Management of obstetrical brachial plexus palsy

Clin Plastic Surg 30 (2003) 289 – 306

Management of obstetrical brachial plexus palsy Evaluation, prognosis, and primary surgical treatment Jeffrey R. Marcus, MD, Howard M. Clarke, MD, PhD, FRCS(C)* Division of Plastic Surgery, Hospital for Sick Children; University of Toronto, 555 University Avenue, Suite 1524 Toronto, Ontario M5R1X8, Canada

Obstetrical brachial plexus palsy is a traction neural injury sustained during the course of the birth process. As is the case in any closed peripheral nerve injury, severity can fall within a wide spectrum and is the key determinant of prognosis and the need for intervention. Ascertaining the severity of a single injured peripheral nerve at the time of presentation is difficult enough; the challenge is far greater when the variable injury involves an intricately patterned and incompletely understood group of nerves. The management of obstetrical lesions of the brachial plexus in the newborn is rich with dissenting opinions. Microsurgical treatment of obstetrical brachial plexus lesions is a relatively young field of surgical expertise. Surgeons engaged in the early effort to improve functional outcome for infants with this potentially lifelong impairment quickly recognized the need to establish appropriate surgical criteria. The determination of surgical indications continues to be the most important, and most difficult, consideration for those caring for patients and families faced with the physical and emotional impact of this condition. The factor that is most responsible for disparity among experts in this regard is the relatively incomplete understanding of the natural history of brachial plexus palsy. In this article, the initial approach to the patient with brachial plexus palsy, the natural history of the condition, the primary surgical treatment, and the expectations for outcome are discussed.

Background and initial approach In modern obstetrical practice, the incidence of obstetrical brachial plexus injury is estimated at 0.5 to 2 per 1000 births [1,2]. Included in this estimate are those who progress expeditiously to complete recovery and those who may improve slowly but incompletely. In reported series, there is wide variation in the rate of complete, spontaneous recovery (30% to 95% in the literature) [3 – 6]. Lower reported rates are reflective, in part, on preselection; those who recover quickly and completely often are not referred to specialists and therefore may be excluded from analyses. In addition, there is some lack of consistency in the definition of ‘‘recovery.’’ There is consensus that a varying degree of improvement is the rule in the majority of cases. However, although many infants with plexopathy recover with minor or no residual functional deficits, a number of children do not regain sufficient limb function and develop functional limitations, bony deformities, and joint contractures. Birch et al stated further that palliative procedures for secondary deformities are unsatisfactory and that the results of musculotendinous transfers in obstetrical palsy are far inferior to those following good nerve regeneration [7,8]. The challenge of statistical prognostication is to identify children most likely to derive benefit from exploration and reconstruction of the plexus. History

* Corresponding author. E-mail address: [email protected] (H.M. Clarke).

Upon presentation, a detailed history is taken, addressing maternal factors, prenatal course, the events of the birth process, and the postnatal course.

0094-1298/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved. PII: S 0 0 9 4 - 1 2 9 8 ( 0 2 ) 0 0 1 0 0 - 1

290

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

An example of the historical data reference in use at the authors’ institution is found in Fig. 1. For the purposes of data collection, historical items in this format have been separated into two categories: maternal factors and child factors. Maternal factors include data pertaining to the prenatal history, such as the incidence of maternal diabetes, pre-eclampsia, duration of labor, and previous obstetrical history. Also included is the natal history described by the parents, such as delivery presentation, use of instrument assistance, and incidence of shoulder dystocia. Child factors include such factors as gestational age, weight, and incidence of complica-

tions—namely, respiratory complications, fractures, diaphragm paralysis, and shoulder subluxation. It is important to document the aforementioned historical data for the purposes of population study. For some centers that receive a larger number of referrals, internal databasing has allowed the analysis and reporting of relatively large series to give us our most current, yet still incomplete, understanding of the disease process. For all centers, and particularly for a smaller center, it may be helpful to participate in currently available, web-based, multicenter data collection programs. Examples include the multicenter database sponsored by Children’s Hospital of Boston and the

Fig. 1. The Hospital for Sick Children Obstetrical Brachial Plexus Lesions database form.

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

Brachial Plexus web site hosted by the Saudi Arabian National Guard Heath Affairs (www.ngha.med.sa/ bp/). These databases compile historical information and serial examination data. With regard to history, attention is given to factors with reported statistical associations with obstetrical brachial plexus injury. The factors of interest are those with suspected causal relationships to the injury, those that might indicate high level of energy at the time of injury, and those that may be suggestive of the anatomic level of injury. From the literature, several associations have been consistent. The average birth weight of affected infants is reported to be 500 to 1000 g heavier than unaffected infants [1]. The reported association of maternal diabetes with obstetrical lesions is likely the result of its effect on birth weight. In addition, subsequent to the first pregnancy, birth weights tend to increase with each pregnancy in succession. Fetal macrosomia is another condition that, although cited separately, refers to the setting of higher average birth weight [9]. In a general sense, these conditions share the same potential delivery-related difficulty: cephalopelvic disproportion [10]. Shoulder dystocia is perhaps the most common association, and the mechanism of vertex brachial plexus injury is most easily understood in this context. It is generally accepted that injury to the brachial plexus occurs as a result of lateral traction on the head of the infant away from the shoulder during the last phase of delivery [1,10 – 12]. Other reported associations include the incidence of brachial plexus injury in previous pregnancies and the use of vacuum or forceps assistance during delivery. It is important to determine the incidence of concurrent delivery-related injuries, such as fractures or dislocations. Some have felt that such injuries provide meaningful clinical suggestions regarding the applied energy at the time of injury. Fracture of the humerus may be indicative of significant applied energy. Humeral fracture may also be associated with pseudoparalysis, in which neurologic impairment occurs as a result of direct impingement of the plexus by the fracture itself. Clavicular fracture has also been associated with obstetrical plexus injury [4], but some suggest that it may actually be protective to a certain degree [12], allowing downward rotation of the shoulder/scapular complex in the setting of dystocia. Shoulder dislocation can also occur with or without an injury to the brachial plexus. The literature has failed to provide a statistical association of prognosis with the incidence of humeral or clavicular fracture [13]. Injury to the brachial plexus is most frequently reported in the setting of vertex delivery but has also been cited in breech delivery and cesarean section.

291

Metaizueau investigated the effects of traction on the brachial plexus in stillborn infants [12]. By immobilizing the shoulder and applying lateral traction to the head, force requirements and patterns of injury were demonstrated. Epineurial rupture occurred at a lower force level than complete rupture. Rupture was more common in the upper roots and avulsion was more common in the lower roots. Metaizueau’s findings coincide with clinical observations and injury patterns in vertex delivery with plexopathy [11]. In the setting of breech delivery with plexopathy, a lower than normal birth weight is more common, as opposed to the converse seen in vertex delivery. The characteristic injury pattern tends to be more severe, particularly with regard to the upper roots. Avulsion of C5 and C6 is frequently seen [14 – 16]. The characterization of brachial plexus injury in the setting of breech presentation has been most clearly documented by European groups. Relatively less experience with breech-related plexopathy is seen in North America, perhaps due to the greater predisposition toward cesarean section under such circumstances. However, authors have noted the incidence of brachial plexopathy after cesarean section [5,11,17]. In reported post-cesarean cases that proceed to exploration, clinical evidence of brachial plexus trauma with characteristic neuroma formation is described. Urgent cesarean section may follow a difficult initial attempt at vaginal delivery. In such cases, the same injury mechanisms occurring in vertex vaginal delivery may still have taken place. The perceived inconsistency of injury after cesarean section with the commonly accepted mechanism for vertex-related plexopathy has prompted some to suggest the possibility of intrauterine injury to the plexus or of congenital malformation. Any conclusive data with regard to this hypothesis have yet to be demonstrated.

Physical examination and motor assessment A thorough examination of the child should be conducted with the following goals in mind: determination of presenting functional severity, identification of other traumatic injuries or complicating factors, and query of possible confounding diagnoses. The child’s head, neck, and shoulders are examined. The position of the head is examined for favored laterality. This could indicate the presence of true or secondary torticollis. Primary (true) torticollis may be demonstrated immediately after birth in infants with brachial plexus injury just as it might be seen in infants after uneventful delivery without brachial plexopathy. In addition, with brachial plexus injury,

292

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

there is a noted tendency for the child to look away from the affected side despite absence of true torticollis. Undiagnosed and untreated with physical therapy, this can result in the gradual development of sternocleidomastoid shortening and secondary torticollis [11,17]. The upper extremities are carefully examined at rest. The position of the upper extremity at rest after brachial plexus injury follows one of a few distinct patterns. The first is characteristic of upper plexus palsy (Erb palsy), in which one would expect deficits primarily at the shoulder and elbow. The level of involvement in this injury is C5,C6, F C7. The rest position, which favors the more richly innervated or least affected muscle groups, is typified in upper root palsy by adduction and internal rotation of the shoulder, extension of the elbow, pronation of the forearm, and flexion of the wrist and fingers. This position is sometimes referred to as the ‘‘waiter’s tip’’ position (Fig. 2). A second common injury pattern is that of total plexus palsy, in which there is complete atonia of the upper extremity (C5,C6,C7,C8,F T1). The baby often gives little or no attention to the affected extremity. Klumpke paralysis, or lower plexus palsy,

Fig. 2. The typical posture of an infant with a right upper trunk palsy. The shoulder is held in adduction with the elbow straight. The wrist, fingers, and thumb are flexed. This is sometimes referred to as the ‘‘waiter’s tip’’ posture.

refers to isolated involvement of C8 and T1 and is generally not seen as a manifestation of obstetrical palsy [18]. Pathologically, intermediate plexus palsy, primarily affecting C7F C8, T1, is a well-described entity [19]. However, it is generally not characterized by a unique or classic position at rest during the early period of evaluation. The appearance would more likely resemble that of upper or total plexus palsy, depending on the extent of involvement of the lessinjured roots. Last, isolated C7 avulsion, although uncommon, presents primarily as a flexion deformity at the elbow. The clavicles, humeri, and ribs are evaluated for possible fractures. The position/posture of the shoulders is evaluated for possible dislocation. The scapulae are often difficult to evaluate in the infant but may suggest proximal injury with impairment of the long thoracic nerve if asymmetry and winging is detected. The phrenic nerve is another proximally based branch, taking contribution from C3-C5 [20]. Breathing is carefully evaluated to determine involvement of the phrenic nerve. Symmetrical movement of the ribcage with coordinated abdominal movement should be seen. In our practice, all preoperative patients undergo ultrasound of the diaphragm. Postoperatively, in some cases, transient hemiparalysis of the diaphragm may result from traction during the course of dissection. If such a patient had a normal preoperative study, one can reasonably expect resolution with expectant management. Only rarely does a paralyzed hemidiaphragm require plication to improve function. Last, the pupils/eyes are examined for signs of Horner syndrome: ptosis, miosis, anhidrosis, and enophthalmos. Again, such findings suggest more proximal injury [16,21,22]; specifically, these findings indicate disruption of T1 proximal to the sympathetic rami communicantes. Preganglionic sympathetic fibers arise from T1 (and often T2). Just after T1 exits its intervertebral foramen, they branch from the spinal nerve to enter the sympathetic chain via the white rami communicantes. They subsequently travel to the superior cervical ganglion and synapse. From there, postganglionic fibers proceed along the carotid plexus ultimately to innervate the dilator pupillae muscle, levator palpebrae muscle, etc. An injury to T1 distal to the rami communicantes does not disrupt the sympathetic fibers and therefore does not result in Horner syndrome. The motor assessment is perhaps the most important component of the systematic examination. Unfortunately, in the neonate and infant, it is also the most difficult. Accuracy and consistency, which require patience and some experience, are important for decision-making on an individual case basis and for

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

systematic evaluation of global treatment plans. Assessment of motor function in the older child and adult is more straightforward and has allowed the use of grading scales that are based on a level of demonstrable voluntary power. The British Medical Research Council’s familiar zero-through-five scale is illustrated in Table 1. Because the infant cannot participate in testing that requires the voluntary demonstration of full power, some authors have sought to develop instruments based on range of motion rather than power. Such is the case for the muscle grading system of Gilbert and Tassin (Table 2), which considers the presence or absence of movement with and without the effect of gravity [23]. The scale is simple, easy to apply, and widely recognized. However, because it must cover a wide range of ability within four possible grades (M0 – M3), discriminatory capability is not one of its strengths. Furthermore, the scale has never been subjected to validation studies. The Active Movement Scale of The Hospital for Sick Children is an eight-point assessment scale for active range of motion (Table 3). It was designed to be highly discriminatory, capturing subtle and significant changes in movement [11,17]. Like that of Gilbert and Tassin, it addresses range of motion in the presence and absence of gravity. Going further, however, it provides a quantitation of the active range under each of these conditions. A number of guidelines have been developed in an effort to instruct others in its use and to standardize application of the scale. In an important example, infants often demonstrate limited movement with and without gravity effects. Using the Active Motion Scale, infants cannot be scored against gravity unless they first achieve a score of 4, reflecting full active range with gravity eliminated. Gravity-eliminated movements are always assessed first to determine if higher scores might subsequently be assigned. The scoring system is applied to the key movements

293

Table 2 Gilbert and Tassin muscle grading system Observation

Muscle grade

No contraction Contraction with movement Slight or complete movement with weight eliminated Complete movement against the weight of the corresponding segment of the extremity

M0 M1 M2 M3

From Gilbert A, Tassin JL. Obstetrical palsy: a clinical, pathologic, and surgical review. In: Terzis JK, editor. Microreconstruction of nerve injuries. Philadelphia: WB Saunders; 1987. p. 529 – 53; with permission.

of each potentially affected joint: shoulder flexion, abduction, adduction, internal and external rotation; elbow flexion and extension; pronation and supination of the forearm; wrist flexion and extension; finger flexion and extension; and thumb flexion and extension. The practical application of the instrument has been described in detail in the literature [11,17,24]; the specific guidelines and suggestions are key to its consistency. Validation studies of Active Motion Scale have demonstrated consistent intra-rater reproducibility and inter-rater reliability among experienced and recently trained physiotherapists. Any variability that is seen is the result of patient factors rather than rater factors [24]. It is just as important for any grading scale to be reliable and discriminatory as it is to be applicable along the full range of a child’s development. One of the major advantages of scales like the Active Motion Scale or that of Gilbert and Tassin is the fact that they can be administered during infancy and repeatedly throughout development, including postoperatively.

Table 3 Hospital for Sick Children active movement scale Table 1 Medical Research Council muscle grading system Observation

Muscle grade

No contraction Flicker or trace of contraction Active movement, with gravity eliminated Active movement, against gravity Active movement against gravity and resistance Normal power

0 1 2 3 4 5

From British Medical Research Council. Aids to the investigation of peripheral nerve injuries. London: His Majesty’s Stationary Office; 1943; with permission.

Observation Gravity eliminated No contraction Contraction, no motion Motion V½ range Motion > ½ range Full motion Against gravity Motion V½ range Motion > ½ range Full motion

Muscle grade 0 1 2 3 4 5 6 7

From Clarke HM, Curtis CG. An approach to obstetrical brachial plexus injuries. Hand Clin 1995;11:563 – 80; with permission.

294

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

This allows consistent analysis of the patient and a statistically sound means to study the natural history of the injury and the postoperative results in those who are treated surgically. In 1972, Mallett established a grading scale that addressed global movement of the extremity and functional adaptations (Fig. 3) [25,26].

It is widely used and takes into consideration the functional utility of the extremity by grading such composite movements as ‘‘hand to nape of neck.’’ Mallet’s classification is limited by the fact that directed complex movements are not reliably provoked in the infant, who often is examined first

Fig. 3. The classification system of Mallet evaluates the global movement of the extremity to identify functional or maladaptive patterns. It requires the cooperation of the patient to perform voluntary movements on command. (From Gilbert A. Obstetrical brachial plexus palsy. In: Tubiana R, editor. The hand. Philadelphia: WB Saunders; 1993. p. 579; with permission.)

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

between 1 and 3 months of age. Because of our strong belief in consecutive application of the same test instrument during development for clinical and analytical purposes, we have not incorporated the Mallet classification into our routine clinical practice.

Prognosis and the decision for surgical exploration Most authors agree that children who will ultimately require surgery should be treated as early as possible to maximize their benefit from the intervention. The difficulty lies in accurately identifying these children at an early time point. There are two circumstances in which, given our current level of understanding, the indication for surgery versus expectant management are quite clear early on. First, as Waters indicates, normal neurologic function can be almost universally expected without surgery for the patient who demonstrates recovery of neurologic function within 1 month after birth [27]. His conclusions, which support the work of Michelow, Tassin, Eng, and most others, were based on the first noted appearance of biceps function by 1 month in 8 of 66 patients who were followed for a minimum of 2 years. In the community, a large proportion of patients would probably fall into this group. Although we would prefer to follow all patients, realistically, many patients in this category may recover without proceeding to specialist referral. On the opposite end of the spectrum, a poor long-term functional result is to be expected if no intervention is undertaken for a baby presenting a total plexus palsy with a Horner sign who fails to improve over the first 3 months. Gilbert, Al Qattan, and Clarke state that the presence of an unequivocal Horner inevitably indicates the need for early exploration [17,21,22]. In Waters’ series, 12% (8/66) demonstrated recovery within 1 month; only 5% (3/66) fell into the latter total palsy group with Horner syndrome. Nearly 80% of the patients in this tertiary referral-based series presented with injury severity between these extremes; therefore, additional data and selection criteria are necessary. Surgery is indicated if there is evidence that the outcome after intervention is likely to result in better long-term function than that resulting from conservative management. To make this judgment, one must first know the natural history of the condition along a full spectrum of severity. If predictors of poor outcome could be identified and used as surgical criteria, then a comparative outcome study could be conducted to evaluate treatment effects. The approach taken by Gilbert and Tassin in 1984 focused on prediction of shoulder outcome in 44 patients with

295

obstetrical palsy. They observed that over a 5-year period, a ‘‘good shoulder’’ (Mallet class IV) was not obtained unless biceps and deltoid contraction began by the third month [28]. ‘‘An average (class III) or bad (class II) shoulder developed if biceps or deltoid recovery began after the third month.’’ Because of the perceived difficulty in evaluating deltoid function, the authors recommended absence of elbow flexion alone at 3 months as the primary indication for surgical exploration. Since their report, absence of elbow flexion at 3 months has become the most widely accepted criterion for surgery in cases that fall between the two aforementioned extremes. Some have felt that this criterion may yield too high a false positive rate, resulting in a surgical recommendation for some children who would otherwise recover satisfactorily with conservative treatment. In our study of the natural history, Michelow reported the progressive changes in active movement over a 12-month period in 66 patients with obstetrical palsy [5]. In the series, patients were evaluated repeatedly at 6-month intervals after birth using an earlier version of the Hospital for Sick Children Active Motion Scale that was later refined. The cardinal movements at the shoulder, elbow, forearm, wrist, and hand were graded. At 12 months, patients’ outcomes were assessed according to the Narakas classification in which poor recovery is defined as elbow flexion of half or less the normal range and shoulder abduction of less than half the normal range [29]. For functional range beyond these limits, recovery is considered to be good. Analysis was undertaken to determine whether the recovery of any one motion or set of motions at 3 months could statistically predict the incidence of poor recovery. In the analysis, absent elbow flexion alone at 3 months incorrectly predicted a poor recovery 12.3% of the time. Using the composite score of elbow flexion plus elbow, wrist, thumb, and finger extension at 3 months, poor recovery was incorrectly predicted 5.2% of the time at a cutoff score of 3.5. Therefore, the authors concluded that an additional 7.1% of children could be spared surgery and yet still be expected to achieve a good recovery if the more comprehensive score is used. Waters also investigated recovery of elbow flexion as an independent criterion for surgery. In his study of 66 patients with obstetrical palsy, recovery of biceps function was evaluated monthly from the time of birth to 6 months [27]. Outcome was based on the Mallet global functional movement scale. Mallet scores were determined at each evaluation, and the minimum follow-up for all patients was 2 years. Waters found that patients who had recovery of biceps function in

296

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

the fourth, fifth, or sixth months after birth did not proceed to normal global function, as did those who recovered biceps function in the first 3 months. For those who underwent surgery at 6 months due to failure of biceps recovery, the outcome was better than for those who regained function at 5 months and did not undergo surgery. His findings give further support to the belief that the specific timing of recovery is an important prognostic variable and that surgical outcome can surpass natural history in carefully selected patients. Therefore, Waters concluded that, in addition to patients with total palsy failing to improve by 3 months, surgery should be offered to patients in whom biceps fails to recover by 5 months. Strombeck studied 247 patients, of whom 54% recovered biceps function within 3 months and later progressed to satisfactory function [2]. Unchanging total plexus palsy at 3 months was seen in 13% of patients, all of whom were given a surgical recommendation. Most of the remaining patients (33%) were observed for ‘‘late recovery,’’ which was defined as biceps recovery at the 6- or 9-month evaluation. Those who did not regain biceps at these later time points were recommended for surgery. The authors found that ‘‘late recovery’’ did occur and reported that there was no significant difference in outcome between the operated group and nonoperated group except for improved shoulder movement in those with upper plexus lesions. The findings led the authors to reject the 3-month biceps recovery criteria altogether in favor of waiting for ‘‘late recovery.’’ Although we agree that some patients regain biceps function after 3 months and proceed to satisfactory recovery, waiting for ‘‘late recovery’’ in all cases is not appropriate either. Although half of the patients in our natural history study with no elbow flexion at 3 months went on to have good function, the other half did not. The test score developed from this study was designed to identify patients with a statistical likelihood for poor recovery while limiting the number of incorrect predictions. In this way, an early surgical recommendation may have been given to some of the patients in the Strombeck study. Despite that fact that the surgical patients in their study (those with more severe injury) had later operations, they still experienced functional recovery that was at least as good as those who had been observed (those with presumably less severe injury). A careful statistical approach to surgical indications, like that of Waters or Michelow, can limit low-yield interventions while maximizing the potential benefits of surgery by making early recommendations when appropriate. In our practice, a Test Score is determined at the 3-month assessment based on the following statistical

discriminants that together were shown to limit the false prediction rate for poor recovery to only 5.3%: elbow flexion and elbow, wrist, finger, and thumb extension. Patients with a Test Score of greater than 3.5 are observed and continue physical therapy to maintain range of motion. Active Movement Scores and the Test Score are repeated every 3 months. Some patients with upper trunk lesions who show good early recovery and have Test Scores greater than 3.5 may still not develop adequate elbow flexion by the end of the first year of life and may have poor shoulder function. In such cases, if at the age of 9 months the child demonstrates elbow flexion less than grade 6 (less than half range of motion against gravity), surgical exploration is offered. At this age, the child is able to cooperate with more complex functional movements; therefore, elbow flexion is assessed during what we term the ‘‘cookie test.’’ The elbow is held in adduction at the trunk, a small cookie is placed in the hand, and the child is encouraged to bring it to the mouth. The child passes the test and is rejected as a surgical candidate only if the cookie is taken to the mouth by elbow flexion against gravity and with less than 45 degrees of neck flexion [17,24].

Primary surgery for obstetrical brachial plexus palsy Reaching the decision for exploration of the brachial plexus in the infant is not easy. Challenging questions of management do not end once the decision is made. Exploration of the brachial plexus is the technically elegant first stage of the operation, exposing and revealing the grossly apparent sequelae of closed traction injury. Once the plexus is exposed, controversies emerge again and include determination of root avulsion and management of the ubiquitous neuroma-in-continuity by neurolysis, neuroma resection and grafting, or neurotization. This section begins with a brief discussion of brachial plexus exploration and the most common findings. We then discuss our current approach to the injury. Exploration of the brachial plexus Positioning The infant is placed in the supine position and is positioned fully at the head of the bed and close to the lateral edge on the injured side (Fig. 4). This allows the surgeon and assistants to work from above, beside, and below the field. The entire upper extremity, upper chest, and neck are prepped into the field. The arm is placed in a towel and clipped at waist

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

Fig. 4. The infant is positioned such that the surgeon may work above, beside, and below the field. The neck is slightly extended, and the head faces away from the field. The hand and arm are accessible for observation of stimulated movements. The patient’s face and the nasotracheal tube are always visible to the anesthetist and surgical team by use of a clear plastic head drape.

level to the drape in a position of internal rotation with the elbow at 90° for the majority of the procedure. It will be accessible for examination when it comes time to stimulate the exposed plexus. The head is turned to face away from the injured side. A small sand bag is placed beneath the shoulder to provide an appropriate amount of neck extension. All pressure points, particularly the scalp, are meticulously padded. Our surgical team and anesthetists favor securing a nasotracheal tube with suture, and we routinely drape the head using transparent material to maintain our ability to monitor the airway throughout the lengthy operation. A rectal temperature probe is used, and appropriate standard techniques for temperature maintenance are used with vigilance. Incision Among experienced plexus surgeons, there is some variation in the choice of incision. Some choose to start with a single transverse incision [30]. We prefer to raise a V-shaped flap for complete exposure of the posterior triangle; the incision follows the posterior border of the sternocleidomastoid, makes a gentle curve just above the clavicle, then follows a line parallel to the clavicle (Fig. 5). We have found that this incision provides optimal exposure of the entire field of interest and avoids the need for adjustments and shifts throughout the procedure that are necessary with a transverse incision. On occasion, we have encountered the need to extend the incision along the deltopectoral groove for exposure of the more distal plexus in long lesions.

297

Dissection The procedure is performed under loupe magnification. The skin is elevated in the subplatysmal plane and is retracted superolaterally. The lateral border of the sternocleidomastoid is identified, and the clavicular head is released from its origin. Along the lateral border of the sternocleidomastoid, the cervical plexus is seen emerging and branching in a characteristic pattern [30]. The branches should be neatly dissected (Fig. 6). Supraclavicular branches are often useful as graft material. As they are followed proximally, larger branches (greater auricular, lesser occipital, and cervical cutaneous) aree identified as one works toward the C3 and C4 roots. Early identification of the C4 root is helpful in proceeding with the exposure and identification of the brachial plexus roots proper (C5-T1). The omohyoid muscle divides the posterior triangle into a superior and inferior portion. The omohyoid should be divided at its tendinous midportion and reflected from the field. The dissection along the lateral border of the sternocleidomastoid can then continue. Beneath the omohyoid, a layer of adipose tissue and lymphatics is seen. This tissue directly overlies the injured plexus. It is best to mobilize this tissue as a flap by dissecting it from the lateral border of the sternocleidomastoid and then sweeping it off the clavicle in a medial-tolateral direction. The flap, which is hinged posterolaterally, can then be returned at the completion of the operation, placing it over the reconstructed plexus. The transverse cervical artery is seen, and it should be divided for further exposure. Because we prefer to

Fig. 5. A V-shaped flap is created and reflected posterolaterally by an incision along the posterior border sternocleidomastoid and the clavicle. Through this exposure, the entire posterior triangle is visible. For longer lesions, the incision may be extended along the deltopectoral groove if necessary.

298

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

Fig. 6. The sternocleidomastoid (8), trapezius (2), and clavicle form the borders of the posterior triangle. The omohyoid (4) divides the triangle into upper and lower portions. The accessory nerve (3) travels obliquely. The branches of the cervical plexus are more superficial and include: lesser occipital (1), greater auricular (5), cervical cutaneous (6), and supraclavicular (7). Supraclavicular branches further divide and can be useful as grafts, providing 3 to 4 cm of material. (From Borrero J. Surgical technique. In: Gilbert A, editor. Brachial plexus injuries. London: Martin Dunitz; 2001. p. 189 – 204; with permission.)

avoid potential instability and having bone callous directly over the reconstruction, we do not divide the clavicle or remove a central segment as some do for the purpose of exposure. We are generally able to proceed with exposure of the distal plexus with good anterior and inferior retraction of the clavicle by an assistant. If one felt it necessary to divide the clavicle, it can be restored as Borrero describes, using a figureof-eight suture passed through drill holes [30]. The plexus exits between the anterior and middle scalenes proximally. These muscles are often scarred and adherent to the neuroma on their respective sides. Starting with dissection of the anterior scalene from the neuroma, one works toward the C5 foramen. Two important structures are visible. The phrenic nerve travels inferiorly and should be carefully preserved, along with any contribution from (and occasionally to) the plexus. It is frequently scarred to the anterior surface of the neuroma. Also directed inferiorly, but passing behind the neuroma, is the long thoracic nerve. The C5 root is dissected proximally all the way to its foramen. In dissecting within the bony foramina at any level, bothersome bleeding is sometimes caused by disruption of small vessels, primarily venous. Time

and pressure generally remedy the situation. The C6 and C7 roots are dissected next in a similar fashion. If a particular root is not found in its expected position, this could indicate that the root has been avulsed from the cord. The foramen should still be completely dissected to confirm this and to perhaps identify a proximal stump in the setting of rupture, rather than avulsion. The only irrefutable proof of avulsion is the identification of a dorsal root ganglion proximally. The upper and middle trunks, which are often difficult to separate, should be in view. Dissection of the lateral border of the plexus in a proximal to distal direction allows the identification of the suprascapular nerve. Depending on the injury, the suprascapular nerve may be seen branching from the upper trunk proximally, distally, or within the zone of injury. The branching pattern of the plexus, in general, is distorted by several factors: the length of the lesion, the presence of ruptures, the position of ruptures, and the extent that the neuroma has contracted. Contraction can sometimes be considerable, and this alters expected branching positions. For example, the division of the middle trunk into anterior and posterior divisions is normally retroclavicular, but this may occur more proximally in a lesion that has contracted [30]. The dissection continues under the clavicle until all branches have been identified and the distal limits of the neuroma have given way to the normal branching network that will ultimately serve as graft targets. The lower trunk, formed by C8 and T1, is dissected. The subclavian artery is found overlying the lower trunk. It should be dissected from the neuroma to be safely retracted for exposure of the C8 and T1 roots and foramina. In addition, the suprascapular artery can be intimately associated with the neuroma. It and the transverse cervical artery (which was more superficial and was already divided) are branches of the thyrocervical trunk from the subclavian artery. Ligation and division of the suprascapular artery is often also a necessary maneuver. Once these vessels have been appropriately dealt with, the C8 and T1 roots can be approached. The T1 root is smaller than C8 and lies posterior and inferior to it. Dissection of T1 is one of the most challenging portions of the dissection because it lies on the parietal pleural surface. From the anterior approach, the T1 foramen is difficult to visualize. The lower trunk is followed distally, taking care to locate and preserve all branches. The dissection again continues until the normal branch network distal to the lesion is identified. Unlike the upper and middle trunk components, which are often densely adherent, the lower trunk components are sometimes more easily dissected from the neuroma. The final portion of the exploration that

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

must be completed is the dissection of the posterior surface of the plexus/neuroma.

Management of the lesion Anatomically, the most common operative finding is neuroma-in-continuity, with variation in length and in the number and extent of involved roots. However, although we see physical continuity, it is difficult to know whether it is accompanied by regeneration of axons through the neuroma (ie, physiologic continuity). Although the use of intraoperative neurophysiologic testing is supported by many authors, we believe that functionally relevant quantification of physiologic continuity is not possible with current technologic measures. This may be the reason that neurolysis of the plexus has not proven to be a useful means of managing the obstetrical brachial plexus lesion. As Borrero indicates, most surgeons agree that the results after neurolysis alone are discouraging. Earlier in our practice, neurolysis was performed when a conducted stimulus across the neuroma produced an appropriate distal muscle contraction. We reported the results of 16 patients who underwent neurolysis, comparing preoperative active motion with that seen at 12-month follow-up [31]. We found improvement in the group of patients with upper plexus lesions but that it is ineffective in patients with total plexus palsy. Neurolysis was abandoned in the latter population in favor of neuroma resection and grafting. Eventually, neurolysis was abandoned in patients with primarily upper plexus lesions when we recognized that infants who were treated with resection seemed to regain baseline function rapidly after resection. In 1998, we reported pre- and postoperative active motion over time in 26 patients treated with neuroma resection [32]. The results demonstrated an initial decrease in function compared with baseline. However, by 3 months, there was no difference in active motion from baseline, and by 6 months patients had reached or surpassed baseline. The difference could not be explained on the basis of regeneration across grafts at these early time points. Furthermore, the early follow-up results (3, 6, and 12 months) were no worse for those who had undergone resection than for patients who had undergone neurolysis. Based on the observation that neuroma resection is not detrimental and on the growing body of evidence from Gilbert, Birch, Kawabata, and others that grafting offers the best opportunity for maximal functional recovery [8,33 – 35], we have strongly advocated neuroma resection and grafting as the primary means for management of the neuroma in obstetrical palsy. Neurolysis can be useful in situations in which a

299

distinct fascicular pattern is demonstrated within surrounding scar. Such can be the case for less severe C8 and T1 lesions in the setting of a predominant upper trunk injury. Considering all data currently available, it is our practice to resect all lesions unless a distinct and normal fascicular pattern is evident.

Evaluation of the proximal roots The exploration and dissection of the plexus ideally would reveal a central traumatized area, represented by neuroma, with normal structures proximally and distally. The distal dissection always continues until grossly normal, identifiable branches are exposed. The length of the neuromatous lesion determines whether this occurs at the level of divisions, cords, or terminal branches. We typically transect the neuroma through its midportion. The distal half of the neuroma is then removed after each of the distal branches is transected at a level that appears grossly normal. These carefully identified ends will be the targets for sural nerve grafts. Depending on its branch point location, the suprascapular nerve is often an independent distal target. The condition, quality, and axonal topography of the distal ends are demonstrable through histologic frozen section using toluidine blue dye. This is always undertaken in our practice. Proximally, however, a major dilemma can be encountered. An empty foramen, or histologic evidence of a dorsal root ganglion outside the foramen with otherwise apparent continuity, irrefutably indicate root avulsion from the spinal cord. With each avulsion, fewer proximal source stumps are available to provide axons to the distal targets. When there is apparent continuity from within a particular foramen along the neuroma, how does one definitively know whether the proximal stump is a suitable source? How can we determine whether an intraforaminal avulsion has occurred because it is not visible? This is the most vexing decision in the operation because grafts from a stump that seems normal (grossly and even histologically) provides no axons if it has been avulsed from the spinal cord. The only way one could know with certainty whether an intraforaminal avulsion has occurred is through laminectomy. However, because this is generally not considered safe in infants, we must weigh all available evidence to reach a most probable conclusion. What evidence is available [30]? 1. Physical examination (in our practice, using the Hospital for Sick Children Active Movement Scale) including presence of Horner sign 2. Radiographic data (ie, CT myelography or MRI in some centres)

300

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

3. Muscle response to direct stimulation of the root (a normal-appearing nerve will not conduct distally if avulsion has occured) 4. Operative findings including gross appearance of the stump and tactile sensation during transection 5. Histologic examination of the proximal stump Several of these are worthy of further discussion. Although the use of MRI in the evaluation of root status has been advocated in recent reports [36], technical difficulty, a reliance on exquisite interpretive skills, and the need for further statistical validation have limited the wide-spread application of MRI for the determination of root avulsion. We continue to investigate all infants before surgery with CT myelogram. The CT myelogram provides useful information depending on the presence of a pseudomeningocele and the ability to visualize rootlets traversing it. Chow

et al studied 281 roots among 58 patients undergoing exploration and reconstruction [37]. Comparing radiographic interpretation to findings at surgery, they determined the sensitivity, specificity, positive predictive value, and likelihood ratio for these two possible indicators of root avulsion. They found that the sensitivity of pseudomeningocele to identify root avulsion was low (0.63); however, the specificity was much higher (0.85). For pseudomeningocele with absent rootlets, the sensitivity was even lower, but specificity was near perfect (0.98), and the likelihood ratio was quite high (18.5). In other words, when rootlets cannot be visualized in association with pseudomeningocele, a false positive (a normal root) is highly unlikely; it is 18.5 times more probable that the involved root is avulsed (Fig. 7). Nevertheless, because sensitivity is low, one must have further evidence to rule out avulsion in the absence of this statistically specific indicator.

Fig. 7. This preoperative myelogram, axially at the level of C8, demonstrates the presence of a pseudomeningocele. Neither dorsal nor ventral rootlets may be seen traversing the pseudomeningocele. This finding is a specific indicator of root avulsion (specificity 0.98), with a likelihood ratio of 18.5 for avulsion.

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

The physical examination is important; the tactile sense of the experienced surgeon as the root is transected can also be suggestive. It should be noted that transection of the root should be performed as far proximal as possible into the foramen while allowing the ability to perform coaptation. Grafting The neuroma has been resected, the targets have been identified, and their proximal and distal ends histologically deemed satisfactory. The usable (decidedly nonavulsed) proximal stumps have been selected. A gap has been created. This gap will be spanned by multiple strands of sural nerve graft. Additionally, if one has carefully dissected the supraclavicular branches of the cervical plexus, these may also serve as graft material. We elect to perform harvest of bilateral sural nerves as the first stage in the operation (ie, before exploration of the plexus). We perform sural nerve harvest using endoscopic assistance [38]. The infant is prone for this portion of the procedure and is then turned supine for exploration and reconstruction of the plexus. In this

301

way, exposure and access are optimal, and only one change in position is required. The endoscopic assisted harvest is performed through three 2-cm transverse incisions (Fig. 8). The sural nerve can be harvested from its most proximal origin from the tibial nerve to the most distal branches supplying the lateral foot. In an 8- to 10-kg infant, an average of 13 to 15 cm can be harvested per leg. The harvested nerves are placed in moist, sealed, sterile containers and are refrigerated until they are needed. A reconstructive plan is then formulated and is individualized to the unique anatomic findings of each patient. If a target is reconstructed using grafts from its developmentally appropriate source root, then the reconstruction is termed ‘‘anatomic grafting’’ for that target. If possible, this is the first choice for reinnervation of target nerves. If, however, the appropriate source root has been avulsed, then associated targets must receive innervation from an alternate source, termed ‘‘neurotization.’’ Neurotization can be accomplished from remaining roots (plexo-plexal or intra-plexal neurotization) or from nerve donors outside the plexus (extra-plexal neurotization).

Fig. 8. Endoscopic sural nerve harvest, as described by Capek et al [38]. (A) A series of three 2-cm transverse incisions centered along the midline of the calf are used in endoscopic sural nerve harvest. The nerve is harvested distally along the lateral aspect of the foot and proximally to its origin from the tibial nerve, where additional length can be obtained by interfascicular dissection. (B) The resultant scars from harvest in this manner are generally acceptable and spare the use of the long ‘‘stocking seam’’ incision.

302

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

From each of the available roots, the distances to every target are measured. The length of available graft material is also measured. The division and allocation of graft material is based on prioritization. Reconstruction must be undertaken with consideration of greatest functional needs. Some authors agree that the use of the upper extremity is most dependent on a functioning hand. Therefore, prioritization of functional motion should be first focused on reinnervation of the hand [22,30]. This is addressed with multi-strand grafting to lower trunk targets. After the hand, prioritization is then directed to elbow flexion, then shoulder motions. Grafts are placed into position in reversed orientation to prevent progressing axons from prematurely

exiting the graft via side branches of the original sural nerve. Graft placement is performed under the operating microscope. In our practice, nerve coaptation is performed using commercially available fibrin glue and without any sutures. Some surgeons prefer to use suture material only, and others use a single suture reinforced with fibrin glue for end-to-end coaptation. A case example of the working plan and final diagram used in the setting of upper plexus palsy with avulsion of C8 is depicted in Fig. 9. On average, the distance from most sources to respective targets is 2.5 to 4.5 cm, as is the case here. In this case, T1 was found to have a distinct and normal-appearing fascicular pattern and was treated with neurolysis. The remaining distal targets, including the suprascapular

Fig. 9. This patient presented with an upper plexus palsy and was found to have an avulsion of C8, demonstrated unequivocally by the presence of the dorsal root ganglion. C5 through C7 were involved in a dense neuroma, and T1 was found to have a normal fascicular pattern with minimal involvement in the neuroma. Anatomic grafting was performed to the suprascapular nerve, anterior and posterior divisions of the upper trunk, and the middle trunk. Intraplexal neurotization to the lower trunk was performed from C7. T1 was treated with neurolysis. The phrenic nerve is shown traversing longitudinally over the plexus. SSN, suprascapular nerve; ANT, anterior division upper trunk; POST, posterior division upper trunk; MT, middle trunk; LT, lower trunk.

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

nerve, were reconstructed with grafts from the three intact stumps. For most targets, grafting was anatomic. Grafts to targets that had received C8 innervation (avulsed) are intraplexal neurotizations because they were grafted from other available stumps. A second case is presented in Fig. 10, in which there was avulsion of C7, C8, and T1. In this case, a longer gap was created after neuroma resection, making strategic graft allocation even more important. The lower plexus targets were given first priority (highest priority to hand function) to receive grafts from the available C5 and C6 stumps. Again, the lower trunk can be said to have been reconstructed using an intra-plexal (nonanatomic) neurotization. The upper trunk targets were selected next to favor elbow flexion and shoulder function.

303

This case also demonstrates the use of extra-plexal neurotization. The suprascapular nerve was reconstructed from the spinal accessory nerve (CN XI). This is the most common extraplexal neurotization. CN XI is dissected distally beneath the trapezius muscle to its major branch point. The dissection is accomplished via the existing neck exposure. The longitudinal branch supplying the thoracic portion of the trapezius is divided distally and transposed into the field where it consistently reaches the cut end of the suprascapular nerve free of tension. Performed in this manner, there is no sacrifice in the function of the cervical/upper thoracic portion of the trapezius, which inserts on the clavicle, acromion, and scapular spine. The suprascapular nerve provides external rotation at the shoulder through the actions of the

Fig. 10. This patient presented with a total plexus palsy and was found to have avulsions of C7, C8, and T1. A long lesion was present. This is a difficult situation due to the paucity of graftable stumps and the long lengths of graft that were required. The lower trunk targets were given first priority to provide for hand function and were grafted from the C6 stump, as was the middle trunk. The anterior and posterior divisions of the upper trunk were grafted from C5. An extra-anatomic neurotization of the suprascapular nerve from the accessory nerve was a useful maneuver to conserve available graft material. SSN, suprascapular nerve; ANT, anterior division upper trunk; POST, posterior division upper trunk; MT, middle trunk; LT, lower trunk.

304

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

infra- and supraspinatus muscles. Marcus et al has demonstrated that neurotization using the spinal accessory nerve produces external rotation at the shoulder equal to that obtained by anatomic reconstruction using sural nerve crafts from C5 [39]. Other extraplexal neurotizations include intercostal nerves (typically 2 to 3 donors are used), and the contralateral C7 root. Before considering extraplexal neurotization, it is important to consider the expected potential gain in function and the donor deficit. At the completion of the operation, a simple dressing is applied over the neck incision, and the child is fitted into a Velpeau garment made with a single length of 2-inch stockinette material (Fig. 11). Properly fitted and maintained, this soft, inexpensive bandage effectively maintains the arm in full internal rotation and shoulder adduction with the elbow at 90° [40]. We have used this technique and have appreciated its efficacy and advantages over plaster immobilization in over 200 primary infant plexus reconstructions. It is tolerated well by the infants and is easy to maintain and adjust by parents and nurses. The Velpeau is worn at all times for 3 weeks.

Complications Minor and major complications can occur with brachial plexus reconstruction. In our recent review of 173 brachial plexus reconstructions for obstetrical palsy, the total combined rate of minor and major complications was 58 (33.5%) [41]. The incidence of complication did not seem to be influenced statistically by the age of the patient (ranging between 3 months and 1 year). There were no mortalities. Major complications included wound infection (2.3%), seroma (1.2%), and the following respiratory complications: postoperative hemidiaphragm paralysis (6.4%), atelectasis (2.9%), pulmonary edema (1.7%), pleural effusion (1.7%), pneumothorax (1.7%), and chylothorax (0.6%). In addition, accidental extubation occurred in five patients in the early portion of the series. With clear plastic head draping and suture fixation of the nasotracheal tube to the membranous septum, this has not occurred in the subsequent 100 patients. Pneumothorax was managed intraoperatively in two of the three affected patients by pleural repair and reinforcement. This was assisted by placement of a rubber catheter to apply negative pressure; repair was confirmed by submerged Valsalva maneuver. In the third patient, pneumothorax was identified by x-ray in the recovery area and was treated by aspiration.

Summary

Fig. 11. A light dressing is placed on the incision, and a soft stockinette Velpeau-style garment is fashioned to maintain the arm securely adducted, with the elbow at 90°. The garment is worn continuously for 3 weeks. We have not found necessary to employ plaster shoulder girdle casting with this technique.

Primary surgery for obstetrical brachial plexus lesions is a young field of surgical expertise that offers the possibility of improved functional ability in carefully selected patients who would otherwise be faced with lifelong impairment and secondary skeletal deformities. One major challenge in this area of peripheral nerve surgery is the selection of patients most likely to derive benefit from surgical intervention. The key to the development of selection criteria and to the resolution of other considerations (such as the determination of root avulsion) is consistency, accuracy, and careful reporting of natural history and outcome data. In particular, we strongly feel that a statistically sound technique of assessment must be consistently applied from the time of presentation through long-term follow-up. Advancement to date has resulted from the application of evidence-based recommendations from large, well-designed, meticulous studies. As the field of obstetrical brachial plexopathy management continues to evolve, we can expect that questions will continue to be answered using such scientific methodology.

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306

References [1] Eng GD, Binder H, Getson P, et al. Obstetrical brachial plexus palsy (OBPP) outcome with conservative management. Muscle Nerve 1996;19:884 – 91. [2] Strombeck C, Krumlinde-Sundholm L, Forssberg H. Functional outcome at 5 years in children with obstetrical brachial plexus palsy with and without microsurgical reconstruction. Dev Med Child Neurol 2000;42: 148 – 57. [3] Greenwald AG, Schute PC, Shiveley JL. Brachial plexus birth palsy: a 10-year report on the incidence and prognosis. J Pediatr Orthop 1984;4:689 – 92. [4] Jackson ST, Hoffer MM, Parrish N. Brachial-plexus palsy in the newborn. J Bone Joint Surg Am 1988; 70:1217 – 20. [5] Michelow BJ, Clarke HM, Curtis CG, et al. The natural history of obstetrical brachial plexus palsy. Plast Reconstr Surg 1994;93:675 – 80. [6] Piatt Jr JH. Neurosurgical management of birth injuries of the brachial plexus. Neurosurg Clin N Am 1991;2: 175 – 85. [7] Birch R. Medial rotation contracture and posterior dislocation of the shoulder. In: Gilbert A, editor. Brachial plexus injuries. London: Martin Dunitz; 2001. p. 249 – 60. [8] Birch R. Invited editorial: obstetric brachial plexus palsy. J Hand Surg (Br) 2002;27:3 – 8. [9] Wikstrom I, Axelsson O, Bergstrom R, et al. Traumatic injury in large-for-date infants. Acta Obstet Gynecol Scand 1988;67:259 – 64. [10] O’Leary JA. Shoulder dystocia and birth injury. New York: McGraw Hill; 1992. [11] Clarke HM, Curtis CG. An approach to obstetrical brachial plexus injuries. Hand Clin 1995;11:563 – 80. [12] Metaizeau JP, Gayet C, Plenat F. [Brachial plexus birth injuries: an experimental study]. Chir Pediatr 1979;20: 159 – 63. [13] Al Qattan MM, Clarke HM, Curtis CG. The prognostic value of concurrent clavicular fractures in newborns with obstetric brachial plexus palsy. J Hand Surg (Br) 1994;19:729 – 30. [14] Blaauw G, Slooff A, Muhlig R. Results of surgery after breech delivery. In: Gilbert A, editor. Brachial plexus injuries. London: Martin Dunitz; 2001. p. 217 – 24. [15] Gilbert A, Khouri N, Carlioz H. [Birth palsy of the brachial plexus – surgical exploration and attempted repair in twenty one cases]. Rev Chir Orthop Reparatrice Appar Mot 1980;66:33 – 42. [16] Gilbert A, Razaboni R, Amar-Khodja S. Indications and results of brachial plexus surgery in obstetrical palsy. Orthop Clin North Am 1988;19:91 – 105. [17] Clarke HM, Curtis CG. Examination and prognosis. In: Gilbert A, editor. Brachial plexus injuries. London: Martin Dunitz; 2001. p. 159 – 72. [18] Al Qattan MM, Clarke HM, Curtis CG. Klumpke’s birth palsy. Does it really exist? J Hand Surg (Br) 1995;20:19 – 23. [19] Al Qattan MM, Clarke HM. A new type of brachial

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33] [34]

[35]

[36]

305

plexus lesion to be added to the classical types. J Hand Surg (Br) 1994;19:673. Al Qattan MM, Clarke HM, Curtis CG. The prognostic value of concurrent phrenic nerve palsy in newborn children with Erb’s palsy. J Hand Surg (Br) 1998; 23:225. Al Qattan MM, Clarke HM, Curtis CG. The prognostic value of concurrent Horner’s syndrome in total obstetric brachial plexus injury. J Hand Surg (Br) 2000;25: 166 – 7. Gilbert A. Indications and strategy. In: Gilbert A, editor. Brachial plexus injuries. London: Martin Dunitz; 2001. p. 205 – 10. Gilbert A, Tassin JL. Obstetric palsy: a clinical, pathologic, and surgical review. In: Terzis JK, editor. Microreconstruction of nerve injuries. Philadelphia: WB Saunders; 1987. p. 529 – 53. Curtis C, Stephens D, Clarke HM, et al. The active movement scale: an evaluative tool for infants with obstetrical brachial plexus palsy. J Hand Surg (Am) 2002; 27:470 – 8. Mallet J. [Obstetrical paralysis of the brachial plexus. B. Treatment of secondary deformities. Principles of treatment of the shoulder. Classification of Results.] Rev Chir Orthop Reparatrice Appar Mot 1972;58 (Suppl 1):166 – 8. Gilbert A. Obstetrical brachial plexus palsy. In: Tubiana R, editor. The Hand. Philadelphia: Harcourt Brace Jovanovich; 1993. p. 575 – 601. Waters PM. Comparison of the natural history, the outcome of microsurgical repair, and the outcome of operative reconstruction in brachial plexus birth palsy. J Bone Joint Surg Am 1999;81:649 – 59. Tassin JL. Paralysies obstetricales du plexus brachial: evolution spontane´e, resultats des interventions reparatrices precoces. Paris: Universite´ de Paris; 1984. Narakas AO. Obstetrical brachial plexus injuries. In: Lamb DW, editor. The paralyzed hand. Edinborough: Churchill Livingstone; 1987. p. 116 – 35. Borrero JL. Surgical technique. In: Gilbert A, editor. Brachial plexus injuries. London: Martin Dunitz; 2001. p. 189 – 204. Clarke HM, Al Qattan MM, Curtis CG, et al. Obstetrical brachial plexus palsy: results following neurolysis of conducting neuromas-in-continuity. Plast Reconstr Surg 1996;97:974 – 82. Capek L, Clarke HM, Curtis CG. Neuroma-incontinuity resection: early outcome in obstetrical brachial plexus palsy. Plast Reconstr Surg 1998; 102:1555 – 62. Gilbert A. Long-term evaluation of brachial plexus surgery in obstetrical palsy. Hand Clin 1995;11:583 – 94. Gilbert A, Brockman R, Carlioz H. Surgical treatment of brachial plexus birth palsy. Clin Orthop 1991;Mar (264):39 – 47. Kawabata H, Masada K, Tsuyuguchi Y, et al. Early microsurgical reconstruction in birth palsy. Clin Orthop 1987;Feb(215):233 – 42. Blaauw G. Can MRI become as precise as CT myelo-

306

[37]

[38]

[39]

[40]

[41]

J.R. Marcus, H.M. Clarke / Clin Plastic Surg 30 (2003) 289–306 gram? XIII International Symposium on Brachial Plexus Surgery. Paris: 2002. Chow BC, Blaser S, Clarke HM. Predictive value of computed tomographic myelography in obstetrical brachial plexus palsy. Plast Reconstr Surg 2000;106: 971 – 7. Capek L, Clarke HM, Zuker RM. Endoscopic sural nerve harvest in the pediatric patient. Plast Reconstr Surg 1996;98:884 – 8. Marcus JR, Curtis CG, Clarke HM. External rotation following suprascapular nerve reconstruction in obstetric brachial plexus palsy: accessory nerve transfer versus C5 grafting. XIII International Symposium on Brachial Plexus Surgery. Paris: 2002. Gilchrist DK. A stockinette-Velpeau for immobilization of the shoulder-girdle. J Bone Joint Surg Am 1967;49: 750 – 1. LaScala GC, Rice SB, Clarke HM. Complications of microsurgical reconstruction of obstetrical brachial plexus palsy. Plast Reconstr Surg, in press.

Further Reading A mathematical model developed by Gonik et al estimated the pressure forces of clinician-applied traction to the fetal head to be 22.9 kPa and the uterine expulsive forces between 91.1 and 202.5 kPa, suggesting that brachial plexus injury is not a priori explained by excessive clinician traction. The susceptibility of neonatal neurons to injury as compared with older neurons is also relevant in any discussion of the etiology of obstetrical nerve palsy. Gonik B, Walker A, Grimm M. Mathematic modeling of forces associated with shoulder dystocia: a comparison of endogenous and exogenous sources. Am J Obstet Gynecol 2000;182:689 – 91. Belin BM, Ball DJ, Bridge PM, et al. The effect of age on peripheral motor nerve function after crush injury in the rat. J Trauma 1996;40:775 – 7.