Keywords
brachial plexus - cervical plexus - nerve transfer - spinal nerve roots
Introduction
Brachial plexus injuries represent 10% to 20% of peripheral nervous system injuries
and are mostly caused by high-energy trauma.[1] They often affect young, economically active people, resulting in important limitations
in daily living and professional activities.[2]
[3]
[4]
[5]
According to lesion level, they are commonly classified as upper injuries, affecting
C5-C6 or C5-C6-C7 roots, lower injuries, affecting C8-T1 roots, or complete injuries.[6] Kaiser et al.[7] reported that a prevalence of complete injuries of 53%, followed by upper plexus
injuries, with 39%, and lower plexus injuries, with 6%. The severity of these injuries
ranges from neuropraxia, usually with spontaneous resolution, to complete avulsion,
with no potential for recovery.[8]
For some authors, such as Verdins and Kapickis,[9] upper trunk involvement (C5-C6) results in significant disability, with loss of
shoulder function (abduction and external rotation), elbow flexion and forearm supination.
Strategies for brachial plexus repair include surgical exploration followed by reconstruction
using nerve grafting or nerve transfer. Graft reconstruction is reserved to post-ganglionic
lesions. Since pre-ganglionic lesions with root avulsion have no proximal stumps available
for graft repair, the surgical approach is based on nerve transfers.[3]
[8]
[10]
[11]
Options for nerve transfer are scarce in cases with upper trunk roots avulsion, including
accessory, phrenic and intercostal nerves. According to Abdouni et al.[12] and Malessy et al.,[13] accessory nerve transfer to the suprascapular nerve, a very common procedure for
brachial plexus reconstruction, has some limitations described in the literature.
In 1991, Yamada et al.[14]
[15]
[16] described C3-C4 root transfer for upper trunk (C5 and C6) avulsions reconstruction
using a nerve graft. In an anatomical study, Yang et al.[17] coaptated the avulsed portion of C5 to C4 root using a laminectomy and intracanal
dissection to approach the affected region. The advantage of this transfer would be
the reactivation of the scapular-humeral muscles contraction, leading shoulder movements
(external rotation and abduction) return with a single neurotization.
This study aimed to evaluate the anatomical features of brachial and cervical plexus
roots to demonstrate the feasibility of performing a C4-C5 nerve root transfer with
no need for a nerve graft.
Materials and Methods
Cadaveric brachial plexuses were analyzed using microsurgical dissection of to identify
and characterize the C4, C5 and C6 cervical nerves. Fifteen plexuses from 10 cadavers
were dissected; eight plexuses were from male cadavers and seven were from female
cadavers. Anthropometric data, such as gender, ethnicity, age, weight, and height,
were recorded.
Cadavers with known neuromuscular disease, other injuries or previous procedures at
the dissection site were excluded.
All dissections were performed by the same researcher aided by a surgical magnifying
glass with a 3.5-fold magnification capacity. Anatomical parameters from the C4, C5
and C6 cervical roots, including length, direction, and distance between them, were
recorded. Measurements were obtained with a millimeter tape and a digital caliper.
Dissection
After placing an interscapular cushion, a path parallel to the posterior border of
the sternocleidomastoid muscle, with approximately 8 cm in length, was created in
its middle third (nervous point) with supraclavicular extension ([Figure 1]). Dissection reached the subcutaneous tissue, platysma muscle and deep fascia. The
external jugular vein was found at the subcutaneous layer, descending superficially
to the sternocleidomastoid muscle. The posterior jugular vein draining into the external
jugular vein was also found in this region and it was ligated to facilitate the approach
to deep structures. The sternocleidomastoid muscle and the external jugular vein were
folded anteromedially, exposing the upper trunk (formed by C5 and C6) located between
the anterior and middle scalene muscles. Immediately above C5 root, C4 root was identified
in a more superficial plane ([Figure 2a]). Branches for the phrenic nerve (located on the anterior face of the anterior scalene)
and the dorsal scapular nerve were identified emerging from the C5 root. Branches
for the phrenic nerve, scalene muscles, levator scapulae muscle and accessory nerve
communicating branches were also identified emerging from C4.
Fig. 1 Right lateral cervical region and schematic representation of the approach used in
this study.
Fig. 2 Dissection of the right brachial plexus. Demonstration of C4, C5, and C6 roots, upper
trunk (TS) and phrenic nerve origin (a). C4 section distal to the phrenic nerve emergence
and simulation of neurotization with C5 root, which was also sectioned (b).
Next, nerve length and diameter, in addition to the distance separating both nerves
after their exit from the intervertebral foramen, were measured. An important parameter
for these measurements was the phrenic nerve origin in C4 and C5. Finally, a transfer
between both roots was simulated ([Figura 2b]), sectioning the distal portion of C4 root at the phrenic nerve emergence point,
taking care to include in it the largest number of branches that would innervate the
scalene and levator scapulae muscles to increase the amount of motor fibers for neurotization.
The origin of the phrenic nerve at C5 root may be released to facilitate its mobilization.
Measurements were made with a Digimess 150 mm quadridimensional digital caliper.
Results
[Table 1] shows anthropometric data from the dissected cadavers. Average age was 62 years
old, ranging from 38 to 86 years old. Maximum height was 182 cm, whereas minimum heigh
was 152 cm; average height was 167 cm.
Table 1
|
#
|
Age (years)
|
Side
|
Gender
|
Height (cm)
|
Weight (kg)
|
|
1
|
52
|
Right
|
Male
|
167
|
61
|
|
2
|
42
|
Right
|
Male
|
182
|
76
|
|
3
|
66
|
Left
|
Male
|
175
|
74
|
|
4
|
58
|
Right
|
Male
|
175
|
86
|
|
5
|
72
|
Left
|
Female
|
164
|
68
|
|
6
|
86
|
Right
|
Female
|
163
|
55
|
|
7
|
86
|
Left
|
Female
|
163
|
55
|
|
8
|
85
|
Right
|
Female
|
157
|
42
|
|
9
|
85
|
Left
|
Female
|
157
|
42
|
|
10
|
38
|
Right
|
Female
|
160
|
62
|
|
11
|
38
|
Left
|
Female
|
160
|
62
|
|
12
|
47
|
Right
|
Male
|
176
|
78
|
|
13
|
47
|
Left
|
Male
|
176
|
78
|
|
14
|
62
|
Right
|
Male
|
168
|
70
|
|
15
|
62
|
Left
|
Male
|
168
|
70
|
|
mean value
|
61.7
|
|
|
167.4
|
65.3
|
|
maximum value
|
86
|
|
|
182
|
86
|
|
minimum value
|
38
|
|
|
157
|
42
|
[Table 2] shows C4 and C5 roots measurements. An important parameter in dissection was the
phrenic nerve origin point in these two roots, as well as the interval between them.
If C4 root were intact, the origin of the phrenic nerve was respected, and only the
segment immediately distal to it was considered useful for mobilization and neurotization.
C5 root was measured in its whole length, both before and after phrenic nerve emergence,
and its entire length was deemed useful for mobilization and neurotization.
Table 2
|
#
|
C4 length after phrenic nerve origin (mm)
|
C5 length after phrenic nerve origin (mm)
|
C5 length before phrenic nerve origin (mm)
|
C5 total length (mm)
|
C6 length (mm)
|
Space (Interval) between C4-C5 (mm)
|
Difference between C4 and C5 length (mm)
|
|
1
|
13
|
19
|
10
|
29
|
18
|
10
|
16
|
|
2
|
14
|
18
|
8
|
26
|
19
|
10
|
12
|
|
3
|
13
|
20
|
7
|
27
|
18
|
10
|
14
|
|
4
|
11
|
20
|
7
|
27
|
20
|
11
|
16
|
|
5
|
11
|
23
|
5
|
28
|
15
|
9
|
17
|
|
6
|
13
|
18
|
7
|
25
|
16
|
9
|
12
|
|
7
|
12
|
18
|
6
|
24
|
15
|
9
|
12
|
|
8
|
13
|
21
|
7
|
28
|
17
|
9
|
15
|
|
9
|
12
|
21
|
8
|
29
|
18
|
9
|
17
|
|
10
|
12
|
19
|
8
|
27
|
18
|
10
|
15
|
|
11
|
11
|
19
|
7
|
26
|
17
|
10
|
15
|
|
12
|
14
|
20
|
8
|
28
|
18
|
11
|
14
|
|
13
|
15
|
20
|
8
|
28
|
19
|
11
|
13
|
|
14
|
12
|
18
|
9
|
27
|
17
|
10
|
15
|
|
15
|
13
|
18
|
8
|
26
|
16
|
10
|
13
|
|
mean value
|
12.6
|
19.5
|
7.5
|
27.0
|
17.4
|
9.9
|
14.4
|
|
maximum value
|
15
|
23
|
10
|
33
|
20
|
11
|
17
|
|
minimum value
|
11
|
18
|
5
|
23
|
15
|
9
|
12
|
Cases 6 and 7 had the shortest distance between roots at the vertebral foramens level
(9 mm), and the smaller difference in length between C4 and C5 roots (12 mm). Cases
4, 12 and 13 had a greater distance between C4 and C5 roots (11 mm), but the length
difference between roots remained greater in all cases (16, 14 and 13 mm). This difference
in length between roots allowed for a tension-free suture.
Discussion
The anatomical features from cervical and brachial plexuses at the cervical region
are known, as well as the distribution of their different branches to the neck and
upper limb.[18] For the cervical plexus, C4 root presents a wide distribution at the cervical region,
with a sensitive portion innervating the supraclavicular area skin via supraclavicular
nerves (with the contribution of C3). Its motor portion contributes to the innervation
of different muscles, especially cervical and appendicular muscles; its most important
branch is the phrenic nerve, formed by contributions from C3, C4 and C5, and responsible
for the motor innervation to the diaphragm. C4 root also provides branches for pre-vertebral
muscles (longus colli and longus capitis muscles), for the scalene and levator scapulae
muscles, and communicating branches for the accessory nerve to innervates the trapezius
muscle.[19] These muscles do not receive exclusive C4 innervation: the adjacent roots of the
cervical plexus also contribute to their innervation. Due to these multiple innervations,
a C4 root section would cause minimum to no neurological deficit, except for the branch
for the phrenic nerve, since C4 contributes with the largest amount of motor fibers
to its formation compared to C3 and C5. This branch is spared by the technique we
propose.
At the brachial plexus, the fibers coming from C5 innervate the shoulder, and their
injury causes a sensitive deficit in the lateral region of the shoulder, in addition
to external rotation and abduction deficits. Branches originating from the proximal
portion of C5 (dorsal scapular nerve and contributions to the phrenic and long thoracic
nerves) are considered compromised by C5 avulsion; therefore, they were sectioned
to facilitate root mobilization.
Today, most neurotization procedures for shoulder movements reactivation include an
accessory nerve transfer to the suprascapular nerve to stabilize the shoulder and
provide some external rotation, and the transfer of a radial nerve branch innervating
the triceps muscle for the axillary nerve, reactivating deltoid contraction to resume
shoulder abduction. Abdouni et al.[12] demonstrated that isolated accessory nerve neurotization often has frustrating outcomes
for shoulder function. In addition, the accessory nerve transfer prevents its future
use in a free muscle flap neurotization. On the other hand, the axillary nerve neurotization
requires an intact radial nerve, and it is restricted to upper (C5-C6) brachial plexus
injuries.
Yamada et al.
[14]
[15]
[16] described the C3-C4 root transfer for upper trunk (C5 and C6) avulsions reconstruction
but the proposed technique required the interposition of a nerve graft for roots connection;
this procedure increased the path for nerve regeneration, and graft placement resulted
in a two-fold increase in suture. In an anatomical study, Tsai et al.
[6] were able to coaptate the avulsed portion of C5 to C4 root with no need for a nerve
graft; however, this procedure required a laminectomy and intracanal dissection, increasing
its technical difficulty and resulting in higher morbidity.
Considering the virtually parallel path between C4 and C5 roots, the greater length
of C5 compared to C4 associated with roots mobilization allows to cover the gap between
these roots with no need for a nerve graft. For instance, at the first case, the space
between roots (at the vertebral foramina level) requiring coverage was 10 mm and the
difference between the length of C4 (after phrenic nerve emergence) and the total
length of C5 was 16 mm, i.e., more than enough to cover the gap and perform a free-tension
neurotization. In all cases, total C5 length was greater than C4 length and proved
to be sufficient to cover the gap between roots and to perform neurotization without
tension. In this study, we mobilized C5 to reach C4, and measurements were made from
the vertebral foramen. In avulsion injuries, the available C5 length is likely to
be even greater, even though the avulsed proximal segment requires resection.
Although neurotization of sensory and motor elements of the cervical plexus to repair
brachial plexus injuries has been described for a long time,[20] C3 and C4 motor branches are usually very thin, with few nerve fibers, and not providing
good outcomes in previous publications. This technique aimed to include a greater
amount of motor fibers in the neurotization, taking the C4 root along with motor fibers
for the scalene, rhomboid, levator scapulae and trapezius muscles (communicating branches
for the spinal accessory nerve).[18]
[21]
[22]
We recently operated on three patients with upper pre-ganglionic injuries confirmed
by magnetic resonance imaging (MRI) scans. Two of these subjects had a C5-C6-C7 lesion,
with no possibility of radial nerve branch transfer to the axillary nerve, whereas
the third patient had a C5-C6 lesion. In addition to C4-C5 neurotization ([Figure 3]), we also performed the Oberlin procedure for biceps brachii reinnervation. In all
cases, direct suture was feasible, with no nerve graft, and no deficit related to
C4 root sacrifice was observed postoperatively. These outcomes are encouraging, although
preliminary, since the time of postoperative follow-up ranges from 4 to 6 months.
Our aim is to carry out a larger series of cases and wait for clinical outcomes to
report them later.
Fig. 3 Intraoperative images of the left cervical region. Dissection showing C4 and C5 roots
and the phrenic nerve. (a). C4 root sectioned distal to the phrenic nerve origin and
C5 root sectioned immediately after its exit from the intervertebral foramen (b).
Neurorrhaphy between C4 and C5 sparing the phrenic nerve (c).
This technique will allow shoulder musculature reinnervation to recover shoulder external
rotation and abduction with a single neurotization, thus reducing the surgical time
and postoperative morbidity. The accessory nerve is spared, and it can be used, if
required, for another neurotization, as in gracilis free functional transfer. In addition,
this technique may assist in the treatment of total, complex brachial plexus injuries.
Conclusions
Simulated C4-C5 transfers were possible in all dissections using direct suture, with
no tension. The technique proved to be safe but required a trained team and an experienced
surgeon with extensive knowledge in regional anatomy.
