There continues to be controversy regarding the care of children with brachial plexus birth palsies. The indications, timing, and type of microsurgical intervention for infants with an acute plexopathy have yet to be clearly defined. In addition, the indications for tendon transfers and osteotomies in a child with a chronic plexopathy and permanent weakness are com-plicated and continue to be debated. Over the past several years, the Hand and Upper Extremity Program at Children's Hospital has been very active in further defining the optimal care of children with these complex palsies.

Background

Infantile Evaluation and Care

Brachial plexus birth palsies occur in 0.1% to 0.4% of live births (1,2) . Perinatal risk factors include large birth-weight, pro-longed labor, multiparous pregnancy, and difficult delivery. Shoulder dystocia in a vertex delivery or difficult upper extremity extraction in a breech delivery increases the risk of neural impairment (3) . Fetal distress with relative muscular hypotonia and thus decreased protection of the plexus may also con-tribute to the development of a plexus injury.

Most infantile injuries to the brachial plexus predomi-nantly involve the upper trunk (C5-C6), the classic Erb's palsy. However, many of these infants also have impairment of the C7 nerve root and this portends a poorer prognosis. Far less frequently, the entire plexus (C5-T1) or the lower trunk (C8-T1, a Klumpke's palsy), may be involved.

The majority of infantile injuries are post-ganglionic stretch injuries. Avulsions of the nerve roots from the spinal cord can occur but are less common. The severity of mechanical stretch and the biologic healing of the neural injury ultimately determine outcome. Unfortunately, both of these are difficult to determine at birth and during the first several months of life. So far, special radiologic studies, such as myel-grams, computerized tomography (CT) – myelograms, or magnetic resonance imaging (MRI) scans, and neurophysiologic studies, have had limited success in predicting long-term outcome.

Serial clinical exams of the infant during the first six months of life have been most frequently utilized to predict Serial clinical exams of the infant during the first six months of life have been most frequently utilized to predict outcome. The timing of recovery of spontaneous shoulder, elbow, wrist, and finger motion is monitored. Neonatal reflexes (e.g. Moro/startle, asymmetric tonic neck) are tested to induce elbow flexion, as well as wrist and digital extension. The presence or absence of a Horner's syndrome –ptosis, miosis and anhydrosis—is also noted. (4)

Gilbert was the first to emphasize the importance of the recovery of biceps function as an important prognostic sign of long-term outcome. In his original work with Tassin and subsequently with Meyer and Whitaker (5,6,7) , Gilbert demonstrated that failure of the biceps to recover by three months of life portended less than normal upper trunk function beyond two years of life. Clarke (8) found that utilizing the return of biceps function alone in predicting poor outcome had a 12% inaccuracy rate. By combining the return of elbow flexion with wrist, digital, and thumb extension, this error rate was reduced to 5%. The presence of a Horner's syndrome indicates a more severe injury with less chance of spontaneous recovery. Most micro-surgeons still rely on physical examination findings to identify appropriate candidates for brachial plexus exploration and reconstruction.

There continues to be debate on the appropriate timing of microsurgical treatment for infantile brachial plexus palsy, with the published recommendations ranging from one month to 18 months of life for infants without signs of recovery. (5,6,7,8,9,10,11,12) Though most surgeons advocate exploration and reconstruction in the first three to six months of life, there is significant disagreement on the optimal timing for microsurgery within this age range. Limited natural history data and the lack of uniform methods of assessing these infants have made this issue difficult to resolve.

In terms of surgical treatment, there is no role for neurolysis alone. Successful surgery requires nerve grafting or nerve transfers. In the case of an upper trunk rupture, grafting is performed from the proximal viable C5 and C6 nerve roots to the unaffected nerve distally. This is most often performed with separate grafts to the suprascapular nerve, lateral cord, and posterior cord. In cases of avulsions, nerve transfers are performed utilizing the intercostal nerves or branches of the spinal accessory nerve (cranial nerve XI).

Chronic Plexopathy

Many infants who fail to recover upper trunk function in the first three months of life are left with residual deficits. Usually this involves limited shoulder abduction, forward flexion and most significantly, external rotation. These children are impaired in above shoulder activities of daily life, such as facial hygiene and hair care, or recreation, such as ball throwing and swimming. Surgical interventions have included tendon transfers, osteotomies, and glenohumeral fusions. (13,14,15) The specific indications for these procedures have, in past literature, been poorly defined, causing difficulty in matching the correct procedure with the appropriate patient. The most difficult decisions have been in choosing between a latissimus dorsi/teres major tendon transfer or a humeral derotational osteotomy in the patient with an internal rotation contracture and external rotation weakness about the shoulder. Most surgeons have agreed in principle that younger patients do well with tendon transfer and older patients benefit from osteotomy.

Our Recent Contributions

Natural History and Microsurgical Indications

Since joining the Department of Orthopaedic Surgery at Children's Hospital, I have been actively involved in clinical research on infantile brachial plexus birth palsies. A database has been created on children who have presented for evaluation and treatment of acute or chronic plexopathy. During each visit, every patient has a thorough clinical evaluation, including a Mallet score of upper extremity and hand function.

Overall, 66 infants with 67 lesions of the brachial plexus noted at birth were evaluated with greater than two years follow- up. This information was the source of a recent Journal of Bone and Joint Surgery publication in May 1999 entitled “Comparison of the natural history, the outcome of microsurgical reconstruction, and the outcome of operative reconstruction in brachial plexus birth palsy.” (16) The infants were grouped according to the age at which biceps function first occurred. Patients with biceps recovery during the first six months of life were treated with physical therapy and followed conservatively. Any infant without biceps recovery by six months of life was treated with microsurgical reconstruction of the plexus, with either exploration and sural nerve grafting or intercostal nerve transfer. All groups were compared for outcome results utilizing the Mallet classification.

Infants with recovery of biceps function within the first two months of life had normal function at greater than two years follow. Infants with recovery at three, four, five, and six months of life had progressively lower functional scores when after two years of age. None of the infants with recovery at four, five, and six months of life were normal in function longterm. Children who underwent microsurgery at age 6 due to lack of biceps return were more functional than similar infants who recovered biceps function between the fifth and sixth months of life. (16)

Many of the children followed beyond two years of life were noted to have shoulder dysfunction. This was particularly true for shoulder abduction, forward flexion, and external rotation. Both a tendon transfer (i.e. latissimus dorsi and teres major to the rotator cuff) and a humeral derotational osteoto-my significantly improved function in respective patients.

The conclusions of this study were as follows: 1) infants with biceps recovery by six weeks of life should not need longterm care or evaluation because of normal subsequent function; 2) infants with biceps recovery between two and five months of life should have regular follow-up; 3) infants with no biceps recovery by five months of life should have microsurgical exploration and neural reconstruction; and 4) those children with internal rotation contractures and external rotation weakness about the shoulder benefit from tendon transfer or humeral derotational osteotomy. It is still unclear from this and other studies whether microsurgery should be performed before five months of life. Although this is common practice in some parts of the world, published data to date does not clearly indicate whether surgery before five months of age is indicated. Unfortunately, the “need” for microsurgery at three months of life or less is being aggressively advocated and advertised by some centers. The internet has become a powerful tool for this form of communication. The lack of consensus has made this confusing for families and frustrating for primary caregivers. Many parents are pressured to act in the “best interest” of their infant by consenting to early microsurgery.

In attempt to increase communication and establish a more uniform approach to infantile brachial plexus injuries, Dr. Vincent Hentz, former President of the America society for Surgery of the Hand, Drs. Borrereo, Gilbert, Meyer and I organized an international one-day discussion group at the 1999 annual meeting of the Hand Society in Boston, MA. Twenty-six attendees presented their work. Although a final consensus was not reached, all attendees were open to sharing their results in a common database if proper security and intellectual copyrights could be maintained. Dr Gilbert's group in Paris has been attempting to develop an internet-based data system. I have been working initially with a group at the Royal Children's Hospital in Melbourne, Australia with a similar goal. Hopefully these efforts will lead to a more uniform classification and data collection and allow for comparative analysis of results between different centers. Ideally, these efforts will lead to a randomized, prospective, multi-center study that can bet-ter ascertain the optimal timing of microsurgical intervention.

Secondary Reconstruction of the Shoulder

Children with chronic upper trunk plexopathy often develop an internal rotation contracture associated with external rotation weakness of the shoulder. The etiology is persistent muscle imbalance due to incomplete recovery from the infantile neural injury. It has been my impression that any persistent muscle imbalance in a growing child should also lead to bone and joint deformity. In conjunction with Dr. Diego Jaramillo of the Department of Radiology at Children's Hospital, we have developed a radiologic protocol to assess the glenohumeral joint by MRI and CT scans in children with chronic plexopathy. Children less than age five years of age had a MRI scan and children older than five years of age underwent CT scan. It was felt that this would give the most reliable structural information about the developing glenoid. This decision was based on a previous study of CT and MRI scans obtained for the evaluation of lung and mediastinal tumors and metastasis in children from the Children's Hospital and the Dana Farber Cancer Institute. That initial study, performed with Dr. Craig Mintzer during his orthopaedic residency at Harvard, defined normal glenoid version values according to age. (17) As a part of a long-standing research protocol, all children under consideration for reconstructive shoulder surgery obtain a preoperative CT or MRI scan to assess glenoid version, the degree of glenohumeral subluxation, and the extent of glenohumeral deformity as compared to the unaffected side.

Analysis of the initial radiographic data was published in the Journal of Bone and Joint Surgery in 1998 with Drs. Jaramillo and Garth Smith, a former Harvard hand surgery fellow, in an article entitled “Glenohumeral deformity secondary to brachial plexus birth palsy.” (18) The degree of glenohumeral deformity was progressive with age. The type of progressive deformity was reproducible and therefore, classifiable. Deformity progressed from type I (normal) to type V. A type II deformity was defined as increased glenoid retroversion by >5 degrees; type III as increased glenoid retroversion and posterior glenohumeral subluxation; type IV as the development of a false glenoid from progressive glenohumeral dislocation; and type V as the development of a deformed humeral head with flattening. Type VI and VII described special cases of infantile dislocation and proximal humeral growth arrest, respectively. The classification of deformity for a particular child with an internal rotation contracture and external rotation weakness was used to determine the appropriate procedure for the child. Children with types I-III were considered candidates for latissimus dorsi/teres major tendon transfer to the rotator cuff insertion at the greater tuberosity. Children with type V defor-mity were treated with a humeral external rotational osteotomy. Patients with a type IV deformity were treated with either a tendon transfer or an osteotomy depending on their age and degree of glenoid deformity. Younger patients, with glenoids that had the capacity to remodel, underwent tendon transfer. Humeral derotational osteotomies were performed in older patients with marked type IV deformity.

Thirty-six children treated with either a tendon transfer or humeral derotational osteotomy for internal rotation contracture and external rotation weakness were reevaluated at minimum two-year follow-up to determine the results of surgical intervention. This data was published in Clinical Orthopaedics and Related Research in a 1999 article entitled “Shoulder reconstruction in patients with chronic brachial plexus birth palsy.” (19) This study was co-authored with Dr. Allan Peljovich, a former Harvard hand surgery fellow. The study demonstrated that both tendon transfer and humeral derotational osteotomy significantly improved function in these children. By comparing pre-operative and post-operative Mallet scores, both operations improved function and external rotation. Shoulder reconstruction is thus strongly advocated for older children with impairment.

It is interesting to speculate on whether rebalancing the muscle forces early can prevent progressive deformity of the glenohumeral joint. In addition, the potential for glenohumeral joint remodeling is unclear. Patients with tendon transfers performed early in life are currently being followed with serial MRI and CT scans to assess the remodeling or progression of deformity. Early analysis suggests that the deformity does not progress and that the glenohumeral joint has the ability to remodel with surgical rebalancing of the muscle forces about the joint.

Future Considerations

We will continue to collect and analyze data from patients with brachial plexopathies in a prospective fashion at Children's Hospital. However, there are limits to the study of any problem at one institution. Hopefully, improved communication and more uniform classification systems will lead to more collaborative efforts and more comparative data analysis amongst different centers. Ideally, this will lead to a better understanding of the best time for microsurgical reconstruction of the injured infantile brachial plexus. In addition, in the present era of aggressive marketing and economic considerations, it is imperative to define clear indications and outcomes for surgical procedures.