition time
19/4.4
15
12.50
120
320 320
0.35 0.35
1
0.35 0.35 0.5
8 min 22 s
FOV: field of view.
1713
Wei Zhang et al.
technique (ASSET) was applied. All MRA examinations
were post-processed with maximum intensity projection
(MIP). Forty-five were separated in steps by 8 (0–360 )
and were reconstructed from all relevant data sets.
MR data sets were evaluated and compared for image
quality and visualization of vascular details. All the MR
images were successful, while technical problems prevented acquisition of the standard protocol in one patient
and was finally overcome. The examination time for MRA
was 8 min 22 s per patient. The scan scope covered the
other four fingers without the thumb, and inflow saturation
was performed at a distance of 10 mm from the finger tip,
thus allowing saturation of vein inflow.
Image review
All image data were transferred to a post-processing
unit (Advantage Workstation 4.4, GE Medical Systems,
Milwaukee, WI, USA) for digital artery analysis, and MIP
reconstructed from all relevant data sets. To avoid any
recognition bias, the identification number of each examination was masked. A consensus readout by two radiologists (8 and 7 years of MR experience, respectively)
included all intradividual comparisons in random order,
respectively, and each reader was blinded to the patient’s
medical history, clinical data and results of other imaging
studies. Separate image-reading sessions were organized
for both readers by the study coordinator, who attended all
reading sessions. The readers were instructed to use the
post-processed data initially and, if required, to use the
source data for interactive reformatting. The number of
arteries in the whole four fingers (not including the thumb)
was counted (total eight in normal), and labelled as 1–8
from the direction of index to the little finger (Fig. 1A and B).
Criteria for image analysis were the overall image quality,
artery visibility (numbers) and the lumen area. To ensure
accurate comparison between subjects, the digital artery
lumen area was measured at the location of the PIP joint as
the average from three consecutive slices. Measurements
were obtained from lumen contours of the proper palmar
digital artery outline by manual work, and figures came out
automatically by the computer. The measured area was
enlarged enough to guarantee accuracy (Fig. 1C and D).
We used the classification to graduate vessel damage
in our patients. Scores for depiction of digital arteries were
as follows:
0 : normal finding, show all the eight digital arteries,
clearly continuous visible vessels with well-defined
boundaries;
1 : total artery number 7–8, clearly continuous vessels
with well-defined boundaries;
2 : total artery number 5–6 and visible vessels with minor
discontinuities;
3 : total artery number 3–4 and poorly visible vessels
with considerable discontinuities; and
4 : total artery number <3 or no visible distant arteries, with
only little short arteries shown in the base of fingers.
Example images of different vascular scores and lumen
area were written down and the primary reviewer’s
1714
measurements were used for clinical evaluation. After
independent interpretations were performed, discrepancies between the two readers in the scoring of the
vessel depiction and measurement of the lumen areas
were resolved by consensus to establish a final value.
Statistical analysis
Analyses were conducted using SPSS 11.0 software
(SPSS, Chicago, IL, USA). Descriptive statistics for all variables were examined, and paired-sample t-tests were calculated to assess inter-reader agreement for all the
difference.
Quantitative data between the SSc patients and healthy
controls were analysed with Mann–Whitney U-test, including the differences in digital artery number count and
lumen area of the arteries presented. P < 0.05 was used
as the criterion to indicate a statistically significant difference for all the statistical tests. Graphs were also plotted
with SPSS 11.0.
Results
Image quality
While all examinations were successfully completed by
the volunteers, two females failed due to vessel spasm
or the motion of fingers while scanning in the SSc
group. In all the other cases, image quality of the digital
arteries was good and no pathological findings were seen
in the volunteer group (Fig. 2).
Inter-reader agreement evaluation
In normal controls, the artery show rate was 100%, and
the lumen area measurements showed no differences by
the two readers (P = 0.426, P > 0.05). In the SSc group, the
P-value of the lumen area was 0.285 (P > 0.05) and no
significant differences were detected. The artery number
count was discussed to get the agreement with -coefficient = 0.920 for comparison between the two readers.
Evaluation of vascular image quality and lumen
area contrast of the two groups
We calculated the presentation rate of the finger arteries,
and determined the lumen area of each. We also carried
out the Mann–Whitney U-test for the lumen area variance
of the SSc group and the volunteers, and the P-values
obtained are listed in Table 3. (The lumen area of the
SSc group mentioned here were those who had presented, not including those not shown.) All P-values of
the eight digital arteries were lower than their corresponding critical values (0.05) other than the digital arteries
No. 1 (P = 0.058, P > 0.05) and No. 3 (P = 0.093, P > 0.05),
which means that only the radialis proper digital artery in
the middle finger and the index finger showed no significant differences between the SSc group and the normal
control group.
The figure saw directly the evident decrease of the
lumen area in the SSc group compared with that in the
control group, especially arteries No. 2, 4, No. 5 and No. 7.
www.rheumatology.oxfordjournals.org
High-resolution MRA of digital arteries
FIG. 1 (A) Illustration of the normal arterial anatomy of the hand. The superficial palmar arch gives rise to the proper
digital arteries (grey) and the deep palmar arch (black) supplies the thumb. (B) Numbers 1–8 refer to the eight digital
arteries from the index finger to the little finger, respectively. (C) The digital artery lumen area was measured at the
location of PIP by manual work, and the measurement was made when the picture was adequately enlarged (D).
Artery No. 5 had the highest presentation rate (70.97%),
followed by arteries No. 2 and 4 (both 64.52%), while
arteries No. 1 (29.03%) and No. 8 (16.13%) had the
lowest presentation rate. The total digital artery visualization ratio was 47.58%, compared with 100% each in the
volunteer group.
www.rheumatology.oxfordjournals.org
Evaluation of the degree of SSc group
According to the above standard, we divided the SSc patients on a 1–4 scale, because all the normal contrasts
belong to Grade 0. We finally got 3 cases in Grade 1,
14 cases in Grade 2, 9 in Grade 3 and 5 in Grade 4 in
the 31 patients (Fig. 3). Patients with a history of digital
1715
Wei Zhang et al.
FIG. 2 MRA. (A) MIP images of a 33-year-old women volunteer show all the eight digital arteries clearly with well-defined
boundaries and (B) MIP images of a 28-year-old female SSc patient show only three consecutive arteries. Comparison of
angiography images without (C) and with (D) vein signal saturation in the fingertip for digital artery reveals that major venous
contamination was observed in the transection image without rest (E) compared with the one with rest (F), with almost no
venous pollution.
TABLE 3 Comparison of the two groups
SSc group
Lumen area,
mean (S.D.)
1
2
3
4
5
6
7
8
0.81
1.01
0.90
0.92
0.88
0.60
0.62
0.48
(0.15)
(0.22)
(0.14)
(0.27)
(0.30)
(0.19)
(0.20)
(0.88)
Presentation
numbers
Artery
show,%
9
20
16
20
22
11
15
5
29.03
64.52
51.61
64.52
70.97
35.48
48.39
16.13
Normal control group
Lumen area,
mean (S.D.)
1.02
1.38
1.05
1.33
1.47
0.85
1.11
0.58
(0.22)
(0.27)
(0.15)
(0.25)
(0.33)
(0.10)
(0.19)
(0.07)
Artery
show, %
P-value
100
100
100
100
100
100
100
100
0.057
0.011
0.093
0.009
0.002
0.015
0.002
0.038
In column 1, numbers 1–8 refer to the eight digital arteries from index finger to the little finger, respectively.
ulcers all belong to Grade 3 (5 cases) and Grade
4 (8 cases), which means they had heavier vessel involvement than those without ulcers or scars on the finger tip
(P = 0.024, P < 0.05).
In the SSc patient group, we found a correlation between
artery graduation and finger abnormality, as shown in
Table 4 (P = 0.03, P < 0.05). Scatter plot analysis between
disease duration since the onset of RP (in years) and the
1716
subject’s artery graduation is shown in Fig. 4. A strong
correlation (Spearman’s correlation coefficient r = 0.850,
P < 0.01) was found between the two parameters.
Discussion
Distal peripheral artery disease is present in the digits
of many SSc patients, showing a high frequency of
www.rheumatology.oxfordjournals.org
High-resolution MRA of digital arteries
FIG. 3 (A) A 31-year-old female in Grade 1 showing all the eight arteries, with a weak signal in artery No. 1; (B) a
36-year-old woman in Grade 3 showing only three digital arteries ambiguously; (C and E) a 28-year-old female in Grade 2
showing only four to five arteries with weak signal and a little venous pollution; and (D and F) a 30-year-old SSc female
patient in Grade 4 showing one digital artery.
digital stenosis and occlusions in the digital arteries of
the patients. Lesions were most frequently found in
the second to fifth proper palmar digital arteries, the
ulnar artery and the superficial palmar arch. As a
consequence, digital ischaemia, ulceration or amputation is a well-known, but feared manifestation of SSc
[12–16].
In contrast to MRA of the organs and head, which gains
more and more clinical application, MRA of the digital
arteries is still exceptional. The vasculature of the hand
is much more complex, with small-vessel diameters
branching in an often unpredictable pattern. High
www.rheumatology.oxfordjournals.org
resolution is necessary to accurately depict the digital
vessels, which are often <1 mm in diameter [17, 18].
MRA of finger arteries with contrast media has already
been published. Allanore et al. [19] used consecutive MR
scans in SSc patients after the injection of MR contrast
media gadolinium-diethylenetriamine penta-acetic acid
(Gd-DTPA), to evaluate the changes and abnormalities
of hand arteries and veins, including the capillary network
dysfunction. This is a useful method to observe the vessels dynamically, but venous contamination poses a major
problem in presenting the image of the arteries due to a
short arteriovenous transit time.
1717
Wei Zhang et al.
Our objective was to introduce an appropriate protocol
for high-resolution, non-contrast MRA of finger arteries. A
fat-saturated 3D-TOF MRA technique with vein signal
saturation can eliminate venous pollution (Fig. 2C–F).
Selective imaging of the hand arteries always demands
TABLE 4 Patient numbers of different graduation
Grade, Colour
Skin
Ulceration Shortening/
degrees change hardening and scaring atrophia
1
2
3
4
3
12
6
2
13
8
5
3
3
1
4
FIG. 4 Scatter plots of disease duration after the onset of
RP vs the digital artery graduation.
a sequence setting that provides maximal spatial resolution and minimal venous overlay. We finally got satisfying pictures by using a voxel size of 0.35 0.35 0.5 mm3
spatial resolution and 1-mm slice thickness. We did not
include the thumb, because the digital arteries are relatively stable in anatomical structure compared with more
variation of the palmar archs (Fig. 1A) [8]. There is one
proper palmar digital artery on each side of the finger
beside the thumb, and the lumen areas are quite even in
the proximal and the distal part of the digital arteries. We
therefore selected the four finger arteries as the object of
our experiment, which made evaluation easier.
Janevski [20] also reported that arteriographic
abnormalities of the vessels, including stenosis and
occlusion, in patients with SSc were primarily observed
in the proper digital arteries and less frequently in
the ulnar artery, the superficial palmar arch and the
common digital arteries. In his series of 12 patients, the
radial artery and the deep palmar arch were never
affected and there was a lesser involvement of the
proper digital artery of the thumb compared with those
of other fingers. That was also one reason we excluded
the thumb in our study. In our outcome of the image, the
arteriographic changes of the vessels in SSc are smooth
and tend to show tapering segmental lesions in the distal
vessels, with smooth vessel walls as opposed to arteriosclerosis, which is similar with the outcome of the Doppler
US [21, 22].
Whole-body 3.0T MR imaging systems generates high
spatial resolution images throughout the body, with the
promise of sufficient SNR. Compared with DSA images
we got from two volunteers (Fig. 5), the latter saw a diluted
artery image unless we used a large volume of iodic contrast media, which is very dangerous for SSc patients who
have already suffered renal damage.
FIG. 5 DSA images of two male volunteers (A, B). The peripheral arteries were not clearly visible with the contrast media
running fast to the capillary network and the vein and terminal vessels were not detected.
1718
www.rheumatology.oxfordjournals.org
High-resolution MRA of digital arteries
Using high-resolution 3D MRA, we have quantified the
lumen area of the SSc patients and established the graduation. We consider it a promising method to identify and
quantify the microvascular involvement in SSc patients,
although it still has limitations concerning spatial and temporal resolution. Besides, the technique is likely to enhance the performance, which means that the lumen
area we measured was sometimes a little larger than
their actual area. We also found that the severity of the
finger artery involvement was correlated with not only the
duration of the disease but also with the activity of the
stage, and images of finger arteries in SSc are entirely
different from that in vasculitis. Further technical refinements and larger study cohorts are necessary to ensure
that diagnostic time-resolved MRA becomes routine clinical practice.
Rheumatology key messages
MRA of finger arteries can identify and quantify the
microvascular involvement in SSc patients.
. The MRA result correlates with the duration and
severity of the disease.
.
Acknowledgements
Qian Jiang, Lifei Ma, Xihai Zhao (presently located
at Washington University) are acknowledged for
technical assistance. Special thanks to S.Y., Qiang Guo
and Xingfang Huang of Department of Rheumatic disease in Shanghai Ren Ji Hospital for their collaborative
efforts.
Funding: This work was generously supported by a grant
from the city’s key subjects of Shang Hai (S30203).
Disclosure statement: The authors have declared no
conflicts of interest.
References
1 Tamborrini G, Distler M, Distler O. Systemic sclerosis.
Med Monatsschr Pharm 2008;31:162–72.
2 Kahaleh MB. Vascular involvement in systemic sclerosis.
Clin Exp Rheumatol 2004;22:S19–23.
3 Herrick AL. Vascular function in systemic sclerosis.
Curr Opin Rheumatol 2000;12:527–33.
4 Herrick AL. Pathogenesis of Raynaud’s phenomenon.
Rheumatology 2005;44:587–96.
5 Serena G, Roberto G, Marco MC. Vascular
complications of scleroderma. Autoimmunty Reviews
2007;6:520–3.
6 Martine G. Raynaud’s phenomenon. Joint Bone Spine
2007;74:e1–8.
www.rheumatology.oxfordjournals.org
7 Ostojı́c P, Damjanov N, Pavlov-Dolijanovic S. Peripheral
vasculopathy in patients with systemic sclerosis:
difference in limited and diffuse subset of disease.
Clin Hemorheol Microcirc 2004;31:281–5.
8 Connell DA, Koulouris G, Thorn DA et al.
Contrast-enhanced MR angiography of the hand.
Radiographics 2002;22:583–99.
9 Brauck K, Maderwald S, Vogt FM et al. Time-resolved
contrast-enhanced magnetic resonance angiography of
the hand with parallel imaging and view sharing: initial
experience. Eur Radiol 2007;17:183–92.
10 Rofsky NM. MR angiography of the hand and wrist.
Magn Reson Imaging Clin N Am 1995;3:345–59.
11 Subcommittee for Scleroderma Criteria of the American
Rheumatism Association Diagnostic and Therapeutic
Criteria Committee. Preliminary criteria for the
classification of systemic sclerosis (scleroderma).
Arthritis Rheum 1980;23:581–90.
12 Hinchcliff M, Varga J. Systematic sclerosis/scleroderma:
a treatable multisystem disease. Am Fam Physician 2008;
78:961–8.
13 Stucker M, Quinna S, Memmel U et al. Macroangiopathy
of the upper extremities in progressive systemic sclerosis.
Eur J Med Res 2000;5:295–302.
14 Hasegawa M, Nagai Y, Tamura A et al. Arteriographic
evaluation of vascular changes of the extremities in
patients with systemic sclerosis. Br J Dermatol 2006;155:
1159–64.
15 Merkel PA, Herlyn K, Martin RW et al. Measuring disease
activity and functional status in patients with scleroderma
and Raynaud’s phenomenon. Arthritis Rheum 2002;46:
2410–20.
16 Chung L, Fiorentino D. Digital ulcers in patients with
systemic sclerosis. Autoimmun Rev 2006;5:125–8.
17 Wang J, Yarnykh VL, Molitor JA et al. Micro magnetic
resonance angiography of the finger in systemic sclerosis.
Rheumatology 2008;47:1239–43.
18 Nael K, Vilablanca JP, Pope WB et al. Supraaortic arteries:
contrast-enhanced MR angiography at 3.0T—highly
accelerated parallel acquisition for improved spatial
resolution over an extended field of view 1. Radiology
2007;242:600–9.
19 Allanore Y, Seror R, Chevrot A et al. Hand vascular involvement assessed by magnetic resonance angiography in systemic sclerosis. Arthritis Rheum 2007;56:
2747–54.
20 Janevski BK. Arteriography of the peripheral vessels of the
extremities: technique and essentials of interpretation. J
Med Imaging 1989;3:197–202.
21 Keberle M, Tony HP, Jahns R et al. Assessment of
microvascular changes in Raynaud’s phenomenon and
connective tissue disease using colour Doppler ultrasound. Rheumatology 2000;39:1206–13.
22 Schmidt WA, Wernicke D, Kiefer E et al. Colour duplex
sonography of finger arteries in vasculitis and in systemic
sclerosis. Ann Rheum Dis 2006;65:265–7.
1719