Brachial Plexus Hand Surgery

Updated: Jun 08, 2021
  • Author: Alan Bienstock, MD; Chief Editor: Joseph A Molnar, MD, PhD, FACS  more...
  • Print


Obstetric brachial plexus palsy is a well known, challenging condition that afflicts 1% of births. The resulting deformity varies greatly in severity, and recovery is difficult to predict. Since the advent of microsurgical techniques and a trend for secondary reconstruction, studies have evolved to assess the surgical treatment of these injuries. Over the last 2 decades, several referral centers have blossomed to treat this complex problem in a multidisciplinary approach by using standardized treatment protocols.

Most obstetric injuries to the brachial plexus involve the upper trunk and cervical nerve roots C5 and C6. Shoulder function is commonly impaired. Secondary residual deformities, such as the internal rotation and adduction deformity of the upper extremity, arise from prolonged conservative management or failed surgical treatment. Several authors advocate early primary exploration and microsurgical repair because of improved long-term outcomes. A large variety of reconstructive modalities involving nerve grafting, plexus neurolysis, nerve transfers, tendon transfers, muscle releases, neurotizations, and free muscle transplantations have surfaced to treat the deficits. The objectives are to restore hand function, elbow flexion, and shoulder abduction.


History of the Procedure

The renowned British obstetrician Smellie is attributed with providing the first medical description of obstetrica brachial plexus palsy. In his 1768 treatise on midwifery, he reported a case of transient bilateral arm paralysis in a newborn after difficult labor.

In 1861, Duchenne invented the term obstetric palsy of the brachial plexus after examining 4 infants with paralysis of identical muscles in the arm and shoulder. He described the physical findings and recognized that traction was the source.

In 1875, Erb studied brachial plexus injuries in adults. He concluded that palsies of the deltoid, biceps, and subscapularis are derived from a lesion at the level of C5 and C6. In the same year, Klumpke described lesions of C8 and thoracic nerve T1 in birth palsy and associated the Horner sign with lower-trunk lesions of the brachial plexus.

Kennedy presaged modern primary brachial plexus reconstruction with his attempts to resect proximal plexus neuromas and to perform primary suture repair in 1903. Fairbank identified the secondary muscle deformities, which evolve from long-standing paralysis and espoused muscle release to combat the characteristic features of internal rotation and adduction.

Because of Seddon's disappointing outcomes with nerve grafts to treat a traction injury, surgical treatment steered away from primary repair and toward muscle releases and transfers. The investigative work continued to focus on pathogenesis, diagnosis, and treatment. However, the initial enthusiasm for surgical treatment of obstetric brachial plexus palsy was not applied until 1984. Gilbert and Tassin implemented modern technologic advances and began a new surgical era by reporting encouraging operative outcomes in 180 patients. [1] Because of their encouraging results with early surgical intervention, primary nerve repair for obstetrica brachial plexus palsy resurfaced. Narakas, [2] Terzis, [3] Kawabata, [4] Boome, [5] Alanen, [6] Slooff, Laurent, [7] and Clarke [8] later supported their conclusions.



Obstetric brachial plexus injuries can be classified into 4 main categories: classic Erb or upper type, intermediate type, Klumpke or lower type, and total brachial plexus injury.

Most brachial plexus injuries are the classic Erb or upper type and occur at spinal nerves C5 and C6. A classic waiter's-tip arm position caused by a muscle imbalance holds the shoulder in an adducted, internally rotated position with the elbow in extension, the forearm in pronation, and the wrist and fingers in flexion because of variable weakness in the wrist and finger extensors.

In a proximal lesion, a winged scapula and weakness of the suprascapularis and infrascapularis muscles are apparent. If the lesion occurs distally and involves the upper trunk, the dorsal scapular and long thoracic nerves are spared, and function of the rhomboid and serratus anterior muscles is preserved. In addition, the upper-type lesion affects the function of the sternocostal portion of the pectoralis major muscle, which accounts for the appearance of forward flexion-contracture of the shoulder.

An upper type of obstetric brachial plexus palsy can be extensive and involve C7, resulting in an injury with the aforementioned findings in addition to slight flexion because of radial nerve involvement. This involvement causes weakness of the triceps and increases involvement of wrist extensors and finger extensors.

An intermediate lesion predominantly involves C7 with occasional involvement of C8 and T1. The affected arm is abducted, the elbow is flexed, and the wrist and fingers are flaccid.

The Klumpke or lower type of brachial plexus palsy is rare. Some authors even question its existence. In this type of injury, paralysis is confined to the hand, with an ipsilateral Horner syndrome if the injury is proximal.

Total brachial plexus palsy, the second most common type of injury, is characterized by complete arm paralysis, decreased sensation, and a pale extremity.

About 73% of cases manifest as injuries to the upper cervical roots, 25% are total plexus injuries, and 2% are isolated lower or Klumpke palsy.




The rate of obstetric brachial plexus palsy is 1-3 cases per 1000 live births. This rate has remained unchanged since the beginning of this century. With the present birth rate, 4000 new cases of obstetrica brachial plexus palsy are estimated to occur each year in the United States.



Risk factors for obstetric brachial plexus palsy are gestational diabetes, forceps delivery, vacuum extraction, and shoulder dystocia. [9]

Obstetric brachial plexus palsy is associated with shoulder dystocia, which most frequently occurs in infants with fetal macrosomia. The incidence is also higher in male individuals than in female individuals. The right side is affected more often than the left because the most common position of the descending fetus is the left occiput anterior position in which the mother's pubis compresses the baby's right shoulder.

Multiparous women are most likely to deliver a neonate with obstetric brachial plexus palsy than women who are nulliparous. Mothers who previously delivered large babies with brachial plexus palsy are predisposed to deliver future children with the palsy. Cesarean delivery should be considered for mothers with this history.



Upper brachial plexus injuries most commonly occur in macrosomic infants during vaginal delivery in which shoulder dystocia results in excessive traction on C5 and C6 by forcibly separating the head and shoulder. Neonates in the breech orientation can also present with rupture. Because the last part extracted is often the baby's head, the C5-C7 roots can rupture, commonly in bilateral fashion.

The lower type of brachial plexus injury may occur during breech extraction in which forcible arm abduction places traction on the lower roots.

Total brachial plexus palsy usually occurs in difficult vertex or breech deliveries. The upper spinal nerves usually rupture, and the lower spinal nerves are avulsed, as signified by the Horner sign. In rare instances, the lower roots may be caught between the clavicle and the first rib; in these cases, the injury appears as a rupture instead of an avulsion.

Severe brachial plexus injuries are often associated with but not predicted because of clavicular fractures, which are the most common bony injuries in birth trauma, especially in the presence of shoulder dystocia. The stretch-injury mechanism of obstetric brachial plexus injury, as proposed above, applies to most cases. However, the injury is a result of excessive traction and not always preventable. The literature includes cases that occurred in uncomplicated vaginal deliveries, in cesarean deliveries, and in smaller-than-average birth weight infants. Authors have recently suggested that congenital brachial plexus palsy is a possible etiology; in this injury, abnormal intrauterine posture induces pressure neuropathy.



Clinical examination

Careful and detailed history taking and physical examination are imperative for diagnosing obstetric brachial plexus injuries. [10] These evaluations can be difficult in young patients and infants because of their inability to cooperate. The surgeon must acquire information about the pregnancy, neonatal period, and delivery as they pertain to gestational age, birth weight, presentation, type of and length of labor, use of forceps, shoulder dystocia or fractures, and Apgar scores, among other factors.

The treatment team must look at the patient's arm at rest and evaluate the chest and for abdominal excursion to ascertain if phrenic nerve injury is present. Fluoroscopy can be performed to evaluate a suspected injury. A Horner sign and/or absent paraspinal muscle activity signifies avulsion of a nerve root.

The surgical team should perform a detailed inspection and test the bulk, range of motion, and strength of all muscles on the patient's back, upper extremity, wrist, and hand (eg, trapezius, rhomboids, and latissimus dorsi muscles). Particular attention should be paid to shoulder abduction, adduction, and internal and external rotation. Contracture of the pectoralis major can be assessed by palpating the anterior axillary fold during external rotation. Likewise, subscapularis contracture can be assessed by palpating the posterior axillary fold during shoulder abduction. The physician should inspect the clavicle, ribs, and humerus to determine if the patient has any fractures or dislocations. If found, such injuries must be confirmed or ruled out by using radiographs.

Grading systems

The information garnered from the patient examination must be quantified and translated to reliable outcome measures to evaluate the postoperative results and to permit consistent comparison with results from other centers.

Several scales for grading muscle function are available, but none are ideal for universal application in evaluating infant motor function.

The British Medical Research Council (MRC) Motor Grading Scale was created to assess brachial plexus injuries in adults, but it has limited application in young children because of the lack of patient cooperation. Gilbert and Tassin modified the MRC scale to decrease the effect of the amount of strength used in the test. [1] However, this simplification also decreases the descriptive value of the results because 1 grade describes a widened range of motor function.

The best measurement of outcome is overall function, as assessed by evaluating joint range of motion rather than motor function of individual muscles. This method is described in the Mallet grading system.



Most patients with obstetric brachial plexus palsy present with a flail arm at birth. The paralysis frequently regresses, and recovery may be complete. The percentage of patients who undergo spontaneous recuperation is somewhat controversial. The discrepancies are likely attributable to inconstant study designs. Reported recovery rates vary in the range of 10-92%.

If the nerve injury is limited to neurapraxia or axonotmesis, recovery begins within a few days and is completed by the time the infant is aged 3 months. However, if the injury is relatively severe, such as in neurotmesis or avulsion, this period may be prolonged and the recovery incomplete or absent. In patients with such injuries, muscular imbalance causes abnormal stresses, and progressive joint contractures and bony deformities develop.

Because early surgical intervention optimizes functional outcome in patients who do not spontaneously recover, a reliable marker is needed to identify these patients. In the ideal case, early prediction of whether surgical intervention will be required with good sensitivity (to not miss patients who require surgery) and good specificity (to avoid unnecessary surgery) are possible. Gilbert and Tassin used biceps muscle function at age 3 months to predict recovery in their surgical decision making. [1] They assume that if deltoid and biceps function is not started by 3 months, ultimate function is likely to be poor.

A 1994 study demonstrated that use of elbow flexion as a criterion resulted in the incorrect prediction of recovery in 12.8% of patients. [11] Combining 2 or more parameters as determinants of recovery was most reliable. Adding an assessment of finger extension to elbow flexion reduced the error rate to 5.2%. Therefore, evaluation has shifted toward multiple muscle groups from a single muscle to determine the timing for surgery.


Relevant Anatomy

Formation of the brachial plexus begins in early development in the fourth week of gestation. Because the sclerotome directs axonal growth, nerve formation follows the dorsal rotation of the upper limb bud. Axons from the ventral column motor cells start to grow toward the sclerotome cell mass, forming the ventral root. The dorsal root similarly forms by axons growing in the opposite direction from the dorsal root ganglion cells. As a consequence, the anterior divisions of the brachial plexus are destined to innervate the ventral muscles, and the posterior divisions innervate the dorsal or extensor muscles.

Anterior and posterior nerve roots emerge from the spinal cord and converge to form spinal nerves C5, C6, C7, C8, and T1 of the brachial plexus. If C4 makes a substantive contribution to C5, the brachial plexus is termed prefixed. If T2 makes a substantive contribution to T1, the plexus is termed postfixed. C5 innervates the deltoid, C6 innervates the biceps, C7 innervates the triceps, and C8 and T1 innervate the hand musculature.

The nerve roots are encased by an endoneurial sheath, then by the dura, followed by a layer of epidural tissue. The endoneurial sheath and dura become the perineurium, and the epidural tissue forms the epineurium of the spinal nerves as they emerge from the intervertebral foramina. These layers serve to provide strength, elasticity, and integrity to the spinal nerve, rendering nerve roots more susceptible to injury than spinal nerves.

The roots of nerves C5 and C6 are anchored to the transverse processes by a dense fibrous sheath, which protect them from avulsion. However, these roots are still prone to rupture. The incidence of avulsion injuries of the lower nerve roots is increased because these roots do not contain ligamentous attachments, a lack which makes them anatomically vulnerable.

The extent of injury to a nerve is inversely proportional to its length. The length of a nerve root increases in a caudal direction, where the upper nerve roots are most susceptible to injury.

Disruption or injury to the nerve root can be classified as supraganglionic or infraganglionic lesions in terms of the dorsal root ganglion. In both types of lesions, the motor and sensory functions are completely lost. In supraganglionic lesions, sensory conduction studies are normal because the disruption occurs proximal to the sensory neuron bodies and because Wallerian degeneration does not occur. The spinal nerves receive gray rami communicantes; the first 4 are from the superior cervical ganglion, the fifth and sixth are from the middle cervical ganglion, the seventh and eighth are from the inferior cervical ganglion, and the thoracic nerves are from the corresponding ganglia. White rami communicantes are preganglionic sympathetic fibers from the spinal cord at T1-T5 that ascend to form the cervical ganglia.

Horner syndrome occurs as a result of disruption of the communicating branch that supplies the sympathetics to the stellate ganglion. Ciliary nerve dysfunction manifests as weakness of the levator palpebrae superioris muscle (ptosis), dilator pupillae muscle (myosis), Müeller muscle (enophthalmos), and decreased sweat secretion (anhydrosis). This usually signifies an avulsion of T1 because the preganglionic sympathetic fibers leave the spinal nerve soon after it appears from the intervertebral foramen.

Spinal nerves divide to provide the posterior rami, which supply the paraspinal muscles, and the ventral rami, which proceed to form the brachial plexus. Therefore, denervation of the cervical erector spinae muscles also indicates a proximal, preganglionic injury. The nerves of the brachial plexus are located in a triangular area of the neck and travel to the axilla. This triangle is demarcated by the posterior edge of the sternocleidomastoid muscle, the anterior border of the trapezius muscle, and the clavicle; the omohyoid muscle bisects this triangle. Its upper portion contains the upper and middle trunks along with the scalene muscles, and its lower portion contains the lower trunk and subclavian vessels.

The narrow posterior scalene space, between the anterior and middle scalene muscles, marks the branching of the upper and middle trunks into posterior and anterior divisions. The lower trunk separates into the posterior and anterior divisions as it travels between the anterior and middle scalene muscles, posterior to the first rib and subclavian artery. As they descend below the clavicle, the 3 posterior divisions unite to form the posterior cord, the upper 2 anterior divisions form the lateral cord, and the lower anterior division continues as the medial cord; these structures are named after their respective locations around the subclavian artery. The cords then divide into terminal branches, forming the intricate framework of the brachial plexus.

The medial and lateral sural cutaneous nerves join to form the sural nerve proper, which runs posteriorly between the heads of the gastrocnemius muscle to innervate the dorsolateral aspect of the calf and foot.



After approximately 18 months of denervation, the muscle atrophies, and reinnervation is not possible because the motor endplate is absent. At this point, only tendon reconstruction or free muscle transfers are indicated.