|Home | About | Journals | Submit | Contact Us | Français|
This article presents our institutional experience with obstetrical brachial plexus injuries including our concepts and approaches and the outcomes of primary surgery and secondary reconstruction.
The first multidisciplinary birth palsy clinic in Saudi Arabia was established 9 years ago and the clinic has assessed over 500 cases of obstetrical branchial plexus injuries.1 The team is led by the reconstructive surgeon (a plastic surgeon) and includes a neonatologist, an obstetrician, an orthopedic surgeon, an occupational therapist, and a physiotherapist.
A data sheet is completed during the first visit to document the following:
Three clinical types of birth palsy are seen: Erb's palsy of the C5 and C6 type (paralysis of the shoulder and elbow only); Erb's palsy of the C5, C6, and C7 type (wrist drop is present in addition to the paralysis of the shoulder and elbow); and total palsy (injury to the C5, C6, C7, C8, and T1 roots resulting in flaccid paralysis of the whole upper limb, including the hand). We have never seen a case of Klumpke's palsy (isolated C8-T1 root injury) or intermediate palsy (mainly involving the C7 root), and we believe that these two types of birth palsy are only of historic interest and are no longer seen with modern obstetric practice.2,3,4,5
Any clinical type of birth palsy can be right sided, left sided, or bilateral. All bilateral cases in our center were seen in newborns delivered by vaginal breech delivery.6 We have assessed 28 cases of birth palsy associated with breech delivery. Six of these 28 cases (21%) were bilateral.
Documentation of risk factors for birth palsy is also done, such as multiparity, maternal height, gestational diabetes, and birth weight. The definition of maternal short stature as a risk factor for birth palsy should vary according to race. We studied 5000 pregnant Saudi women and compared them with North American women. On average, Saudi women were shorter and gave birth to smaller babies.7 Fetal macrosomia (increased birth weight) is a well-known risk factor for birth palsy. However, it is important to note that the mean birth weight in newborns with birth palsy following breach delivery is low (only 2.3 kg in our center).6 Finally, our center was the first to document clearly that multiparous mothers who previously had delivered large infants with branchial plexus injury are at extremely high risk for giving birth to babies with birth palsy in subsequent vaginal deliveries.8,9
Documentation of the method of delivery (vaginal cephalic, vaginal breech, or caesarean section) is done as well as documentation of the use of instrumentation (forceps or vacuum) during delivery. It is important to note that although delivery by caesarian section has a protective role, it is not a guarantee against the occurrence of birth palsy. In our center, 1% of birth palsy cases were delivered by caesarian section.10 Finally, documentation of any difficulties with the delivery such as shoulder dystocia (in cephalic cases) and difficulty of delivery of the “after-coming” head (in breech cases) is important for medicolegal reasons. It is also important to know that birth palsy can be of intrauterine origin and this can be confirmed by electromyography. It takes about 7 to 10 days for muscle to develop electromyographic evidence of denervation. Therefore, such findings in the first few days of life support the intrauterine onset of palsy.
Asphyxia, fractures, shoulder subluxation, and torticollis are commonly seen in association with birth palsy. Phrenic nerve palsy may be seen in both Erb's and total palsies, and Horner's syndrome (ptosis, myosis, enophthalmos, and anhidrosis on the ipsilateral side of the face) may be associated with total palsy.
We do not perform any objective sensory assessment in infants or children with birth palsy. However, sequelae of lack of sensation of the hand (such as burns and self-mutilation, usually by biting the fingers) should be documented and treated.11,12 We use the Toronto Muscle Grading System13 (Table 1) to assess the motor power of the limb in infants and very young children because it is impossible to test power against resistance or to ask such children to perform a specific task. In older children, we use our own muscle grading system14 as follows:
Most centers use the Mallet grading system, which is a measure of both shoulder abduction and external rotation. We tried to apply this grading system but found that several patients with Erb's palsy did not fit into the Mallet scale, mainly because they had poor or no external rotation although they had good shoulder abduction. We, therefore, use a modified Mallet scale for assessing shoulder external/internal rotation and use a modified Medical Research Council System for assessing other shoulder movements (Table 2). Active shoulder abduction (if present) is also assessed in degrees of motion.
We use the modified Medical Research Council Score (similar to the scale for shoulder abduction, see Table Table2)2) to assess elbow flexion/extension and wrist flexion/extension. Two special functional scores were developed to assess forearm and hand functions in our center (Table 3).
We do not routinely perform electromyography and nerve conduction studies in our infants with birth palsy because of the difficulty in the interpretation of results. We believe that the only way to make neurophysiological testing useful is to have an interested pediatric neurologist who slowly acquires experience from the analysis. An excellent example is the useful analysis of Smith from the United Kingdom.15 Similarly, we believe that computed tomographic myelography is too invasive and magnetic resonance imaging (MRI) requires general anesthesia in these babies. Furthermore, we do not depend on the radiological findings for the decision regarding primary brachial plexus exploration.
We believe that the two most important prognostic factors for spontaneous recovery are the type of palsy (Erb's palsy has a better chance for spontaneous recovery than total palsy) and the spontaneous recovery of biceps by 4 months of age.16 Another important prognostic factor for total palsy is the presence of persistent Horner's sign. We studied this factor in 22 infants and none of them had a satisfactory spontaneous recovery.17 Our studies also showed that the presence of concurrent phrenic nerve palsy or clavicular fractures in newborns with birth palsy had no prognostic value for spontaneous recovery.18,19
The most controversial issue in Erb's palsy is the indication for primary brachial plexus exploration. We have previously reviewed the literature20 and classified the various approaches into three main groups:
Gilbert et al21 relied on the spontaneous recovery of the biceps as the indication for surgery. If the recovery of the biceps had not begun at 3 months of age, the functional prognosis was considered poor and surgical repair of the plexus was warranted. In our center, we use the lack of active elbow flexion against gravity at 4 months of age as the indication for surgery.16
An example is the Toronto group, who relied on a limb motion score calculated from multiple joint motions as the indication for primary surgery.13
An example is the study by Nehme et al,22 who concluded that primary branchial plexus reconstruction is justified when there was initial C7 involvement associated with increased birth weight and poor elbow flexion at 6 to 9 months of age.
A standard supraclavicular incision is made in Erb's palsy cases and is extended along the deltopectoral area in total palsy cases. I frequently perform clavicular osteotomy in total palsy cases to improve exposure. In our center, the second step (after skin incision) in the exploration procedure is the identification of the phrenic nerve.23 The transverse supraclavicular incision transects the supraclavicular nerves. The technique is based on the fact that both the phrenic and supraclavicular nerves take their origin from the C4 root. Therefore, retrograde dissection of the supraclavicular nerve will end at the C4 root and hence identify the phrenic nerve. The omohyoid muscle is then cut and the transverse cervical vessels are cauterized. Finally, the nerve roots are identified and stimulated.
Surgical findings may include a variable combination of the following:
There is a small bulge in the root, with a distinct fascicular architecture and strong muscle contraction on electrical stimulation. This lesion is treated by neurolysis only.
Intraoperative electric stimulation proximal to the neuroma shows weak muscle stimulation and there is lack of a distinct fascicular architecture across the neuroma. This should be treated by resection of neuroma and nerve grafting because neurolysis alone does not provide useful functional recovery.24
There is no fascicular continuity across the scarred area and there is no response on electric stimulation. Reconstruction with nerve grafts is done using intraplexus donor nerves (i.e., using the same root).
There is loss of contact with the foramen or there is an exteriorized ganglion. Reconstruction with nerve grafts is done using other ruptured roots or using extraplexus motor donors.
There is a soft pale root that can be traced up to the foramen, without presentation of the ganglion, and there is no response on electrical stimulation. This is a unique common finding in breech cases and we approach this lesion either with a “wait and see” attitude (hoping for spontaneous recovery) or by end-to-side/side-to-side nerve grafts. Using this approach, we have obtained good functional recovery of the motor function of the root in 85% of breech cases.6
Several other technical points during the reconstruction procedure are worth mentioning. The accessory nerve is almost always utilized to neurotize the suprascapular nerve. Several options exist regarding the extraplexus neurotization of the musculocutaneous nerve, such as the intercostals, the phrenic nerve, the medial pectoral nerve, and the ulnar nerve. The intercostal nerve transfer is contraindicated in cases with phrenic nerve palsy and complete transection of the phrenic nerve (for nerve transfer) is contraindicated in infants because postoperative respiratory problems and recurrent pneumonias are frequent consequences. Instead, we occasionally do a partial neurotomy in the phrenic nerve and suture a nerve graft to the neurotomy site in an end-to-side fashion.6 In isolated C5-C6 avulsion or in patients with Erb's palsy who present late with no spontaneous recovery of the elbow, we prefer direct neurotization of the biceps nerve (in the upper arm) utilizing a fascicle of the ulnar nerve6,25 (Fig. 1). This transfer has several advantages including simplicity, good size match between the transferred fascicle of the ulnar nerve and the biceps nerve, lack of notable donor nerve deficit, no need for nerve grafting, and the short distance between the neurorrhaphy site and the motor end plate.
In difficult cases with limited motor donors, we started doing end-to-side and side-to-side neurorrhaphies.6 We still do not have enough numbers followed for a long enough time to give final conclusions regarding this motor neurotization technique, although we have documented its success both experimentally and clinically for sensory nerve reconstruction.26,27,28 It should be noted that, in clinical cases, end-to-side neurorrhaphies are done with a perineurial window for sensory nerves (such as the superficial radial nerve) and with partial neurotomies for the reconstruction of motor function (such as the roots of the brachial plexus).
Neurorrhaphy is performed using sutures or fibrin glue. After surgery, the limb is immobilized in a custom-made brace or a sling. Physiotherapy is started 3 weeks after surgery. The results of surgery largely depend on the severity of the lesion and the number of avulsed roots. We have reported our results in breech cases6 and we are currently studying our results in cephalic cases. Generally, very good to excellent results are obtained in the shoulder and elbow following repairs in Erb's palsy (Fig. 2). However, the result of primary plexus reconstruction in total palsy depends on the number of roots available as donor nerves. In cases of multiple avulsions, we favor the hand and elbow in the distribution of nerve grafts.
We29 modified Zancolli's classification30 of secondary shoulder deformities and included Birch's classification31 of joint abnormalities following internal rotation contracture of the shoulder (Tables 4 and and5).5). Global paralysis is not associated with joint deformity, is almost always associated with breech deliveries (usually C5, C6 avulsion), and is best treated by trapezius muscle transfer.6 Pure paralysis of external rotators is not uncommon and is treated with transfer of latissimus dorsi to the rotator cuff. Pure paralysis of internal rotators, external rotation contractures, and pure abduction contractures are extremely rare shoulder deformities. The latter two contracture deformities were commonly seen over 20 years ago as iatrogenic contractures secondary to prolonged splinting of the arm in abduction and external rotation.30 These iatrogenic deformities are no longer seen with modern physiotherapy and rehabilitation of birth palsy.
The most frequent and most important secondary deformity of the shoulder in birth palsy is the internal rotation contracture. These children should undergo a detailed clinical and radiological examination of the shoulder including an MRI to assess the sphericity of the humeral head (the humeral head is predominantly in young children and is best visualized with MRI, and this is illustrated in Fig. Fig.3).3). Every birth palsy center has its own approach for the correction of internal rotation contracture deformity. In our center, children with a round (spherical) humeral head with no joint deformity undergo release of the subscapularis and/or anterior release of the shoulder joint. Later, latissimus muscle transfer is done for reconstruction of active external rotation.32 If the humeral head is not round or there is joint subluxation, external rotation osteotomy of the humerus is done33 (Fig. 4). We noted that reconstruction of active external rotation of the shoulder will also improve both shoulder abduction and the flexion contracture of the elbow.32,33
The last type of shoulder deformity in our classification29 (Table 5) is scapular winging. We have treating this deformity conservatively but are currently considering treating it with muscle transfer following the impressive results recently presented by Terzis and Papakonstantinou.34
The most common elbow deformity seen in birth palsy is the elbow flexion contracture (also known as the extension deficit). We noted that this deformity is frequently associated with the internal rotation contracture of the shoulder and we have our own theory regarding this association.33 Children with shoulder internal rotation contracture always stand with the shoulder slightly abducted (and hence the elbow falls into slight flexion) to place the limb away from the body in a better functional position. This constant flexed posture of the elbow contributes to elbow flexion contracture. We treat mild cases with splinting and severe cases with surgical release.
Another difficult problem in birth palsy is lack of elbow flexion. We treat this problem by a reconstructive ladder approach (i.e., from simple to complex procedures). If the child has strong wrist extension and finger flexion, a Steindler procedure (relocation of the flexor pronator origin to the midanterior surface of the humerus) is performed. If the child is not a candidate for the Steindler procedure, we prefer reconstruction of elbow flexion by a bipolar latissimus dorsi pedicle muscle transfer. If this fails, we then perform a functional muscle free flap transfer.
Erb's palsy patients frequently present with mild pronation contracture, and this does not require any treatment because the posture puts the hand in a functional position.
Children with total palsy may present with a supinated forearm causing a major functional and cosmetic disability, and we have called this deformity the “beggar's posture” or the child with the “unshakable hand.”35 We treat the supination deformity with either biceps rerouting or rotational osteotomy of the radius.35 The former procedure is performed in children with good passive forearm pronation, strong biceps, and no radial head dislocation. The osteotomy procedure is performed if there is a supination contracture, weak biceps, or dislocation of the radial head (Fig. 5).
The most frequent wrist deformity in birth palsy is lack of wrist extension. Our preferred muscle transfer for this deformity is either the flexor carpi ulnaris or radialis.36
We assessed our results in 20 patients utilizing this transfer according to our modified Medical Research Council Muscle Grading System (Table 2). A good result was obtained in 18 patients (grade 4) and a fair result (grade 3) in the remaining 2 patients (Fig. 6).
Finally, paralysis of long flexors, paralysis of long extensors, claw deformity, and lack of thumb opposition are treated with the usual transfers, and we have outlined our approach previously with total palsy.35 In these patients, correction of the supination deformity of the forearm can be done simultaneously with tendon transfers to the hand.