|Home | About | Journals | Submit | Contact Us | Français|
Early surgical intervention in obstetrical brachial plexus palsy started with sophisticated microsurgery techniques about 20 years ago. However, the indication for the operation is still controversial. A new classification of obstetrical brachial plexus palsy is advocated to establish systematic surgical planning. It is based on the mode of delivery and type of paralysis at 1 month old. It consists of four main types: type I, vertex delivery–upper type palsy; type II, vertex delivery–total type palsy, incomplete; type III, vertex delivery–total type palsy, complete; type IV, breech delivery–upper type palsy. Two types are added to cover the whole spectrum of this condition. Type O is neurapraxia and is clinically unimportant. Type V is a miscellaneous type, which should be considered individually. Most of the cases are classified into one of the four main types. The first priority for reconstruction is given to upper root function in types I and IV, posterior cord function in type II, and hand function in type III. The surgical procedure is intraplexal neurotizations in type I, and a combination of intraplexal and extraplexal neurotizations is a choice in types II and III. In type IV, extraplexal neurotizations are necessary. Surgical results are also discussed briefly in this article.
Brachial plexus palsy in newborns was first recognized by Smellie in 1768 as having an obstetrical origin. Danyau (1851) and Duchenne (1861) suggested that the cause was traction of the brachial plexus, and Erb (1874) described the lesion of the nerve trunk at C5 and C6 roots, Erb's point. Since then it has been well accepted that it is associated with birth trauma, specifically traumatic manipulations at delivery. Recently, however, there have appeared reports that insist on the importance of propulsion rather than the artificial traction force given by a midwife or an obstetrician, although they at least agree that the impairment results from stretching of the brachial plexus.1
Alder and Patterson stated that the incidence of obstetrical brachial plexus palsy decreased steadily from 1.56 per 1000 births in 1938 to 0.38 in 1962.2 However, the tendency of decrease has stopped since then, and the reported incidence in the last three decades of the twentieth century varied considerably from 0.6 to 2.5 per 1000 births.3,4,5,6,7
Kennedy first reported surgical repair of the damaged brachial plexus in 1903.8 He excised a neuroma at the C5 and C6 nerve roots and sutured directly. Sever in 1925 reported in his large series that most of obstetrical brachial plexus palsies recovered spontaneously and could be further improved by secondary musculoskeletal reconstruction.9 According to the prevalence of the conservative attitude, early surgical intervention disappeared from the literature until the 1970s. During that period, the attention of surgeons was aimed to invent new palliative surgical techniques.10,11,12 In the 1980s, interest in the surgical repair of the brachial plexus reappeared, encouraged by results of microsurgery of the adult brachial plexus.13,14,15,16,17,18,19,20
In Osaka, Japan, the exploration of the brachial plexus in babies started in 1979. In the memorial case, the brachial plexus was explored and electrical continuation of the nerves was confirmed. Hence, no reconstructive procedure was done, which resulted in poor recovery. In the following 10 years, 16 cases were operated. They were very conservative. Neurotizations with the spinal accessory nerve and intercostal nerves were used for reconstruction of the suprascapular and musculocutaneous nerves while other nerves remained untouched. Neuroma resection and sural nerve grafting were done only in two cases. The results were very encouraging and instilled in us a more aggressive attitude. So far, the present author has experience in more than 100 early reconstruction cases and shows here the concepts and the results.
Obstetrical brachial plexus injuries are normally classified into upper type, lower type, and total type. Upper type palsy is known as Erb's palsy involving C5, C6 or C5, C6, C7 roots. The typical arm position of the patient is extension in the elbow, pronation in the forearm, and flexion in the wrist, called the “waiter's tip” position. Lower type palsy is called Klumpke's palsy and the paralysis is confined to the hand. Total type palsy involves the whole plexus. Typically, the limb is entirely flaccid. Upper type palsy is the most common and total type palsy follows. The existence of lower type palsy is thought to be doubtful.21 The present author also agrees with this opinion. Lower type palsy is extremely rare, and my personal experiences are limited to cases with cesarean section, which implies that the mechanism of paralysis is somewhat different. Cases of lower type palsy in some reports may simply represent cases of total type palsy in which the upper root lesion showed recovery while the lower root lesion did not.
Before introducing a new classification (Table 1), the present author insists on the importance of mode of delivery. Distribution and severity of paralysis and recovery patterns from paralysis are quite different between patients associated with vertex delivery and patients with breech delivery. We compared 67 cases of obstetrical brachial plexus injuries associated with breech delivery with 136 cases with vertex delivery. The average birth weight was 2895 g in breech cases and 4186 g in vertex cases. Total type palsy was dominant in vertex cases, whereas most of the breech cases had upper type palsy. Bilateral involvement was not infrequent in breech cases, but it was rare in vertex cases. The most important difference was the injury pattern, which required different treatment strategies. The brachial plexus in breech cases was apparently normal, and somatosensory evoked potentials could not be recorded from C5 and C6 roots, which implied isolated avulsion injuries of the upper roots. Because the lower roots were completely normal, intraplexal neurotizations were not indicated.
Excluding rare types in the brachial plexus injuries, such as isolated C7 injuries proposed by Brunelli and Brunelli,22 the present author classified obstetrical brachial plexus injuries mainly into four types. The classification is done according to mode of delivery and type of paralysis at 1 month of age. It is helpful in deciding surgical strategies.
Patients who showed almost full recovery at 1 month of age are categorized here. Only this type of paralysis recovers completely.23
C5, C6, and sometimes C7 suffered postganglionic injuries. The shoulder is frail and the biceps function is absent. Although wrist and finger extension is possible if the C7 root injury is incomplete, some weakness is usually present.
The entire limb is flaccid except for finger motion. Some active flexion of the fingers always starts before 1 month of age. The wrist flexion accompanied by ulnar deviation is sometimes present but weak. This type corresponds to C5, C6, C7, C8 injuries. Occasionally one or more preganglionic avulsion injuries are associated, although C5 root avulsion is extremely rare. No Horner's sign is observed.
The entire limb is flaccid including finger flexion. This is the most severe form of brachial plexus injuries. Preganglionic avulsion injuries are more common but not always present. Horner's sign may be observed.
Isolated C5, C6, and sometimes C7 avulsion injury. Spared lower nerve roots are completely normal. Avulsions are sometimes incomplete but sufficient recovery can rarely be expected. Phrenic nerve palsy may be present, which means that the lesion extends more cephalad. There is always a clear dividing line between function damaged and spared; that is, no intermediate recovery is observed.
Atypical types of paralysis are actually present. The prognosis and indication for operation should be individually discussed. For example, isolated C7 type, pure Lower type, and breech delivery–total type are included in this type.
Our concept for the treatment of obstetrical brachial plexus injuries during the period before the timing for an early operation is as follows. Start occupational therapy as early as possible: Assist parents with care of the baby at home and provide psychological support to cease anxieties of the parents derived from their baby's impairment. After a couple of weeks, start range-of-motion exercise to prevent joint contracture. Support and help normal development of the baby, and let the baby pay attention to the injured limb to use it properly. As early as possible, select poor cases in which surgical repair of the brachial plexus is indicated to prevent severe sequelae of paralysis in the future.
The second period is between the operation and the beginning of the sensory recovery. Motor recovery is observed, but the changes are not always useful. Some motion appears to be automatic motion. Continue occupational therapy in the first period; in addition to that, sensory stimuli should be given to the patient to reeducate integration of motor and sensory functions.
The third period usually starts about 1 year after the operation. Bimanual activities are most important in this period. The affected limb should work actively along with the other limb in daily play activities of the patient. Subjects who are suitable for this purpose should be suggested by occupational therapists. They teach parents how and how much the parents should help their baby in various circumstances.
The recovery reaches a plateau in the fourth period: usually 2 years after operation for the shoulder and elbow functions and 4 years for the hand function. This period corresponds to the preschool age. The activities that are required for school life will increase in quality and in quantity. Because no further recovery can be expected in this period, it is important to increase activities that the patient can manage. Also, palliative surgery may be indicated in this period because the patient is old enough to cooperate with clinical evaluation and postoperative rehabilitation.
Throughout all the periods, maintenance of range of joint motion is important and release operation is indicated at any age if joint contracture becomes conspicuous.
The indication for early reconstruction of the brachial plexus is most controversial in this type of obstetrical brachial plexus injuries because at least some recovery takes place in all patients and the residual disability can be managed by musculoskeletal operations. The results of neurosurgery should be compared with the results of palliative surgery in view of not only final functions but also risk at operation, economic efficacy, and psychological impacts on patients and their family. No such data, however, are available so far.
The indication for and timing of operation in the literature vary considerably. Gilbert24 recommended microsurgical repair in patients who had no evidence of recovery of biceps function within 3 months after birth. Waters25 concluded that microsurgical repair for the shoulder and elbow functions should not be accepted if the patient is younger than 6 months of age. Chuang and his colleagues26 showed excellent results of palliative surgery for shoulder impairments and mentioned that early neurosurgical intervention is rarely indicated in upper type palsy.27 At the beginning, the present author performed brachial plexus surgery when the biceps function was absent at 6 months of age, in the 1990s operated earlier, and now has settled again at 6 months as the timing for the first decision. Key joint motions are elbow flexion with the biceps muscle and shoulder abduction with the deltoid muscle. Trick motion should not be accepted. Both motions should be more than 20 to 30 degrees against gravity. Otherwise the operation is indicated.
In these types of paralysis, reconstruction of the hand function is most important. Hence, the present author recommends earlier intervention. If there is not sufficient recovery in fingers at 3 months of age, early reconstruction of the brachial plexus is indicated. Al Qattan et al stated that the presence of Horner's sign indicates a poor prognosis.28 From my experience, more than 90% of the patients with Horner's sign showed poor recovery and those patients received early operation.
Most obstetrical palsies associated with breech delivery are either neurapraxia, which is classified in type O, or complete upper type paralysis. Although intermediate type cases (incomplete paralysis) are rare, some exceptional cases experienced by the present author recovered at 3 to 6 months. Hence, the decision for operation should be postponed to 6 months of age. The other reason for waiting until 6 months is that results of intercostal nerve neurotizations are very satisfactory even if the operation is done at 6 months of age.29 The indication for the operation is the same as in type I.
Surgical procedures for brachial plexus reconstruction can be divided into intraplexal neurotizations, extraplexal neurotizations, and a combination of both. Intraplexal neurotizations are applied to postganglionic lesions and most frequently a bundle of the sural nerve is interposed as a free cable graft. Extraplexal nerve neurotizations are applied to more severe cases in which the cervical nerve roots are avulsed from the spinal cord and the proximal nerve stumps are not available for intraplexal reconstruction. Some surgeons have exclusively used intraplexal neurotizations and reported good recovery of elbow and shoulder functions.14,15,17,18,19,20,30 In severe cases, however, one or more cervical nerve roots are often avulsed and the proximal nerve stumps in the brachial plexus are not enough to reconstruct all the upper extremity functions. In such circumstances, extraplexal neurotizations with intercostal nerves or spinal accessory nerve are used as supplementary procedures.16,29,31,32,33 Clarke and his colleagues concluded that neurolysis for neuromas was useless.34 The present author has found that neurolysis cases, in which the brachial plexus was explored but no nerve grafts or neurotizations were performed because intraoperative examinations showed at least some electrophysiological continuity in all nerves, showed less favorable results than nerve graft cases35 and now agrees with their opinion.
In this type, intraplexal neurotizations are indicated. Proximal nerve stumps are healthy and sural nerve grafts are placed mainly in the upper trunk. Pure avulsions of C5 and C6 nerve roots are very rare in vertex delivery cases. In such circumstances, the surgical procedure for type IV is indicated.
Spontaneous recovery of flexion of the fingers is expected in this type of paralysis. The first priority for reconstruction is the posterior cord, especially the radial nerve. Qualities of the proximal nerve stumps are evaluated, and the best one or two stumps should be used for reconstruction of the posterior cord. The lateral cord will receive reinnervation from the other nerve stumps through nerve grafts. If concomitant avulsion injuries make nerve graft to the lateral cord difficult, the intercostal nerves are used to reinnervate the lateral cord. The present author harvests five intercostal nerves for reconstruction of the entire lateral cord. The medial cord will recover spontaneously from the T1 root. Spinal accessory nerve neurotization to the suprascapular nerve is planned unless intraplexal neurotization to it is done.
The first priority for reconstruction in this type of paralysis is hand function, especially flexion of the fingers. Nerve grafting to the C8 root or the medial cord is done. The lateral cord is reconstructed with intercostal neurotizations because there are often one or more root avulsion injuries and the quality and quantity of the proximal nerve stumps affected by postganglionic injuries are not so good either. The suprascapular nerve is reinnervated with spinal accessory nerve. The posterior cord reconstruction is considered last. In some instances, intraplexal neurotization to the posterior cord is impossible and some surgeons may consider contralateral C7 transfer. Palliative tendon transfers will be necessary in the future.
Isolated upper root avulsion injuries in this type mean that spared lower nerve roots are completely normal. There is no indication for intraplexal neurotizations. The present author transfers the intercostal nerves to the musculocutaneous and axillary nerves. The spinal accessory nerve is transferred to the suprascapular nerves. The results are very satisfactory.36
In moderately severe cases of types I and II, cocontraction of the biceps and triceps muscles will take place. Intercostal nerve neurotization to the musculocutaneous nerve is indicated in such circumstances. After the biceps muscle is reinnervated by the intercostal nerves, the biceps and triceps muscles can be activated separately and the cocontraction clinically disappears. Of course, this theoretically creates cocontraction between the biceps and intercostal muscles, but it is less important clinically.
If a patient is already 1 year old when he or she is first seen showing poor recovery, the present author is not courageous enough to cut nerves recovered in the brachial plexus. The attitude is conservative and considers neurosurgical procedures that will never result in deterioration of functions already recovered.
About 75% of the cases obtained more than 120 degrees of shoulder abduction in type I; in other types, 20 to 30% of the cases obtained 120 degrees of abduction. Shoulder abduction of more than 90 degrees was obtained in all of type I, 80% of type II, 60% of type III, and 50% of type IV.
About 60 to 80% of the cases obtained more than 30 degrees of shoulder external rotation in all types. Ten percent in type II and 20% in types III and IV showed poor results.
Elbow flexion of more than 90 degrees was obtained all cases unless the operation was done after 1 year of age. Most of the cases reached 120 degrees of flexion. Otherwise, deterioration in active flexion of the elbow joint caused disability in daily life.
The grading system for hand function consists of ability to grasp and pinch, spreading fingers, and thumb functions. Before the operation, an average score was 1.5 in type II and 0 in type III. The final result showed that the patients in type II were largely improved to 9.2, but the improvement in type III was limited to 3.0 on average. The forearm and wrist function had the same tendency, better in type II than in type III.
Obstetrical brachial plexus palsy is a traction injury that occurs during difficult delivery. Spontaneous recovery can be expected in most of the cases, but there are severe forms of injuries in which a neurosurgical approach is essential to prevent disabilities in the future. Surgical procedures should be a combination of intraplexal neurotizations and extraplexal neurotizations, according to the status of the damaged nerve roots. A new classification that is proposed by the present author is useful for determining surgical strategies. After the surgery, recovery of the shoulder and elbow functions was very satisfactory, but recovery of the hand function was limited.