AJR:175, August 2000

417
Original Report

OBJECTIVE.



The purpose of this report is to describe the normal MR anatomy of the thoracic
outlet and its modification after postural maneuvers using an anatomic–MR imaging correlation.

CONCLUSION.

MR imaging appears to be a useful technique to study the thoracic outlet
and its contents because of its excellent soft-tissue depiction and its multiplanar capabilities. T1-
weighted images obtained in the sagittal plane clearly depicted the different compartments of the
cervicothoracic–brachial junction. Hyperabduction maneuvers may have potential applications in
the assessment of the thoracic outlet syndrome by showing the location of compression.
he thoracic outlet or cervicotho-
racic–brachial junction consists of
several confined spaces extending
from the cervical spine and the mediastinum
up to the lower border of the pectoralis minor
muscle. It is divided into three tunnels: the in-
terscalene triangle, the costoclavicular space,
and the retropectoralis minor space [1].
Symptoms and signs of thoracic outlet
syndrome result from the compression or irri-
tation of the neurovascular bundle at various
levels of the cervicothoracic–brachial



pas-
sages. Compression usually occurs as a result
of congenital or acquired changes in the sur-
rounding fibroosseous and fibromuscular
structures [2]. There is potential for static or
dynamic compression or both. Moreover, in
an already “tight” thoracic outlet, dynamic
movements such as holding the arm overhead
and backward (hyperabduction) can further
compress the enclosed structures and bring
on symptoms [2–4].
The usefulness of MR imaging for the as-
sessment of the thoracic outlet [5] and bra-
chial plexus [6–8] has yet to be fully defined.
To the best of our knowledge, only two case
reports about MR imaging of patients with
arms in hyperabduction have been published
[9, 10]; no MR imaging and anatomic corre-
lation of the thoracic outlet spaces with arms
alongside the body and with arms hyperab-
ducted has been reported. The purpose of
this report is to describe the normal MR
anatomy of the thoracic outlet and its modifi-
cations after postural maneuvers using an an-
atomic–MR imaging correlation.

Subjects and Methods

Five fresh



cadavers (two men and three women;
age range, 76–85 years; mean, 81.4 years) were ex-
amined. Two were injected bilaterally into the bra-
chial artery with a mixture of warm gelatin,
gadolinium (Dotarem; Laboratoire Guerbet, Aul-
nay-sous-Bois, France), and red stain to visualize
arterial structures during MR and anatomic studies.
All cadavers were examined with a 1.5-T imager
(Magnetom Vision; Siemens, Erlangen, Germany)
and a body coil. Sagittal images were obtained bi-
laterally using a T1-weighted spin-echo sequence.
Imaging parameters were as follows: TR/TE, 500/
14; slice thickness, 3 mm; interslice gap, 0.3 mm;
matrix, 366

×

512; and field of view, 175

×

200
mm. Thereafter, the specimens were frozen and
sawed into 3-mm-thick contiguous sagittal sections
with a band saw. Four cadavers were positioned
with the arms alongside the body; one cadaver,
with the arms hyperabducted.
Twelve volunteers (three men and nine women;
age range, 22–40 years; mean, 31.5 years) were also

X. Demondion

1,2

N. Boutry

1

A. Drizenko

2

C. Paul

1

J. P. Francke

2

A. Cotten

1

Received July 12, 1999; accepted after revision
January 11, 2000.

1

Department of Musculoskeletal Radiology, Roger
Salengro Hospital, 59037, Lille Cedex, France. Address
correspondence to X. Demondion, Service de Radiologie
Ostéo-Articulaire, Hôpital Roger Salengro, Blvd. du Pr. J.
Leclercq, 59037, Lille Cedex, France.

2

Anatomy Department, Faculty of Medicine, University of
Lille 2, Pl. de Verdun, 59037, Lille Cedex, France.

AJR

2000;175:417–422
0361–803X/00/1752–417
© American Roentgen Ray Society
T

Thoracic Outlet:

Anatomic
Correlation with MR Imaging

418

AJR:175, August 2000

Demondion et al.

examined bilaterally with a 1.5-T imager (Magne-
tom Vision; Siemens) and a body coil. Sagittal T1-
weighted spin-echo sequences were performed in all
volunteers. Imaging parameters were as follows:
500/14; slice thickness, 3 mm; interslice gap, 0.3
mm; imaging matrix, 224

×

256; and field of view,
263

×

350 mm. These sequences were performed
first with the arms alongside the body and then with
the arms hyperabducted (135°). The study was ap-
proved by our institutional review board and in-
formed consent was obtained from each volunteer.
To determine a radioanatomic correlation, first
the gross anatomic sections and the corresponding
MR images were evaluated in consensus by two
musculoskeletal radiologists. The reviewers were
asked to identify the clavicle, the first rib, the sub-
clavian vein and artery, the dorsal scapular artery,
the trunks and cords of the brachial plexus, the
scalene muscles, the subclavius muscle, the serra-
tus anterior muscle, the subscapularis muscle, and
the pectoralis minor and major muscles. Then the
same radiologists identified these structures on
MR images of the volunteers in both arm posi-
tions. They were also asked to assess the compres-
sion of vessels, the presence of fat surrounding the
vascular or nervous structures, and the modifica-
tions of the different tunnels after hyperabduction.

Results

In two volunteers, the MR images were
slightly blurred because of large respiratory
movements during the examination. However,
CB
A
Fig. 1.—Interscalene triangle. 1 = clavicle, 2 = subclavian artery, 3 = subcla-
vian vein, 4u = upper trunk of brachial plexus, 4m = middle trunk of brachial
plexus, 4l = lower trunk of brachial plexus, 5 = first rib, 6 = anterior scalene
muscle, 7 = middle scalene muscle, 8 = dorsal scapular artery, 9 = lung.
A, Photograph of sagittal gross anatomic slice shows interscalene triangle
in 76-year-old male cadaver with arms positioned alongside body. Inter-
scalene triangle is bordered by anterior scalene muscle anteriorly and by
middle and posterior scalene muscles posteriorly.
B and C, Sagittal T1-weighted MR images of 32-year-old male volunteer with
arms positioned alongside body (B) and with arms hyperabducted (C) show
interscalene triangle. In C, note narrowing of space between posterior side
of clavicle and anterior side of anterior scalene muscle (prescalene space)
when compared with B.