Congenital Anomalies of the Trunk and Upper Extremity

Congenital Elevation of the Scapula (Sprengel Deformity)

First described by Eulenberg in 1863, Sprengel deformity is characterized as a congenital upward elevation of the scapula in relation to the thoracic cage. The scapula is commonly hypoplastic and misshapen ( Fig. 31.1 ). Other congenital anomalies may be present, such as cervical ribs, malformations of ribs, and anomalies of the cervical vertebrae (Klippel-Feil syndrome); rarely, one or more scapular muscles are partly or completely absent. The presence of this deformity can often indicate abnormalities in other organ systems. The severity of the functional impairment typically is related to the severity of the deformity ( Table 31.1 ). If the deformity is mild, the scapula is only slightly elevated and is a bit smaller than normal and its motion is only mildly limited; however, if the deformity is severe, the scapula is very small and can be so elevated that it almost touches the occiput. The patient’s head is often deviated toward the affected side. The primary limitation of shoulder motion is abduction secondary to diminished scapulothoracic motion. In about half of patients, an extra ossicle, the omovertebral bone, is present; this is a rhomboidal plaque of cartilage and bone lying in a strong fascial sheath that extends from the superior angle of the scapula to the spinous process, lamina, or transverse process of one or more lower cervical vertebrae. Recognition of this abnormality is essential to surgical management. A similar osseous structure has also been reported extending from the medial border of the scapula to the occiput. Sometimes a well-developed joint is found between the omovertebral bone and scapula; sometimes it is attached to the scapula by fibrous tissue only. A solid osseous ridge between the spinous processes and the scapula is rare.

Radiographic workup is essential to surgical planning. Plain radiographs should be obtained to assess the level of the scapula in relation to vertebrae and in comparison to the contralateral side. Radiographs also can help recognize the presence of associated abnormalities such as the omovertebral bone.

In a morphometric analysis using three-dimensional CT, Cho et al. found that most of the affected scapulae in 15 patients with Sprengel deformity had a characteristic shape, with a decrease in the height-to-width ratio. An inverse relationship was found between scapular rotation and superior displacement; no significant difference was found in glenoid version. Cho et al. suggested that the point of tethering of the omovertebral connection, when present, may determine the shape, rotation, and superior displacement of the scapula and that three-dimensional CT can be helpful in delineating the deformity and planning scapuloplasty.

If deformity and impairment are mild, no treatment is indicated; if they are more severe, surgery may be indicated, depending on the age of the patient and the severity of any associated deformities. Because the deformity is more than just simple scapular elevation, the results of surgical treatment of Sprengel deformity can vary. The long-term function of the shoulder and cosmetic appearance must be carefully measured against the surgical risk and natural history of the deformity. A 26-year review of 22 patients with Sprengel deformity treated by either observation or surgical repair suggested that surgically treated patients had almost 40 degrees more abduction than their nonsurgical counterparts, with a subjective improvement in cosmesis.

An operation to bring the scapula inferiorly to near its normal position is ideally attempted soon after 3 years of age, because the operation becomes more difficult as the child grows. In older children, an attempt to bring the scapula inferiorly to its normal level can injure the brachial plexus.

Numerous operations have been described to correct Sprengel deformity. Green described surgical release of muscles from the scapula along with excision of the supraspinatus portion of the scapula and any omovertebral bone. The scapula is moved inferiorly to a more normal position, and the muscles are reattached. Other modifications include suturing the scapula into a pocket in the latissimus dorsi after rotating the scapula and moving it caudad to a more normal position, and avoiding dissection of the serratus anterior muscle so that mobilization is begun immediately postoperatively. Wada et al. performed a morphometric analysis and reported 23 scapulae in 22 patients treated with the modified Green procedure. At 4-year follow-up the patients demonstrated a 63-degree increase in range of motion.

Woodward, in 1961, described transfer of the origin of the trapezius muscle to a more inferior position on the spinous processes. Greitemann et al. recommended the Woodward procedure for patients with impaired function; for patients with only cosmetic problems, resection of part of the superior angle of the scapula was preferred. They suggested that better results are obtained with the Woodward procedure because (1) the muscles are incised farther from the scapula, which lowers the risk of formation of a scar-keloid that may fix the scapula in poor position; (2) a larger mobilization is possible; and (3) the postoperative scar is not as thick as with Green’s procedure. Borges et al. added excision of the prominent superomedial border of the scapula to the Woodward procedure. In a series of patients with long-term follow-up at an average of 14.7 years, Walstra et al. demonstrated improvement of Cavendish grade 3 to 1 or 2 and significant improvement in overall shoulder abduction and improved contrast; Disability of the Arm, Shoulder, and Hand (DASH) and Simple Shoulder Test (SST) scores also improved. No long-term complications were reported. We generally prefer the Woodard procedure (see later) ( Fig. 31.2 ).

To improve function of the shoulder and the cosmetic appearance, Mears developed a procedure that includes partial resection of the scapula, removal of any omovertebral communication, and release of the long head of the triceps from the scapula. In the eight patients in whom this technique was used, average flexion improved from 100 to 175 degrees and abduction improved from 90 to 150 degrees. In two patients, hypertrophic scars formed at the curvilinear incision; this problem was eliminated by the use of a transverse incision in subsequent patients. Mears observed that a contracture of the long head of the triceps seems to represent a significant inhibition to full abduction in patients with Sprengel deformity and that release of this contracture allows increased abduction. Early postoperative active and active-assisted motion exercises of the shoulder are used to improve function.

Brachial plexus palsy is the most severe complication of surgery for Sprengel deformity. The scapula in this deformity is hypoplastic compared with the normal scapula. During surgery, attention should be directed to placing the spine of the scapula at the same level as that on the opposite side, rather than aligning exactly the inferior angles of the scapulae. To avoid brachial plexus palsy, several authors recommended morcellation of the clavicle on the ipsilateral side as a first step in the operative treatment of Sprengel deformity. This is not a routine part of surgical treatment but is recommended in severe deformity or in children who show signs of brachial plexus palsy after surgical correction. In patients older than 8 years of age, a 2-cm midclavicular osteotomy is recommended to decompress the brachial plexus and first rib before scapular resection. In a younger patient, a 1-cm resection osteotomy is considered for severe deformity. Others have suggested the use of intraoperative somatosensory evoked potentials to monitor brachial plexus function during surgical correction.

Congenital Muscular Torticollis

Congenital muscular torticollis (CMT), present in as many as 3.92% of neonates, is caused by fibromatosis within the sternocleidomastoid muscle. A mass either is palpable at birth or becomes so, usually during the first 2 weeks. Congenital muscular torticollis is more common on the right side than on the left side. It may involve the muscle diffusely, but more often it is localized near the clavicular attachment of the muscle. The mass attains maximal size within 1 or 2 months and may remain the same size or become smaller; usually, it diminishes and disappears within 1 year. If it fails to disappear, the muscle becomes permanently fibrotic and contracted and causes torticollis, which also is permanent unless treated ( Fig. 31.4 ).

Although CMT has been recognized for centuries, its cause remains unclear. Clinical studies have shown that infants with CMT are more often the product of a difficult delivery and have an increased incidence of associated musculoskeletal disorders, such as metatarsus adductus, developmental dysplasia of the hip, and talipes equinovarus. There is a reported incidence of congenital dislocation of the hip or dysplasia of the acetabulum ranging from 7% to 20% in children with CMT. Careful hip screening and, if necessary, ultrasound evaluation are indicated.

Various hypotheses of the cause of CMT include malposition of the fetus in utero, intrauterine constraint, birth trauma, infection, and vascular injury. Davids et al. found that MRI of 10 infants with CMT showed signals in the sternocleidomastoid muscle similar to signals observed in the forearm and leg after compartment syndrome. Further investigation included cadaver dissections and injection studies that defined the sternocleidomastoid muscle compartment; pressure measurements of three patients with CMT that confirmed the presence of this compartment in vivo; and clinical review of 48 children with CMT that showed a relationship between birth position and the side affected by contracture. These findings led the authors to postulate that CMT may represent the sequela of an intrauterine or perinatal compartment syndrome ( Fig. 31.5 ).

A palpable nodule is present in the affected sternocleidomastoid muscle at birth or within the first few weeks of life. The patient may also have associated plagiocephaly and facial asymmetry. The presence of the characteristic fibrotic nodule typically confirms the diagnosis, rendering further radiographic evaluation unnecessary in most cases. When the diagnosis remains in doubt, or in the presence of neurologic findings, cervical spine radiographs or advanced imaging is appropriate. Ultrasonography also has been advocated for the evaluation and management of congenital muscular torticollis.

When CMT is seen in early infancy, it is impossible to tell whether or not the mass causing it will disappear spontaneously. Lin and Chou reported that ultrasonography was useful in predicting which infants would require surgical treatment. Those patients in whom fibrotic change was limited to only the lower third of the sternocleidomastoid muscle recovered without surgery, whereas 35% of patients with whole muscle involvement required surgical release. Han et al. found several differences between patients with a sternocleidomastoid lesion identified with ultrasound and those without a lesion. Those without a lesion generally were seen later but had a better prognosis and a shorter treatment duration. A head rotation/tilting pattern in which both occurred in the same direction was observed only in patients without a lesion, and head tilting was more limited than head rotation. These findings suggest that pathophysiologic mechanisms may differ between patients with CMT with and without a sternocleidomastoid lesion.

Only conservative treatment is indicated during infancy. The parents should be instructed to stretch the sternocleidomastoid muscle by manipulating the infant’s head manually. The child’s chin is rotated toward the shoulder on the side of the affected sternocleidomastoid muscle while the head is tilted toward the opposite shoulder. Excising the lesion during early infancy is unjustified; surgery should be delayed until evolution of the fibromatosis is complete, and then, if necessary, the muscle can be released at one or both ends. CMT typically resolves with a home stretching program during the first year of life. Some authors suggest a strong correlation between sternocleidomastoid (5-cm) thickness to duration of and response to stretching exercises. Canale et al. found that CMT did not resolve spontaneously if it persisted beyond the age of 1 year. Children who were treated during the first year of life had better results than children treated later, and an exercise program was more likely to be successful if the restriction of motion was less than 30 degrees and there was no facial asymmetry or if the facial asymmetry was noted only by the examiner. Nonoperative therapy after age 1 year was rarely successful. Regardless of the type of treatment, established facial asymmetry and limitation of motion of more than 30 degrees at the beginning of treatment usually precluded a good result.

Early intervention for infants with CMT, initiated before 3 to 4 months of age, results in excellent outcomes, with 92% to 100% achieving full passive neck rotation and 0 to 1% requiring surgical intervention. The earlier intervention is started, the shorter the duration of intervention, and the need for later surgical intervention is significantly reduced. Petronic et al. found that when treatment was initiated before 1 month of age, 99% of infants with CMT achieved excellent clinical outcomes (no head tilt, full passive cervical rotation), with an average treatment duration of 1.5 months, but if initiated between 1 and 3 months of age, only 89% of infants achieved excellent outcomes with treatment duration averaging 6 months. When initiated between 3 and 6 months of age, 62% of infants achieved excellent outcomes with treatment duration averaging 7.2 months. When intervention was initiated between 6 and 12 months of age, 19% of infants achieved excellent outcomes, with an average treatment duration of 8.9 months. In contrast to recommendations to provide stretching instruction to the parents when CMT is identified at birth and only refer to a physical therapist at 2 months of age if the condition does not resolve, more recent studies suggest that early physical therapy reduces the time to resolution compared with parent-only stretching, that stretching becomes more difficult in infants as they age and develop neck control, and that earlier intervention can negate the need for later surgery.

For more rigid deformities that persist beyond the age of 1 year or despite 6 months of physical therapy, surgical correction has historically been the treatment of choice. Continued physical therapy alone in more resistant CMT, even when performed carefully, can result in injury. Botulinum toxin (BTX) type A injection has been suggested to improve the results of physical therapy. The goal of BTX injections in CMT is to relax the tight soft-tissue structures enough so that a trained physical therapist can gain further correction of the contracture and also work on strengthening to maintain correction. Limpaphayon et al. reported that guided stretching of resistant CMT along with BTX type A injection resulted in resolution of deformity and maintenance of correction, with clinical improvement noted by 92% of caregivers. Sinn and Renaldi described successful treatment with BTX in 82 (61%) of 134 children.

Any permanent torticollis slowly becomes worse during growth. The head becomes inclined toward the affected side and the face toward the opposite side. If the deformity is severe, the ipsilateral shoulder becomes elevated and the frontooccipital diameter of the skull may become less than normal. Such severe deformity could and should be prevented by surgery during early childhood. Ideally, surgery is performed just before school age so that sufficient growth remains for remodeling of facial asymmetry while giving enough time for the growth of the structures to make surgical dissection and release easier. Many patients are first seen only after the deformities have become fixed, and the remaining growth potential is insufficient to correct them ( Fig. 31.6 ). Nevertheless, many authors have suggested that surgical release in older children can be successful and should be attempted even if the child presents later. The clinical results are significantly less successful in children who have finished growth than in children who still have growth remaining; however, most patients have marked improvement in neck motion and head tilt, with satisfactory functional and cosmetic results. Lee et al. reported marked improvement in craniofacial deformity after surgical release in 80 patients with CMT. Improved results were demonstrated if the release was performed before the patient reached 5 years of age. In their study of 31 patients older than 7 years with late-presenting CMT, Lepetsos et al. obtained 84% excellent and 16% good results with operative treatment, and Seyhan et al. reported marked improvements in neck motion and head tilt in patients between the ages of 6 and 23 years who had bipolar release.

Several operations have been devised to release the sternocleidomastoid muscle at the clavicle. Unipolar release of the muscle distally is appropriate for mild deformity. Bipolar release proximally and distally may be indicated for moderate and severe torticollis. Endoscopic release of the sternocleidomastoid muscle has been described, with suggested advantages of precise division of the muscle fibers, preservation of the neurovascular structures, and an inconspicuous scar; we have no experience with this technique and no large series have been reported.