This article is more than 5 years old. Information may no longer be current.

Panel discusses nonoperative treatment, surgical intervention for children with brachial plexus palsy

Issue: October 2013

Add topic to email alerts

Receive an email when new articles are posted on

Please provide your email address to receive an email when new articles are posted on .

Subscribe

Added to email alerts

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

Back to Healio

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

Back to Healio

The primary risk group for brachial plexus birth palsy includes large babies, patients with a history of gestational diabetes, difficult delivery with shoulder dystocia and forceps delivery. The secondary risk group includes patients with associated clavicle and humerus fractures, with a differential diagnosis of pseudoparalysis related to fracture alone, sepsis or other conditions. There are few natural history studies, and the range of recovery is 50% to 90%. Newborns who recover at 1 month of life tend to have normal function. Recovery at 2 months and 3 months old has near-normal recovery. A negative modifier is the presence of a Horner’s sign. While there may be evidence of slight muscle function returning, the prolonged effect of neuromuscular imbalance may lead to significant bony changes and function limitations.

In this Orthopedics TodayRound Table, we convene a team of pediatric orthopedic upper extremity surgeons to share their opinions on nonoperative vs. early surgical intervention of brachial plexus birth palsy.

Alvin H. Crawford, MD, FACS
Moderator

Alvin H. Crawford, MD, FACS: If you are an orthopedic surgeon seeing a child with brachial plexus birth palsy (BPBP), under what circumstances and what age should you refer the child to a specialist?

Michael L. Pearl, MD: Ideally, every child with a brachial plexus birth injury should be evaluated by someone with a dedicated interest and experience in this problem. Perhaps the only exception would be an infant who spontaneously recovers within a few weeks. Even then, some children with early recovery (by 3 weeks) have been shown to manifest late sequelae that can limit motion. Children with persistent neurological deficits need surveillance to identify those who would benefit from neurologic surgery within the first year of life. After 1 year of age, and even as early as 5 months in some cases, many children develop contractures that can have profound consequences on skeletal development. Foremost among the problematic contractures is the internal rotation contracture of the shoulder (Figure 1a). Most of these contractures result in glenohumeral dysplasia (Figure 2). The child’s function (Figure 1b) and deformity (Figure 3) can be improved with early and appropriate treatment.

 

Figure 1a. Passive external rotation intraoperatively under general anesthesia demonstrated a fixed, internal rotation contracture of 30° to 40°.

 

Figure 1b. Active external rotation is shown at 2-year follow-up.

Images: Pearl ML:

Alain Gilbert, MD: The orthopedic surgeon, upon seeing a child with an obstetric brachial plexus (OBP) injury, should send the child immediately to the specialist, whatever the age and stage of recovery. It is important to see the child early as the evolution is important in the decision. It is generally necessary, except in severe cases, to see the child on two or three occasions before the decision is made.

 

Figure 2. Preoperative MRI shows a convex pseudoglenoid deformity with the entire humeral articular surface posterior to the centerline of the scapula (HH – humeral head, GL – glenoid).

 

Figure 3. Postoperative MRI at 2-year follow-up shows restoration of the glenoid concavity, improvement in the sphericity of the humeral head and normalization of the glenohumeral alignment.

Peter M. Waters MD: Age of referral is dependent on functional status of the infant, the practice patterns of referring physicians and the organizational structure of the subspecialty practice. In general, the plan for the first 6 weeks of life is initial rest and therapy. Infants who begin to recover function by 6 weeks of age have the best prognosis. They will progressively improve motor function from either an initial flail extremity or more commonly a classic “waiter’s tip” position (shoulder adduction, internal rotation, elbow extension and wrist and finger flexion), to active wrist and digital extension, elbow flexion and shoulder abduction. These infants will not need nerve surgery and most will recover fully by 6 months to 12 months of life. Most often these infants are managed by the primary physician.

Infants with a Horner’s syndrome and flail limb are most at risk for poor outcome. It is our practice to see these infants in the first month of life to examine the child and meet the parents. The decision making about nerve reconstruction is complex and ample time for discussion, question answering and independent research is important for the parents.

The infant who fails to recover any elbow flexion by 3 months of life needs to be seen then. These infants are at risk for failure of recovery of C5, C6 and even C7 function and frequently come to nerve surgery. It is helpful to examine the infant, review recovery and therapy, and answer questions before nerve surgery.

Roger Cornwall, MD: I prefer to see children as early as possible, typically around 3 weeks of age. It is easier to make decisions about surgical repair of injured nerve roots if I have had the opportunity to examine the child on at least two occasions to see the progression of recovery. A single snapshot at 3 months does not tell the whole story. Also, the indications for surgical exploration and reconstruction of the brachial plexus in infants are evolving, so there is not one simple rule of thumb (e.g., absent biceps at 3 months of age) that can suffice in the decision to operate and thus, the decision to refer to a specialist.

We have learned in recent years that secondary problems such as contractures and shoulder dysplasia begin shortly after birth, and developmental dislocation of the shoulder can happen as early as 3 months of age, even in children who do not meet conventional criteria for surgical exploration and repair of the brachial plexus. I believe early specialty evaluation of these children is important, even if it means we see some children who resolve completely and need no intervention.

Scott H. Kozin, MD: The timing and referral for an infant with BPBP is dependent upon the extent of injury. In children with global injuries, we recommend immediate referral as microsurgical intervention is more likely. In children with BPBP affecting the upper trunk (aka Erb’s palsy), referral can be delayed. We recommend that if the child has not recovered from elbow flexion by 2 months, then referral is required. The natural history of brachial plexus birth palsies has been well defined over the last decade. In children who have recovery of elbow flexion within the first 2 months, then complete resolution is likely and the injury has been classified as a neurapraxia. In children who recover elbow flexion between 3 months and 6 months, the injury is axonotmesis and residual impairment is likely. Lastly, in children who do not recover elbow flexion by 6 months or so, then microsurgical intervention is indicated in children with BPBP.

Figure 4. Shoulder passive range of motion into external rotation with the scapula stabilized is shown.

Images: Kozin SH

Crawford: How has recent knowledge of the effects of the nerve injury on shoulder development changed the approach to the care of the child with a brachial plexus birth injury? What are the roles of physical and occupational therapy?

Pearl: Nerve injury affects the development of the shoulder by altering the normal balance of forces that surround the growing glenohumeral joint. The subsequent development of contractures imposes additional aberrant forces on the various growth centers. There is also a generalized diminution of growth of the entire extremity from the neurological insult. A recent hypothesis posits that contractures result from less growth of one opposing muscle group over another. My experience suggests otherwise, but either way, the result is the same and dedicated stretching and splinting can help lengthen the shortened soft tissues. Nonsurgical modalities, unfortunately however, have definite limits. I have seen the most dedicated parents, even parents who happen to be trained physical therapists, make only modest improvements to contractures that have a neurological origin. One should also not underestimate the difficulties in enlisting a toddler in such treatment. Once children have reached an age where they can understand the purpose of the exercises and have a vested interest in achieving their goals, gains are more likely. This varies from child to child and may not appear until adolescence.

Gilbert: The effect of imbalance in shoulder muscle recovery has been known for some time. We had the idea that excessive internal rotation was a cause of joint dystrophy and posterior subluxation. It was treated early with physiotherapy and sometimes splints, and if this failed (and it failed often), by surgical release after 18 months. We have learned recently the importance of treating early imbalance (essentially with onabotulinumtoxinA [Botox, Allergan] and the necessity to keep a good relationship between the humeral head and the glenoid to maintain a good anatomical position.

Waters: It is clear the shoulder joint is at risk of deformity and dislocation if there is failure of recovery of balanced active movement about the shoulder and restoration of full passive motion of the glenohumeral joint. The risk is high and the onset is early in life, as young as 6 months to 18 months of age. Prevention of an internal rotation and adduction contracture is critical. Formal therapy is important, as are daily home exercises. In children who fail to maintain or regain full passive motion, consideration of Botox injections, spica casting or bracing, or operative release is important before the development of shoulder joint deformity and dislocation.

Cornwall: The development of the shoulder joint depends on normal muscle growth, development and function, which in turn depend on normal nerve function. Thus, disruption of nerve function in the newborn period can lead to dysplasia of the shoulder similar to neuromuscular dysplasia of the hip in cerebral palsy. This process begins soon after birth and can lead to dislocation of the shoulder as early as 3 months of age. While it used to be thought that such an early dislocation was the result of birth trauma, it is becoming clear that the dislocation results from abnormal growth and balance of the muscles that stabilize the humeral head in the glenoid. Thus, early physical and occupational therapy aimed at improving strength and maintaining length of denervated, tight muscles is important to prevent potentially permanent secondary musculoskeletal problems.

Kozin: The expansion of knowledge regarding shoulder development has changed our approach to the management of children with BPBP. The shoulder has a propensity for internal rotation compared to external rotation. This propensity is exaggerated in children with BPBP, especially those injuries affecting C5, C6 and/or C7. For the shoulder to develop in a normal fashion, there must be internal and external rotation. This is analogous to the development of the hip.

The role of physical and occupational therapy is to maintain the passive range of shoulder motion. The passive range of motion must be performed with the scapular stabilized. This movement promotes normal glenoid and normal humeral head development. Failure to maintain external rotation past neutral results in glenohumeral joint dysplasia. This is defined as glenoid retroversion and posterior humeral head subluxation.

We stress to the therapist that the main role of therapy is to maintain the supple joint. A supple joint will respond to early re-innervation of muscles and allows active movement. Contracture of the joint needs to be avoided as this will lead to glenohumeral joint dysplasia. Nerve recovery after contracture is difficult to detect and the reinnervated muscles cannot move the contracted shoulder joint.

Crawford: What major types of surgical intervention can help children with BPBP recover function? How do you decide when to do them?

Pearl: One of the confusing aspects in explaining treatment options to parents and sometimes colleagues is that even though the problem began as a neurologic injury, neurosurgical options are limited both in the timing of their indications and efficacy. Hence, there is a broader role over a longer time frame for musculoskeletal surgery options. For the severely affected infant, reanimation of the limb with nerve transfers and neurotization procedures can produce some functions for the limb that will have no potential for natural recovery. Patients with root avulsions or complete post-ganglionic ruptures will not regain muscle function in the target muscles of these nerves and need early consultation with a surgeon who can perform these procedures. After the first year of life, neurological procedures are less indicated because they lose their effectiveness and consideration turns toward releases and muscle transfers.

The most common procedures of this type are directed toward releasing the internal rotation contractures of the shoulder. Various methods have been described, with and without associated transfer of the latissimus dorsi and teres major. My preferred approach is a trial of Botox injection for the incipient internal rotation contractures in the first 2 years of life, then arthroscopic release of the joint capsule (and tenotomy of the subscapularis when necessary to effect the release) for children who fail injection or present too late for this treatment. For children older than 3 years with deficient active external rotation or those who failed a primary release, I add transfer of the latissimus dorsi tendon to the posterior rotator cuff (at the insertion point of the infraspinatus tendon). Other procedures I find effective are rotational osteotomies of the humerus for older children (older than 8 years) with advanced glenohumeral dysplasia, and of the radius for children with excessive and fixed supination. For children with more distal involvement of the wrist and hand, tendon transfers can help augment wrist and finger, flexion and extension.

Gilbert: There are different situations. If at 3 months, the child has a complete palsy, or at least a complete hand paralysis, with a Horner’s syndrome, then it is an indication for repair with grafts associated sometimes with neurotizations. If at 3 months to 4 months the child has a C5-C6 or C5-C6-C7 palsy and has not recovered wrist extension and triceps, I will do a graft. If the child has a wrist extension and triceps, it is an option to do a triple neurotization (suprascapular, axillary, musculocutaneous). In late cases (up to 2 years old) where there are no external rotators, I will do an accessory to suprascapular nerve transfer. In late cases of complete palsies with good recovery of the shoulder and elbow and no hand, I now do a contralateral C7 transfer by the prespinal route. But in all these cases, the patient should be operated.

Waters: The options for surgery are in two major categories: primary nerve surgery and secondary tendon transfers, joint reductions, and/or osteotomies. Nerve surgery is indicated for two typical situations: avulsion in which prognosis is poor and there is high risk of limited long-term hand function; and injuries in the plexus in which too much neural disruption has occurred to heal with healthy anti-gravity motor function. Shoulder abduction and external rotation, elbow flexion, and at times wrist and digital extension are the most at risk movements. Avulsion injuries have nerve surgery ideally by 3 months and usually involve a combination of nerve grafts from any viable nerve roots and nerve transfers from functioning expendable nerve fascicles. Extra-foraminal ruptures have similar nerve reconstructions usually at 5 months to 6 months of life, and most of us are using more nerve transfers now when we can. Nerve transfers have the advantage of motor-to-motor nerve direct connection close to the motor endplate. Recovery is faster and specific to the peripheral nerve repaired.

Secondary surgery usually involves reconstruction about the shoulder. An important discriminating factor is the status of the glenohumeral joint, which can be assessed by ultrasound, arthrogram, CT or MRI scans. We most commonly use MRI scans of the glenohumeral joint for surgical decision making regarding joint reduction, tendon transfers and lengthenings, or osteotomies. We intervene as soon as the joint clearly shows signs of deformity and subluxation that has not responded to nonoperative means. We most often perform a combined open reduction of the joint and tendon rebalancing through an axillary incision that is extensile and aesthetic with healing. Surgery is often between 18 months and 3 years of life.

Cornwall: The two main categories of surgical intervention involve treatment of the nerve injury and treatment of the secondary effects of the nerve injury. To treat the nerve injury, excision of the injured nerve roots and reconstruction with sural nerve grafts remains the gold standard. However, distal nerve transfers are gaining popularity, whereby uninjured nerves are detached from their target muscle and sutured to the distal end of the injured nerves as they enter the paralyzed muscle. For instance, shoulder abduction and external rotation can be restored by transferring the spinal accessory nerve to the suprascapular nerve and transferring a triceps branch of the radial nerve to the axillary nerve. Minimal function is lost in the donor muscle, and overall function is improved, although motor relearning is required much like after a tendon transfer. Most importantly, nerve transfers can be performed later than nerve grafting, as the nerve anastomosis is performed more distally and thus has a shorter regeneration time to get to the muscle before irreversible muscle atrophy takes place.

Many surgical interventions are designed to address the secondary problems, including surgical release of contractures, reduction of shoulder dislocations, tendon transfers to augment shoulder and hand function, and osteotomies to reposition the arm and hand in space if muscle function or joint range of motion cannot be improved. The decision between different types of surgical intervention is based on the age of the child, the degree and distribution of the neurological deficit, and the presence of secondary problems. However, the standard mantra of nerve surgeries in the first year of life and secondary surgeries later no longer holds true, as surgical reduction of a dislocated shoulder may be performed prior to surgical nerve reconstruction by nerve transfer. Therefore, it is helpful for the physician or team to consider the problem comprehensively from the beginning.

Kozin: There are two major types of surgical intervention. The priority is nerve surgery that is time-dependent and performed in children within the first year of life. The parameters that guide intervention are the natural history of recovery and the basic nerve and muscle reinnervation issues. From a nerve standpoint, the nerves regenerate somewhat between 1 mm and 3 mm a day. From a muscle standpoint, irreversible motor endplate demise is established between 18 months and 24 months following injury. Therefore, primary surgery must occur in a timely fashion to allow adequate time for nerve regeneration to prevent motor end plate demise. Therefore, global injuries require earlier surgery than injuries delegated to the upper and/or middle trunk.

Secondary surgical intervention is defined as procedures that do not involve the nerve. These interventions include joint reduction, muscle/tendon transfer and osteotomy. Joint reduction is primary centered on the shoulder joint. A shoulder joint contracted in internal rotation will result in glenohumeral joint dysplasia. Once there is established deformity, then open joint reduction is necessary. Early open joint reduction is better than delayed joint reduction. Early joint reduction harnesses the power of youth and allows for joint remodeling. We do not know when a child is too old for open reduction. However, our results indicate that joint reduction prior to 2 years old performs better than those performed after the age of 2 years. Muscle/tendon transfers are valuable about the shoulder joint. The standard transfer of the latissimus dorsi and teres major promotes overhead motion and external rotation. There were numerous studies that report positive outcomes following this tendon transfer. Other secondary procedures such biceps rerouting and humeral osteotomy have been beneficial. The exact procedure needs to be individualized to the particular patient, with respect to the status of the joint, muscles and tendons available and patient age.

 

Figure 5. The left shoulder undergoes transfer of the latissimus dorsi and teres major tendons for external rotation.

 

Figure 6. A humeral osteotomy with plate and screw fixation is shown.

Crawford: Where are the frontiers in BPBP research? How might they change our treatments in the future?

Pearl: The Holy Grail in the treatment of BPBP is the ability to regenerate nerves. Unfortunately, this frontier may be a ways off. One area for which the technology is robust but has yet to be best used in my opinion, however, is in advanced imaging. This applies equally to the imaging of the injured brachial plexus and to the secondary consequences to musculoskeletal tissues. One would hope that we will soon be able to define the extent of the neurological injury and delineate its prognosis from MRI/fMRI and other advanced imaging technology. The status quo of monthly neurological exams waiting for biceps function to return seems crude in the context of modern imaging.

Similarly, if we could make better use of 3-D modeling of the skeletal structures, then we could enhance our understanding of the deformities and contractures, enabling a more nuanced approach to our patients’ musculoskeletal deficits. For example, not all internal rotation contractures of the shoulder are the same. Some present in the context of a pseudoglenoid and others do not. Yet we tend to treat these different entities the same. A better understanding of the relative contributions of the joint incongruence and the shortened soft tissues to the contracture will allow a more targeted approach to treatment.

Gilbert: There is no specific direction of research in OBP — all concern the nerve recovery. The most important is awareness. For 30 years, I have seen more sophistication in the understanding and treatment, but there is a growing reluctance in the medical community to accept the scientific results. The consequence is now the larger number of untreated patients who have severe sequelae not because they were not previously seen, but because somebody from the medical community told them to wait and gave them hope of spontaneous recovery.

Waters: The recent animal models are fine-tuning our understanding of the glenohumeral joint deformity associated with a BPBP and focusing our indications and treatment options with secondary reconstructions. In addition, clinical effectiveness human studies, particularly long-term prospective multicenter analytics, are necessary. Our Treatment and Outcomes Brachial Plexus Injuries (TOBI) study should prove beneficial to deepen our understanding of the indications and outcomes of primary and secondary reconstructions. The TOBI study is a prospective multicenter study initially funded by the American Society for Surgery of the Hand and the Pediatric Orthopaedic Society of North America. Researchers in the TOBI study are now presenting and publishing initial data with follow-up of 1 year to 5 years on more than 800 enrollees.

We still need a new frontier in nerve reconstructions. The severity of some of these injuries, and the limits of viable nerves for reconstruction, make the results in multiple level avulsion injuries less than ideal. The basic science laboratory will hold the key for improving neural recovery, especially to the hand, and improve the quality of life for these severely effected children.

Cornwall: Several major questions remain unanswered. First, the best solution to the problem of BPBP is prevention, yet identifiable risk factors for the injury are present in half the patients. We need a better understanding of the pathogenesis of the condition in order to maximize preventative strategies. Also, a debate currently rages about the optimal timing and type of nerve reconstruction. Theoretically, immediate reconstruction of the injured nerves could prevent many of the secondary problems in muscles and joints, but it is currently impossible to know at birth which children will fail to recover spontaneously.

Research regarding imaging and other diagnostic modalities may allow early detection or prediction of permanent injuries. Finally, the interaction between nerve function and muscle and skeletal development is complex and incompletely understood at a biological level. It is possible that research into the biological interactions between nerves and muscles during the neonatal period could allow the development of medical treatments to preserve normal muscle growth and development following neonatal nerve injury.

Kozin: There are two main areas of research that will directly impact the care and treatment of children with BPBP. The first area is nerve-oriented. Research will find ways to enhance nerve regeneration and that will distinctly change outcome, as nerve regeneration can occur quicker and earlier with prevention of irreversible end plate demise. Nerve research will also identify the extent of injury in the newborn. Currently, there is no way to discriminate between axonotmesis and neurotmesis imaging studies or testing. Once we are able to define a neurotmesis, then early intervention will be performed which will simply the algorithm.

The other area of research is in muscle. Cornwall and colleagues have shown that pediatric muscle reinnervates differently than adult muscle. The muscle tends to “tighten” which limits motion and promotes contracture, even after reinnervation. Research will find a way to prohibit this reinnervation property of pediatric muscle. In addition, research will find ways to prevent irreversible and plate demise. Therefore, ongoing re-innervation can occur for a longer time which will promote better outcomes.

Crawford: I would like to thank the panel for taking the time from their busy schedules to assist me in providing this very stimulating and up-to-date information on BPBP for our readers. I look forward to the results of the TOBI study to illustrate imaging capabilities to define nerve and muscle involvement and the results of direct and indirect surgeries to improve motor function and prevent deformity.