Abstract
Background: The axillary nerve is out of the field of view during
shoulder arthroscopy, but certain procedures require manipulation of capsular
tissue that can threaten the function or integrity of the nerve. We studied
fresh cadavers to identify the course of the axillary nerve in relation to the
glenoid rim from an intra-articular perspective and to determine how close the
nerve travels in relation to the glenoid rim and the inferior glenohumeral
ligament.
Methods: We dissected nine whole-body fresh-tissue shoulder joints
and exposed the axillary nerve through a window in the inferior glenohumeral
ligament. Then we cut coronal sections through the glenoid fossa of ten
unembalmed, frozen shoulder specimens after the axillary nerve had been
stained with Evans blue dye. All specimens were studied with the joint secured
in the lateral decubitus position used for shoulder arthroscopy.
Results: Microsurgical dissection through the inferior glenohumeral
ligament from within the joint capsule revealed the axillary nerve as it
traversed the quadrangular space. In each dissection, the teres minor branch
was the closest to the glenoid rim. The coronal sectioning of the unembalmed
shoulder specimens demonstrated that the closest point between the axillary
nerve and the glenoid rim was at the 6 o'clock position on the inferior
glenoid rim. At this position, the average distance between the axillary nerve
and the glenoid rim was 12.4 mm. The axillary nerve lay, throughout its
course, at an average of 2.5 mm from the inferior glenohumeral ligament.
Conclusions: We used two novel approaches to map the axillary nerve
from an intra-articular perspective. Our analysis of the position of the nerve
with use of these methods provides the shoulder arthroscopist with essential
information regarding the location, route, and morphology of the nerve as it
passes inferior to the glenoid rim and shoulder capsule.
Clinical Relevance: Orthopaedic surgeons have long known that the
axillary nerve is vulnerable to damage during repair of the shoulder joint
capsule. Knowledge of the precise relationship of the axillary nerve and its
branches to the inferior glenohumeral ligament can be of benefit in shoulder
arthroscopy.
The axillary nerve is out of the surgeon's visual field during shoulder
arthroscopy, but arthroscopic procedures such as inferior capsulolabral
reconstruction, inferior thermal capsulorrhaphy, and inferior capsular release
require manipulation of capsular tissue that threatens the function and even
the integrity of the nerve. The purpose of this study was to identify the
anatomy of the axillary nerve in cadaveric specimens with use of techniques
that minimally distort the nerve and simulate conditions encountered by the
shoulder arthroscopist. It is important for the shoulder arthroscopist to know
exactly where the axillary nerve lies in relation to both the glenoid rim and
the shoulder capsule.
Eakin et al.1,
Uno et al.2, and
Bryan et al.3
described the relationship of the axillary nerve to the capsule or to the
glenoid rim, but they used methods that required extensive extra-articular
dissection of the nerve. The capsule may have a tethering effect on the nerve,
and its removal may distort the normal relationship between the axillary nerve
and the glenoid rim.
We used two techniques in fresh, unembalmed cadaver shoulders to determine
the course of the axillary nerve and its distance from the glenoid rim and the
inferior glenohumeral ligament. In the first approach, a rectangular window
was cut into the inferior glenohumeral ligament so that the axillary nerve
could be seen within the shoulder joint cavity. In a second group of
shoulders, the axillary nerve was stained, the shoulder specimen was frozen,
coronal sections were cut through the glenoid fossa and the humeral head, and
the distances of the nerve from the glenoid rim and the inferior glenohumeral
ligament were measured.
Window Approach
Nine unembalmed cadaver shoulders from five female and four male donors,
ranging in age from seventy-eight to ninety-one years at the time of death,
were used. None of the shoulders had had previous surgery, and none had
advanced osteoarthritis. Whole-body cadavers were positioned in the standard
lateral decubitus position, and each arm was placed in traction with a 5-lb
(2.3-kg) weight applied through a pulley system to maintain the arm in 45°
of abduction and 20° of flexion.
After the shoulder was positioned, a T-shaped incision was made with a 6-cm
incision running parallel and 2 cm proximal to the lateral border of the
acromion and another 6-cm incision running distally from the lateral border of
the acromion (Fig. 1).
Identical cuts were made through the exposed deltoid muscle to expose the
superior muscles of the rotator cuff. A 2 × 1-cm piece of the acromion
was removed to better expose the rotator cuff muscles and the underlying
shoulder joint capsule. The supraspinatus and subscapularis muscles were
detached from the humerus to expose the humeral head and the long head of the
biceps tendon. The biceps tendon was excised, and approximately 2 cm of the
proximal part of the humeral head also was excised to allow visualization of
the glenoid face and inferior glenohumeral ligament.
A window (dotted lines in Fig.
2) was cut through the inferior glenohumeral ligament from inside
the capsule with use of microsurgical dissecting instruments. With the
insertion of the biceps tendon considered to be the 12 o'clock position on the
glenoid face, the inferior glenohumeral ligament was detached from the glenoid
labrum from the 4 o'clock to 8 o'clock positions, corresponding to the
anterior and posterior bands of the inferior glenohumeral ligament,
respectively (Fig. 3). The
inferior glenohumeral ligament was separated from the underlying connective
tissue and was excised as a rectangular piece, leaving a 2 × 1-cm window
in the capsule. Careful microdissection of the tissue exposed in the window
revealed the axillary nerve and its branches
(Fig. 4).
Coronal Section Approach
Ten additional unembalmed cadaver shoulders from five female and five male
donors, ranging in age from sixty-two to ninety-two years at the time of
death, were used. Through an anterior incision over the deltopectoral groove,
the axillary nerve was identified at its point of origin from the posterior
cord of the brachial plexus. Microsurgical dissection through the epineurium
of the most proximal part of the nerve exposed its fascicles. A saturated
aqueous solution of Evans blue dye (approximately 100 mg in 500 mL) was
injected into the perineurial space with a 30-gauge needle and syringe. This
technique was repeated for as many fascicles as possible
(Fig. 5). The arm then was
positioned in approximately 45° of abduction and 20° of flexion with 5
lb (2.3 kg) of traction and was placed in a freezer at 5° to 15°F
(—15° to —9°C) for forty-eight hours.
A fluoroscope was used to determine the plane of section. The initial
section ran perpendicular to the surface of the glenoid fossa, passing through
the 12 o'clock and 6 o'clock positions on the glenoid and through the center
of the humeral head and shaft (Fig.
6). It was marked with ink on the skin of the specimen, and then
the frozen specimen was divided along this center line with a band saw.
Parallel cuts were made 5 and 10 mm anterior to the first cut (toward the
anterior portion of the inferior glenohumeral ligament) and 5 and 10 mm
posterior to it (Fig. 7).
Figure 8 shows a typical
coronal section with the blue-stained axillary nerve in cross section. The
distances from the axillary nerve to the glenoid rim and from the axillary
nerve to the inferior glenohumeral ligament were measured at the 6 o'clock
position and 5 and 10 mm anterior and posterior to that position
(Fig. 9). The nerve branch
closest to the glenoid rim was used for the measurements reported here.
Window Approach
Dissection through the inferior glenohumeral ligament and the underlying
connective tissue revealed an axillary nerve branching pattern as depicted in
Figure 4. The same branching
pattern was seen in all specimens, and the borders of the subscapularis and
teres minor muscles were visualized anteriorly and posteriorly, respectively.
The branch of the axillary nerve supplying the teres minor lay closest to the
glenoid labrum in each dissection. This branch arose from the posterior
deltoid branch of the axillary nerve. The superior lateral brachial cutaneous
nerve arose from the posterior deltoid branch of the axillary nerve, parallel
to the teres minor branch.
Coronal Section Approach
The distance between the axillary nerve and the glenoid labrum in each
specimen is listed in the Appendix. In the centerline position, corresponding
to a perpendicular cut through the 12 o'clock and 6 o'clock positions of the
glenoid fossa, the average distance was 12.4 mm (95% confidence interval =
11.6 to 13.2). Anterior to the center line, the average distance was 14.5 mm
at 10 mm (95% confidence interval = 13.3 to 15.7) and 13.1 mm at 5 mm (95%
confidence interval = 12.2 to 14.0). Posterior to the center line, the average
distance was 13.9 mm at 10 mm (95% confidence interval = 13.0 to 14.8) and
12.9 mm at 5 mm (95% confidence interval = 12.1 to 13.7).
Statistical analysis of these data that fit a quadratic equation with the
center-line measurement as its center demonstrated that the distance of the
nerve from the glenoid rim was significantly smaller (average, 12.4 mm) at the
center-line (6 o'clock) position than it was at the other positions (p <
0.002). At the center-line position, the nerve was 1.78 mm closer to the
glenoid rim than it was 10 mm anterior and posterior to that position.
The individual distances between the axillary nerve and the inferior
glenohumeral ligament are also presented in the Appendix. In the center-line
position, the average distance was 2.3 mm (95% confidence interval = 1.7 to
2.9). Anterior to the center line, the average distance was 2.8 mm at 10 mm
(95% confidence interval = 2.3 to 3.3) and 2.6 mm at 5 mm (95% confidence
interval = 2.1 to 3.1). Posterior to the center line, the average distance was
2.8 mm at 10 mm (95% confidence interval = 2.1 to 3.5) and 2.5 mm at 5 mm (95%
confidence interval = 1.9 to 3.1). Analysis of these data showed no
differences, with the numbers available, between the measured distances at the
different cuts; thus the nerve traveled at a relatively fixed distance in
relation to the inferior glenohumeral ligament.
The course of the axillary nerve is often described with reference to the
quadrangular space. The quadrangular space is a helpful guide for
investigators performing dissections, who can identify its borders, but it is
less helpful for arthroscopic surgeons who cannot expose its borders.
Previous investigators of the anatomy of the axillary nerve used methods
that could have compromised the validity of their findings. Eakin et
al.1 placed a row of
stitches in the capsule parallel to the glenoid rim, froze the specimen, made
a single cut in the sagittal plane, and identified the axillary nerve with
subsequent dissection. With this approach, there is a risk that the capsule
will be distorted and the nerve will be displaced before the critical
measurements are made. Bryan et
al.3 performed an
entirely extracapsular axillary dissection to trace the nerve and then
measured only the distance from the axillary nerve to the inferior
glenohumeral ligament; they did not measure the distance from the glenoid rim.
Uno et al.2 used an
indirect measurement technique by first exposing the axillary nerve through an
extracapsular dissection, placing marker needles that pierced the nerve and
passed through the capsule, and then observing the position of the needles in
relation to the glenoid rim through an arthroscope.
We used two techniques that minimized distortion and displacement artifacts
and that placed the upper extremity in the position most often used in
arthroscopic surgery, 45° of abduction and 20° of flexion. The window
technique provided qualitative data on the course of the nerve and its
branching pattern as viewed from within the joint. With the coronal section
technique, carefully positioned intact frozen shoulder specimens were used to
provide quantitative data on the distances between the nerve and the glenoid
rim and the inferior glenohumeral ligament. The sections were made in a plane
perpendicular to the glenoid face and parallel to its long axis. Staining the
axillary nerve with Evans blue dye permitted its easy identification in the
sectioned tissues. As a result, dissection of the nerve was unnecessary, and
the risk of artifact was reduced. The nerve-staining technique reported here
is new, and future studies will be performed in an effort to advance its
application.
With regard to the branching pattern of the nerve, our findings agree with
those of previous studies. Zhao et
al.4 described four
branches in the quadrangular space: motor branches to the anterior and
posterior portions of the deltoid muscle, a sensory branch (the superior
lateral brachial cutaneous nerve), and a motor branch to the teres minor
muscle. We found that the branch to the teres minor and the branch supplying
the lateral cutaneous innervation lie closest to the glenoid rim and are
therefore the most vulnerable during surgical manipulation of the inferior
joint capsule. This observation corroborates the clinical finding by Wong and
Williams5 that 182
of 196 patients with postoperative axillary neuropathy sustained only sensory
deficits. Damage to the teres minor branch, however, may be difficult to
assess. Figure 10 shows the
typical branching pattern and course of the axillary nerve from an
intra-articular perspective.
Statistical analysis showed that the axillary nerve travels at a fixed
distance from the inferior glenohumeral ligament throughout its course. On all
cuts, the axillary nerve traveled adjacent to the capsule and was separated
from it by an average distance of 2.5 mm. Previously, Bryan et
al.3, using
extra-articular dissection of the axillary nerve, found its average distance
from the capsule to be 3.2 mm.
Eakin et al.1
reported that the nerve was closest to the glenoid (12.5 mm) at the
four-thirty position. In contrast, we found that the nerve lies closest to the
glenoid at the 6 o'clock position. (Fig.
11).
There were two major limitations to our anatomic study. First, we were
unable to exactly simulate the shoulder arthroscopy environment. As our
methods do not allow joint distention, we could not quantify the effect of
joint distention on the distance of the axillary nerve from the glenoid rim.
Additionally, we used cadaveric shoulder specimens with normal capsulolabral
anatomy. Thus, we could not replicate the effect that a patulous capsule or a
detached labrum could have on the nerve and the measured distances. The second
limitation is related to the distance of the smaller branches of the axillary
nerve—i.e., the branch to the teres minor and the branch to the superior
lateral cutaneous nerve—in relation to the glenoid rim. These smaller
branches may not have been adequately stained with the Evans blue dye with the
coronal section technique; thus, the true distance of the nerve or some of its
branches from the glenoid rim may be closer than what we measured. Because the
microdissection of the capsule with the window technique may subtly distort
the relationship of the axillary nerve to the inferior glenoid rim, we do not
believe that it is a reliable method for measuring the distances of these
smaller branches.
Despite these limitations, we believe that we have provided an improved
understanding of the branches of the axillary nerve and their relationship to
the inferior glenohumeral ligament and the glenoid rim. The window approach to
the axillary nerve provides direct visualization, from an arthroscopic
perspective, of the course and branching pattern of the nerve as it traverses
the quadrangular space. The coronal section approach provides accurate
morphometric data about the relationship of the axillary nerve with the
shoulder capsule and the glenoid rim. Our study provides the shoulder
arthroscopist with more accurate data regarding the vulnerability of the
axillary nerve.
Tables showing the distances of the axillary nerve from the glenoid rim and
from the inferior glenohumeral ligament are available with the electronic
versions of this article, on our web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM).
EakinCL,
Dvirnak P, Miller CM, Hawkins RJ. The relationship of the axillary nerve
to arthroscopically placed capsulolabral sutures. An anatomic study.
Am J Sports Med.1998;26:
505-9.26505
1998
[PubMed]
UnoA, Bain GI,
Mehta JA. Arthroscopic relationship of the axillary nerve to the shoulder
joint capsule: an anatomic study. J Shoulder Elbow
Surg.1999; 8:
226-30.8226
1999
[CrossRef]
BryanWJ,
Schauder K, Tullos HS. The axillary nerve and its relationship to common
sports medicine shoulder procedures. Am J Sports Med.1986; 14:
113-6.14113
1986
[PubMed][CrossRef]
ZhaoX, Hung
LK, Zhang GM, Lao J. Applied anatomy of the axillary nerve for selective
neurotization of the deltoid muscle. Clin Orthop.2001; 390:
244-51.390244
2001
[PubMed][CrossRef]
WongKL,
Williams GR. Complications of thermal capsulorrhaphy of the shoulder.
J Bone Joint Surg Am.2001;83 Suppl 2 Pt 2:
151-5.83151
2001
[PubMed]