|Home | About | Journals | Submit | Contact Us | Français|
Stability of the shoulder joint and restoration of abduction are important following obstetrical brachial plexus paralysis, as more distal functions depend on having a stable and functioning shoulder. Both deltoid and supraspinatus muscles are active and play a significant role during arm abduction. Along with the suprascapular nerve reinnervation, it is our policy to also neurotize the axillary nerve. The purpose of this report is to present our experience of suprascapular nerve reconstruction in cases of obstetrical brachial plexus palsy. Our overall results were good or excellent in 96% of the patients for the supraspinatus muscle and 75% of the patients for the infraspinatus muscle. Suprascapular nerve neurotization from distal spinal accessory nerve produced similar results to those obtained after reconstruction from the C5 proximal stump. Early surgery (at less than 6 months of age) yielded significantly better results than late surgery (at more than 6 months of age). When microsurgical reconstruction of the brachial plexus is performed before the age of 3 months, it is unlikely for the patient to need a secondary procedure for correction of results obtained after nerve repair.
Patients with obstetric brachial plexus palsy and upper root involvement (C5–C6±C7) have no shoulder abduction and external rotation because of paralysis of the deltoid, supraspinatus, infraspinatus, and teres minor muscles. Over time, without treatment, muscular imbalance produced by paralysis of the external rotators and abductors will cause permanent and significant deformities at the shoulder joint, with articular surfaces sequelae that result in a flattened glenoid fossa and a deformed humeral head that subluxates posteriorly.1,2,3,4,5 Limb discrepancy as well as lateral displacement and underdevelopment of the ipsilateral scapula have been described.6,7,8
Stability of the shoulder joint and restoration of abduction are important, as more distal functions depend on having a stable and functioning shoulder. Shoulder abduction involves a very important function dependent on a group of muscles that must be absolutely coordinated and working together. This function can be divided into the motion of the scapula on the chest wall and the motion of the arm on the scapula.8 There are three joints in the shoulder complex. The main one is the glenohumeral joint, which is the most mobile joint in the human body. The other two joints are the sternoclavicular and acromioclavicular joints. Movement at the glenohumeral joint requires motion at the other joints of the shoulder complex. The coordinated movement of these joints is referred to as scapulohumeral rhythm and allows the shoulder to move through its full range of movement with the head of the humerus centered within the glenoid fossa. Most investigators indicate a ratio of 2:1 for overall glenohumeral:scapulothoracic motion, so for 180 degrees of shoulder abduction, 120 degrees occur at the glenohumeral joint and 60 degrees at the scapulothoracic joint.9
External rotation at the glenohumeral joint during arm elevation is also a necessary function to clear the greater tuberosity from the coracoacromial arch, to accommodate the retroverted articular surface in an optimal position for glenoid contact, and to allow for maximum shoulder abduction.9
During abduction of the arm, both deltoid and supraspinatus muscles are active, and their specific role is debated. The supraspinatus muscle is a key component of the rotator cuff group of muscles and an effective stabilizer of the glenohumeral joint, especially at 90 degrees of shoulder abduction, and it must exert its power before the deltoid can act as an abductor.10 It has been found that both supraspinatus and deltoid muscles are equally responsible for generating torque during arm abduction, and suprascapular and axillary nerve palsy each produced a similar 50% reduction in torque compared with the nonparalyzed shoulder.11
Different secondary procedures such as tendon and muscle transfers have been prescribed to improve shoulder abduction and external rotation function, but the experience in late obstetrical cases has shown that these procedures do not result in as good function as that produced after suprascapular nerve reconstruction and successful reinnervation of the supra and infraspinatus muscles.12,13,14,15,16,17,18
Thus, early reinnervation of the suprascapular nerve offers a good opportunity to restore stability at the shoulder joint and to produce adequate shoulder abduction and external rotation.19,20,21,22,23,24,25
A detailed clinical examination is necessary for accurate preoperative diagnosis of the extent and the level of injury. Upper brachial plexus palsy (C5–C6 involvement) is characterized by adduction, internal rotation, and pronation of the limb, while the elbow is held in extension. If the C7 nerve root is also involved in the lesion, then the elbow is slightly flexed. In cases with total brachial plexus involvement, the entire upper extremity is atonic and flaccid, with signs of sensory disturbances present. In most of the cases with upper nerve root lesion, it is possible to palpate a neuroma in the supraclavicular area. A plain x-ray of the shoulder, clavicle, and affected humerus should be obtained to rule out a possible fracture. A humeral fracture will indicate that a significant force was applied to the affected arm. The shoulder joint is evaluated for possible dislocation. In our center, all patients preoperatively undergo an inspiratory–expiratory chest x-ray and chest fluoroscopy to rule out a phrenic nerve lesion. A paralytic hemidiaphragm indicates phrenic nerve palsy and high plexus involvement. In addition, scapula winging will indicate a proximal lesion with long thoracic nerve involvement.7,8
Computed tomography myelography and myelography of the cervical spine is done preoperatively in all the patients to identify the presence of meningoceles. EMG-NCV study is performed routinely to access the brachial plexus lesion. The presence of normal sensory potentials in one or several fingers while the patient has no sensory perception will indicate a preganglionic lesion.
Skin incision starts proximally along the posterior border of the sternocleidomastoid muscle, with the neck turned away from the affected side, and curves inferiorly over the superior border of the clavicle and then extends along the deltopectoral groove. The upper brachial plexus spinal nerves are identified in the space between the scalenus anterior and medius. Care must be taken to preserve the phrenic nerve, which is located along the scalenus anterior muscle. A nerve stimulator at 0.5, 1.0, and 2.0 mÅ is used to confirm the identity of motor branches throughout the surgical exploration.
The suprascapular nerve usually is found as it exits the upper trunk neuroma. The spinal accessory nerve is found as it emerges along the lateral border of the sternocleidomastoid muscle, cranial to the C4 spinal nerve. After its exit from the sternocleidomastoid muscle, the spinal accessory nerve proceeds distally and laterally in the posterior triangle, under the anterior border of the trapezius muscle. The transverse cervical vessels make up a landmark for detection of the spinal accessory nerve on the anterior surface of the trapezius.26 These vessels enter the muscle at the base of the neck. After identification of these vessels, a nerve stimulator is used to identify the distal part of the spinal accessory nerve. It is imperative not to downgrade the existing trapezius muscle function. For this reason, emphasis is placed on dissecting the spinal accessory nerve as far distally as possible and protecting the proximal two to three branches that go to the upper part of the muscle. The distal nerve is divided and moved to the supraclavicular fossa for transfer.
For suprascapular nerve repair, the senior author has used the following motor donors: the distal spinal accessory nerve, cervical plexus motor nerves, intraplexus donors, and dorsal scapular nerve.27 When the suprascapular nerve is reinnervated from the C5 or C6 nerve roots, this procedure is termed reconstruction. When extraplexus motor donors are used, this procedure is named neurotization.
An intraoperative severity score assessment of the brachial plexus lesions has been developed by the senior author (J.K.T.). Each root is graded as follows: 0=avulsion, 1=avulsion/rupture, 2=rupture, 3=rupture/traction, 4=traction, and 5=normal. A normal brachial plexus severity score is equal to 25.28
The clinical examination includes testing of all upper extremity musculature in comparison with the normal side, using the British Medical Research Council Grading System, but expanded further, with intermediate grades for the final postoperative results. The following scale is used for postoperative functional assessment: a poor result was a grade of M0 to M2, M2+to M3 was fair result, M3+to M4−was good, and M4 to M5−was an excellent result. The Mallet scale is used for preoperative and postoperative shoulder evaluation of the obstetrical brachial plexus patients.29
The most common finding in the obstetrical brachial plexus patients is neuroma-in-continuity located at the upper trunk level. Some authors advocate neurolysis as the only therapy for the neuroma-in-continuity.21,30 Laurent et al31 evaluate the neuromas by measuring the amplitude of the compound muscle action potential in the distal muscle and then resecting the neuromas if there is more than a 50% drop in amplitude. When there is less than a 50% decrease in the amplitude, than only neurolysis is performed. Marcus and Clarke,32 after the poor results they obtained with neurolysis alone, advocate neuroma resection and grafting as the primary management of the obstetrical brachial plexus lesion. In general, the results obtained when only neurolysis is performed are not encouraging.33,34,35 Other authors resect all neuromas regardless of conduction in a distal muscle.23,25 Our policy is to resect all neuromas in early cases (less than 3 months of age), despite some contraction in the distal muscles, and to bridge the gap with nerve grafts. Although neuroma resection will cause some functional loss in the early postoperative period, in all our patients treated with neuroma resection and nerve grafting we have seen significant improvement in overall function 12 months after the nerve repair. For late cases with established deformities, our policy is to perform microneurolysis with the diamond knife under high magnification, along with “jump” grafting of selected targets. If there is perineurial continuity and bulging of fascicles on decompression, along with contraction of the distal target, then that component of the neuroma is not resected. Regardless of the condition of the upper trunk, our treatment strategy always involves reconstruction of the suprascapular nerve.
In the obstetrical brachial plexus lesion, the upper roots are mostly involved in a rupture injury, while avulsion is seen more often in the lower roots.20,23,36,37 Hentz and Narakas38 found avulsion of the upper roots in 27% of their patients. In Gilbert's series, 20% of the upper roots were found to be avulsed.23 Laurent et al31 showed 34% root avulsion in upper brachial plexus palsy. Forty percent of the patients in Slooff's series39 had one or more nerve roots avulsed. In our series,27 29% of upper nerve roots were found to be avulsed, and 58% of all patients had at least one root avulsed from the spinal cord. The C5 nerve root was found to be avulsed only in one patient, and the most commonly avulsed roots were C7 and C8 (in 42% of the patients). Avulsion of all five roots was never found in our series with obstetrical brachial plexus patients. In almost all cases, the C7 root was involved in the plexus injury in association with either C5–C6 nerve roots (40% of the cases) or C8–T1 roots (60% of the cases). The frequent involvement of the C7 root reflected the paralysis of finger extensors in our cases.
Upper plexus injuries with involvement of C5, C6, or C7 roots account for 60–80% of cases in published series,19,20,21,23,31,32,33 whereas global plexus injury is seen in 20–40% of all obstetric brachial plexus palsy cases. In Shenaq's series,25 the upper plexus involvement accounted for 70% of the cases, followed by total plexus involvement in 20% of the patients. In Clarke and Curtis's40 series, the upper plexus was involved in 30% of patients, whereas 70% had a total brachial plexus palsy. In the senior author's patient population, upper brachial plexus injury accounted for 40% of the patients, whereas global plexopathies were encountered in 60% of the cases.27
The extent of injury (upper or total brachial plexus involvement) is one of the potential prognostic factors determining the functional results in obstetric brachial plexus palsy patients. Different series reported that patients with C5–C6 root involvement have better prognosis than those with C5, C6, and C7 involvement.22,36,40,41 Nehme42 showed that C7 nerve root involvement is associated with an 80% risk for a poor result. In the Gilbert and Tassin series,43 patients with C5–C6 nerve roots lesion, if not operated on, reached at most a grade III on the Mallet scale, whereas more than 60% of operated patients reached a grade IV on the Mallet scale. Of patients with a C5–C6–C7 nerve root lesion, treated conservatively, none reached a Mallet scale of IV, whereas only 28% of operated patients reached a scale IV. In our series, for patients with C5 and C6 nerve root involvement after nerve reconstruction, evaluation of shoulder abduction using the Mallet scale showed that 75% of them postoperatively achieved a grade IV; this percentage dropped to 60% for patients with C5, C6, and C7 nerve root lesions.27
The timing of nerve reconstruction still remains a controversial issue in the treatment of obstetrical brachial plexus palsies. Gilbert advocated early nerve repair for optimal results.23 In Laurent and Shenaq's31 series, the indication for surgery was the lack of improvement in upper plexus palsy by the fourth month of life. Others suggest waiting for late recovery until 6 or 9 months of age.44,45 Waters,46 comparing the results obtained after microsurgical repair with those treated conservatively, concluded that the functional results obtained after microsurgical repair 6 months after the birth were better than those for patients with spontaneous biceps recovery at the fifth month of life. In the Strombeck series,47 the outcomes were not related with the patient's age when surgery took place (before or after the age of 6 months), and the most important prognostic factor was the number of roots avulsed. Our indication for nerve reconstruction in cases of upper plexus injury is no measurable deltoid or biceps function by 3 months of age. For global palsy patients, if there is a positive Horner sign and no function in the upper extremity, in combination with strong indications by EMG-NCV study and computed tomography myelography that root avulsions are present, then primary exploration and brachial plexus reconstruction should be performed before 3 months of age.
It is generally accepted that the functional results obtained after nerve microreconstruction are less successful the later it is performed. In the author's experience, the functional results obtained in the early group of patients (denervation time < 6 months) were significantly better than those obtained in the late group (denervation time > 6 months). Furthermore, in cases of early repair, the need for secondary reconstruction (shoulder arthroplasty, scapula stabilization procedure, latissimus dorsi, and teres major rerouting) is significantly minimized. Our data show that when surgery is performed under the age of 3 months, it is rare for the patient to have a tendon or muscle transfer to improve the results of nerve reconstruction.27
The nerves used for suprascapular nerve neurotization in our patient population included intraplexus and extraplexus motor donors. The intraplexus donors have a larger number of axons than extraplexus, and theoretically this will increase the chances for successful nerve repair. However, it has been shown that, using the spinal accessory nerve for suprascapular nerve neurotization, the results obtained are similar to those after using intraplexus motor donors.18,19,20,21,22,33,36,38 In our series, spinal accessory nerve was used as donor nerve to neurotize the suprascapular nerve in 46% of the patients, whereas other intra or extraplexus donors were used in 54%.27 In the majority of patients, if intraplexus donors were used, the procedure involved the C5 proximal nerve root stump. A strong argument for using the spinal accessory nerve for neurotization is that it is a pure motor nerve, and direct repair of the suprascapular nerve can be achieved without the need for an interposition nerve graft. Our data are in accordance with those reported by Marcus and Clarke32 and show that neurotization of the suprascapular nerve from the distal spinal accessory nerve produced similar results with reconstruction from the C5 root. Also, to achieve better functional results in the shoulder, it is our policy to reconstruct the axillary nerve along with the suprascapular nerve.
Our experience with adult posttraumatic brachial plexus cases showed that nerve grafts used and their length had a strong effect on the end results.28 Nerve grafts in our obstetrical paralysis series were used in 25 patients, and the average graft length was 5.1±1.5 cm. Although the functional results obtained after direct suprascapular nerve repair were better than those obtained after repair via nerve graft, there was not a statistically significant difference between the two groups in the obstetrical population.27
Narakas20 noted that suprascapular neurotization by the spinal accessory nerve yields a muscle grading of M3 or better and will allow shoulder abduction of more than 60 degrees.
In Gilbert's23 series, 50% of the patients with lesion of the upper roots reached a shoulder abduction range of 90–140 degrees, with external rotation in neutral. Boome and Kay48 showed external rotation restoration to muscle grading M4 or better in 25% of patients, and Waters46 found that according to the Mallet scale, the average score for shoulder abduction was 3.5, and for external rotation it was 2.7. In our series, overall functional results for shoulder abduction were good and excellent in 96% of the patients, and for external rotation, they were good and excellent in 75% of them.27
This girl was born to a G2P2A0 mother after an uncomplicated pregnancy at 40 weeks of gestation, vaginal delivery and vertex presentation. The labor lasted 10 hours, and no suction or forceps were used. No fractures or shoulder dislocation were noted. The Apgar scores were 9 at first minute and 10 at 5 minutes. Baby's birth weight was 3650 g. Left obstetrical brachial plexus palsy was noted immediately after birth. She was presented at our institution at 3 months of age. During the clinical evaluation, the infant had a Mallet's scale of grade I as far as shoulder abduction and external rotation. At the age of 5 months, the patient was operated on. The supraclavicular plexus was explored. The phrenic nerve was stimulated and was found to be intact. A huge neuroma was found in the upper and middle trunk. The neuroma was stimulated, and no distal contractions were seen. The neuroma was resected to healthy fascicles, and the suprascapular nerve was reconstructed from the C5 proximal stump, using a 6-cm sural nerve graft. The posterior cord and lateral cord were reinnervated from C5 and C6 proximal stumps, using 5 × 4 cm nerve grafts, whereas the middle trunk was reconstructed from C7 nerve root. Postoperatively, the patients reached grade IV of shoulder abduction and external rotation by the Mallet scale.
This 32-month-old girl had been born after 40 weeks of gestation, vaginal delivery and vertex presentation. No forceps or suction were used. The baby's birth weight was 5200 g. During the preoperative examination, the right shoulder displayed 15 degrees of abduction and no external rotation. The patient was operated on at 32 months of age. Supra and infraclavicular brachial plexus exploration showed findings compatible with C5 rupture, C6 rupture, C7 avulsion, C8 rupture, and T1 traction.
The suprascapular nerve was directly neurotized by the distal spinal accessory nerve. The posterior cord was reconstructed from the proximal stump of the C5 root, whereas the C6 root was used for lateral cord reconstruction.
Two years after the surgery, despite her late presentation for treatment, the results of brachial plexus reconstruction were rewarding. The patient could abduct the shoulder 90 degrees and had 30 degrees external.
The patient was a 26-month-old boy born after 40 weeks of gestation, vaginal delivery and vertex presentation. No forceps or suction were used. The baby's birth weight was 3900 g. During the preoperative examination, the right arm had no abduction or external rotation. The patient was operated on at 26 months of age. Supra and infraclavicular brachial plexus exploration showed findings compatible with C5 rupture, C6 and C7 avulsion, and C8 and T1 rupture. The suprascapular nerve was directly neurotized by the distal spinal accessory nerve. The proximal stump of the C5 root was used to reconstruct the distal C8 and T1 roots. In addition, additional grafts from the C5 root were given to the median, lateral pectoral, and musculocutaneous nerves.
Three years after the surgery, the patient is able to demonstrate 80 degrees of shoulder abduction and 15 degrees of external rotation.
Suprascapular nerve neurotization is a worthwhile procedure for restoration of glenohumeral joint stability and shoulder abduction in obstetrical brachial plexus paralysis. Neurotization of the suprascapular nerve from the distal spinal accessory nerve yields similar results with reconstruction from the C5 nerve root. Concomitant neurotization of the axillary nerve yields improved outcomes in shoulder abduction function. The best results are seen in patients with only the involvement of upper roots and when direct neurotization of the suprascapular nerve was performed in the early postnatal period.