Content uploaded by Tommy Nai-Jen Chang
Author content
All content in this area was uploaded by Tommy Nai-Jen Chang on Feb 17, 2016
Content may be subject to copyright.
The Journal of Hand Surgery
(European Volume)
XXE(X) 1 –10
© The Author(s) 2014
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1753193414553753
jhs.sagepub.com
JHS(E)
Introduction
Despite the numerous classifications that have been
proposed for adult brachial plexus injuries (BPI)
(Alnot, 1987; Chuang, 1999; Millesi, 1988; Narakas,
1993; Terzis et al., 1999), there is no single system
that yields both clinically and surgically useful infor-
mation. Chuang (2009) described a new classification
system for the different levels of BPI. This uses
numerical descriptions that both simplify the injury
types and add useful clinical information (Table 1).
This classification system expands the popular
descriptions of ‘supraclavicular’ and ‘infraclavicular’
BPI, with Levels I–III encompassing supraclavicular
and retroclavicular injuries, and Level IV limited to
infraclavicular injuries.
This study focuses on Level IV injuries. This is
the second most common BPI after Level I injuries
(Chuang, 2013). They usually occur in isolation, and
there is rarely any upward extension of Level IV
injuries to Level I or II. There are unique features of
this injury, both anatomical and surgical, which
justifies its separate description. Owing to the
juxtaposition of the subclavian and axillary vessels
with the plexus at this level, there is a high inci-
dence of concomitant vascular injuries, presenting
either with rupture or segmental occlusion of the
arteries. The infraclavicular plexus is also situated
immediately anterior to the scapula and the gleno-
humeral joint, and there is a high incidence of
associated fractures of the scapula or dislocations
of the shoulder (Alnot, 1988). During surgery, the
surgeon dissecting the infraclavicular plexus often
encounters badly scarred tissues in multiple
planes. Finally, multiple long nerve grafts (8 cm)
Management of infraclavicular (Chuang
Level IV) brachial plexus injuries: A
single surgeon experience with 75 cases
W. L. Lam, D. Fufa, N.-J. Chang and D.C.-C. Chuang
Abstract
Infraclavicular brachial plexus injuries (Level IV in Chuang’s classification) have special characteristics,
including high incidences of associated scapular fractures, glenohumeral dislocations, and vascular injuries. In
addition, there are specific difficulties in surgical dissection and nerve repairs, especially if surgery is delayed
(3 months). A total of 153 patients with Level IV brachial plexus injuries underwent surgery between 1987
and 2008 with 75 patients (average age 29 years) available for a minimum of 4 years follow-up. Accompanying
fractures/dislocations were suffered by 48 (64%) patients, and 17 (23%) had associated vascular injuries.
The most common nerves to be injured were the axillary and musculocutaneous nerves. Nerve grafts to the
axillary, musculocutaneous, and radial nerves achieved impressive results, but less reliable outcomes were
achieved with the median and ulnar nerves. Decompression and/or external neurolysis were also beneficial
for nerve recovery. Some surgical tips are presented, and the use of the C-loop vascularized ulnar nerve graft
and functioning muscle transfers are discussed.
Level of Evidence: IV
Keywords
Infraclavicular, Level IV, brachial plexus injury
Date received: 1st July 2013; revised: 5th June 2014; accepted: 1st September 2014
Division of Reconstructive Microsurgery, Chang Gung Memorial
Hospital, Chang Gung University, Tao-Yuan, Taiwan
Corresponding author:
D. C.-C. Chuang, Department of Plastic Surgery, Chang Gung
Memorial Hospital, Chang Gung University, 5 Fu-Hsing Street,
Kuei-Shan, Tao-Yuan 333, Taiwan.
Email: dccchuang@gmail.com
553753JHS0010.1177/1753193414553753Journal of Hand SurgeryLam et al.
research-article2014
Full Length Article
at Chang Gung Mem.Hosp. Linco on December 29, 2014jhs.sagepub.comDownloaded from
2 The Journal of Hand Surgery (Eur)
are often required to reconstruct long segments of
nerve injury in this region.
This retrospective study analyses our experience
with the different injury patterns and methods of
reconstruction of Level IV injuries. Over the years,
improved understanding of the characteristics of the
injury, anatomy, and surgical approaches has allowed
the development and refinement of the methods for
managing such injuries to produce optimal outcomes.
Patients and methods
From 1987 to 2008 (a 22-year period), the senior
author (DCCC) operated on 1328 patients with differ-
ent levels of BPI. A total of 153 were Level IV BPI
(153/1328, 12%). Of these, 75 cases were selected
with a minimum of 3 years follow-up (Figure 1). A
total of 67 of the surgically treated patients were
male and the remainder were female, with an aver-
age age of 29 years (range 16–61). Most patients were
referred after a delay of more than 3 months. The
majority of injuries (63 patients) were caused by
motor-vehicle accidents (Table 2).
Outcome assessments were focused on the
proximal muscle groups innervated by the five ter-
minal nerves of the infraclavicular plexus: muscu-
locutaneous, axillary, radial, median, and ulnar
nerves (Table 3). Recovery of function in the intrin-
sics was rarely achieved, and therefore did not con-
stitute a component of our assessment. The
modified Medical Research Council (MRC) scale
system (M0–M5) was used (Chuang, 2008). Recovery
of motor function was assessed by the strength of
joint movements and not on individual muscles,
making the assessment simple and practical. A
score of M 3 was considered a successful result,
except for the axillary nerve, since patients with
deltoid paralysis in Level IV injuries will usually
present with good shoulder abduction because of
compensation from the supraspinatus muscle
(innervated by the suprascapular nerve). The recov-
ery of deltoid muscle power was therefore meas-
ured by muscle palpation and overall degrees of
shoulder abduction. Palpation of muscle contrac-
tion and shoulder abduction of more than 90° were
considered a success.
Table 1. The Chuang classification of levels of BPI.
Chuang levels of BPI Analogous to Anatomical and surgical characteristics
Level I Preganglionic (or supraganglionic)
root injury.
Inside the bone (and canal).
Requires laminectomy to visualize the nerves (roots).
Level II Postganglionic spinal nerve injury. Inside the scalene muscle.
Requires segmental resection of muscle.
Level III Trunk and division injury. Under the clavicle.
Requires osteotomy of clavicle.
Level IV Cord and terminal branches injury. Infraclavicular (after divisions) – see text.
BPI: brachial plexus injuries.
Figure 1. Distribution of patients in our series of BPI and infraclavicular injuries with at least 3 years follow-up.
at Chang Gung Mem.Hosp. Linco on December 29, 2014jhs.sagepub.comDownloaded from
Lam et al. 3
Surgical technique
An incision in the deltopectoral groove is made rou-
tinely for Level IV plexus exploration. The incision is
usually extended downwards through the axillary
fossa to the upper medial arm (Figure 2). The cephalic
vein, clavicular part of the pectoralis major muscle
inferiorly, deltoid muscle superiorly, clavicle bone
medially, and pectoralis major insertion laterally are
fully exposed. In the exposed pectoralis major mus-
cle, an intermuscular septum, called the ‘white line’,
located between the clavicular (C-PM) and sternal
(S-PM) parts of the pectoralis major is identified and
separated (Figure 3). This white line is a relatively
avascular zone, and allows the exposure of the
underlying pectoralis minor muscle once the two
parts of the pectoralis major muscle are separated.
The pectoralis minor muscle is then retracted medi-
ally with seldom any need for division. Once the
underlying adipofascial tissue is removed, the infra-
clavicular brachial plexus and juxtaposed subclavian
artery and veins become visible. The surgeon often
encounters extensive scars in the subpectoral space,
which renders identification of the different cords and
branches extremely tedious and difficult. Therefore,
if no identifiable structures can be seen, the dissec-
tion should start from the identification of healthy
nerves, either proximally, distally, or both.
Bidirectional dissection then makes the complex dis-
section easy and safe. Proximally, the dissection can
commence in ‘Chuang’s triangle’ (Chuang, 2006,
2013) (Figure 4) to identify the three cords and ves-
sels. This is a space bounded by two limbs: one is the
cephalic vein and C-PM muscle, and the other is the
clavicle. Some of the C-PM and deltoid muscle
attachments can be detached from the clavicle to
increase this space. A relatively avascular space, the
triangle contains the subclavius muscle, underneath
which lies the lateral and posterior cords. Distally,
the dissection starts in the upper arm or axillary
fossa to find the terminal branches. It is advisable to
detach the insertion of S-PM through a Z-lengthening
incision to facilitate good exposure of all components
of the distal plexus.
Management of vessel injuries depends on the
acuteness of presentation. In an open injury, immedi-
ate exploration is carried out and any arterial injuries
that are encountered are repaired. If segmental
artery occlusions are observed, a by-pass vein graft-
ing procedure is done.
Management of nerve injuries depends on the
condition of the nerves as well as the presentation.
Cleanly divided nerves in open wounds as a result
of sharp penetrating objects are repaired primarily.
Table 2. Demographics for 75 patients with level IV bra-
chial plexus injuries.
Gender (M:F) 67:8
Average age (years) 29 (range 16–61)
Mechanism of injury Motor-vehicle accident: 63
Hit by falling object: 2
Falling accident: 3
Penetrating injury: 7
Time to surgery (months) 5 (range 0–14)
Length of operation (hours) 9 (range 4–17)
Arterial Injuries (n = 17,
23%)
Axillary artery with
segmental occlusion: 13
Open injury with arterial
rupture: 4
Fractures or dislocations
(n = 48, 64%)
Humerus (n = 21)
Clavicular (n = 12)
Scapular or glenoid (n = 8)
Radius/ulna (n = 7/7)
Shoulder dislocation (n = 3)*
*This only includes the patients that underwent surgery.
Table 3. Assessment of outcome after nerve repair.
Nerve involved and repaired Motor functional evaluation Definition of success
Musculocutaneous nerve Elbow flexion strength (biceps and/or brachialis) M 3
Axillary nerve Deltoid Palpation (+) and
shoulder elevation 90°
Radial nerve 1. Elbow extension strength (triceps) M 3
2. Wrist extension strength (ECR and/or ECU) M 3
3 MPJ extension strength (ED) M 3
Median nerve 1. Wrist flexion strength (FCR and PL) M 3
2. Finger flexion strength (FDS) M 3
Ulnar nerve 1. Wrist flexion strength (FCU) M 3
2. Finger flexion strength (ulnar three FDP) M 3
ECR, extensor carpi radialis; ECU, extensor carpi ulnaris; ED, extensor digitorum; FCR, flexor carpi radialis; FCU: flexor carpi ulnaris;
FDP, flexor digitorum profundus; FDS, flexor digitorum superficialis; MPJ: metacarpophalangeal joint; PL, palmaris longus.
at Chang Gung Mem.Hosp. Linco on December 29, 2014jhs.sagepub.comDownloaded from
4 The Journal of Hand Surgery (Eur)
In open wounds, any reconstruction for nerve rup-
tures or obviously stretched nerves is delayed for a
few weeks. We do not routinely carry out nerve
resection and grafting at the primary exploration.
In delayed explorations, an injured nerve in conti-
nuity is treated by neurolysis. If the injured nerve is
ruptured with formation of a significant neuroma,
nerve grafting is done after resection of the nerve
ends to healthy fascicles. Once the extent of injury
is recognized, the lengths of nerve graft require-
ment are determined. Owing to the length of grafts
needed, donor nerve grafts from both sural
and saphenous nerves, as well as the lateral ante-
brachial cutaneous nerve and/or medial brachial
cutaneous nerve from the injured limb, are all
potential donors and their sites are marked.
Recently, consistently poor results after nerve
grafts to the ulnar nerve have been recognized,
which has encouraged the harvest of the ulnar
nerve as a vascularized ulnar nerve graft (VUNG).
This is especially useful in devastating total cord
injuries (all three cords damaged), each with a gap
of more than 15 cm. The VUNG can be used either
as a single graft to bridge defects in the median
nerve or as a ‘C’ looped graft to reconstruct the
median and radial nerves (Chuang, 2013; Terzis
1999). In the latter situation, the VUNG is harvested
on a proximally based pedicle via the superior col-
lateral ulnar artery. The nerve is then looped and
reversed, and the proximal and distal ends coapted
to the proximal ends of the median and radial
nerves in the upper arm. The loop is then posi-
tioned at the distal ends of the median and radial
nerves and divided, maintaining the vascular con-
nections via the connective tissues. The cut ends
are then coapted to the distal ends of the two
nerves (Figure 5). After repair of all nerves and/or
vessels, the divided S-PM is then repaired in a
lengthened fashion to cover the plexus.
Post-operative management and
rehabilitation
Immediate post-operative neck splinting for 3 weeks
after nerve grafts or nerve transfer is used in all
patients, together with a light shoulder sling for
4 weeks if the S-PM has been detached and repaired.
Thereafter, rehabilitation with muscle stimulation
and physiotherapy is started.
Figure 2. Incisions for Level IV plexus exploration: This
begins at the deltopectoral sulcus and extends downwards
to the upper medial arm.
Figure 3. Operative dissection showing the ‘white line’,
which allows separation of the clavicular and sternal part
of the pectoralis major muscles to access the underly-
ing pectoralis minor muscle and infraclavicular brachial
plexus and subclavian vessels.
Figure 4. Operative dissection showing ‘Chuang’s trian-
gle’ (upper arrow) and separation of the white line (lower
arrow). On opening the triangle, the underlying lateral and
posterior cord of the brachial plexus can be seen.
at Chang Gung Mem.Hosp. Linco on December 29, 2014jhs.sagepub.comDownloaded from
Lam et al. 5
Results
The timing of surgery varied from immediate explo-
ration in cases of penetrating wounds, to 14 months
when patients were referred from other hospitals,
with an average time of 5.2 months. Direct repairs
were only possible in five patients with open injuries,
with three axillary and two radial nerve divisions. The
average length of surgery was 9 hours (range 4–17).
Of the 75 cases, 23% (17/75) of patients had an
associated vascular injury, either an open injury with
axillary artery rupture requiring immediate explora-
tion and repair (four cases), or a closed injury with
subsequent segmental occlusion (12 cases) requiring
arterial reconstruction with prosthetic or vein grafts.
One patient with strong pectoralis muscles sustained
a penetrating injury, and was subsequently found to
have a spontaneously sealed injury of the subclavian
artery. This was discovered intraoperatively and
repaired 1 month after the injury.
A total of 64% (48/75) had an accompanying frac-
ture or joint dislocation (Table 2). This series did not
include those patients who had an initial plexopathy
secondary to shoulder dislocation but received no
subsequent surgical intervention. Only three patients
with continued plexus symptoms after shoulder dis-
location underwent exploration after an observation
period of 6 months. All of them were treated by
neurolysis.
Most of the injured terminal branches involved
the axillary nerve (52 patients), followed by the mus-
culocutaneous (38 patients), the radial (21 patients),
the median (20 patients), and the ulnar nerves (nine
patients). The commonest pattern of injury was a
combination of axillary and musculocutaneous
nerve palsies (34 patients), followed by an isolated
axillary nerve injury (17 patients). The injury pattern
can probably be explained by the more superficial
location of certain nerves (the lateral cord or mus-
culocutaneous nerve), or the close location to a bone
(the posterior cord or axillary nerve), rendering
these nerves more susceptible to injuries. There
were four patients with a total cord injury (all cords
and branches) who presented with a flail limb.
Conventional nerve grafts were most commonly
used for reconstruction, followed by neurolysis, and
finally, VUNGs (Table 4).
At a minimum of 3 years follow-up, the number of
patients who achieved a recovery of M3 after conven-
tional nerve grafts was excellent for the musculocuta-
neous and radial nerves, but more disappointing for
the median and ulnar nerves (Table 5) (Figure 6). Most
achieved a good result after axillary nerve reconstruc-
tion. Recovery after nerve grafting in the axillary,
musculocutaneous, and radial nerves was evident at
1 year follow-up in most patients (Table 5). In contrast,
only one patient recovered function in the median
nerve at 1 year and no patient achieved recovery of the
ulnar nerve within the same period. The average time
for recovery of elbow flexion (lateral cord, musculocu-
taneous nerve) was 14 months, compared with
19 months for elbow extension (posterior cord, radial
nerve) and 25 months for wrist flexion (medial cord,
median, and ulnar nerves). A total of 13 patients had
decompression and/or external neurolysis, and all
except one achieved good recovery within 1 year.
No ulnar nerve or medial cord was injured in isola-
tion; such injuries were usually part of a total cord
injury. Therefore, the VUNG was considered as a
reconstructive option whenever the ulnar nerve was
extensively injured (more than 15 cm), and when
there was an accompanying median or radial nerve
Figure 5. The C-looped VUNG.
Table 4. Different methods for reconstruction used in the
series.
Methods of reconstruction Numbers
Conventional nerve grafts 105
Vascularized ulnar nerve grafts 7
Neurolysis 13
Direct repair 5
Nerve transfer 2
Free functioning muscle transfers 10
at Chang Gung Mem.Hosp. Linco on December 29, 2014jhs.sagepub.comDownloaded from
6 The Journal of Hand Surgery (Eur)
injury also with a large defect (more than 15 cm). Of
the seven patients who were treated by the VUNG,
four (with total cord injuries) were for reconstruction
of both median and radial defects using a ‘C’ loop,
and three were for reconstruction of a median nerve
gap (with extensive median and ulnar nerve injuries).
Of the four patients who received a ‘C-looped’ VUNG
for reconstruction of both radial and median nerve
defects, three showed good radial nerve recovery by
1 year, whereas only two showed good median nerve
recovery by 2 years.
Only two patients underwent nerve transfers. In
these patients, an intercostal (T3-5) nerve to muscu-
locutaneous nerve transfer was done after it was
found that the nerve gap measured more than 15 cm
after resection of the damaged ends.
A free functioning muscle transplant (FFMT)
was carried out in ten patients to increase func-
tion. Eight patients had a FFMT after failed nerve
reconstruction (M < 3 after 3 years follow-up)
and the remaining two were late presenters to
the unit (more than 1 year after injury). All the
muscles used were gracilis myocutaneous flaps.
The donor nerves included the spinal accessory
nerve (four for elbow flexion, two for extensor
digitorum reconstruction), and the intercostal
nerves (one for elbow flexion and three for flexor
digitorum profundus reconstruction).
Discussion
Level IV injuries present with a wide degree of sever-
ity. Open Level IV injuries are usually caused by a
knife or other sharp objects, producing significant
motor and sensory deficits, which are immediately
apparent in the conscious patient. These require an
immediate or at the very least an early exploration
(within 2 weeks) to achieve optimal results. The ideal
scenario is immediate exploration and primary repair
of the cleanly divided nerves, in which case excellent
results can usually be expected (Chuang, 1999). In
gunshot wounds, unless a clear transection of the
nerves is found during exploration, the majority of
nerve palsies in Level IV injuries will improve sponta-
neously within 6 months. If there is no improvement
by 6 months, exploration and reconstruction can be
undertaken (Chuang, 1999). Similarly, when there are
isolated vessel ruptures or occlusions with a seem-
ingly intact plexus at exploration but accompanying
clinical nerve palsies, the nerve injury can be treated
expectantly, as most are likely to improve spontane-
ously (Table 6).
If there is any suspicion of traction involved with
an open wound, then a high likelihood of nerve rup-
ture is expected. This is different from simple nerve
division, as there is usually a longer length of dam-
aged nerve that requires resection until healthy
Table 6. Optimal repair time for Level IV BPI and artery injury.
Presentation Management recommendation
Penetrating injury: sharp knife or objects with nerve
and arterial severance
Immediate vessel and nerve repair
Penetrating injury: sharp knife or objects with nerve
severance only (no vessel rupture)
Immediate nerve repair or repaired as soon as
possible (2 weeks)
Open wound: traction with arterial and nerve ruptures 1. Artery repair first and nerve inspection
2. Nerve grafts 2–3 weeks after
Closed injury: arterial occlusion and nerve palsy but
perfused upper limb
Repair artery and nerve simultaneously 3–4 months
after accident if nerve shows no sign of recovery
Closed injury: nerve palsy only Exploration and repair at 3–4 months after injury if no
recovery (unless shoulder dislocation – wait 6 months)
Table 5. Recovery rates in the first 3 years (number of patients achieving good results) after nerve grafts for different nerve
injuries.
Nerve Year 1 Year 2 Year 3 Overall
Poor Good
Axillary 33 (70%) 7 0 6 40/46 (87%)
Musculocutaneous 22 (64%) 6 3 2 31/33 (94%)
Radial* 7 (54%) 4 1 1 12/13 (92%)
Median** 1 (9.1%) 4 1 5 6/11 (55%)
Ulnar 0 3 0 2 3/6 (50%)
*Four cases with radial nerve treated by vascularized ulnar nerve graft.
**Seven cases with median nerve treated by vascularized ulnar nerve graft.
at Chang Gung Mem.Hosp. Linco on December 29, 2014jhs.sagepub.comDownloaded from
Lam et al. 7
fascicles are reached in the nerve stumps. The gaps
then require reconstruction with nerve grafts with-
out any tension. The treatment for such injuries fol-
lows that of the closed traction injury (see below).
Debridement and nerve grafts at the time of the ini-
tial injury are generally not recommended. Instead,
re-exploration of the wound after waiting for at least
3 weeks allows the extent of nerve injury to be better
demarcated and better defined, facilitating better
judgment of the length of nerve to be resected and a
higher chance of obtaining healthy nerve ends
(Chuang, 1999).
The debate continues about the management of
closed traction injuries (Birch, 2014; Hems, 2014). In
Asian countries, the most common aetiology of BPI
remains motorcycle accidents, usually presenting as
closed traction injuries. Extensive experience has been
gathered because of the unfortunately high volume of
these injuries, which has allowed the development of a
reliable protocol. When a patient presents with a Level
IV closed traction injury, the following staged manage-
ment strategy is recommended (Chuang, 1999):
Stage 1, stabilization stage (the first month)
Stage 2, diagnostic stage (second month)
Stage 3, surgical stage (third and fourth months)
Stage 4, rehabilitation stage (at least 2 years)
Stage 5, late reconstruction (third and fourth years
post-operatively)
Figure 6. (A) and (B) A 24-year-old man with a total Level IV BPI, involving ruptures of all cords and terminal branches of the
left infraclavicular brachial plexus and subclavian artery after a car accident 1.5 months previously. He presents with a flail
left upper limb, except for shoulder abduction. (C), (D) and (E) Four years after nerve grafts to the musculocutaneous nerve
(8 cm nerve graft length x 2 nerve grafts), axillary (12 cm nerve graft length x 2 nerve grafts), radial (4 cm nerve graft length x
3 nerve grafts), median (4 cm nerve graft length x 4 nerve grafts), and ulnar (10 cm nerve graft length x 2 nerve grafts) nerves,
there is good restoration of elbow flexion and extension, wrist flexion and extension, and flexion of the fingers.
at Chang Gung Mem.Hosp. Linco on December 29, 2014jhs.sagepub.comDownloaded from
8 The Journal of Hand Surgery (Eur)
Some authors, however, have disagreed and rec-
ommended urgent exploration and nerve repair
within weeks, so as to avoid difficult dissections at a
later stage (Birch, 1997; Burge et al., 1985; Narakas,
1978). We have found the judgment of both the degree
and extent of nerve injuries at such an early stage to
be difficult and uncertain, whether in open or closed
wounds. The injured nerve suffers extensive bruising
and often appears oedematous in the early days after
a traction injury. In the weeks and months that follow,
it undergoes a fibrosis reaction, which allows much
easier clinical determination of the length of nerve
for resection and grafting. There is only one chance
to obtain a successful nerve repair, and selection of
healthy stumps for nerve coaptation is crucial for a
successful outcome. Close supervision of the patient
with a closed injury for a period of 3–6 months has
therefore been recommended by other authors as
well as ourselves (Chuang 2006, 2013; Millesi, 1988;
Terzis et al., 1999), and we remain of the opinion that
the benefits of waiting outweigh the advantage of
early exploration.
Accurate diagnosis of the extent of a Level IV injury
can be difficult at times. An accurate localization of
the injured elements in the brachial plexus avoids
unnecessarily big incisions (from neck to upper
medial arm) with extensive dissection and trauma to
tissues. This also shortens the operative time,
decreases post-operative wound pain and therefore
speeds up rehabilitation. Distinguishing a total Level
IV injury (all cords and branches) from a total root
avulsion injury (Level I) is clinically relatively straight-
forward, especially with the aid of nerve conduction
and imaging studies that can usually diagnose relia-
bly a Level I root avulsion injury, and therefore rule
out a Level IV injury (Chuang, 1999). However, in a
Level II or III injury, which presents with weakness or
paralysis of shoulder and elbow function but preser-
vation of hand function (e.g. a ruptured upper trunk
lesion with incomplete traction of C8/T1), differentia-
tion from a partial Level IV injury may be difficult in
the early stages. Neuro-imaging studies generally
have a limited role in diagnosing Level IV BPI. The use
of Tinel’s sign as an examination aid is helpful, but is
often subjective and therefore not 100% accurate.
Clinically, the preservation of function in proximal
muscles innervated by branches given off supraclav-
icularly or retroclavicularly (such as the supraspina-
tus/infraspinatus, serratus anterior, clavicular part
of pectoralis major, teres major, and latissimus dorsi
muscles) should alert the surgeon to the likelihood of
a Level IV lesion. However, the mere palpation of
muscle contractions does not guarantee 100% accu-
racy in diagnosis. This is due to the complexity of
anatomy, degree of injury, and mixed innervations of
muscles from the brachial plexus. A study into the
reliability of various predictors for pre-operative dif-
ferentiation between Level IV and Level III/II lesions is
currently proceeding within our department.
A clue to diagnosing a Level IV injury may be the
higher frequency of accompanying injuries. In both
Asian and Western countries, Level IV injuries are
associated with a higher incidence of shoulder dislo-
cation than other levels. This is especially true in
older patients (50 years). As shoulder dislocation is
usually a low energy traction injury, the potential for
recovery is high. However, the timing of treatment
remains crucial, especially the time between disloca-
tion and reduction (Chuang, 1999). If the reduction
occurs within 2 hours, the palsy usually recovers
within 2–3 months. However, if it is delayed for longer
than 3 hours, recovery usually takes at least 6 or
more months. We agree that conservative treatment
is indicated for any brachial plexus palsy caused by
shoulder dislocation (Hems and Mahmood, 2012;
Leffert and Seddon, 1965). However, if recovery is
stagnant after 6 months, surgical exploration should
be considered.
Once the decision had been made to explore a
Level IV injury, the surgeon may face a difficult and
tedious dissection due to the dense scarring and the
proximity of the major vessels. In an injury involving
multiple nerves, the need to decide which nerves to
be reconstructed and the method of nerve recon-
struction further complicates these operations. To
facilitate dissection, the exploration should always
start from areas where there is little or no scarring.
The use of well-defined landmarks helps in this: the
proximal stumps are reliably exposed through
Chuang’s triangle, and the distal stumps through the
medial part of the upper arm after division of the pec-
toralis major muscle. Once these are identified, bidi-
rectional dissection should make the anatomy much
clearer and the operative field wider, allowing identi-
fying of the lesions and functional preservation of the
pectoralis major and minor muscles. Finally, con-
comitant harvest of donor nerves from all four
extremities using a two-team approach saves time
and allows the harvest of multiple long nerve grafts.
Choosing the optimal method of nerve reconstruc-
tion remains a challenge. Recently, there is an
increasing popularity with the use of distal nerve
transfer techniques in treating BPIs. These tech-
niques are most commonly used for musculocutane-
ous (Mackinnon and Novak, 1999; Midha, 2004;
Oberlin et al., 1994) or axillary nerve reconstruction
(Garcia-Lopez and Perea, 2012; Witoonchart et al.,
2003). With such techniques, there is now a gradual
move towards the avoidance of dissection in the injury
zone, especially in Level I and II injuries. In Level IV
at Chang Gung Mem.Hosp. Linco on December 29, 2014jhs.sagepub.comDownloaded from
Lam et al. 9
injuries, however, surgical exploration of the injured
plexus should still be strongly advocated. The nature
of Level IV traction injuries is such that there is often
a higher percentage of nerve-in-continuity lesions
than in supraclavicular injuries. These cases in our
series (n = 13) were managed by simple neurolysis
with good clinical outcomes. Distal nerve transfers
without exploration of the site of injury may lead to
unnecessary iatrogenic sacrifice of donor nerves and
a result that is less optimal than neurolysis. In our
opinion, distal nerve transfer is only indicated for dis-
tal nerve avulsion at the point of entry into muscles,
which can only be revealed after surgical exploration
(Chuang, 1999).
Nerve grafts are needed for neuromas and rup-
tures with gaps of varying distances. Results follow-
ing nerve grafting are usually excellent for the
axillary, musculocutaneous, and radial nerves, per-
haps because of the less aberrant reinnervation at
this level of the plexus. Restoration of function after
injury to the median and ulnar nerves is less satisfac-
tory. There have been almost no reports of intrinsic
hand function recovery in most series (Alnot, 1988;
Kim et al., 2004). In the present series, only 55% and
60% of patients obtained recovery of the median and
ulnar nerves, respectively, and usually after a delay
of more than 12 months. The recovery markers were
limited to the proximal muscle groups (wrist and fin-
ger flexors). This is why in a total Level IV lesion the
use of the VUNG in the primary operation is an option
to improve the prognoses of median and radial nerve
reconstructions. Reconstruction of the ulnar nerve is
a futile operation for recovery in the intrinsic muscles
of the hand, and the vascularized nerve graft may
allow quicker and more reliable recovery of the
important functions of the median and radial nerves
(Terzis and Kostopoulos, 2009).
Functioning free muscle transplantation remains
a valuable option to restore or improve elbow flexion
when nerve reconstruction has failed or is not possi-
ble (Chuang, 2010). Level IV injuries usually spare the
spinal accessory nerve and/or intercostal nerves,
which remain available to power a free muscle trans-
fer at a later stage. Additional strategies to improve
function include tendon transfers, arthrodeses, and
tenodeses, or a combination of different methods.
For example, a combination of wrist and thumb car-
pometacarpophalangeal joint arthrodesis, in con-
junction with a subsequent free muscle transfer to
finger flexors, may permit the hand to regain some
degree of useful prehension (Chuang, 2010; Terzis
and Kostopoulos, 2009).
Conflict of interests
None declared.
Funding
This research received no specific grant from any funding
agency in the public, commercial, or not-for-profit sectors.
References
Alnot JY. Traumatic paralysis of the brachial plexus: pre-
operative problems and therapeutic indications. In:
Terzis JK (Ed.) Microreconstruction of nerve injuries.
Philadelphia, WB Saunders, 1987: 325–45.
Alnot JY. Traumatic brachial plexus palsy in the adult.
Retro- and infraclavicular lesions. Clin Orthop Relat
Res. 1988, 237: 9–16.
Birch R. Infraclavicular lesions. In: Boome RB (Ed.) The
brachial plexus. New York, Churchill Livingstone, 1997:
79–88.
Birch R. Timing of surgical reconstruction for closed trau-
matic injury to the supraclacicular brachial plexus. J
Hand Surg Eur. Epub ahead of print 30 June 2014. DOI:
10.1177/1753193414539865.
Burge P, Rushworth G, Watson N. Patterns of injury to the
terminal branches of the brachial plexus. The place
for early exploration. J Bone Joint Surg Br. 1985, 67:
630–4.
Chuang DC. Management of traumatic brachial plexus
injuries in adults. Hand Clin. 1999, 15: 737–55.
Chuang DC. Adult brachial plexus injuries. In: Mathes S,
Hentz VR (Eds.) Plastic surgery, 2nd Edn. Philadelphia,
Saunders- Elsevier, 2006, VII: 515–38.
Chuang DC. Nerve transfer with functioning free muscle
transplantation. Hand Clin. 2008, 24: 377–88.
Chuang DC. Adult brachial plexus reconstruction with the
level of injury: review and personal experience. Plast
Reconstr Surg. 2009, 124: 359–69.
Chuang DC. Brachial plexus injury: nerve reconstruction
and functioning muscle transplantation. Semin Plast
Surg. 2010, 24: 57–66.
Chuang DC. Brachial plexus injury: Adult and pediatric. In:
Neligan PC (Ed.) Plastic surgery, 3rd Edn. Philadelphia,
Saunders-Elsevier, 2013, VI: 789–816.
Garcia-Lopez A, Perea D. Transfer of median and ulnar
nerve fascicles for lesions of the posterior cord in infra-
clavicular brachial plexus injury: report of 2 cases. J
Hand Surg Am. 2012, 37: 1986–9.
Hems TEJ. Timing of surgical reconstruction for closed
traumatic injury to the supraclacicular brachial plexus.
J Hand Surg Eur. Epub ahead of print 30 June 2014. DOI:
10.1177/1753193414540074.
Hems TEJ, Mahmood F. Injuries of the terminal branches
of the infraclavicular brachial plexus: patterns of injury,
management and outcome. J Bone Joint Surg Br. 2012,
94: 799–804.
Kim DH, Murovic JA, Tiel RL, Kline DG. Infraclavicular
brachial plexus stretch injury. Neurosurg Focus. 2004,
16: E4.
Leffert RD, Seddon H. Infraclavicular brachial plexus inju-
ries. J Bone Joint Surg Br. 1965, 47: 9–22.
Mackinnon SE, Novak CB. Nerve transfers. New options for
reconstruction following nerve injury. Hand Clin. 1999,
15: 643–66.
at Chang Gung Mem.Hosp. Linco on December 29, 2014jhs.sagepub.comDownloaded from
10 The Journal of Hand Surgery (Eur)
Midha R. Nerve transfers for severe brachial plexus inju-
ries: a review. Neurosurg Focus. 2004, 16: E5.
Millesi H. Brachial plexus lesions: classification and opera-
tive technique. In: Tubiana R (Ed.) The hand. Philadelphia,
WB Saunders, 1988: 645–55.
Narakas A. Surgical treatment of traction injuries of the
brachial plexus. Clin Orthop Relat. 1978, 133: 71–90.
Narakas AO. Lesions found when operating traction inju-
ries of the brachial plexus. Clin Neurol Neurosurg.
1993, 95(Suppl): S56–64.
Oberlin C, Beal D, Leechavengvongs S, Salon A, Dauge MC,
Sarcy JJ. Nerve transfer to biceps muscle using a part
of ulnar nerve for C5-C6 avulsion of the brachial plexus:
anatomical study and report of four cases. J Hand Surg
Am. 1994, 19: 232–7.
Terzis JK, Kostopoulos VK. Vascularized ulnar nerve graft:
151 reconstructions for posttraumatic brachial plexus
palsy. Plast Reconstr Surg. 2009, 123: 176–91.
Terzis JK, Verkris MD, Soucacos P. Outcomes of brachial
plexus reconstruction in 204 patients with devastating
paralysis. Plast Reconstr Surg. 1999, 104: 1221–40.
Witoonchart K, Leechavengvongs S, Uerpairojkit C,
Thuvasethakul P, Wongnopsuwan V. Nerve transfer
to deltoid muscle using the nerve to the long head of
the triceps, part I: an anatomic feasibility study. J Hand
Surg Am. 2003, 28: 628–32.
at Chang Gung Mem.Hosp. Linco on December 29, 2014jhs.sagepub.comDownloaded from
