|Home | About | Journals | Submit | Contact Us | Français|
Treatment strategy in infants with obstetric brachial plexus palsy (OBPP) largely depends on clinical neurological examination and the degree of improvement in the first 3 months. Usually, less severe lesions show early recovery within a few months, and prognosis for spontaneous recovery is good. However, sometimes the time course of improvement may differ and will not allow clinical differentiation from more severe types of lesions with less favorable prognosis. Ancillary preoperative investigations assist in determining the type, level, and extent of the nerve lesion. The role of electrodiagnosis in preoperative assessment of OBPP is depreciated by many, whereas others stress the importance of combining results from electromyography and nerve action potential recording to discriminate between a nerve conduction block and a root avulsion or to predict the severity of axonal injury or degeneration. There is no role for motor-evoked potentials in OBPP yet. For imaging of the brachial plexus in infants, magnetic resonance imaging has surpassed computed tomography–myelography as modality of choice. High-strength magnetic resonance scanners, applying different techniques in a noninvasive way, allow imaging of plexus structures with great detail. Detection of different nerve lesion types is possible, such as root avulsions or nerve ruptures, formation of pseudomeningoceles, neuromas, or scarring as well as deformities of the shoulder joints. Magnetic resonance imaging is becoming a great aid as a preoperative investigation in determining treatment strategy in infants with severe OBPP.
Obstetric brachial plexus palsy (OBPP) is caused by a stretch injury to the brachial plexus. Reliable studies concerning prognosis are still lacking and, as a consequence, so are those concerning treatment. Most (75%–90%) infants will recover spontaneously, but babies with severe lesions will need microsurgical reconstruction of their damaged brachial plexus. Although the timing of and criteria for surgery are still a matter of debate, some consensus has been reached.1,2,3,4,5 Preoperative ancillary investigations ideally should provide information on the type, extent, and location of the nerve injury and should be reliable indicators of prognosis. In OBPP neurophysiological and imaging studies help to delineate the lesions and in many cases allow for an accurate diagnosis. When children meet the more or less accepted criteria for surgery, preoperative neurophysiological and imaging studies are performed, usually around the fourth month of age.4,5 However, especially in babies and small infants, these investigations have their limitations.
The brachial plexus supplies sensory and motor innervation to the upper limb. It forms from the ventral rami of the C5–T1 spinal nerves. These nerves coalesce into three trunks along the posterolateral margin of the anterior scalene muscle. The upper trunk is derived from C5 and C6, the middle trunk from C7, and the inferior trunk from C8 and T1. Toward the periphery, the trunks divide into six divisions, finally coalescing into three cords, which are named the medial, lateral, and posterior cords.
There are four basic types of nerve injury.6 Neurapraxia is a temporary conduction block. In axonotmesis the axon is severed but the surrounding nonneural elements are intact. In both types of injuries, spontaneous recovery is very likely. The other two, much more severe, types of nerve injury are usually managed by surgery at some stage. Neurotmesis is a complete postganglionic disruption of a nerve, and avulsion is a preganglionic disconnection of a spinal nerve from the spinal cord. In OBPP, most lesions are at the level of the supraclavicular plexus structures (i.e., proximal to the cords).
In OBPP in general there is still a lack of reliable indicators of prognosis on which treatment strategy can be based. As the success of nerve-grafting procedures in OBPP depends on the presence of regenerating axons in the proximal nerve stumps, the central diagnostic problem in severe OBPP for which a decision to perform surgery must be made is discriminating between neurotmesis and avulsion and especially reliably demonstrating spinal root avulsions in preoperative investigations.7,8 When spinal roots are avulsed from the spinal cord, regenerative capacity is lost, as there is no connection with the central cell body. During surgery, a root avulsion is evident when the spinal ganglion is retracted extraforaminally (Fig. 4C). However, if an actually avulsed nerve root is not retracted, it may seem deceivingly normal at its exit from the foramen. Repair from such a nerve is doomed to be ineffective because there are no regenerating axons. The same holds true for intradural and intraforaminal ruptures of spinal nerve roots. With the demonstration of other types of lesions in preoperative ancillary examinations in OBPP peripheral to the foramina, it may be possible in the future to relate these to indications for surgery or to outcome. As yet it seems of less clinical importance, because most of the severe lesions necessitating surgery can readily be recognized and analyzed at operative exploration of the brachial plexus. An exception may be the electrodiagnosis of a prolonged nerve conduction block in which late spontaneous recovery can be expected.9
Clinical neurological examination in OBPP is the base for further treatment decisions. Ancillary investigations as yet delineate different types of nerve lesions and support treatment strategies. In the following pages, preoperative electrodiagnosis and different imaging techniques will be discussed.
There are serious doubts about the value of electromyographic (EMG) studies in babies with OBPP, as a severe clinical picture is often in contradiction with optimistic EMG findings.3,10,11,12,13,14 The EMG in babies with OBPP differs from adults in that denervation occurs and disappears much earlier. It can be found already on the fourth day and may have disappeared at 4 months. The very early denervation, as a sign of axonal degeneration, can be explained in infants by short distances and small diameter of the nerves. As compared with nerve size in adults, a 7.5–10 times faster denervation can be expected in infants. For adults, a 10–14-day period is normal, whereas a period of 1–2 days should not be too surprising in babies. Denervation activity usually disappears between day 10 and day 60 of age.15 In many centers, EMG studies are performed at age 3–4 months.16 By then, even children with completely avulsed C5 and C6 roots and a paralytic biceps muscle show a normal EMG on reflex-activated contraction of the biceps muscle. For this discrepancy between lack of functional muscle activity and EMG findings with motor unit potentials (MUPs), van Dijk and Vredeveld offer several explanations.13,14,17
An inadequate clinical examination may fail to register a slight muscle contraction. Because muscle fibers and corresponding motor units are 11-fold smaller in infants than in adults, the number of active MUs in infants is easily overestimated if the same-size EMG needle is used for both adults and infants.
Another explanation is the concept of luxury innervation. At birth there may still be a polyneural innervation in which nerves from multiple segments supply extra innervation to muscles. This pattern is reported to disappear and to be replaced by mononeural innervation from week 16 to 25 of gestation, up to 3 months of age.18,19 In the absence of the original dominant innervation, as in root avulsions, this pattern probably persists, explaining the overly optimistic, near normal EMG findings. In contrast, and at the same time, the functional discrepancy with severe paretic muscles is striking. The reason for this may be that the central motor pathways do not primarily project to the anterior horn cells of these luxury nerves.
During infancy, central motor programs for different movement patterns are developed, the quality of which is dependent on afferent impulses. In OBPP, serious sensory deafferentiation interferes with the development of motor programs, resulting in inadequate, abnormal movement patterns and MUPs without effective movement, as is frequently observed.
If misdirection of axonal outgrowth occurs after a nerve lesion, different muscles—agonists as well as antagonists—may receive reinnervation from the same nerve, which in turn may lead to abnormal nonfunctional movement patterns, such as cocontractions, but MUPs in the EMG are readily registered.
In the EMG of infants with OBPP, the recording of MUPs is not a reflection of functional muscle activity based on a developing central motor program but merely an indication of some axonal continuity from the spinal cord to the muscle.
Despite the restrictions of EMG in infants, according to some investigators, the combination of EMG with recording nerve action potentials from the median and ulnar nerves at the elbow after stimulation at the wrist may significantly add to the clinical assessment in determining the nature and level of a lesion. In individual cases, it is thus possible to predict the prognosis and the need for surgery.7,20 With this technique, a nerve conduction block or avulsion can be revealed.7,20 If a nerve action potential is comparable to the healthy limb, a lesion is not degenerative, but there is either a conduction block or avulsion in which the sensory ganglion has at least remained intact. In avulsions, the EMG usually shows signs of denervation. If denervation activity is lacking in the EMG, then the lesion is a prolonged conduction block. In such a case, surgery should be deferred because spontaneous recovery is to be expected. In combining EMG and nerve action potential recordings, some investigators have even been capable of distinguishing further between axonotmesis and neurotmesis.7,20
After electric stimulation of a peripheral sensory (e.g., the digital) nerve, somatosensory evoked potentials (SEPs) are recorded over the somatosensory cortex. A normal cortical response in OBPP indicates that a fair amount of sensory fibers in the brachial plexus and in the dorsal roots are conducting normally. However, it lacks qualitative, quantitative, and localization precision regarding the type and extent of a lesion. Direct root SEPs during surgery are more informative, but they nevertheless have comparable restrictions and do not offer information on the condition of the ventral, motor roots.17,20
Preoperative adequate evaluation of the functional integrity of anterior cervical spinal nerve roots by motor-evoked potentials (MEPs) in OBPP is not possible, as direct root registration is “not done.” Transcranial electrical MEPs with the usual registration of muscle activity by EMG is not diagnostic for single or multiple anterior root lesions because most muscles have a polyneural innervation.13,14,20 Only intraoperative MEP and direct registration at the nerve roots offer information on the proximal continuity of anterior spinal roots, but they also do not reflect the quality and recovery potential of an individual nerve root.21 The technique is not reported in babies with OBPP. Because of the relative lack of myelin in the central white matter in infants compared with adults, it is presumed that cortical depolarization by transcranial electrical stimulation is more difficult to achieve. Therefore, MEPs have not been introduced as a preoperative or intraoperative diagnostic tool in babies with OBPP.
In infants with OBPP who have severe enough lesions to meet the more or less accepted criteria for performing surgery, imaging is part of the preoperative analysis and is usually performed in or around the fourth month.1,2,3 As mentioned before, the key issue in severe OBPP is to reliably exclude or demonstrate preganglionic spinal nerve root lesions and, even more demanding, to show precisely the combinations of intact or avulsed anterior and dorsal roots as well as complete or incomplete nerve root avulsions. Hemilaminectomy and intradural inspection of root entry zones of relevant spinal cord levels offer the most reliable control for root avulsions but are only feasible in adults.22,23 Even with direct inspection and combined intraoperative, direct root SEP recordings, some uncertainty remains regarding the functional integrity or recovery potential of anatomically intact roots. In babies, such operative procedures are not performed, which is why a golden standard to assess the accuracy of any preoperative diagnostic study remains elusive.
As many associated lesions are described in OBPP, x-ray studies are of value in the newborn to show fractures or luxations of the cervical spine, humerus, or clavicle. Additional radiographs of the shoulder are of interest in a long-term follow-up scheme to assess osseous deformities of the glenoid fossa and humeral head.24
A chest x-ray, or ultrasound as an alternative, should be a standard procedure for the diagnosis of a diaphragmatic paralysis caused by a C4 or phrenic nerve lesion. To prevent postoperative respiratory problems, especially during bottle or breast feeding, plication of a paralytic diaphragm is recommended before exploring the brachial plexus and is even more important when intercostal nerve transfers are considered.25
In the past decade, computed tomography (CT)–myelography was the preferred diagnostic tool in OBPP in most centers. Many investigators experienced a higher reliability in the assessment of intradural root avulsions than with magnetic resonance imaging (MRI) techniques.22,25,26,27,28,29 Disadvantages of CT-myelography, however, include the need for anesthesia, intrathecal contrast application, and radiation. Several shortcomings of CT-myelography are also well known.22,23,28,29 Although contrast-filled pseudomeningoceles (Fig. 1), subarachnoid space deformity, or missing root shadows in CT-myelography are associated with root avulsions, they are known to show false-positive or false-negative results. Because there is a discrepancy between the root exit zone at the spinal cord and the corresponding intervertebral exit foramen, uncertainty about the determination of the correct spinal level may prevail even when using digital reformation techniques based on 1.5-mm transverse slices. An avulsion can only be ruled out when relevant anterior and posterior root shadows can be traced from the spinal cord to the exit foramina.25,29 In their intra- and extraforaminal course, CT-myelography fails to demonstrate ruptures or other types of lesions of spinal nerves.
MRI has become the preferred examination in many centers for imaging the brachial plexus in infants because of its noninvasive character.27,30 High-strength (at least 1.5 Tesla) MR machines create high-resolution images that enable visualization of different types of nerve lesions. Detection of nerve root avulsions and intraspinal nerve lesions is most valuable for treatment strategy, as mentioned. For imaging of intradural spinal nerve segments, three-dimensional (3D) constructive interference in steady-state MRI provides excellent-quality imaging.16,25 This is a heavy-weighted T2 sequence with a strong and constant signal for cerebrospinal fluid. With 3D fast spin echo T2 MR imaging with gray-scale inversion, fine-quality imaging can also be achieved.27 Studies of babies can be performed under mild sedation. Different MR techniques ideally even allow the demonstration of other types of lesions in the extraforaminal, peripheral brachial plexus nerves, such as scarring, edema, or formation of neuromas.30,31
Secondary deformities of the shoulder in OBPP can also be demonstrated with MR and are of interest in follow-up schemes. However, around the fourth month of age, these deformities rarely have direct orthopaedic surgical consequences and thus have limited value as preoperative diagnostics.32,33
MR studies avoid radiation, are noninvasive, can be performed under mild sedation in most infants, are less time consuming than CT-myelography studies, do not put a claim on anesthesia personnel, and are thus more cost effective.
Avulsions can be complete, affecting both anterior and posterior roots, or incomplete, with selective avulsion of either anterior or posterior roots or even of part of their contributing rootlets. Avulsions are often associated with pseudomeningoceles, which occur after disruption of nerve root sleeves, thus allowing cerebrospinal fluid to extrude from the subarachnoid space (Fig. 2A). Pseudomeningoceles can be isolated injuries with intact or at least not avulsed nerve roots (Fig. 2B). When root shadows of spinal nerves can be traced from the spinal cord to the respective exit foramina, there is no avulsion.22,25,29,34 If there is an interruption of a root shadow on an axial image, the coronal image should be checked for confirmation, because continuity is easily missed in the axial 2-mm slices. Avulsion is likely when images in both planes fail to demonstrate continuous roots. Demonstration of partial disruption of rootlets or avulsion of either an anterior or posterior root is incidentally possible. Sometimes recoil of a root, showing a blunt stump at the spinal cord, is suspected after preganglionic disruption in the absence of a pseudomeningocele (Fig. 3) In pseudomeningoceles, which are isointense to cerebrospinal fluid, delineation and tracing of root shadows can be difficult. Intact, avulsed, compressed, or scarred intradural spinal nerves are possible.
There are no specified reports on imaging of spinal nerve lesions from the dural exit and along their intervertebral intraforaminal course. Roots can be traced to their exit foramina on axial and coronal views, using 3D constructive interference in steady-state MRI or 3D fast spin echo T2, but a study on differentiation and correlation of lesion type is still lacking, as intraoperative control in this trajectory in babies (and even adults) is troublesome, requiring more or less extensive foraminotomies. Some conclusive, though not in all cases, additional evidence on continuity of a spinal nerve may be gained by intraoperative monitoring techniques (SEP and MEP), which have their own restrictions and limitations, as mentioned before.
Imaging of roots, trunk divisions, and cords is possible using high-strength MRI scans with T1- and T2-weighted different imaging techniques in axial, coronal, and sagittal planes individually adapted and planned parallel to the angle of the brachial plexus with additional use of reformatting techniques.30,31,35 Precise imaging of the extraforaminal brachial plexus requires scanning times of more than 1 hour, posing problems to maintaining protocols with only sedation, thus requiring anesthesia anyway. Most reports describe posttraumatic plexus injuries in older children and adults. Birchansky elegantly describes MR imaging in babies.30 Normal or even pathologic nerves are isodense to adjacent muscle but hypodense to surrounding fat. On T2-weighted and short tau inversion recovery images, the plexus is slightly hyperintense to muscle. Posttraumatic sequelae such as neuromas or fibrosis mainly affect the configuration of plexial nerves but do not usually alter signal intensity in T1 or T2 imaging. Some increased signal intensity in T2-weighted images may be observed. After complete or incomplete nerve ruptures, neuromas are formed. In a neuroma, the normal fascicular anatomy is lost and will be replaced by a thickened mass of disorganized proliferating axons and fibrous tissue, which can be recognized using MRI as a fusiform mass. Depending on the severity of the OBPP, the localization and number of nerve ruptures and consequent formation of neuromas may vary. Posttraumatic perineural fibrosis of the brachial plexus can be focal or diffuse. It may be recognized as thickening of plexial structures with ragged borders.30
Nevertheless, exact discrimination of all possible combinations of nerve lesions in OBPP is not possible with MRI. Although at present, demonstration of neuromas or fibrosis of the brachial plexus in infants with insufficient recovery of function would not generally influence treatment strategy, the more and more precise imaging quality adds to the understanding of different plexus lesions and may in the future have a growing effect on treatment decisions.
In adults, pathology of peripheral nerves can be studied by MR neurography—a technique enabling demonstration of nerve fascicles in cross sections of larger nerves in adults.31 As MR neurography is not suited to study smaller nerves, even those of sizes comparable with the trunks or cords in babies, it cannot yet be applied in infants with OBPP.
The newest generation of ultrasonography equipment allows imaging of supraclavicular plexus structures in infants. In individual cases, it is possible to demonstrate different pathology; for instance, neuromas or missing nerve roots. To check the accuracy of this imaging technique, preoperative images should be compared with intraoperative findings (Fig. 4). With only preliminary experience, valuation of the method is premature, and as yet, studies are not reported on this issue in OBPP.
In the developing field of treatment of OBPP, skillful and creative combined use of electrodiagnosis and imaging techniques not only allow characterization and differentiation of nerve lesions with ever-increasing precision but also may provide growing insight into largely unexplored central mechanisms of functional plasticity.