OBSTETRIC TRACTION INJURIES OF THE BRACHIAL PLEXUS

NATURAL HISTORY, INDICATIONS FOR SURGICAL REPAIR AND RESULTS

R. S. BOOME, J. C. KAYE

From Groote Schuur Hospital, Cape Town

Of a consecutive series of 70 babies with obstefric fraction injury to the brachial plexus we were able to
review 40 who had had only conservative treatment. We found that if recovery in the upper roots had not
started by three months of age then, at follow-up there was a significant residual functional deficit.
In a further 22 babies who showed no recovery of the upper roots by three months, exploration and sural
nerve grafting were performed; 20 of these children were reviewed at one year. Good recovery of the deltoid
was seen in 80% and of the biceps in 55%, while 25% had good external rotation at the shoulder. We conclude
that if there is disruption of the upper roots with no sign of recovery at three months, grafting of these roots
provides the best chance of useful recovery.

Obstetric palsy of the upper limb is caused by a traction already be impaired, is further compromised by the
injury to the brachial plexus during birth (Taylor 1907; proximal weakness.
Sever 1925). The first description was by Smellie in 1764 The first attempt to repair extra-foraminal ruptures
and Duchenne in 1872 implicated traction on the arm as ofthe CS and C6 roots was made by Kennedy in 1903. He
the cause. In 1874 Erb described a similar palsy in adults reported three cases with one excellent result, but had no
and suggested that traction or compression of the CS and follow-up in the other two. Taylor (1920) reported 70
C6 roots could produce the injury (Rang 1966). The explorations, and obtained good improvement in many,
prognosis of such injuries has been reviewed by Bennet especially the very young, but had three postoperative
and Harrold (1976) and Hardy (1981), who reported deaths. Direct repair of the upper roots remained
complete recovery in 94% and 84% of patients respec- controversial (Sever 1925 ; Jepson 1930 ; Wickstrorn et a!.
tively. Earlier reports indicated complete recovery in as 1955) and eventually ceased to be standard practice.
few as 13% (Wickstrom, Haslan and Hutchinson 1955) More recently, modern microsurgical techniques
but, since recovery varies with the severity of injury, this and safe anaesthesia have renewed interest in the repair
discrepancy may be due to recent improvement in of extra-foraminal root injuries, especially of the upper
obstetric practice. roots. Solonen, Telaranta and Ryoppy (1981) reported
In half the patients, only the upper roots of the three cases in which the upper trunks were repaired at
brachial plexus are affected, but the C7 root also is three months of age with good recovery. Gilbert, Khouri
involved in a further quarter, with injury to the whole and Carlioz (1980), in a larger series with a wider age
plexus in the remaining quarter. The Klumpke lower range, had the best results in patients under six months of
root injury is very rare (Tassin 1983). age. If children with extra-foraminal injuries with little
Permanent damage to the upper roots is uncommon, chance of recovery could be identified early, these would
but causes a deformity and a severe functional deficit. clearly merit exploration and grafting of the upper roots.
Classically there is a loss of abduction and external This paper examines the natural history, the
rotation of the shoulder, of flexion at the elbow and indications for, and the results following repair of the
supination of the forearm. Hand function, which may brachial plexus in babies after traction injuries at birth.

PATIENTS
R. S. Boome, FRCS, FCS(Ortho)SA, Consultant Orthopaedic Surgeon
City Park Medical Centre, 181 Longmarket Street, Cape Town 8001,
South Africa. All the patients had sustained birth injuries of the
J. C. Kaye, FRCS, Orthopaedic Registrar brachial plexus. The Groote Schuur Hospital brachial
Princess Margaret Rose Hospital, Fairmilehead, Edinburgh EH1O
plexus clinic is a secondary referral centre, so that
7ED, Scotland.
transient palsies were not normally seen. Between 1981
Correspondence should be sent to Mr R. S. Boome.
and 1985, 70 children under six months old were seen
1988 British Editorial Society of Bone and Joint Surgery
0301-620X/88/41 12 $2.00 with 72 brachial plexus palsies. Eight children were
J Bone Joint Surg [Br] 1988;70-B:571-6.
excluded as having inadequate records. The remaining

VOL. 70-B, No. 4, AUGUST 1988 71

572 R. S. BOOME, J. C. KAYE

64 palsies in 62 babies were considered in two groups:

Non-operated group. Spontaneous recovery was seen of
42 palsies in 40 children. These were followed either to Suprascapular nerve

full recovery or until the maximum expected recovery.

Average follow-up was 17.9 months (range, one month to
five years). In those not attaining full recovery the
Musculocutaneous nerve
minimum follow-up was one year.
Operated group. In all, 22 infants had an exploration of
the brachial plexus. This was considered at three months
if there was no evidence of recovery in the muscles
Radial
innervated by CS or C6. Ti

Median nerve Ulnar nerve

METHODS Fig. 1

Clinical assessment. All children were examined by one of Diagram of the brachial plexus. Neurolysis (indicated by black roots)
the authors (JCK). The strength of individual muscles or
was used alone in two patients.
muscle groups was recorded at each clinic visit, and
mothers were instructed in stretching exercises for the
indicate a more proximal injury and are not suitable for
shoulder, performed at every nappy change, to prevent
grafting. Where a neuroma is present, it is excised.
joint contractures.
The gap between the nerve ends is grafted, using the
CS root was considered to innervate the deltoid
sural nerve (Fig. 2). The mean length ofgraft was 2.5 cm,
muscle and the external rotators of the shoulder, C6 to
and on average four grafts are placed on each root stump.
innervate the biceps, C7 the extensors of the wrist and
The suprascapular nerve is always grafted separately to
the fingers and Tl the intrinsic muscles of the hand.
the CS root. In two cases nerve roots were in continuity
Muscle strength was assessed by observation during play
but severely constricted by dural fibrosis at the foramen;
and by clinical examination. Muscle power was graded
only neurolysis was performed. The lower roots were
according to the MRC scale (Medical Research Council
explored when this was indicated by the pre-operative
1943). Accurate grading is difficult in infants and young
findings.
children, so this was simplified. MRC Grades 1 and 2
Postoperatively, the child is nursed for two weeks in
were considered together, and scored as Grade 1/2 when
a plaster cowl to prevent distraction between the head
there was definite muscle contraction but no useful
and shoulder. The arm is strapped to the body for a
movement could be made against gravity. In Grade 3
further four weeks, after which time the childs mother is
there was some movement against gravity but with
encouraged to resume shoulder exercises to prevent joint
limited range and not sustained. Grades 4 and S were also
contracture.
considered together, including those showing some
weakness after repetitive movements with slowness on
reflex testing, and those which were clinically normal. RESULTS
External rotation of the shoulder is essential for full
Non-operated group. The age at presentation of this
shoulder abduction and is difficult to test in an infant.
group was between two weeks and three months (mean,
Weakness was diagnosed when placing the hand to the
seven weeks). There were 26 boys and 14 girls, with 24
mouth necessitated shoulder abduction to 90#{176}
to compen-
right, 14 left and two bilateral palsies. There were 21
sate for the weak rotators ; power Grade 3 was recorded.
palsies involving CS and C6 (50%), 16 involving CS, C6
Inability to put the hand to the mouth despite good
and C7 (38%) and five involving all five roots (12%).
power in deltoid and biceps was recorded as Grade 2 or
In the CS and C6 palsies the age at which muscle
less for external rotation.
contraction was first seen to start was related to the
eventual recovery (Table I). If muscle contraction in the
deltoid or biceps had started by three months of age, then
TECHNIQUE OF OPERATION
a Grade 4/S recovery ensued. Of six cases with a four-
The brachial plexus is approached above the clavicle month delay, five had residual weakness at the final
through a horizontal incision. If a wider exposure is review. The other infant had been admitted for
required, the clavicle is divided. Once exposed, damaged exploration at four months of age, but deltoid activity
roots are followed proximally and neurolysis performed was noted on admission and operation was deferred;
as far into the foramen as possible (Fig. 1). The viability there was a full recovery. External rotation was weak or
of nerve roots is assessed ; those that are thin and absent in nine of the 21 children (42%); this required an
attenuated or soft and oedematous, with minimal distal external rotation transfer to restore good function to the
neuroma, suggesting empty fascicles, are considered to shoulder.

THE JOURNAL OF BONE AND JOINT SURGERY

 OBSTETRIC TRACTION INJURIES OF THE BRACHIAL PLEXUS 573

Suprascapular
C5

:c6

C7

. C8

Ti

Fig. 2a - Ten cases Fig. 2b - Four cases Fig. 2c - Three cases

C5

C6

C7
Patterns of placement of sural nerve grafts
in 20 patients.
C8

. Ti

Fig. 2d - Two cases Fig. 2e - One case

In those with involvement of the lower roots, all was clinically evident only in the CS and C6 roots in two
except one recovered full function. In that child, power cases, in 1 3 C7 was also involved and seven had
was Grade 3 in the C7 and C8 innervated muscles. involvement of the whole plexus. Immediately before
Operated group. Twenty-two infants had exploration of operation none had clinical evidence of muscle contrac-
the brachial plexus, all operations being performed by tion in either the deltoid or biceps. The average age at
the same surgeon (RSB). At first presentation the palsy exploration was 5.3 months (range three to 20 months).
The operative findings and procedures are shown in
Table I. Non-operated group - the age in months at which recovery Figures 1 and 2 and in Table II. The only complication
first started in CS and C6 innervated muscles related to muscle strength was one case of wound infection. Two patients (RC, SC)
at final review
had a constriction within the foramen from dural
Number of scarring ; they had only neurolysis of the roots. In one of
patients Muscle strength at review these there was biceps recovery on the first day after
Age at first
sign of recovery Recovery operation.
(months) Total as shown External rotators Deltoid Biceps Two patients (CQ, MN) were lost to follow-up,
1 8 7 4/5 4/5 4/5 leaving 20 patients with an average follow-up of 1 .8 years
1 2/3 4/5 4/5 (range, nine months to four years). Of these, external
2 10 5 4/5 4/5 4/5 rotation had recovered to Grade 4/S in five cases (25%)
5 2/3 4/5 4/5 and the remainder were Grade 3 or less and required
3 18 11 4/5 4/5 4/5 transfer of latissimus dorsi and teres major to attain
7 2/3 4/5 4/5 good shoulder function.
4 6 1 4/5 4/5 4/5 In 16 patients (80%) there was Grade 4/S recovery in
1 0 2 3 the deltoid. In one patient (YE) with no deltoid recovery,
I 3 3 5
1 2 3 0 the CS root had been grafted to the lateral cord to supply
1 0 4/5 3 biceps. Unfortunately, the CS root was attenuated ; this
1 3 3 3
was the only case without recovery in any grafted root.

VOL. 70-B, No. 4, AUGUST 1988

574 R. 5. BOOME, J. C. KAYE

Table II. Operated group - details of findings and muscle strengths at final review in 22 patients

Muscle strength at review

Extensors of
Patient External wrists and
(see text) Injury Comment rotators Deltoid Biceps fingers

RC C56 Neurolysis 4/5 4/5 4/5 (5)5

Fibrosis only

SC C567 Neurolysis 0 4/5 4/5 2

Fibrosis only

NR C56 - 4/5 4/5 4/5 (5)

EFRt

RJ C56 Cl 3 4/5 4/5 5

EFR Neurolysis

BLR C56 Cl 0 4/5 4/5 5

EFR Neurolysis

SG C56 - 4/5 4/5 4/5 (5)

EFR

SA C567 EFR - 0 4/5 4/5 0

C8 avulsion

IS C56 - 3 4/5 4/5 (5)

EFR

ES C567 C8, Tl 3 4/5 4/5 3

EFR Fibrosis

AVN C56 C6 root 4/5 4/5 3 (5)

EFR attenuated

CC C56 Cl 4/5 4/5 0 5

EFR Neurolysis

OF C56EFR - 3 4/5 1/2 3

CT avulsion

TB C56 C7 0 4/5 3 (4)

EFR Fibrosis

JK C56 Cl 3 4/5 3 (5)

EFR Fibrosis

LM C56 C7 0 3 4/5 (3)

EFR Fibrosis

JL C567 - 0 3 3 3
EFR

YE C5 EFR CS root 0 0 0 (5)

C6 avulsion attenuated

CQ CS EFR Wound No follow-up

C6 avulsion infection

MN C567 C7 root No follow-up

EFR attenuated

NM C56EFR C8,Tl 3 3 0 0
Cl avulsion extensive
fibrosis

NT C5 EFR Cl 0 4/5 3 W3, F5

C6 fibrosis

VB C5EFR - 3 4/5 4/5 (5)

C6 avascular

a Parentheses indicate that recovery had occurred before operation

t EFR, extra-foraminal rupture

THE JOURNAL OF BONE AND JOINT SURGERY

 OBSTETRIC TRACTION INJURIES OF THE BRACHIAL PLEXUS 575

Biceps recovery was Grade 4/S in 1 1 patients (55%) and and cervical myelography to help define which palsies
Grade 1/2 or less in four (20%). In the latter four patients, are surgically correctable. We have not used EMG
the viability of the grafted root had been doubtful in one studies ; myelography in nine of our patients correlated
(YE); in two patients (NM, OF) the C6 root was grafted poorly with the operative findings and added no useful
to provide wrist and finger extension as well as biceps, information regarding selection or operative decisions.
and the fourth patient (CC) had a standard graft. We no longer use it.
Grafts of C7 to provide wrist and finger extension In Gilberts patients (Tassin 1983) all injuries to the
were placed in five patients, of whom only three (JL, OF, CS and C6 roots were extra-foraminal, but a functionless
ES) gained Grade 3 power after slow recovery in the C7 root had always been avulsed. Our findings were that
wrist and finger extensors. In four patients where C7 although the CS root was always damaged extra-
function was weak or absent pre-operatively and the root foraminally, the C6 root had been avulsed in four cases
was found to be intact but fibrosed, neurolysis was and the C7 root had been ruptured extra-foraminally in
performed. In three of these patients (BLR, CC, RJ) three. Furthermore, in some cases with an extra-fora-
Grade 4/S power eventually returned. minal rupture the root was thin and attenuated or soft
In two cases the C8 and Tl roots were explored. In and oedematous suggesting that there had also been
one patient (SA), C8 root had been avulsed ; the medial proximal damage. We would not now graft such a root;
cord was grafted to the C7 root with no significant we believe grafting to be valueless.
recovery. In the other, there was intense fibrosis and root In every case in which the posterior cord was grafted
viability was in doubt, so no action was taken. This to provide deltoid function, power of Grade 3 or better
patient (NM) had no recovery of wrist or finger flexor was achieved ; 16 patients (80%) had effectively normal
function and was the only patient to develop trophic deltoid strength.
ulcers on the knuckles. In 10 of the 20 operated patients The suprascapular nerve was always grafted separ-
there was full, or near full, functional recovery of both ately, but although the CS root had been viable, as shown
the deltoid and biceps muscles (Table II). by deltoid recovery, only 20% had normal recovery of
external rotation. In the non-operated group external
rotation was also the most common residual weakness.
DISCUSSION This has been noted throughout the literature. Taylor
(1907) emphasised this and suggested the presence of a
In the non-operated group, 50% of the palsies were proximal lesion. A distal lesion at the level of the
confined to the CS and C6 roots. There was a high ratio#{149} suprascapular notch is also possible, but in one case we
of C7 injuries compared to C8 and Tl injuries, which in have explored the suprascapular nerve as far as the
Tassins (1983) series were equally distributed. Spon- suprascapular notch, but found no lesion. However, an
taneous recovery of the lower roots may account for the external rotator muscle transfer, in our experience, is
difference. By contrast, in the operated group only 7% of very rewarding and simple to perform ; we consider it as a
injuries were confined to the CS and C6 roots ; this second-stage procedure after the nerve repair.
reflects a greater severity of injury. Biceps did not recover as successfully as deltoid;
Recent reports have demonstrated that the overall only 55% became functionally normal. In the four
prognosis for recovery is good, but early identification of patients with no useful biceps recovery, the proximal
the infant with a bad prognosis has not been discussed root stump had also been used for grafts for either
(Bennet and Harrold 1976; Hardy 1981). Tassin (1983) shoulder or wrist function ; this may have affected the re-
reviewed Gilberts patients and showed that failure of innervation. Recovery of C7 function was disappointing,
recovery in the deltoid and biceps within two months of and this parallels the experience of other authors. In
age meant that rupture of the CS and C6 roots was three patients with some wrist recovery, that recovery
probable, and no useful recovery could be expected. In was much slower in returning than the more proximal
our series all babies that had started to recover some muscles. Only one child finally showed dominance on the
function in either the deltoid or biceps before three side operated on, which reflects the fact that true
months of age gained full recovery, except in the external normality is unlikely after a grafting procedure. How-
rotator muscles. When there was four months delay ever, at the early age of our follow-up, continued
before recovery started, all but one child had significant improvement in strength and co-ordination can be
residual weakness in deltoid or biceps. If the delay was expected.
greater, as in some of those undergoing exploration, Conclusions. Most babies who suffer traction injuries of
recovery did not start. Furthermore, the surgical find- the brachial plexus at birth have an excellent prognosis if
ings, in our opinion, would have precluded useful recovery has started within three months. If recovery in
recovery. On this basis, operation should be considered the muscles innervated by the upper roots is delayed
from three months of age if there is no evidence of beyond three months, then root disruption is likely.
recovery in either the deltoid or biceps. Exploration and nerve grafting then offers the best
Other authors have used electromyography (EMG) prospect of a useful arm.

VOL. 70-B, No. 4, AUGUST 1988

576 R. S. BOOME, J. C. KAYE

REFERENCES

Rang M. Anthology oforthopaedics. Edinburgh, etc : E & S Livingstone

Bonnet GC, HarroW AJ. Prognosis and early management of birth
injuries to the brachial plexus. Br Med J 1976;! :1520-1. Ltd. 1966.
Gilbert A, KbOUrI N, Carlioz H. Exploration chirurgicale du plexus Sever JW. Obstetric paralysis. JAMA 1925 ;85 : I 862-5.
brachial dans Ia paralysie obst#{233}tricale : constatations anatomiques
chex 21 malades op#{233}r#{233}s.
Rev Chir Orthop l980;66:33-42.
Solonen KA, Telaranta T, Roppy S. Early reconstruction of birth
injuries ofthe brachial plexus. J Paediatr Orthop 1981 ;(4):367-70.
Hardy AE. Birth injuries of the brachial plexus: incidence and
Tassin JL. Paralysies obstetricales du plexus brachial : evolution
prognosis. J Bone Joint Surg [Br] 1981 ;63-B:98-lOl.
spontanee, resultats des interventions reparatrices precoces.
Jepson PN. Obstetrical paralysis. Ann Surg l930;9l :724-30. Thesis. Paris, Universite Paris VII, 1983.
Kennedy R. Suture of the brachial plexus in birth paralysis of the upper Taylor AS. Results from the surgical treatment of brachial birth palsy.
extremity. Br Med J l903;I:298-30l. JAMA l907;48:96-l04.
Leffert RD. Brachial plexus injuries. New York, etc : Churchill Taylor AS. Brachial birth palsy and injuries of a similar type in adults.
Livingstone, 1985. Surg Gynecol Obstet 1920;30:495-502.
Medical Research Council. Aids to the investigation ofperipheral nerve Wickstrom J, Haslam ET, Hutchinson RH. The surgical management
injuries. War Memorandum No. 7 revised 2nd ed. London : HMSO, of residual deformities of the shoulder following birth injuries of
1943. the brachial plexus. J Bone Joint Surg [Am] l955;37-A:27-36.

THE JOURNAL OF BONE AND JOINT SURGERY