Received: 13 December 2015

| Accepted: 14 August 2016

DOI 10.1111/vsu.12590

ORIGINAL ARTICLE

C8 cross transfer for the treatment of caudal brachial plexus

avulsion in three dogs

Pierre Moissonnier1 | Claude Carozzo2 | Jean-Laurent Thibaut3 |

Catherine Escriou4 | Antoine Hidalgo1 | St
ephane Blot3,5

1
Unit
es de Chirurgie, Ecole nationale
v
et
erinaire d’Alfort, Universit
e
Paris-Est, Maisons-Alfort, France
2
Department of Surgery,
VetAgroSup, Campus v
et
erinaire de
Lyon, Marcy l’Etoile, France
3
Unit
es de Neurologie, Ecole
nationale v
et
erinaire d'Alfort,
Universit
e Paris-Est, Maisons-Alfort,
France
4
Unit
es de M
edecine Interne,
VetAgroSup, Campus v
et
erinaire de
Lyon, Marcy l’Etoile, France
5
Inserm, L’Institut Mondor de
Recherche Biom
edicale, University
Paris Est Cr
eteil, Cr
eteil, France
Correspondence
Pierre Moissonnier, Ecole Nationale
V
et
erinaire, 7 avenue du G
en
eral de
Gaulle, Maisons-Alfort, F-94704,
France.
Email: pierre.moissonnier@vet-alfort.fr
Abstract
Objective: To evaluate the cervical nerve 8 cross-transfer technique (C8CT) as a
part of surgical treatment of caudal brachial plexus avulsion (BPA) in the dog.
Study Design: Case series.
Animals: Client-owned dogs suspected to have caudal BPA based on neurological
examination and electrophysiological testing (n 5 3).
Methods: The distal stump of the surgically transected contralateral C8 ventral
branch (donor) was bridged to the proximal stump of the avulsed C8 ventral branch
(recipient) and secured with 9-0 polypropylene suture under an operating microscope.
A carpal panarthrodesis was performed on the injured limb after C8CT.
Results: Surgical exploration confirmed avulsion of nerve roots C7, C8, and T1 in
all cases. There was no evidence of an iatrogenic effect on the donor forelimb. Grad-
ual improvement in function of the affected forelimb occurred in all dogs, with
eventual recovery of voluntary elbow extension. Reinnervation was evident in EMG
recordings 6 months postoperatively in all three dogs. Stimulation of the donor C8
ventral branch led to motor evoked potentials in the avulsed side triceps brachialis
and radial carpus extensor muscles. Variable functional outcome was observed in the
3 dogs during clinical evaluation 3-4 years after surgery. Digital abrasion wounds,
distal interphalangeal infectious arthritis, and self-mutilation necessitated distal pha-
lanx amputation of digits 3 and 4 in 2 dogs.
Conclusion: C8CT provided partial reconnection of the donor C8 ventral branch to
the avulsed brachial plexus in the 3 dogs of this series. Reinnervation resulted in
active elbow extension and promoted functional recovery in the affected limb.

1 | INTRODUCTION and because of a fundamental lack of axonal regrowth from

Brachial plexus avulsion (BPA) is a traumatic condition
resulting from traction injury in which nerve roots of the
brachial plexus are torn away from the spinal cord.1–4 Fol-
lowing simple nerve transection (neurotmesis), peripheral
nerves may regenerate with good functional outcome after
surgical repair, depending on the distance between the lesion
and the muscular effector.5 In contrast, rootlet avulsion was
until now considered beyond surgical repair because of the
size of the gap separating the spinal cord from avulsed roots
damaged neurons in the central nervous system.6 BPA
results in permanent paralysis of affected myotomes and cor-
responding dermatomes.
Functionally, the canine brachial plexus can be divided
in 2 parts.4,7–9 The cranial part of the plexus originates from
the ventral branches of cervical spinal nerves C5, C6, and
C7, which control shoulder mobility and elbow flexion. The
caudal part of the plexus originates from the ventral branches
of C7, C8, T1, and sometimes T2, and innervates muscles
involved in elbow extension and carpus and digit mobility.

136 | V
C 2017 The American College of Veterinary Surgeons wileyonlinelibrary.com/journal/vsu Veterinary Surgery 2017; 46: 136-144
MOISSONNIER ET AL.
The caudal part of the plexus is extremely important for
locomotion because elbow extension is essential for weight
bearing.8
In the dog2–4,7 and in people,10 the treatment of BPA is
usually palliative. However, recent studies have reported that
rootlet reimplantation into the spinal cord in people can
promote motor reinnervation of denervated muscles.10,11 In
the dog, this potential surgical therapy for brachial plexus
injuries has unsatisfactory results because of the large dis-
tance of neurotization from the spinal cord to the muscular
effector, and because of the low number of motor neurons
that can regenerate through the reimplanted rootlets.12,13
Rootlet reimplantation has not been widely used in clinical
cases by veterinary surgeons for multiple reasons: the need
for weight bearing in dogs (which is not the case in peo-
ple); signs of reinervation take many months during which
continuous postoperative care is necessary; and the rela-
tively good quality of life after forelimb amputation in
dogs.3
We hypothesized that the poor results for surgical treat-
ment of BPA in the dog could be improved by using an
anastomosis between the avulsed part of the brachial plexus
and the distal stump of a donor nerve. A technique of cross-
transfer, using the contralateral brachial plexus, has been
described and applied in people.14–18 The C7 spinal nerve,
collected on the healthy side, was anastomosed to the ulnar
nerve in the affected limb. Three months later, the same
ulnar nerve could be connected to another injured peripheral
nerve, chosen according to the patient’s main functional def-
icit (ie, musculocutaneous, median, radial, or thoracodorsal
nerves).
We selected the C8 spinal nerve for a similar procedure
in dogs. The C8 ventral branch contributes to the radial,
median, and ulnar nerves in the dog. However, none of
these nerves originate solely in the C8 ventral branch, per-
mitting the C8 nerve to be used without loss of radial,
median, or ulnar nerve function. In a preliminary experi-
mental study (unpublished data), we performed C8 ventral
branch transection in 6 dogs. Over a period of 6 months a
clinical survey showed no functional consequences of this
procedure. We experimentally evaluated this new strategy in
the treatment of a caudal brachial plexus lesion and per-
formed the contralateral transfer of the right, eighth ventral
nerve branch (C8) in cats.19 C8 cross-transfer (C8CT)
enabled reinnervation of the contralateral brachial plexus in
cats and allowed the establishment of new functional neuro-
muscular units.19 Brachial plexus reinnervation can enable
the restoration of function, and could potentially lead to par-
tial recovery after caudal BPA in the cat.19 The aim of this
study was to demonstrate that the C8 ventral branch could
be used to neurotize the contralateral avulsed caudal plexus
in dogs.
| 137
2 | CLINICAL REPORT
2.1 | Selection of cases
All dogs included in this study were young adults in which
right sided caudal BPA was suspected by neurological exam-
ination and electromyographic testing, and confirmed by sur-
gical exploration of the brachial plexus (C7, C8, and T1
nerve roots found to be avulsed). Case 1 was an 8-month-old
female Labrador Retriever weighing 19 kg. Case 2 was an 8-
month-old female Rottweiller weighing 28 kg. Case 3 was a
1-year-old female mongrel weighing 14 kg. All dogs had
sustained trauma from being hit by a car 15 days (cases 1
and 3) or 1 month (case 2) before being examined at the
Centre Hospitalier Universitaire d’Alfort. No spontaneous
recovery could be expected in these cases and we suggested
to the owners that a C8CT could be performed. The owner
of each dog signed an informed client consent form before
the potential inclusion of their dog in the study.
At the time of examination, all dogs showed typical
lower motor neuron clinical signs of the affected limb con-
sistent with caudal brachial plexus dysfunction. Partial flex-
ion of shoulder and elbow was observed but neither elbow
nor carpus extension nor carpus flexion could be detected.
Tricipital, extensor carpi radialis, withdrawal and ipsilateral
pannicular reflexes were absent. Lack of cutaneous sensation
(using skin pinch) was observed over radial, ulnar, and
median nerve cutaneous distributions. Dogs 1 and 3 demon-
strated a right sided Horner’s syndrome while there were no
ocular abnormalities in Dog 2. Severe muscular atrophy was
obvious in all muscles of the affected limb at the time of
examination, but appeared particularly severe in the triceps
muscle and the antebrachial muscles.
Under general anesthesia, EMG recordings20,21 were
made before surgery with concentric bifilar and monopolar
needle electrodes connected to an EMG device (Viking IVD,
Nicolet Instruments Corporation, Madison, Wisconsin). These
studies were conducted in all muscles of the right forelimb.
EMG recordings were taken in the corresponding muscles of
the left (unaffected) forelimbs. Spontaneous electrical activity
following denervation was quantified using a subjective
scale22: 0 5 electrically silent muscle; 1 5 rare spontaneous
activity recorded at few sites; 11 5 diffuse occasional spon-
taneous activity or frequent spontaneous activity recorded at
few sites; and 111 5 diffuse abundant spontaneous activity
in the muscle. The level of spontaneous activity was variable.
The most severe spontaneous activity was recorded in the
muscles innervated by C7, C8, and T1, while muscles inner-
vated by C5 and C6 showed no spontaneous activity (Table
1). The results were consistent with avulsion and/or axonotm-
esis of the caudal part of the brachial plexus.
Compound muscle action potentials (CMAP) were
recorded (in mV) in the biceps, triceps, and carpus extensor
138 | MOISSONNIER ET AL.

T A B L E 1 Spontaneous electrical activities recorded before tanyl (5 mg/kg IV) and midazolam (0.2 mg/kg IV) for preop-
surgery in forelimb muscles innervated by the avulsed roots erative analgesia. Induction of general anesthesia was
C7, C8, and T1 achieved with thiopental (10 mg/kg IV). General anesthesia
was maintained with isoflurane in oxygen administered
Muscles Dog 1 Dog 2 Dog 3
through a semi-closed circuit. During surgery, dogs received
Interosseous 111 111 111 an IV infusion of 10 mL/kg/h lactated Ringer’s solution.
Analgesia was continued with a constant rate infusion of fen-
Carpal flexor 111 111 11
tanyl (2-5 mg/kg/h as needed), lidocaine (50 mg/kg/min), and
Carpal extensor 111 111 111 ketamine (10 mg/kg/min).
Triceps brachialis 111 111 11 The right brachial plexus was initially examined using
a ventral approach to the cervical spine (Figure 1). The
Biceps brachialis 11 11 11 conventional approach was modified to improve bilateral
Infra/supra spinatus 11 1 1 brachial plexus visualization. A midline incision was made
with a #10 scalpel blade between the sternocephalicus
Deltoid 11 1 11 muscles at the insertion on the manubrium sterni. The ster-
Pectoralis 11 1 1 nohyoidus and sternothyroidius muscles were not separated
to protect the esophagus and the trachea during mobiliza-
Paravertebral 111 11 11
tion. Caspar retractors were used to hold the trachea and
Scale for spontaneous electrical activity following denerva- the surrounding muscles apart, taking care to protect the
tion22: 0 5 electrically silent muscle; 1 5 rare spontaneous carotid sheath, vagus nerve, and the jugular vein from
activity recorded at few sites; 11 5 diffuse occasional spon-
inadvertent damage. The longus colli and the scalenius
taneous activity or frequent spontaneous activity recorded at
few sites; and 111 5 diffuse abundant spontaneous activity muscles were bluntly separated to locate the ventral
in the muscle. branches of spinal nerve C7, C8, and T1. In all dogs, C7,
C8, and T1 nerve roots (the dorsal ganglia could be identi-
fied), spinal nerves, and ventral branches were found under
muscles following supramaximal stimulation of C7 and C8 the scalenus muscles in connection to the brachial plexus,
ventral branches by a 0.45 3 50 mm stimulating anode elec- confirming a caudal BPA (Figure 2). Nerve roots of the
trode introduced percutaneously, caudal to the transverse spinal nerves and their corresponding ventral branches
process of the sixth cervical vertebrae. The stimulating cath- were isolated from a well-established hematoma (Dogs 1
ode electrode was positioned under the skin dorsal and ros- and 3) or from a moderate fibrous scar (Dog 2). The left
tral to the scapulohumeral joint. The recording needle was eighth cervical ventral branch (donor side) was exposed as
positioned in the muscles (active electrode in the belly of the distally as possible and severed before it reached the left
muscle and reference electrode in the tendon distal to the plexus. The proximal extremity of the right (avulsed) C8
active electrode). Both sides were tested with the unaffected was debrided and cut sharply. The distal stump of the left
side being tested first for a frame of reference. No CMAP C8 was bridged to the proximal stump of the avulsed right
were recorded in muscles innervated by the radial, ulnar, or C8, and anastomosed with 6-8 epiperineural simple inter-
median nerves on the right side (side of avulsion). During rupted sutures using 9-0 polypropylene under an operating
this examination, no late potentials (H or F waves) were microscope (Figure 3). The incisions of the surgical site
present in the affected muscles. Sensory nerve conduction were closed routinely. The sternohyoideus muscle and the
was present in the right ulnar and radial nerves (conduction mastoid part of the sternocephalicus muscle were apposed
velocity 5 62-75 m/s; latency 5 2.9-3.5 m/s) for all dogs. along the midline by separate cruciate sutures. The subcu-
The absence of CMAP but presence of a sensory nerve taneous fascia and the skin were closed by continuous pat-
action potential for the same nerve supported a diagnosis of tern sutures.
nerve root avulsion. All electrophysiological testing was con- One (Dog 3), 2 (Dog 2), or 6 months (Dog 1) after
sistent for severe axonotomesis and/or neurotmesis of the C8CT, right side carpal panarthrodesis were performed. A
caudal part of the brachial plexus. dorsal midline approach of the distal part of the radius, the
carpus, and the metacarpus was performed. The radiocarpal,
middle carpal, and carpometacarpal joint surfaces were
2.2 | Surgical procedure
exposed and articular cartilage was removed. A bone plate
Dogs were anesthetized and routinely prepared for aseptic was applied as a compression plate to the dorsal surface of
surgery of the ventral cervical region. Dogs initially received the radius, the carpus, and the metacarpus. The skin was
diazepam (0.2 mg/kg IV) as a preanesthetic medication, fen- sutured with interrupted cruciate sutures.
MOISSONNIER ET AL.
| 139

FIGURE 1 Semischematic perioperative view and cross section describing the C8CT technique. The muscles and vasculature
are not represented to improve visibility. The left C8 ventral branch is dissected as distally as possible before dividing into the left
brachial plexus peripheral branch, then transected and pulled between the trachea and longus coli muscle. This distal stump of left
C8 ventral branch (donor side) is anastomosed to the proximal stump of the right C8 ventral branch (avulsed from the spinal cord;
recipient side). C5-8 5 ventral branchs of cervical spinal nerves 5-8. T1 5 ventral branch of thoracic spinal nerve 1. *Avulsed ven-
tral branches. 1 5 Longus coli muscle. 2 5 First cervical vertebra. 3 5 First rib. 4 5 Trachea. 5 5 Esophagus. X 5 Distal section of
normal C8. O 5 Anastomosis of the distal stump of left C8 ventral branch and the proximal stump of the right C8 ventral branch

2.3 | Postoperative treatment 2.4 | Postoperative evaluation

Following anesthetic recovery, pain was controlled by mor-
phine chlorydrate injection (0.2 mg/kg subcutaneously every
6 hours for 24 hours), and by ketoprofen (2 mg/kg IM every
once daily) given over 4 days. Postoperative treatment con-
sisted of passive mobilization of the limb 2 times a day for
20 minutes, muscle electrostimulation,23,24 and digit protec-
tion using a homemade leather brace covering the entire
limb. Isometric electrostimulation (50 Hz; Genesy 600,
Globus, Italy) was applied 30 minutes per day (during the
first 6 months) and twice a week (until EMG revealed
CMAP). The stimulator generated a constant biphasic charge
voltage (80 mV) with stimulation currents of 250 mA
applied by cutaneous electrodes positioned at the level of the
brachial and antebrachial muscles. In practice, the stimula-
tion current was progressively increased until the dog exhib-
ited discomfort and then decreased to a lower intensity.
Stimulation was performed by the surgical staff during the
hospitalization period, and afterwards, by the owner.
The effects of C8CT on the left and right limbs were moni-
tored during clinical exam (Dog 1 for 4 years; Dog 2 for 5
years; and Dog 3 for 5 years) and EMG (Dog 1 for 3 years;
Dog 2 for 5 years; and Dog 3 for 5 years) evaluation. Dogs
maintained their ability to walk on the donor forelimb (left)
immediately after surgery without any obvious functional
deficits. There was no evidence of severe pain during the
postoperative period.
With time, muscle atrophy of the right limb improved
and the brachial triceps progressively regained muscle mass.
Six months to 1 year after surgery, gradual improvement in
the function of the right forelimb occurred in all dogs. Vol-
untary elbow extension was observed and dogs could bear
their weight on the repaired limb with the help of external
restraint with a modified Robert Jones bandage, but there
was no carpal extension at this time. After carpal panarthrod-
esis, the dogs regained partial use of their limb with external
protection of the toe. At the end of the survey period, the
140 | MOISSONNIER ET AL.

and not in contralateral muscle. These MUAP are consistent

FIGURE 2 Perioperative view of the avulsed ventral roots
and branches during exploration of the right brachial plexus in
Dog 2
functional outcome appeared to be variable from 1 dog to
another. Dog 1 showed good functional recovery with walk-
ing, running, and jumping with the right forelimb (Figure 4),
while persistent moderate lameness was evident in Dogs 2
(grade 3) and Dog 3 (grade 2).
Two to 3 months after surgery, EMG recordings of rein-
nervation appeared in the muscles of the grafted limb. Spon-
taneous activity on the right side in the brachial triceps,
radial extensor, and interosseous muscles was only observed
in Dog 2 whereas spontaneous activity was absent in all
muscles of the right frontlimb of Dogs 1 and 3. These obser-
vations could be consistent with reinnervation or the forma-
tion of fibrous tissue in the denervated muscle. In Dog 1, 6
months after surgery, long duration and polyphasic (5 turns
or more) motor unit action potentials (MUAP) were observed
in the right triceps brachialis during voluntary contraction
with a reinnervation process.25
Stimulation of the right caudal brachial plexus and direct
stimulation of the left C8 ventral branch led to the recording
of CMAP in the right triceps brachialis 6 months postopera-
tively in Dog 1 and was still observed around 4 years after
surgery in Dogs 2 and 3, showing that a reinnervation process
was occurring and that the muscle was not fibrotic (Table 2).
Progressively the left triceps muscle became free of fibrilla-
tion potentials and of positive sharp waves. There was no
spontaneous activity on the left side 6 months after surgery in
any dogs. Low sensory nerve conduction was recorded in the
radial and ulnar nerves. The loss of sensation and trophic dys-
function led to late complications such as toe abrasion
wounds (3 dogs), distal interphalangeal infectious arthritis (1
dog), and self-mutilation (1 dog). Phalanx amputation of dig-
its 3 and 4 was necessary in Dogs 1 and 3.
3 | DISCUSSION
Our case series shows that in dogs with caudal BPA, C8CT
leads to the reconnection of the donor plexus to the avulsed
brachial plexus. This reinnervation promotes a partial func-
tional recovery of the avulsed caudal brachial plexus. Vari-
ability in the functional results seen in the 3 dogs of our
study could be explained by a fiber misrouting phenomenon
or by the variability in the anatomy of the brachial plexus.
The small sample size is a major limitation of the study,
nevertheless, C8CT should be considered as an adjunctive
possibility in the treatment of caudal BPA in dogs.
The brachial plexus arises from the ventral branches of the
C5 to T2 spinal nerves. These branches contribute in a

FIGURE 3 Perioperative views under operating microscope. A, The left eighth cervical ventral branch (*) has been severed and
is transferred over to the right side by pulling it between the longus colli muscles and the trachea, to the level of the right brachial
plexus. B, Sutures between the distal stump of the left C8 to proximal stump of avulsed right C8
MOISSONNIER ET AL.
| 141

FIGURE 4 Clinical evaluation of locomotion in Dog 1, 2 years after C8CT and carpal panarthrodesis. These sequential pictures,
extracted from a video recording, show that the dog is using its right limb to run with a moderate lameness. Contraction of the
brachial triceps muscle leads to voluntary elbow extension and weight bearing on the limb

variable manner to the formation of the brachial plexus.7,9 The
brachial plexus is made up of C6, C7, C8, and T1 in 58.62%
of dogs, C5 to T1 in 20.69% of dogs, C6 to T2 in 17.24% of
dogs, and C5 to T2 in 3.4% of dogs. The origins of peripheral
nerve roots have also been reported to be variable.7,26,27 The
radial nerve usually originates from C7 to T1. The median and
the ulnar nerves typically originate from C8 and T1. In the sit-
uation of partial BPA, functional recovery could be the result
of collateralization of remaining axons originating from intact
branches or from less severely damaged branches (demonstrat-
ing a neurapaxia or an axonotmesis).
The decision to perform C8CT on the dogs in our series
was made after avulsion of the C7, C8, and T1 nerve branches
was confirmed by surgical exploration. Based on the anatomic
confirmation of avulsion, postoperative recovery of the triceps
brachialis muscle function in the dogs of our study could be
attributed exclusively to C8CT. Functional recovery of the
radial carpal extensor muscle, which is mainly innervated by
C7, did not occur. The C7 nerve was not directly involved in
cross transfer and would require axonal regrowth over a long
distance for recovery. Thus, in caudal BPA (C7, C8, T1),
functional recovery of the limb muscles depends exclusively
(ulnar nerve) or partially (contribution of C6 to the radial
nerve) on the C8CT.
The current treatment of complete BPA in dogs is only
palliative and limb amputation is often proposed.1–5,21 Various
surgical techniques have been experimentally designed to treat
BPA in people. These include avulsed spinal nerve root reim-
plantation into the spinal cord,28–31 neurotization of the plexus
with either the phrenic nerves,32 the intercostal nerves,33–35 or
the accessory nerves,36–38 cross transfer,39 and, more recently,
C7-cross-transfer.14–18 The choice among these techniques for
human patients remains controversial and depends on the
exact location and extent of the avulsion.
In dogs with partial BPA, limited treatment can be per-
formed by various surgical treatments including arthrodesis,
and the use of various external devices, such as casts. C8CT
offers an alternative to amputation in the case of caudal
BPA. However, the complexity of the surgical procedure, the
length of postoperative care, and potential complications
must be considered when selecting patients as well as the
owners. Indeed, avoiding amputation in the dogs of this
series necessitated constant postoperative care even after
motor recovery occurred. The axonal regrowth speed after
anastomosis is estimated to be 1-2 mm per day. It may
require 1 year or more for reinnervation of a large breed
dog’s forelimb. During this time, the distal limb remains
insensible (dermatome anesthesia), and constant monitoring
of the paw is necessary to avoid formation of secondary
trophic or traumatic injuries (abrasion, infection). Owners
were informed of this in the consent form.
From the technical point of view, C8CT appears simpler
than root reimplantation. Moreover, there is no risk of
the dog developing spinal cord trauma secondary to

T A B L E 2 Motor evoked potentials (mV) recorded after supramaximal nerve stimulation preoperatively (preop) and during
the last EMG test

Dog 1 Dog 2 Dog 3

Tested Sites of right
nerve side recording Preop 6 months Preop 4.3 years Preop 3.6 years
Radial Triceps brachialis 0 0 0 3.5 0 30.8

Extensor carpi radialis 0 14.2 0 19.9 0 21.9

Ulnar Interosseus 0 0 0.5 10.4 0 8.3

Left C8 Triceps brachialis – 5.9 – 14.4 – 27.4

The left (donor) C8 ventral branch was stimulated at the left axilla to evaluate reinnervation of the right (avulsed) plexus by the
donor branch. The right peripheral nerves (radial and ulnar) near the right humeral shaft were stimulated to evaluate the reinner-
vation of these nerves after C8CT.
142 | MOISSONNIER
reimplantation into the spinal cord.12 The low morbidity we
observed after C8 transection in the donor limb is similar to
the low morbidity observed in human patients14 after C7
transection. Furthermore, the distance separating the site of
anastomosis from the targeted muscle is shorter, reducing the
time necessary for neurotization and increasing functional
recovery.40 Indeed the result of the neurotization procedure
is mainly affected by the regenerative ability of reinnervated
muscles and the remaining amount of muscular fiber.41,42
Muscle atrophy is transitory if the denervation period is
short.43 In our study, reinnervation was documented in the
triceps muscle 6 months after surgery and the function of
this muscle was adequate for limb extension and weight
bearing at the end of the survey period. The radial extensor
muscle showed EMG signs of reinnervation 3 years after
C8CT; however, this muscle did not recover functionality.
Carpal panarthrodesis was considered as soon as possible to
allow dogs to regain weight bearing when reinnervation of
the triceps brachialis muscle provided functional elbow
extension. Our observations were similar to those of clinical
trials in human patients after contralateral transfer of C7.14–17
The choice of C8 in the dog was made after a comparative
anatomical study demonstrating that C8 plays an important
role in the constitution of all peripheral nerves in the fore-
limb and after an experimental study showed that C8 transec-
tion did not lead to a severe deficit in the cat.19
Fiber misrouting could explain inconstant results after
peripheral nerve (or plexus) surgery, and this could be one of
the technique limitations.44 It can be defined as the amount
of regenerating nervous fibers that reconnect to the wrong
target. Before avulsion, a peripheral nerve is made up of var-
ious fiber types including autonomic fibers (going, for exam-
ple, to an artery), motor fibers connected to flexor muscles,
motor fibers connected to extensor muscles, sensory fibers
(coming, for example, from a Meissner corpuscule), and sen-
sory nociceptive fibers that terminate in the skin. During neu-
rotization, the wrong pathway can be taken by the nerve and
the new distribution pattern can connect, for example, to a
vasomotor fiber, or to the end-plate of a flexor muscle, or
any other possible combination. Current methods are insuffi-
cient to demonstrate the extent of misrouting. However, it
could explain the variable results of neurotization. To
improve the results of C8CT, one of the surgical aims would
be to correctly select the fascicle (motor vs. sensory) used
for neurotization. Another aim would be to increase the
speed of axonal regrowth with substances such as nerve
growth factors or stem cells, which have been recently eval-
uated in experimental studies in dogs.45,46
The number of axons regenerating in the avulsed plexus
and the small amount of fiber misrouting is decisive in the
functional outcome of the surgical procedure. In root reim-
plantation the number of regenerated axons is low12 and
ET AL.
these axons have to follow complex pathways (ventral root-
lets, ventral root, ventral branch, plexus, and peripheral
nerve) to the muscle. In C8CT, regenerating axons are
directly injected into the caudal plexus. In a postoperative
survey of 10 human patients,31 recorded CMAP 9-12 months
after root reimplantation for BPA, and three people showed
satisfactory recovery 2 and 3 years after surgery. We believe
that C8CT provides a close viable nerve source, thus ensur-
ing a more rapid muscle reinnervation. However, it necessi-
tates a learning phase for the patient, which requires
contralateral spinal neurons to command the avulsed plexus
as has been demonstrated in other neurotization techniques
in the cat.47
Our study suggests that an active therapeutic approach
for BPA could enhance the functional outcome of this trau-
matic condition. Therefore, C8CT appears to be an attractive
candidate for caudal BPA treatment. However, before it can
be used routinely, many problems have to be addressed such
as better correlation between reinnervation and functional
recovery, preserving muscle structure during the denervation
period, attaining more selective neurotization to decrease
axonal misrouting, and stimulating axonal regrowth.
ACKNOW L E DGM E NT
The authors thank Pr. JA Flanders (Cornell University,
Ithaca, NY) for his critical reading and his invaluable help
for rewriting this article, and E Josi
e for his drawings.
C O NFL IC T O F IN T E RE S T
The authors declare no conflicts of interest related to this
report.
R EFE RE NC ES
[1] Griffiths IR, Duncan ID, Lawson DD. Avulsion of the
brachial plexus-2. Clinical aspects. J Small Anim Pract.
1974;15:177–183.
[2] Wheeler SJ, Clayton-Jones DG, Wright JA. The diagno-
sis of brachial plexus disorders in dogs: a review of
twenty-two cases. J Small Anim Pract. 1896;27:147–157.
[3] Rodkey WG, Sharp NJH. Surgery of the peripheral nerv-
ous system. In: Slatter D, ed. Textbook of Small Animal
Surgery. Philadelphia, PA: Saunders; 2003:1218–1219.
[4] Platt SR, da Costa RC. Brachial plexus trauma. In: Tobias
KM, Johnston SA, eds. Veterinary Surgery: Small Animal.
St. Louis, MO: Elsevier Saunders; 2012:424–430.
[5] Gibson KL, Daniloff JK. Peripheral nerve repair. Comp
Cont Educ Pract Vet. 1989;11:938–945.
[6] Aguayo AJ. Axonal regeneration from injured neurons in
the adult central nervous system. In: Cotman CW, ed.
Synaptic Plasticity. New York, NY: Guilford Press;
1985:457–484.
MOISSONNIER ET AL.
[7] Sharp JW, Bailey CS, Johnson RD, et al. Spinal nerve
root origin of the median, ulna and musculocutaneous
nerves and their muscle nerve branches to the canine
forelimb. Anat Histol Embryol. 1990;19:359–368.
[8] Sharp NJH. Neurological deficits in one limb. In:
Wheeler SJ, ed. Manual of Small Animal Neurology.
Shurdington, UK: British Small Animal Veterinary Asso-
ciation; 1995:159–178.
[9] Allam MWD, Lee DG, Nulsen FE, et al. The anatomy of
the brachial plexus in the dog. Anat Rec. 1952;114:173–180.
[10] Friedman AH. Neurotization of elements of the brachial
plexus. Neurosurg Clin North Am. 1991;2:165–174.
[11] Chang KW, Justice D, Chung KC, et al. A systematic
review of evaluation methods for neonatal brachial
plexus palsy. J Neurosurg Pediatr. 2013;12:395–405.
[12] Moissonnier P, Duchossoy Y, Lavieille S, et al. Evalua-
tion of ventral root reimplantation as a treatment of
experimental avulsion of the cranial part of the brachial
plexus in the dog. Revue M
ed V
et. 2001;152:587–596.
[13] Moissonnier P, Duchossoy Y, Lavieille S, et al. Lateral
approach of the dog brachial plexus for ventral root reim-
plantation. Spinal Cord. 1998;36:391–398.
[14] Gu YD, Zhang GM, Chen DS, et al. Seventh cervical
nerve root transfer from the contralateral healthy side for
treatment of brachial plexus root avulsion. J Hand Surg.
1992;17:518–521.
[15] Liu J, Pho RWH, Kour AZ. Neurologic deficit and
recovery in the donor limb cross-C7 transfer in brachial-
plexus injury. J Reconstr Microsurg. 1997;13:237–343.
[16] Gu YD, Chen DS, Zhang GM. Long-term functional
results of controlateral C7 transfer. J Reconstr Microsurg.
1998;14:57–59.
[17] Terzis JK, Kokkalis ZT, Kostopoulos E. Contralateral C7
transfer in adult plexopathies. Hand Clin. 2008;24:389–400.
[18] Lin H, Sheng J, Hou C. The effectiveness of contralateral
C7 nerve root transfer for the repair of avulsed C7 nerve
root in total brachial plexus injury: an experimental study
in rats. J Reconstr Microsurg. 2013;29:325–330.
[19] Carozzo C, Cuvilliez V, Escriou C, et al. Cross-neuroti-
zation of the caudal brachial plexus with the controlateral
C-8 ventral nerve branch in the cat: potential surgical
applications, effects of graft collection on the healthy
donor limb and results. Neurochirurgie. 2005;51:89–105.
[20] Duncan ID. Electromyography and nerve conduction
studies. In: Wheeler SJ, ed. Manual of Small Animal
Neurology. Shurdington, UK: British Small Animal Vet-
erinary Association; 1995:50–52.
[21] Sims MH. Electrodiagnostic evaluation. In: Braund KG,
ed. Clinical Syndromes in Veterinary Neurology. St.
Louis, MO: Mosby; 1994:349–362.
[22] Le Chevoir M, Thibault JL, Labruyère J, et al. Electro-
physiological features in dogs with peripheral nerve
sheath tumors: 51 cases (1993-2010) J Am Vet Med
Assoc. 2012;241:1194–1201.
| 143
[23] Kanaya F, Tajima T. Effect of electrostimulation on
denervated muscle. Clin Orthop. 1992;283:296–301.
[24] Pshenisnov KP, Pulin AG. The use of preoperative muscle
denervation and postoperative electrostimulation to max-
imalize functional results in microneurovascular muscle
transplantation. J Reconstr Microsurg. 1994;10:65–75.
[25] Wijnberg ID, Back W, de Jong M, et al. The role of
electromyography in clinical diagnosis of neuromuscular
locomotor problems in the horse. Equine Vet J. 2004;36:
718–722.
[26] Sharp JW, Bailey CS, Johnson RD, et al. Spinal root ori-
gin of the radial nerve and nerves innervating shoulder
muscles of the dog. Anat Histol Embryol. 1991;20:205–
214.
[27] Oliver JE, Hoerlein BF, Mayhew IG. Brachial plexus
avulsion and compression. Vet Neurol. 1987;366–369.
[28] Carlstedt T, Anand P, Hallin PV, et al. Spinal nerve root
repair and reimplantation of avulsed ventral roots into the
spinal cord after brachial plexus injury. J Neurosurg.
2000;93:336–338.
[29] Carlstedt T, Hallin RG, Hedstr€om KG, et al. Functional
recovery in primates with brachial plexus injury after spi-
nal cord implantation of avulsed ventral roots. J Neurol
Neurosurg Psychiatry. 1993;56:649–654.
[30] Hems TEJ, Clutton RE, Glasby MA. Repair of avulsed
cervical nerve roots-an experimental study in sheep.
J Bone Joint Surg (Br). 1994;76-B:818–823.
[31] Hoffmann CFE, Thomeer RTWM, Marani E. Reimplan-
tation of ventral rootlets into cervical spinal cord after
their avulsion: an anterior surgical approach. Clin Neurol
Neurosurg. 1993;95:112–118.
[32] Gu Y-D, Wu M-M, Zhen Y-L, et al. Phrenic nerve trans-
fer for the treatment of root avulsion of the brachial
plexus. Chin Med J. 1990;103:267–270.
[33] Isla A, Bejarano B, Morales C, et al. Anatomical and
functional connectivity of transected ulnar nerve after
intercostal neurotization in cats. J Neurosurgery. 1999;
90:1057–1063.
[34] Malessy MJA, Thomeer RTWM. Evaluation of intercos-
tals to musculo-cutaneous nerve transfer in reconstructive
brachial plexus surgery. J Neurosurg. 1998;88:266–271.
[35] Tomita Y, Tsai T-M, Burns JT, et al. Intercostal nerve
transfer in brachial plexus injuries; an experimental
study. Microsurgery. 1983;4:95–104.
[36] Allieu Y, Cenac P. Neurotization via the spinal accessory
nerve in complete paralysis due to multiple avulsion inju-
ries of the brachial plexus. Clin Orthop Relat Res. 1988;
237:67–74.
[37] Nikkhah G, Carvalho GA, Samii M. Nerve transfer (neu-
rotization) for functional reconstruction of arm functions
in cervical root avulsion. Orthopade. 1997;26:606–611.
[38] Kawai H, Kawabata H, Masada K, et al. Nerve repairs
for traumatic brachial plexus palsy with root avulsion.
Clin Orthop Relat Res. 1988;237:75–86.
144 | MOISSONNIER
[39] Fisch U. Cross-face grafting in facial paralysis. Arch
Otolaryngol. 1976;102:453–464.
[40] Frykman GK, Mc Millian PJ, Yegges S. A review of
experimental methods measuring peripheral nerve regen-
eration in animals. Orthop Clin North Am. 1988;19:209–
219.
[41] Pellegrino C, Franzini C. An electron microscope study
of denervation atrophy in red and white skeletal muscle
fibers. J Cell Biol. 1963;17:327–349.
[42] Anzil AP, Wenig A. Muscle fibre loss and reinnervation
after long-term denervation. J Neurocytol. 1989;18:833–
845.
[43] Gulatia K. Restoration of denervated skeletal muscle
transplants after reinnervation in rats. Restorative Neurol
Neurosci. 1990;2:23–29.
ET AL.
[44] de Medinaceli L. Functional consequences of experimen-
tal nerve lesions: effects of time, location and extent of
damage. Exp Neurol. 1988;100:154–166.
[45] Huang J, Xiang J, Yan Q, et al. Dog tibial nerve regener-
ation across a 30-mm defect bridged by a PRGD/
PDLLA/b-TCP/NGF sustained-release conduit.
J Reconstr Microsurg. 2013;2:77–87.
[46] Ren Z, Wang Y, Peng J, et al. Role of stem cells in the
regeneration and repair of peripheral nerves. Rev Neuro-
sci. 2012;23:135–143.
[47] Liss AG, Af Ekenstam FW, Wiberg M. Re-organisation
of primary afferent nerve terminals in the brainstem after
peripheral nerve injury. An anatomical study in cats.
Scand J Plast Reconstr Surg Hand Surg. 1995;29:185–
187.