|Home | About | Journals | Submit | Contact Us | Français|
Among various congenital disabilities following birth trauma, brachial plexus palsy has remained one of the most devastating for many years. The debates about physical therapy alone versus surgical intervention, as well as the proper timing for surgery if indicated, are still open. In our institute, brachial plexus palsies with hand involvement and Horner's sign are surgically treated at the third month of age, and infants with insufficient elbow flexion undergo surgery at the fifth month. Although early neural reconstruction decreases the need for future reconstructive surgical procedures, the children can still have upper extremity deformities that may need to be treated by future muscle releases and transfers. We believe that in patients who missed the chance of primary neural reconstruction, satisfying shoulder function can still be achieved by palliative surgery before glenohumeral joint deformity occurs.
As social standards have improved, the rate of unattended home deliveries has subsided. In addition, antenatal evaluation has become commonplace. Yet the incidence of obstetrical brachial plexus palsy (OBBP) has not decreased during the past 20 years. According to the World Health Organization, the incidence of OBBP is generally 1 to 2 in 1000 live births worldwide.1,2 Although there are no detailed statistical studies, because the birth rate is 1.2% in Turkey and the population is about 70 million, approximately 1000 to 1500 new OBBP cases are added to the society every year.
In our country, the physicians dealing with obstetrical palsy have believed in the assets of nonsurgical treatment for many years. Even surgeons did not seem convinced that neurosurgery of the brachial plexus would be rewarding. The dogmatic belief of many physicians in the success of physical therapy alone has prevented many patients from benefiting from early neural repair, leaving many children with heavy sequelae.
Initially, late secondary interventions such as tendon transfers, contracture release, and deformity corrections dominated our surgical spectrum in obstetrical palsy patients who have had no histories of prior nerve surgery. Subsequently, as we have had the opportunity to experience the treatment algorithms of pioneer clinics and monitor the results of early neural surgery, a common understanding of the current concepts regarding treatment modalities was formed. Since 2001, we have evaluated obstetrical palsy patients in a multidisciplinary fashion along with pediatricians, electromyography (EMG) specialists, and rehabilitation therapists on a monthly basis.
Although many clinical and laboratory tests have been recommended, the methods used to assess the severity of the plexus lesions, as well as predict candidates for neurosurgical intervention, are still controversial. We perform EMG studies and magnetic resonance imaging (MRI) preoperatively along with clinical examination using the Hospital for Sick Children active movement scale for evaluation (Table 1).
We believe that EMG findings in OBPP might be overly optimistic in some cases. Vredeveld et al compared obstetrical and older traumatic brachial plexus patients with C5, C6 versus C5, C6, C7 involvement.3 Infants with C5, C6 lesions at 4 months showed nearly normal motor unit potentials in the biceps and deltoid muscles with little or no denervation, whereas the older brachial plexus group with the same lesion exhibited complete denervation of both muscles. Infants with C5, C6, C7 lesions had complete denervation of both muscles. This indicates that C7 largely contributes to the innervation of both muscles at the time of birth but this intervention is lost later in normal individuals.
In a study conducted by a group of pediatric specialists in our institution, the role of clinical and laboratory approaches for the assessment of the severity of plexus lesions was determined. MRI revealed pseudomeningoceles in two of the five patients with a poor prognosis (total plexus involvement) and in two of the eight patients with a good prognosis (C5, C6, ±C7 involvement).4 EMG implied root avulsion in three of five patients with a poor prognosis. The clinical evaluation was determined to be the most predictive method overall for final prognosis.
There is still a lack of consensus on the timing and absolute indications for nerve surgery in OBPP. It is accepted by most of the pioneers in the field of brachial plexus surgery that a baby with hand involvement and Horner's sign should be operated on as early as 3 months of age.5,6,7 Terzis and Papakonstantinou recommend surgery for global paralysis prior to the third month if the patient is physically able to withstand general anesthesia.8 Major controversies remain regarding treatment options in the upper palsy group. The most common lesion is neuroma-in-continuity between roots C5, C6 or C5, C6, C7. This type of lesion allows some fibers to regenerate distally toward the targets. The question is whether or not they are sufficient in number to reinnervate the affected muscles.
Gilbert and Terzis and their colleagues advocate that upper root paralysis should be operated on by the third month if there has been no recovery of the biceps muscle.6,8 Marcus and Clark offers surgical exploration for babies who have biceps function less than half the range of motion against gravity at the age of 9 months.9
In our institute, in the first 3 months of life the infant is evaluated by conventional radiograms to rule out fractures of the clavicle or humerus. Neurological consultations are also obtained. At 3 months of age MRI and EMG studies are obtained. Patients with hand involvement, especially with a positive Horner's sign, undergo nerve surgery at this time. Babies who present with deltoid and biceps insufficiency and who have been treated only with physical therapy are operated on by the fifth month. We use the Hospital for Sick Children active movement scale for pre- and postoperative evaluation. According to this scale, elbow flexion less than grade 5 (which is less than half the normal range of motion against gravity) is accepted as a surgical indication.
In some cases it is observed that biceps and/or deltoid recovery may start as late as 5 to 6 months of age. This often results in optimistic expectations by the families and leads to refusal for surgical intervention. At this point, we often experience difficulties in convincing them of the need for surgery. As Gilbert also emphasized, parents in Mediterranean or Middle Eastern environments are prone to reject surgical treatment options if some improvement occurs on the affected muscles, even though they are told that these muscles will never recover.10 Our clinical observations parallel Gilbert's on this subject. These parents are afraid of losing what the infant has already gained.
One of the significant problems we often experience is related to patients who have passed the optimal allotted time for surgery. Although the benefits of surgery are unknown for the late cases, patients younger than 1.5 years of age are accepted for treatment. Of course, the earlier the patient undergoes nerve surgery, the quicker regeneration takes place and the less motor end plates irreversibly atrophy.
Our clinical experience with OBPP can be divided into two groups: primary nerve surgery and treatment of sequelae of late OBBP, which consists of secondary procedures such as tendon transfers, soft tissue release, and bone procedures.
In the 2001–2003 period, 92 patients with obstetrical palsy were evaluated within the pediatric neurology department. Twenty-four of these patients had surgical repair of the brachial plexus. Twenty-three of the 24 patients had a cephalic presentation with spontaneous delivery; one patient with cephalic presentation had vacuum extraction. The average birth weight was 4097 g (2750–4750 g). The condition was equally distributed between boys and girls. The side predominantly affected was the left (54%) more than the right (46%).
In Gilbert's series comprising 241 patients, C5, C6 involvement was 39%; C5, C6, C7 involvement was 33%; and C5-T1 involvement was 26%.11 Concerning the Baylor experience, this proportion turned out to be 73% for C5, C6, ±C7 and 20% for C5-T1.12 However, the most common clinical picture of our surgical cases was complete palsy with C5, C6, ±C7 rupture and C8, T1 avulsion at 70%, followed by upper palsy C5, C6 at 17% and C5, C6, C7 palsy at 13%. This is because physicians from other hospitals who still question surgical intervention, especially for those with Erb's palsy, mostly refer the infants with global palsy for surgery.
The patient is placed in the supine position. The head is turned to the noninjured side and the chest is brought forward and supported with a back roll vertically placed between the scapulae. The neck and entire upper extremity of the injured side and both lower extremities are prepared. Long-acting muscle relaxants should be avoided to allow intraoperative nerve stimulation.
An incision is made parallel to the posterior border of the sternocleidomastoid muscle and the upper border of the clavicle. This incision should be extended to the deltopectoral groove in some cases if there is an insufficient amount of exposure of the injury. The omohyoid muscle, transverse cervical vessels, and phrenic nerve are major landmarks during the exploration. The phrenic nerve, which consists of fibers of C3, C4, C5 roots, lies on the anterior scalene muscle and follows a path to the C5. Then it is easy to reach the C6 root. Usually a neuroma involves the upper trunk, which is composed of the C5 and C6 roots. The middle trunk, which is a continuation of C7, sometimes participates in neuroma formation as well. If the paralysis is total, C8 and T1 roots should be identified. Care must be taken during this part of the exploration to avoid injury to the parietal pleura. The dissection of the plexus is continued until grossly normal structures have been reached proximally and distally. A careful nerve stimulation is performed with 0.5, 1, and 2 mA to observe muscle responses.
All data collected by physical examination, EMG, radiology, and nerve stimulation should be taken into consideration for a final diagnosis of the injured nerves. Of all investigations, physical examination and observation of the infant have the greatest value in establishing the strategy of reconstruction.
Lower roots tend to avulse from the spinal cord, and upper roots rupture more distally after they traverse the foramen. Horner's sign and hand involvement indicate C8 and T1 root avulsion even if these roots and lower trunk have a normal appearance. Neuroma formation or ruptures are very rare in lower roots. In our series, all cases with lower trunk involvement were determined to be avulsions from the spinal cord. On the other hand, upper roots most likely show a neuroma-in-continuity, which is indicated by deficient shoulder±elbow function and cocontractions. Neuroma or segments involved with lesions are resected until a normal-appearing fascicular pattern is observed. After neuroma resection, available proximal stumps are appraised as to whether or not they are suitable sources for distal targets. Frozen sections are obtained and examined histologically to observe for ganglion cells because their availability indicates root avulsion or to be able to visualize healthy axons at levels out of the injury zone for possible coaptations.
Neurolysis, neuroma excision, nerve grafting, and neurotization are the techniques used for nerve surgery. One of the greatest controversies in the management of OBPP has been decision making for a lesion that maintains its continuity. Clarke et al stated that neurolysis of neuroma-in-continuity in Erb's palsy improves both muscle grade and functional ability of patients but it does not provide useful functional recovery in patients with total plexus palsy.13 In a later study, they compared 26 patients with OBPP who underwent resection of the neuroma–in-continuity and interpositional nerve grafting and 17 patients who underwent neurolysis only.14 The results showed that resection of neuromas-in-continuity is not harmful, and the alternative, resection followed by interpositional grafting, has the potential for providing superior results as opposed to neurolysis alone.
In our clinic, neurolysis is performed for nerves that have nearly normal gross appearance, have nearly normal function, and show bulging after removing the surrounding scar tissue, epineurium, and perineurium. This concept has been utilized only if procedures such as grafting and/or neurotization are performed in other areas of the same plexus at the same time. All neuromas-in-continuity were treated by resection followed by interpositional grafting. After resection of the neuroma, autologous nerve grafts are used to bridge available roots to distal targets. Possible donors for graft material are the sural nerves of both lower limbs and, if further needed, the superficial radial nerve of the affected side and supraclavicular nerves that are already in the surgical field may be used. To avoid stretching traumas we prefer harvesting the surals via a longitudinal incision beginning from the lateral malleolus to the popliteal fossa.
We routinely use fibrin glue, which is a biologic tissue adhesive found to be an effective sealant and topical hemostatic agent. It consists of concentrated fibrinogen and factor XIII combined with thrombin and calcium. It has been proved that this sealant material does not act as a barrier between the coaptation sites and therefore does not interfere with the regeneration of sprouting axons.15,16 Gaps are measured carefully and grafts are divided into pieces accordingly, then glued together to form cables for easier handling. Next, grafts are glued to the roots and to the targets followed by suturing each graft with only one stitch to enhance the strength and security. The use of fibrin glue considerably decreases operation and anesthesia time, which is of utmost importance, especially for infants.
Neurotization or nerve transfer is needed when the original proximal stump is not available. The spinal accessory nerve serves as an effective source that leaves less morbidity for neurotizations.17 If needed, other sources such as intercostal nerves, the phrenic nerve, and contralateral C7 should be kept in mind. The spinal accessory nerve is typically used to neurotize the suprascapular nerve. The most proximal branches to the upper portion of the trapezius are preserved, and the distal part of the nerve is coapted either directly or using a nerve graft to the suprascapular nerve.
If only one root is available, it is used for transfer to the upper±middle trunks. More grafts should be directed to the upper trunk because it supplies major targets at the shoulder and elbow. However, the middle trunk should not be ignored because it may play an important role in wrist extension, which is the sine qua non for a functioning hand. In this case the spinal accessory nerve can be transferred to the suprascapular nerve. If two or three roots are available, the suprascapular nerve is repaired from C5 using grafts. C5, again, is directed to the posterior division and C6 to the anterior division, which is the main supplier of the biceps muscle. Either root or, if suitable, C7 is used for the middle trunk.
The worst case we faced was a four-root avulsion injury (Fig. 1A). In this case, the use of the spinal accessory for the suprascapular is a must (Fig. 1B); then the remaining available root is used for the rest of the distal targets. Lower trunk or C8 and T1 roots are grafted first because the hand always has priority in OBPP reconstruction. Upper and middle trunks or lateral and posterior cords are repaired depending upon the distance needed for grafting (Fig. 1C and D). If the length of the surals is insufficient, the superficial radial nerve of the affected extremity is our second choice in infants. Although they are already in the surgical field, the supraclavicular sensory nerves are too short to overcome relatively long gaps. If there are two or more available roots, the strategy of repair is the same as noted for upper root palsy.
Following the surgery, the infant is placed in a thermoplastic custom-made brace in a position with the arm adducted and the elbow at 90 degrees flexion. A halo is also added to the brace to prevent uncontrollable head movements, which may disrupt the coaptations. The brace is worn for 3 weeks, and physical therapy and electrical stimulation are started right after removal of the brace.
The main concern of the parents is “which functions of the extremity will my child lose after the operation” rather than “which functions will he gain in the future?” Even with total palsy, many patients have at least some anterior flexion in the shoulder, which leads to optimism for the parents and causes unnecessary delay for the eventual operation. Sometimes it is difficult for us to explain that we may need to use these functioning nerves for other targets and in the immediate postoperative period the child might not move his shoulder. We studied postoperative regeneration time for shoulder and elbow flexion and found out that the children operated on have gained and exceeded preoperative shoulder abduction and elbow flexion values at about the fifth month postoperatively (Figs. 2 and and3).3). This study helped us to assure parents concerned the surgery. Our early outcomes of nerve surgeries are satisfactory for shoulder function and elbow flexion. The postoperative period has still not been long enough to determine hand functions, but our very first cases show encouraging improvements.
OBPP is not a static process. Therefore, as the child grows up, secondary deformities of the extremity occur, necessitating palliative surgery both in patients who received early primary neural reconstruction and in patients who did not have any surgical treatment at all.
The extent and severity of the deformity in obstetrical palsy may vary from patient to patient. Although a variety of elbow, forearm, and hand deformities are commonly seen in children who have sustained an obstetrical brachial plexus injury; the shoulder is the most frequently affected joint.
Between 1998 and 2003, 66 (41 male and 25 female) children with OBPP who had no surgical treatment before were operated on to restore shoulder abduction and external rotation (66 cases) and also other deformities of the elbow (16 cases), forearm (14 cases), and hand (29 cases). The average age at surgery was 7.75 years (2–16 years).
Involvement of the right side was seen in 37 patients and of the left side in 29 patients. No patient had bilateral involvement. All of the patients were vaginally delivered with vertex presentations. Obstetrical history revealed that most mothers of patients were multiparous and 16 of the patients were delivered with the assistance of forceps or vacuum. The mean birth weight of the patients was 4509.4 g (2500–6600 g).
The pool of the patients consisted of 18 patients with C5, C6 spinal root involvement, and 14 of the patients had additional C7 involvement. Finally, 34 of the patients had total brachial plexus root involvement. Accompanying birth complications were fractures of the clavicle in one case, fracture of the humerus in two cases, fracture of the scapula in one case, injury of the sternocleidomastoideus muscle in one case, and Horner's sign in six cases.
Internal rotation contracture causing limitation in abduction and external rotation with dislocation of the shoulder was the most common deformity observed during follow-up. The shoulder instability was not caused directly by obstetrical trauma but rather was related to a dynamic phenomenon of muscle imbalance, the most accepted factor for shoulder internal rotation contracture. It was discovered that the subscapularis muscle usually recovers sooner than the external rotators and abductors.18,19
Under these conditions, the main goal of treatment is to reestablish muscle equilibrium. The first step is to treat the muscle retraction and capsular contracture. Afterward, muscle transfer is performed to restore missing external rotation and abduction function.
Muhlig et al described a common policy followed by most of the centers.20 According to this, if passive external rotation of the shoulder remains less than 30 degrees, surgical treatment is indicated. If there is no posterior displacement of the humeral head, a subscapular slide is used. However, if there is posterior displacement of the humeral head, subscapular lengthening by an anterior approach is preferred. Indications for tendon transfer for improving external rotation and abduction are determined as well. If the infraspinatus muscle does not show signs of reinnervation by the age of 2 years, a muscle transfer should be added to the subscapularis lengthening to avoid recurrence. If there is a fixed medial rotation contracture and posterior subluxation of the humeral head with deformities of the glenoid, derotational osteotomy of the humerus should be added to the subscapularis lengthening.
In the literature there are several reports showing remarkable improvement in external rotation of the shoulder as well as abduction by using the Sever-L’Episcopo procedure and its modified forms.18,21,22,23,24,25,26
In our series we modified the Hoffer technique by performing extended dissection, rigid fixation, and insertion of the conjoined tendon of the latissimus dorsi and teres major to the selected insertion point.27 Several descriptions had been made to determine the center of rotation. According to Fischer, the center of rotation is located in the inferior part of the humeral head and to obtain elevation of the arm, the tendon transfers should pass outside the center of rotation.28 Thus, we inserted the conjoined tendon in front of the greater tuberosity of the humerus. We evaluated the results by measuring the preoperative and postoperative range-of-motion degrees of abduction and external rotation and using the Mallet classification and Gilbert-Raimondi assessment (Table 2).
Preoperative active and passive range-of-motion degrees of abduction and external rotation were measured, videos were taken during shoulder abduction and external rotation, and Mallet scores were noted. Abduction degree is measured in the standing position and external rotation is measured in the prone position with 90 degrees shoulder abduction and at 90 degrees elbow flexion. The overall abduction was 71.24 degrees (10–170 degrees) and external rotation was 23.36 degrees (0–80 degrees).
The Gilbert-Raimondi grading system was used to group the patients. Most of our patients were grouped in grade I (12 patients) and grade II (41 patients). We had one patient with grade 0, four with grade III, and eight with grade IV (Table 3). Radiograms of the shoulder in adduction and 90 degrees abduction and axial MRI of the shoulder were performed. Shoulder deformity was classified according to Waters-Peljovich grading system29(Table 4) and the patients with type I and type II deformities were included in the series.
The mean duration of follow-up was 32.7 months. The average abduction improved from 71.24 degrees (10–170 degrees) to 132 degrees (90–170 degrees) and external rotation from 23.36 degrees (0–80 degrees) to 80 degrees (45–95 degrees). The overall gain in abduction was 60.6 degrees (85.1%) and external rotation was 57.6 degrees (246.5%) (Fig. 4). Mallet scores showed increases from 3 to 3.9 for global abduction, from 2.5 to 3.9 for global external rotation, from 2.2 to 3.7 for hand to head, and from 2.5 to 3.4 for hand to mouth; the Mallet score for hand to back decreased from 2.6 to 2.1. All results were found to be statistically significant according to Student's t-test (p<0.01). According to the Gilbert-Raimondi grading system, all of the patients who were grouped as grade IV improved to grade V with a mean 36.2 degrees abduction and 54.6 degrees external rotation gain. In the grade III group two of the patients improved to grade IV, and the other two advanced to grade V (Table 3). The mean abduction gain was 28.7 degrees and the mean external rotation gain was 47.2 degrees.
In the grade II group two of the patients improved to grade IV and 39 patients showed grade V performance. The mean abduction gain was 62.1 degrees and the mean external rotation gain was 57.2 degrees. Eight patients in the grade I group rose to grade IV, four patients to grade V. The mean abduction gain was 73.2 degrees and the mean external rotation gain was 54.9 degrees. There was only one patient in the grade 0 group, which improved to grade V with 110 degrees abduction gain and 94 degrees external rotation gain.
Elbow deformities and dysfunctions are frequent sequelae of OBPP. The most commonly encountered problems include weakness of elbow flexion, weakness of elbow extension, elbow flexion contracture, and forearm supination contracture.
Although total absence of elbow flexion is rarely seen in this population, weakness in elbow flexion is a frequent problem. Although there are several reports of secondary neurosurgical reconstruction, also called “axonal supercharging,” such as utilizing the partial ulnar nerve Oberlin technique30 or medial/lateral pectoral nerves,31 reconstruction of elbow flexion relies mainly on classical muscle-tendon and functional free muscle transfers.
Reconstruction of elbow flexion requires more complex techniques than other standard transfers and reeducation is arduous following transfers such as triceps, latissimus dorsi, or pectoralis.32
Steindler flexor-tendoplasty, latissimus dorsi, pectoralis major, triceps, sternocleidomastoid transfers, and functional free muscle transfer are preferred methods in the literature.32 We performed the Steindler flexor-tendoplasty technique in nine cases that had good wrist extension. Reeducation after the Steindler procedure is relatively simple. The latissimus dorsi was used previously to restore shoulder abduction and external rotation. We achieved good results with an average of grade 4 according to the Mallet classification. In two patients transient median palsy occurred, but this complication subsided in a few months.
Numerous procedures for soft tissue release, including Z plasty of antecubital fossa skin, Z-lengthening of the biceps tendon, and release of lacertus fibrosis, are used to address this deformity.32
Most of the parents of these patients complain about passive elbow extension deficits because of the cosmetic problem. There is no absolute indication for surgery unless it is a severe flexion contracture because weakening biceps and diminishing elbow flexion power cannot be a price to pay to gain a few degrees of elbow extension. We have performed soft tissue contracture release operations without biceps lengthening in three cases.
Elbow extension weakness has been considered of little practical importance because gravity helps to extend the elbow in different positions. However, there are certain tendon transfers for other functions of the upper extremity that can be performed only if the patient has good triceps power. There are very well known techniques in the literature for the functional restoration of triceps function. Transfer of the posterior head of the deltoid muscle to triceps,33 transfer of the biceps to triceps,34 and pedicled latissimus dorsi or free gracilis functional muscle transfers are options to restore elbow extension.35,36 During rehabilitation of patients after the latissimus dorsi plus teres major tendon transfers who also had elbow extension deficit, we observed that they were more willing to use their hands above their heads if their elbow extension was supported with an elbow extension splint. Therefore we planned bipedicled transposition of the brachioradialis muscle to the triceps muscle as an “internal splint” in patients who have at least M4 brachioradialis and no elbow flexion contracture.
When the early results were evaluated, patients who had triceps strength less than M3 but not equal to zero showed an increase in angles for elbow extension and shoulder abduction. In one case who had zero triceps function preoperatively, the triceps function reached a value of M2. However, no differences in shoulder abduction and elbow goniometric measurements in the postoperative period were observed. Despite the fact that the force provided by the brachioradialis muscle is not excellent, gains at elbow extension using this procedure easily allow the individual to perform daily tasks.
Neither the strength nor the excursion of the brachioradialis muscle is ideal for the replacement of triceps. However, other alternatives in triceps reconstruction are associated with considerable morbidities. In selected cases, bipedicular transposition of the brachioradialis muscle onto the triceps is promising.
Paralytic supination deformity of the forearm is usually due to the imbalance between active supinator muscles (m. biceps, m. supinator) and paralyzed pronator muscles (m. pronator teres, m. pronator quadratus). In brachial plexus birth palsy, passive correction of the deformity is possible initially, but as the child's skeleton grows, the deformity becomes a fixed one because of contracture of the interosseous membrane. Once the deformity is fixed, bowing of the radius and ulna follows, which may eventually lead to volar subluxation of the distal end of the ulna and/or proximal head of the radius. Initial reports regarding the correction of the deformity focused on the osseous deformities. Blount introduced closed osteoclasis of the middle third of both bones of the forearm.37 Zaoussis indicated osteotomy of the proximal end of the radius.38 Later, soft tissue releases and tendon transfers were described.37 In 1967, Zancolli described a treatment that used a combination of biceps rerouting and interosseous membrane release.39 Later Manske et al reported combined application of biceps tendon rerouting with percutaneous osteoclasis in the treatment of supination deformity.40
One of the most commonly preferred current surgical techniques is rerouting of the biceps tendon around the radial head and releasing the interosseous membrane. This is usually done at the early stage to restore active pronation and to correct the deformity. However, the procedure is not recommended in cases in which the triceps is paralyzed because it can cause elbow flexion contracture.41 Biceps rerouting is also not applicable in cases in which the radial head has been resected because of dislocation.
We performed extensor carpi ulnaris to extensor carpi radialis longus (ECRL) transfer in six cases, brachioradialis rerouting around the radius with interosseous membrane release in four cases, and radius rotation osteotomy in four cases for forearm supination deformity.
We would like to mention our new surgical technique, brachioradialis rerouting and interosseous membrane release, which has advantages compared with the conventional methods of reconstruction. Brachioradialis rerouting provides an average of 78 degrees net gain in active pronation of the forearm, it is easier to perform than alternative procedures because biceps rerouting is technically demanding, and brachioradialis rerouting does not affect the integrity of the elbow joint. These results suggest that brachioradialis rerouting and interosseous membrane release may be a viable alternative to biceps rerouting procedures in brachial plexus birth palsy with supination deformity of the forearm with a congruent proximal and distal radioulnar joint.
An example of brachioradialis rerouting pronatoplasty is presented in Figure Figure66A–C.
Weak or absent wrist extension, metacarpophalangeal (MP) joint extension and/or interphalangeal (IP) joint extension, weak or absent finger flexion, and thumb opposition deficit are the major problems in OBPP patients with hand involvement.
If the forearm flexor-pronator group is powerful, these muscles can be transferred for wrist and finger extension with the traditional technique of tendon transfer for radial nerve palsy.
We performed flexor carpi ulnaris (FCU) to ECRL transfer in three cases for wrist extension, flexor digitorum superficialis (FDS4) in one case, FCU to extensor digitorum communis transfer in another for finger extension deficit; and FDS and palmaris longus transfer to extensor pollicis longus in another case for thumb extension.
If the forearm flexors are not powerful enough to be transferred, functional free muscle transfer for finger extension, ECRL to extansor digitorum communis (EDC) transfer in an end-to-side fashion, plication of EDC and/or abductor pollicis longus (APL), and wrist arthrodeses in older patients can be utilized to maintain finger extension or wrist stability.42 We performed APL plication in nine cases and wrist fusion in three cases.
In total plexus palsy, weakness or partial absence of finger flexion is almost always present. This problem can be managed by traditional extensor to flexor transfer when the patient has powerful wrist extension. We utilized ECRL through the interosseous route to FDP tendons in seven patients. The results are evaluated as moderate because it is difficult to achieve a useful grasp in an intrinsic minus hand. If the wrist extensor donors are not suitable; functional muscle transfers can be performed using either intercostal nerves or contralateral C7 as donor nerves.42
We have used the extensor indicis proprius in two cases and FDS4 in one case to restore thumb opposition.
There are questions regarding the timing of surgery for palliative reconstruction of the hand and the priority of surgical treatment in deformities effecting the shoulder, elbow, and hand.
It is accepted that the corrective procedures for the shoulder to rebuild the muscle equilibrium are best undertaken before permanent bone deformity occurs at 3 to 4 years of age.21 Active cooperation, easy clinical evaluation, and cooperation in the rehabilitation program are other factors that most authors favor after 4 years of age or greater as optimal timing for palliative surgery. In many centers, muscle release procedures are performed before the age of 2 years, but for older children tendon transfers to restore abduction and external rotation are added at an age when they can cooperate with postoperative physiotherapy. Muhlig et al reported that they could not achieve external rotation only with subscapularis lengthening procedure in 57 (68%) of 84 patients.20 Bone deformities, inactive external rotator muscles, and cocontraction of the internal rotator and external rotator muscles are thought to be the possible reasons for this situation.
All of our patients were older than 2 years, and latissimus dorsi plus teres major transfer was combined with subscapularis and pectoralis major release. Although the treatment must be individualized, shoulder and elbow reconstructions are usually performed before forearm and hand. It was stated that regardless of etiology, if the child cannot lift the arm because of internal rotational contracture, he or she is more likely to avoid flexion of the elbow. That is why shoulder deformity must be corrected before management of the elbow. We prefer to keep the proximal to distal route as well in terms of surgical intervention and perform shoulder and elbow reconstruction at 4 to 6 years and forearm and hand reconstruction at 6 to 10 years.
In total brachial plexus palsy, multiple stages are necessary for reconstruction of the total extremity. The severity of the skeletal deformity, age of the child, social stature of the child's family, and cooperation with the child affect the surgical plan. At times we had to perform reconstructive procedures for the shoulder and forearm at the same session because of the patient's social problems. When we performed tendon transfers for shoulder reconstruction in the same session with static procedures (e.g., osteotomy, tenodesis) of the distal forearm, we did not observe any difficulty in the postoperative physiotherapy. However, when we performed tendon transfers for both shoulder and distal forearm (e.g., restoration of pronation-supination of forearm and finger flexion-extension) reconstruction, our physiotherapists’ comments noted that adaptation of the child to the rehabilitation of multiple tendon transfers was difficult.
Chen et al asserted the need for an additional muscle transfer to the shoulder in patients who had less than 90 degrees abduction to increase the success of the classical latissimus dorsi plus teres major transfer.43 Usually they preferred the trapezius muscle. However, in our series, patients with ≤ 90 degrees abduction and no external rotation improved to grade IV and V. Therefore we are not totally convinced about adding trapezius transfer at the same stage with latissimus dorsi plus teres major transfer. This muscle is preserved for the patients who could not achieve enough shoulder abduction after the first operation.
In the past, grade III and IV shoulder function was considered to be sufficient. For these patients today, most of the authorities have accepted that a considerable amount of improvement could be obtained by combined muscle release and muscle transfer techniques.44 In our series even in grade IV shoulders, we obtained 36 degrees (30%) shoulder abduction and 54 degrees (186%) external rotation gain, which were better than the results of the grade III group, who had abduction gain of 29 degrees (31%) and external rotation gain of 47 degrees (106%). We state that in selected grade IV shoulders, tendon transfer can still improve shoulder function especially in terms of external rotation.
Although early neural reconstruction decreases the need for future reconstructive surgical procedures, the children can still have upper extremity deformities that should be treated by muscle releases and transfers. We believe that in children who could not receive primary early neural reconstruction, almost near-normal shoulder function can still be reached by combined muscle release and muscle transfer techniques that are performed before severe glenohumeral deformities occur. If the child is suffering from C5, C6, ±C7 palsy, utilization of muscles innervated by the lower trunk and elbow flexion and extension deformities can be managed by tendon transfers, but in total palsy cases, because of limited donor muscles, little can be achieved by means of tendon transfer for the hand and elbow. Other procedures, including osseous surgery such as wrist fusion and radius osteotomy for supination deformities, is used to improve cosmetic appearance.