Etiology

Infantile scoliosis occurs roughly in 1 of 10,000 births. Possible causes are thought to occur from intrauterine molding or postnatal pressure on the spinal column from supine positioning while sleeping. Other etiologic factors that have been considered in idiopathic scoliosis include dysfunction in proprioception to maldevelopment in central pattern generators in the spinal cord^9–11 and connective tissue, hormonal, and muscle structural changes.¹² More recent reports in the literature strongly suggest a genetic link. The growth spurt noted among adolescents seems to play a role in progression, as a critical buckling load is reached on the existing curve as the spine grows. In a review of the literature, Kouwenhoven and Castelein¹³ concluded many factors may play a role in the initiation and progression of adolescent idiopathic scoliosis at a certain age. The literature suggests, however, that in the observed deformation of the spine, genetics and the unique mechanics of the fully erect posture, which is exclusive to humans, play an important role.

Genetics

The genetics of idiopathic scoliosis seem to be similar in all idiopathic groups. The incidence rate is 11% among first-degree relatives, 2.4% among second-degree relatives, and 1.4% among third-degree relatives.¹⁴ Concordance among monozygotic and dizygotic twins has been reported to range from 73% to 92% and 36% to 63%.^15,16 In 2007, Anderson and colleagues¹⁷ published a population-based study taken from the Danish Twin Registry in which 46,418 twins were registered. From the 34,944 respondents, the concordance rate for monozygotic twins was 13% versus 0% for dizygotic twins. In 2007, Gao and colleagues¹⁸ published a report providing evidence of linkage and association with 8q12 loci. Through further investigation, they discovered the first gene (CHD7) associated with a susceptibility to idiopathic scoliosis. In 2008, Kulkarni and colleagues¹⁹ mapped a developmentally critical CHD gene to 15q26.1 and when taken together with CHD7, this suggested a possible role for CHD2 in the embryonic development of the spine. More recently in 2009, Gurnett and colleagues²⁰ published a report of a single multigenerational family in which adolescent idiopathic scoliosis and pectus excavatum segregated as an autosomal dominant condition. Through linkage analysis, the investigators identified a genetic locus for the two conditions on chromosome 18q. Although genetic mapping for idiopathic scoliosis is still in its infancy, great strides are being made to help clinicians better understand the etiology and prevalence of this condition.

History and Physical Examination

A complete history and physical examination is completed, and any family history of scoliosis is noted. With infantile and juvenile cases in particular, a thorough prenatal and birth and developmental history is obtained. In adolescent cases, growth spurt history, if any at the time of presentation, is noted. This information is imperative in determining peak growth velocity and its implications on curve progression. Symptoms of pain or weakness and how the patient perceives his or her appearance relative to the deformity are especially important with adolescent idiopathic scoliosis.²⁵ Age at onset of menarche and voice changes in boys are noted as well because they are likely predictors of growth potential and possible curve progression.

During the examination, height, weight, and age (years plus months since last birthday) are recorded. The head is examined with special attention to torticollis and plagiocephaly because the latter has been associated with higher incidence in infantile scoliosis. Possible conditions and anomalies that might be present are buccal and palatal anomalies; café au lait spots; and midline dimples or hair patches or both over the lumbodorsal spine, which can be important clinical clues that an intraspinal pathologic process might be present. Limb laxity is also checked, and genetic counseling and testing is requested when laxity is present.

Trunk shift is evaluated with the patient standing and the hips and knees fully extended. The relationship of the patient’s head to the pelvis is also noted in evaluating the overall coronal and sagittal balance. Any shoulder, breast, or pelvic asymmetry is noted.²⁶ Curve rotation is assessed by performing an Adams forward-bend test and is quantified with a scoliometer. This assessment is modified in infants by laying the patient on the examiner’s knee. This test also helps assess the rigidity of the curve, which is an important factor in terms of prognostication. Leg-length discrepancy and pelvic obliquity are evaluated. Alternatively, a sitting forward test can be performed. The latter maneuver can also help rule out plagiocephaly and developmental hip dysplasia, especially in infants. When leg-length discrepancy is the likely cause of the deformity, a shoe lift is used to reevaluate the patient to determine if the curve corrects.

A thorough neurologic examination is performed. The neurologic examination includes all cranial nerves; motor strength; reflexes (including abdominal reflexes), often associated with Chiari malformations; sensory modalities; and gait.²⁷ Finally, other possible causes of scoliosis, such as congenital, neuromuscular, and syndromic types, must be ruled out. Infection, neoplasms, and spondylolisthesis also must be discounted.

Radiographic Evaluation

Posteroanterior and lateral 36- × 14-inch long cassette views including bending films—with appropriately placed bolsters—for further curve classification and planning bracing or surgical intervention are obtained, and the Cobb angles are measured.⁵ Curves greater than 20 degrees in infants and children, neurologic symptoms in all patients with idiopathic scoliosis, and left-sided, sharp angular or irregular curve patterns require further investigation, including screening total spine magnetic resonance imaging (MRI).^28,29 When anomalies of the nervous system are present on MRI, a neurosurgical consultation is indicated.

Three-Dimensional Classification

Although the Lenke classification system has criteria in the coronal and the sagittal planes that have an impact classification, it still has shortcomings because scoliosis is a known three-dimensional deformity, and the axial or transverse plane is not a component of this system. A task force of the Scoliosis Research Society has been charged with developing a clinically useful three-dimensional analysis and ultimate classification of scoliosis.⁴⁰ The key factor in this assessment would be the plane and maximal curvature, which is the three-dimensional deformity that occurs as the spine translates and rotates out of the normal sagittal profile in scoliosis deformities. It is expected that in the next decade much more information will be provided so that three-dimensional analysis and classification will become a standard for all practitioners.

Observation

Most infantile curves are left-sided; these curves have been known to resolve spontaneously up to 90% of the time, but they can progress.⁴¹ Deciphering which curves will progress can be guided by the RVAD and the relationship of the apical rib head to the vertebral body (i.e., phase I or II).²² Typically, curves less than 20 degrees are expectantly followed every 6 to 8 months. Infants with curves less than 25 degrees and RVAD less than 20 degrees and children with curves less than 25 degrees should be followed clinically and radiographically every 6 months. Treatment is instituted for curves greater than 25 degrees. Treatment is also started for a progression of 5 degrees or greater in two consecutive visits or 10 degrees or greater in one follow-up visit. Juvenile curves more often require operative intervention, however.

Bracing and Casting

Bracing⁴⁴ is the nonoperative treatment of choice in small but progressive scoliosis in growing children and teens. In about 75% of cases, bracing can control the curve and avoid progression, rendering the curve small enough so that the risk of progression after growth is unlikely.⁴² In a younger child, whose growth potential remains a significant issue, bracing allows for continued growth until the patient requires eventual operative treatment because of curve progression. With infantile cases, molded casting followed by bracing used to be the mainstay in nonoperative management. Bracing and casting of these patients comes with potential consequences, however, that include pulmonary restriction, which can have future ramifications.^45,46 Sanders and colleagues⁴⁷ found serial casting to be beneficial in the treatment of infantile scoliosis. They reported that curves less than 60 degrees often fully corrected in infants if casting was started before age 20 months.

Adolescents with curves less than 20 degrees at presentation are observed and followed at 4-month and 6-month intervals. For curves between 20 degrees and 30 degrees, bracing is started if a curve progresses 5 degrees or more in two consecutive visits or 10 degrees or more in one visit.

Bracing is usually started the first office visit when the patient is skeletally immature (Risser ≤2) and presents with a 25- to 40-degree curve. Several brace options exist, and deciding which brace to use depends on the apex of the curve and physician preference. Curves with an apex above T6 would likely require the use of a Milwaukee (cervicothoracolumbosacral orthosis) brace.⁴⁸ Conversely, curves with apices at T7 or below and above L2 do well in a Boston underarm thoracolumbosacral orthosis brace, and these braces are more socially acceptable because of the lack of a cervical extension. The Charleston bending brace is an option if the child is noncompliant to brace-wear. This brace is typically worn at night, and some studies have shown its efficacy.^49,50 The efficacy of a brace seems to depend on the length of time the brace is worn.⁵¹ When bracing is initiated and pad placement is deemed appropriate, patient follow-up occurs every 4 to 6 months, with in-brace radiographic evaluation and appropriate fitting adjustments made when necessary.

Modifications to the standard thoracolumbosacral orthosis include variations of the Chêneau brace (Jacques Chêneau, Münster, Germany) and the SpineCor dynamic brace (SpineCorporation, Chesterfield, UK). The Chêneau 2000 orthosis allows for a greater amount of initial correction by using a hypercorrected mold and pads, which provide derotational forces.⁵² This brace is the first that uses the theory of expansion to allow for active correction by respiratory movements.⁵³ The SpineCor⁵⁴ and TrIAC (Boston Brace International, Avon, MA) are nonrigid braces. They work by using straps, which correspond to a specific correcting movement depending on the curve pattern, producing a progressive positional change, dynamic curve correction, and appropriate muscle balance.

Surgical Techniques

Currently, anterior-only, posterior-only, and circumferential procedures are the mainstay of surgical treatment options.⁵⁷ The prevalence of anterior-only and circumferential procedures has declined with a concomitant increase in posterior-only procedures. Surgeons have become aware more recently that early intervention with a definitive anteroposterior fusion for progressive infantile and juvenile curves leads to loss of trunk height development, which can lead to chest wall and lung underdevelopment.^58,59 This problem has promoted innovative techniques to try to control progressive curves surgically without definitive fusion, which include epiphysiodesis,^60,61 growing rod placement (Figs. 23-2 and 23-3),⁶² intervertebral stapling (Fig. 23–4),⁸⁶ spinal tethering (Fig. 23–5),⁶⁶ and the vertical expandable prosthetic titanium rib⁶⁷; the last-mentioned is used more in progressive early-onset scoliosis in which rib and chest wall deformities can be quite severe.⁶⁴

FIGURE 23–2 A and B, A girl age 2 years + 1 month who presented with 77-degree left thoracic infantile idiopathic scoliosis. C and D, She was treated with a posterior single hook growing rod with only partial correction of her deformity and increasing sagittal kyphosis. E and F, Ultimately, she underwent posterior spinal fusion ending up with overall acceptable coronal and sagittal balance 5 years after fusion.

FIGURE 23–3 A and B, A girl age 2 years + 9 months who presented with severe infantile-onset idiopathic scoliosis. Her left thoracic curve measured 122 degrees. C and D, She was placed in halo-gravity traction, and underwent a short apical anterior release and fusion and was prepared for a growing rod construct. E and F, She had a dual rod, pedicle screw growing rod construct placed, and at 5 years + 6 months after initiation of growing rod construct, she continues to be lengthened with overall good coronal and sagittal balance and acceptable lung fields produced via this treatment. G and H, Clinical photographs of preoperative and latest lengthening show maintenance of trunk alignment and growth.

FIGURE 23–4 A and B, A boy age 7 years + 9 months who presented with progressive juvenile-onset right thoracic idiopathic scoliosis. His main thoracic curve measured 60 degrees and was progressive despite bracing. C and D, He had anterior thoracic stapling performed with slow progression of deformity with growth. E and F, A posterior dual screw-rod growing construct was placed, and 5 years after insertion, his deformity correction has been maintained to 18 degrees of the main thoracic curve with good sagittal profile. G-I, Clinical photographs before surgery, status post stapling, and 5 years status post growing rod construct show improvement of truncal deformation.

FIGURE 23–5 A and B, A girl age 8 years + 8 months who presented with progressive left thoracic scoliosis. She had a positive family history of scoliosis with her mother requiring scoliosis fusion as a child. Her left thoracic curve progressed to 25 degrees with normal sagittal profile. C and D, She was treated with a single left thoracic mobile tether, with slow progressive correction of her deformity to 6 degrees with a normal sagittal profile 4 years after treatment. She had only one surgery and did not wear a brace postoperatively.

Many surgeons today still prefer to perform an anterior approach in younger patients when there is risk of crankshaft development and especially for thoracolumbar and lumbar major curves (Lenke type 5). With the introduction of pedicle screws, a posterior approach has shown numerous benefits over an anterior procedure, such as better maintenance of the obtained correction, providing more powerful corrective forces, three-column control, and obviating the need for anterior releases and thoracoplasties.^68–74 In addition, a posterior approach avoids the negative consequences of chest cage disruption and pulmonary compromise that can result from an open anterior approach (Fig. 23–6).^75–77 In addition, the thoracic aorta is at risk if an anteriorly placed screw penetrates the vertebral cortex on the opposite side.^78,79

FIGURE 23–6 A-D, A girl age 12 years + 8 months with severe right thoracic adolescent idiopathic scoliosis deformity. Her curve progressed to 110 degrees bending to 66 degrees with overall 44 degrees of thoracic kyphosis—a 4A+ classification. E and F, She underwent open anterior release and primary posterior hook construct from T2-L3. At 6 years postoperatively, adequate coronal and sagittal alignment and balance is shown. G and H, Preoperative and postoperative clinical photographs show improved truncal correction with right thoracotomy, posterior midline and iliac crest scars visible.