BRACHIAL PLEXUS PALSY
Birth brachial plexus injury occurs in between 1 and 2 per 1,000 live births in the United States. Babies with increased birth weights, multiparous mothers, and shoulder dystocia are at the highest risk for brachial plexus palsy (48,49). The most widely described mechanism of action for this is lateral stretch, which is logical secondary to the location of the brachial plexus, the high correlation with shoulder dystocia, and the positioning of the mother and infant (49). It has been described that between 50% and 95% of these infants will recover spontaneously. The overarching goal of treatment of brachial plexus injuries is maximizing arm and hand function. Goals are normalization of limb function, with optimization of nerve regeneration and mechanical increase of elbow flexion and shoulder stabilization. This can be achieved through aggressive rehabilitation and surgical intervention (48).
For any nerve that is injured, classification makes evaluation and comparison clearer. The Seddon Classification of Nerve Injury is commonly used. Neurapraxia occurs with no lasting anatomic changes, with fibers preserved. This is exemplified by a football "stinger" injury. Complete resolution is expected. In axonotmesis, there is an interruption of neural continuity to some degree. There is an extremely variable level of deficit that is difficult to evaluate and predict the degree of recovery. Neurotmesis is the most severe injury, with total disruption of the elements of the nerve, and this will not independently recover. If it is preganglionic, or proximal to the dorsal root ganglion, it is called an avulsion. If it is postganglionic, or distal to the dorsal root ganglion, it is called a rupture (50). Both of these require surgical intervention for recovery.
There are also descriptors for the levels of brachial plexus palsy. Injury at C5 and C6 is called Erb's palsy, sometimes called Erb-Duchenne palsy. This is the most common level of involvement, present in approximately three-fourths of those with birth brachial plexus palsy (BBPP). Involvement of C8 and Tl is Klumpke's palsy. It is debated whether Klumpke's can occur in a birth brachial plexus injury, though it definitely occurs in other types of brachial plexus injury. The reason for this question is whether it is anatomically possible to have a C8 and Tl lesion alone without the involvement of C5 and C7. It appears that if there is an anatomic variation—for example, a rib, tendon, bony, or another anomaly that leads to the compromise of C8 and Tl—this can occur in a birth brachial plexus injury. Otherwise, it appears that it cannot. Therefore, if a child presents with a C8 and Tl birth brachial plexus injury, it may be from anatomic anomaly, but there are two other options to consider. Most likely, it was initially a complete brachial plexus involvement but there was quick recovery of C5 and C7. This is likely, since the upper cervical root levels are relatively protected anatomically, so C8 and Tl may end up with the most severe injury. It is also possible that a spinal cord injury has been mistaken for brachial plexus palsy. All of these are important to consider during evaluation. There also may be complete brachial plexus palsy, including C5 and Tl, with total motor and sensory loss. There also can be a variety of levels involved between upper plexus and total plexus palsy.
Evaluation of patients with brachial plexus palsy includes clinical findings, electrodiagnosis, and MRI. There is debate about which of these is the most effective. MRI is expensive and requires sedation to perform on infants. It has been found to correlate with surgical findings 70% of the time, electromyography (EMG) 87% of the time, and clinical findings 60% of the time. The correlation was highest when all three of these were combined. MRI was effective only in those with C5, C6 root involvement (51).
Clinical exam consists of a history and physical examination. The history includes the birth number of the child, the birth weight, and presence of maternal diabetes during the pregnancy, along with the size of previous infants and the birth size of the parents. The motor and sensory findings at birth, along with any change up to the time of evaluation, are important. The use of vacuum or forceps may be indicative of any difficulty with delivery. The most common association is shoulder dystocia. Other useful information is whether there were signs of bruising or other injuries, or whether there was involvement of the contralateral arm or the legs at delivery.
Physical examination begins with visualization of the arm to include the size and bulk. A cool temperature may be noted in those with severe involvement. Sensory evaluation is critical to determine the extent and levels of involvement. Muscle stretch reflexes will be decreased or absent in the distribution of a brachial plexus injury.
The primitive reflexes are also important. Since the upper plexus has more frequent involvement, the Moro reflex, which shows shoulder abduction and elbow flexion, is valuable in assessing those active movements. Torticollis is frequently seen, and usually this is with the face turned away from the involved arm. ROM is an important part of the evaluation since contractures are commonly seen in shoulder adduction and internal rotation, wrist flexion, forearm pronation, and even at the elbow into flexion commonly in later months and years.
A key goal of the electrodiagnostic evaluation is to find subclinical nerve and muscle responses. The study must be specific depending on the clinical deficits noted, with studies performed that are pertinent to each individual's examination. Sensory nerve conduction studies, motor nerve conduction studies, and EMG are performed. Diagnostic evaluation should include nontraditional nerve conduction studies, and frequently not the classic median and ulnar nerves, due to frequent involvement of only the upper brachial plexus. Axillary, musculocutaneous, and radial nerves are among those useful for electrodiagnostic study. Sensory nerve action potentials (SNAPs) are important, as these are most sensitive to axonal loss (52). The presence of SNAP responses in an insensate area is indicative of a preganglionic lesion, due to the location of the sensory cell bodies in the dorsal root ganglion. EMG may show activation of motor unit potentials in muscles with no clinical motor activity. Electromyographic evaluation is reported as being of some benefit, but it underestimates the severity of lesions (53). It has been recommended to be performed early in the first few days, then with a repeat evaluation after several months to more accurately identify cases where there is reinnervation occurring and therefore providing earlier determination of the need for surgical intervention (54). EMG at 1 month has been shown to have the best prediction for recovery in babies (55,56).
Plain x-rays are useful in some clinical circumstances. Some abnormalities may mimic a brachial plexus palsy, including a fracture of the clavicle or humerus. Osteomyelitis may also mimic this, and has actually been reported as inciting temporary brachial plexus palsy (57). Neurofibromatosis or other tumors may also damage the brachial plexus.