Between 1980 and 1992, 170 patients with neonatal brachial plexus palsy
presented to our institution. With the approval of our institutional review
board, data regarding age, sex, side of injury, obstetric presentation,
assisted extraction, birth weight, associated fractures, and shoulder bruising
were prospectively collected on initial presentation.
The extent of the brachial plexus palsy was assessed with repeated
standardized clinical examinations, which included documentation of the active
range of motion of the upper limb, sensation in the hand, presence of a Horner
syndrome, scapular winging, and contractures at the shoulder or elbow. The
active range of motion of the limb segments was characterized qualitatively as
0 (no function), 1 (no palpable contraction), 2 (less than antigravity
strength), 3 (strength to move the joint against gravity), 4 (good but not
normal strength), and 5 (normal strength). Five types of movement were
evaluated in the shoulder (abduction, forward flexion, internal rotation,
external rotation, and extension), and two each were assessed in the elbow
(extension and flexion), forearm (supination and pronation), wrist (extension
and flexion), and fingers (extension and
flexion)31.
The patients were grouped according to the above findings. Group 1 had a
C5-C6 lesion with weakness of shoulder abduction, elbow flexion, and forearm
supination. Group 2 had a C5-C7 lesion with additional weakness of elbow and
wrist extension. Group 3 had a pan-plexus lesion (C5 through T1) with
additional weakness of finger extension and intrinsic muscle function, with or
without a Horner syndrome. This classification is widely
accepted27 and
similar to that described by
Narakas10,26,
except that we placed children in group 3 without distinguishing between a
pan-plexus lesion with a persistent Horner syndrome and absent Tl function
(Narakas Group 4) and a pan-plexus lesion without such findings (Narakas Group
3).
Of the 170 patients who presented with neonatal brachial plexus palsy,
twenty-nine had been observed by one of the senior authors (P.R.C. or M.E.) to
have absent biceps muscle function at the age of three months or older. Eight
patients who had a history of biceps muscle function returning after three
months of age but who presented at a later age were not included in the study.
Of the twenty-nine patients with documented absence of biceps muscle function
at three months of age, twenty-eight were available for long-term follow-up,
at a mean age of eleven years and one month (range, five years and seven
months to sixteen years).
At the time of follow-up, patients and parents responded to a questionnaire
(Table I). Details of any
brachial plexus or orthopaedic surgery were obtained. Each patient's shoulder
function was evaluated and scored according to the system of
Mallet32, as
modified by
Tassin14 and
reported by Gilbert et
al.18,20.
This system, which was designed specifically to evaluate shoulder function
following neonatal brachial plexus injury, has been shown to be reliable for
measuring active motion of the shoulder in patients with brachial plexus
palsy33. The
original Mallet classification assigns a grade of 1 to 5 in four categories:
global abduction, external rotation, bringing the hand to the mouth, and
putting the hand on the head. A grade of 1 indicates absent function, with
gradations of abnormal up to a normal shoulder, grade 5. The modified Mallet
score has an additional category of internal rotation
(Fig. 1). Subsequently, Gilbert
and Whitaker15
reported grading the function of the shoulder on the basis of a single
parameter, abduction, with class IV being >90°.
Patients also underwent manual muscle strength testing of elbow, wrist, and
finger flexion and extension and of the intrinsic muscles of the
hand31. In addition
to a subjective assessment of sensation with the questionnaire, evidence of
protective sensation in the hand was evaluated.
Six of the patients had had brachial plexus surgery. The remaining
twenty-two represented a long-term cohort of patients with neonatal brachial
plexus palsy who did not have biceps muscle function at three months of age
and had had no prior brachial plexus surgery. The shoulders of these patients
were classified according to the modification of the Mallet scoring system,
and the numbers of class II, III, and IV shoulders were determined for each
level of injury15.
Patients were grouped according to the level of the injury, and the level was
then related to whether biceps muscle contraction had occurred between three
and six months of age or at six months of age or older.
Each of the final modified Mallet scores was related to the level of the
brachial plexus injury; first for all of the patients, second after exclusion
of patients who had had shoulder surgery, third after exclusion of patients
who had had brachial plexus surgery, and fourth after exclusion of patients
who had had brachial plexus and shoulder surgery.
For analysis of the results, the patients were also grouped according to
which month of life a biceps muscle contraction was first demonstrable. These
groups were subdivided according to whether brachial plexus surgery or
shoulder surgery had been performed. The modified Mallet scores were analyzed,
and the groups were compared.
When the groups were compared according to the modified Mallet scores,
differences were detected with use of an unpaired Student t test, with p <
0.05 considered to indicate a significant difference.
Initial Data
Of the twenty-eight patients with absent biceps muscle function at three
months of age, sixteen were boys and twelve were girls. Nineteen lesions were
on the left side, and nine were on the right. There was one bilateral case,
with recovery of biceps muscle function on the right side by six weeks of age
but function still absent on the left side at three months of age. All of the
infants were delivered vaginally, with twenty-five vertex and three breech
presentations. Shoulder dystocia was documented in twelve (43%) of the
twenty-eight infants, and the delivery of five infants (18%) had been
difficult. The mean birth weight of the infants with a vertex delivery was
4495 g (range, 3492 to 5600 g), and the mean birth weight of those with a
breech delivery was 2360 g (range, 2041 to 2676 g). There were three
ipsilateral clavicular fractures and three humeral shaft fractures, two of
which were ipsilateral and one of which was contralateral to the side of the
brachial plexus palsy.
Serial Examination
Neurological examination demonstrated a C5-C6 lesion in thirteen (46%) of
the twenty-eight infants, a C5-C7 lesion in five (18%), and complete plexus
(C5-TI) involvement in ten (36%). Four of the patients with complete brachial
plexus involvement had a Horner syndrome.
Of the twenty-eight patients with absent biceps muscle function at three
months of age, twenty had a biceps muscle contraction by six months of age:
ten had it by four months; five, by five months; and five, by six months.
Patients who regained biceps muscle function by six months of age were
distributed unequally according to the level of injury. All thirteen patients
with a C5-C6 lesion, three of the five patients with a C5-C7 lesion, and four
of the ten patients with a C5-T1 lesion regained biceps function by six months
of age (see Appendix). None of the four patients with a Horner syndrome
regained biceps function by six months of age.
Of the twenty-eight patients with absent biceps muscle function at three
months of age, twenty-seven had at least antigravity biceps muscle function at
the time of final followup, at a mean age of eleven years and one month
(range, five years and seven months to sixteen years). These patients included
seven of the eight with absent biceps muscle function at six months of age
(see Appendix). Biceps function was absent at the time of final follow-up in
only one patient, who had a C5-T1 lesion and Horner syndrome.
Intervening Procedures
Our indication for exploration of the brachial plexus was absence of elbow
flexion at six months of age with a plateau in neurological recovery. Brachial
plexus surgery was discussed and offered to eight infants' families. Six
patients underwent brachial plexus surgery, three at our institution and three
elsewhere. One of these patients subsequently had a release of the
subscapularis origin at the age of thirteen months, followed by an open
reduction of the shoulder with a latissimus dorsi transfer at three years and
six months of age.
Nine of the twenty-two patients who did not have brachial plexus surgery
had subsequent orthopaedic procedures. These included two releases of
abduction, external rotation contracture in the shoulder and four
subscapularis and pectoralis major lengthening procedures for internal
rotation contracture of the shoulder, with latissimus dorsi transfer
concurrently in two and subsequently in another. Two patients with a C5-T1
brachial plexus lesion had tendon transfers to improve wrist or forearm
position, and one had an unsuccessful Steindler procedure to treat persistent
absence of elbow flexion. The effects of the shoulder surgery on abduction and
external rotation are shown in Table
II.
Shoulder Function at the Time of Follow-up
Biceps Muscle Recovery by Four Months of Age
Seven patients who had had biceps muscle function at four months of age and
no subsequent shoulder surgery had a mean modified Mallet grade of 3.6 for
global abduction, 3.6 for external rotation, 3.9 for hand-to-neck function,
3.7 for hand-to-mouth function, and 2.7 for hand-to-back function. Two
patients with biceps muscle recovery at four months of age who had shoulder
surgery had a mean grade of 4.0 for global abduction, 3.5 for external
rotation, 4.0 for hand-to-neck function, 3.0 for hand-to-mouth function, and
3.5 for hand-to-back function. No patient with biceps muscle recovery by four
months of age had brachial plexus surgery.
Biceps Muscle Recovery by Five Months of Age
One patient recovered biceps muscle function at five months of age and had
no subsequent shoulder surgery. The modified Mallet scores were 3 for
abduction, 2 for external rotation, 2 for hand-to-neck function, 1 for
hand-to-mouth function, and 2 for hand-to-back function. Two patients
recovered biceps muscle function at five months of age and had subsequent
shoulder surgery. Their mean scores were 4.0 for global abduction, 2.5 for
external rotation, 4.0 for hand-to-neck function, 4.0 for hand-to-mouth
function, and 4.0 for hand-to-back function. Two patients recovered biceps
muscle function at five months of age and underwent brachial plexus surgery at
another institution. One patient underwent neurolysis alone, and the other
patient had nerve grafts. The average scores for those patients were 3.0 for
abduction, 3.5 for external rotation, 3.5 for hand-to-neck function, 3.5 for
hand-to-mouth function, and 2.0 for hand-to-back function.
Biceps Muscle Recovery by Six Months of Age
Five patients regained biceps muscle function at six months of age; four
had no subsequent shoulder surgery and one was lost to follow-up. Their mean
modified Mallet scores were 3.25 for abduction, 3.5 for external rotation,
3.75 for hand-to-neck function, 3.25 for hand-to-mouth function, and 3.0 for
hand-to-back function. One patient regained biceps muscle function at six
months of age and had shoulder surgery. The follow-up scores were 3 for
abduction, 4 for external rotation, 4 for hand-to-neck function, 4 for
hand-to-mouth function, and 2 for hand-to-back function. One patient who
regained biceps muscle function at six months of age underwent plexus
exploration alone. The follow-up Mallet scores were 4 for abduction and 2 for
all of the other categories.
Biceps Muscle Recovery After Six Months of Age
Four patients regained biceps muscle function after six months of age and
had no subsequent surgery. Their mean scores were 3.0 for abduction, 2.75 for
external rotation, 2.75 for hand-to-neck function, 2.75 for hand-to-mouth
function, and 2.0 for hand-to-back function. One patient who regained biceps
muscle function after six months of age had only shoulder surgery. The scores
were 4.0 for abduction, 2.0 for external rotation, 3.0 for hand-to-neck
function, 2.0 for hand-to-mouth function, and 2.0 for hand-to-back function.
Two patients who regained biceps muscle function after six months of age had
brachial plexus surgery (one neurolysis and one graft procedure). Their mean
modified Mallet scores were 3.5 for abduction, 2.5 for external rotation, 3.5
for hand-to-neck function, 3.0 for hand-to-mouth function, and 2.0 for
hand-to-back function. One patient regained biceps muscle function after six
months and had subsequent neurolysis of the brachial plexus followed by two
separate surgical procedures on the shoulder. The follow-up Mallet scores were
2 for abduction, 4 for external rotation, 3 for hand-to-neck function, 4 for
hand-to-mouth function, and 2 for hand-to-back function.
Comparison of Three-to-Six-Month Group with Group Older Than Six
Months
When all patients who regained biceps function between three and six months
of age were compared with all patients who regained biceps function after six
months of age, the three-to-six-month group had significantly better scores
for global abduction (p = 0.04), hand-to-neck function (p = 0.05), and
hand-to-back function (p < 0.001). The same two groups were also compared
after exclusion of (1) patients who had brachial plexus surgery, (2) patients
who had shoulder surgery, and (3) those who had both brachial plexus and
shoulder surgery. The only significant differences detected, with the numbers
available, were a decreased score for hand-to-back function for the patients
who were six months of age or older (p < 0.01), with use of all three
exclusion criteria, and decreased hand-to-neck function (p = 0.03) after
exclusion of only the patients with brachial plexus surgery.
Level of Plexus Palsy Compared with Modified Mallet Scores
With the numbers available, no significant difference in any of the
modified Mallet scores could be detected between the group with a C5-C6 injury
and the group with a C5-C7 injury. However, compared with the group with a
C5-T1 plexus injury, the group with a C5-C6 lesion had significantly better
scores for external rotation (p = 0.04), hand-to-neck function (p = 0.05),
hand-to-mouth function (p < 0.01), and hand-to-back function (p <
0.001). With the numbers available, no significant difference was found for
global abduction.
The shoulder class based only on shoulder abduction (IV indicating
>90°, III indicating 30° to 90°, and II indicating
<30°), as defined by Gilbert and
Whitaker15, was
evaluated in the group of patients who had not regained biceps muscle function
by three months of age and had not had brachial plexus surgery. Of the twelve
patients with C5-C6 plexus palsy, eight had class-IV shoulder abduction, three
had class-III, and one had class-II. Of the four patients with C5-C7 brachial
plexus palsy, two had class-IV abduction and two had class-III abduction. Of
the six patients with C5-T1 plexus palsy, two had class-IV abduction, three
had class-III, and one had class-II.
Extent of Injury and Time of Biceps Recovery
The mean age at which biceps muscle function spontaneously returned in the
twelve patients with a C5-C6 injury was 4.8 months, (range, four to six
months), with no patients older than six months of age when it returned. The
mean age at which biceps muscle function spontaneously returned in the five
patients with a C5-C7 injury was 5.2 months (range, four to seven months),
with two patients older than six months of age. The patients with C5-T1 plexus
injury were a mean of 8.1 months of age (range, five to thirteen months) when
they regained biceps muscle function, with six patients older than six months
of age, including one who never regained elbow flexion.
With the numbers available, the C5-C6 group did not differ significantly
from the C5-C7 group with regard to the time of return of biceps muscle
function, whereas the C5-T1 plexus group was significantly different from both
the C5-C6 group (p = 0.01) and the C5-C7 group (p = 0.05).
Distal Function in the C5-T1 Group
The findings of the follow-up clinical examinations of patients with a
C5-T1 lesion are shown in Table
III. All but two patients had useful hand function, and one of the
two patients who did not had had brachial plexus surgery. Wrist extension was
graded as 0 for all of the patients who had had exploration of the brachial
plexus. Of the three patients who had not had brachial plexus surgery and who
had less than grade-3 wrist extension, two had tendon transfers resulting in
grade-2 or 3 extension. Biceps muscle function was grade 0 in one patient who
had not had plexus surgery, and triceps muscle function was grade 0 in one
patient who had had brachial plexus surgery.