Summary of Key Points
Pediatric spinal deformity can be a complex condition involving and affecting other physiologic systems along with the spine.
Obtaining thorough prenatal and perinatal history, detailed examination, assessment for other associated conditions, and pulmonary function are important for the diagnosis, treatment planning, and prognosis of pediatric spinal deformity.
Skeletal maturity may be predicted with greater accuracy by using more than one method. Monitoring and anticipating the deformity progression during the pubertal growth spurt is important. Conservative management is primarily used to delay curve progression and may help to delay surgical intervention until skeletal maturity.
Bracing in Adolescent Idiopathic Scoliosis Trail showed that bracing significantly reduced the progression of high-risk curves.
Pediatric spinal deformity (PSD) can result from congenital anomalies, neuromuscular disorders, genetic conditions, connective tissue disorders and skeletal dysplasia, and developmental (idiopathic) causes. The complexity of each category is dictated by the combination of underlying pathophysiology, associated comorbidities, and growth-related changes.
Scoliosis, kyphosis, and lordosis refer to deviations from normal spinal alignment. In the coronal plane, the spine is normally straight. In the sagittal plane, the thoracic region is kyphotic (range, 20 to 40 degrees), the lumbar region is lordotic, and the transition over the thoracolumbar region is relatively straight ( Fig. 158-1 ). Scoliosis, curvature in the coronal plane, is also associated with transverse rotation, as well as with pathologic lordosis or kyphosis ( Figs. 158-2 and 158-3 ). Therefore, the terms lordoscoliosis and kyphoscoliosis are frequently used to characterize the three-dimensional nature of a deformity. When more than one pathologic curvature exists along the length of the spine, the primary (or major) curve is designated on the basis of its size and rigidity. The secondary (or minor) curve(s), even if compensatory, may be rigid or have a “structural” component. Surgical planning should account for the magnitude and flexibility of all the curves in all three planes.
Growth of the Spine
Pediatric spinal deformities may become evident during growth periods. Therefore, anticipating and modifying the growth potential of the vertebral elements composing the deformity is essential. Two periods of rapid growth occur in children: the first between birth and 3 years and the second during the adolescent years. The timing and duration of the adolescent growth spurt can be determined by monitoring the growth velocity ( Fig. 158-4 ). The spine grows heterogeneously—that is, during the adolescent growth spurt, the thoracic spine grows 1.2 cm per year and, in contrast, the lumbar spine grows 0.6 cm per year. Measuring, estimating, and monitoring the changes in sitting height (thoracic curve growth) during growth spurts can be helpful for treatment planning. The presence of scoliosis and other vertebral anomalies (mainly apical) can exacerbate the multiplanar deformity during growth spurts. Other associated conditions, including neuromuscular disorders, that may affect the progression of the curve should also be considered in monitoring and management.
Predicting growth around the time of puberty is based on physical and radiographic examinations. In girls, Tanner stage 2, the development of pubic hair and breast buds marks the onset of the growth spurt and typically precedes menarche. Skeletal age at this stage is approximately 11.5 years. The growth spurt ends at a skeletal age of 14 years, or approximately 1.5 years after menarche. For boys, Tanner stage 3, when the pubic hair becomes curly, corresponds to the onset of the growth spurts. The skeletal age is approximately 13 years and continues until 18 years. A growth rate chart is the ideal means of monitoring growth, but realistically this is not always feasible to obtain. Otherwise, the surgeon should consider physical findings, such as the Tanner stage, in conjunction with historical information regarding the onset of menarche or the appearance of axillary hair in boys.
Several methods for the assessment of skeletal maturity have been reported, including Risser stage, presence of triradiate cartilage, and hand films. The Risser stage is a method based on the degree of iliac ossification, with stages 1 through 4 corresponding to sequential ossification of each quarter of the iliac crest from ventral to dorsal ( Fig. 158-5 ). Stage 4 reflects completion of spinal growth, and stage 5 is defined as fusion to the ilium. Alternatively, hand films can be obtained for the assessment of skeletal maturity, without the need to expose the pelvis to radiation, as is required for both the Risser stage and assessment of the triradiate cartilage.
The rib–vertebral angle difference (RVAD) is another measure that can guide decision making in pediatric spinal deformity. This measure is an important prognostic indicator for infantile scoliosis. The RVAD is the difference between the angles formed by a line along the rib head and perpendicular to the base of the apical vertebra on the right and left sides of the spine. Spontaneous resolution of the scoliosis is expected in 85% to 90% of the cases if the RVAD is less than 20 degrees, but progression is expected with an RVAD greater than 20 degrees. Use of more than one method in assessing skeletal maturity can provide a more accurate estimate. This is more important in situations where spinal deformity may have developed before the growth spurt to help prevent a possible crankshaft phenomenon with fusion surgery.
Initial evaluation should begin with a detailed history regarding the prenatal and perinatal events and cognitive and motor development progression since birth. Details of the suspected spinal disorder should be documented, including symptoms, deficits, onset, and progression, as well as disability and impact on quality of life. Past medical history can be a significant contributor, especially with congenital spinal disorders, which can be associated with other anomalies.
Physical examination should include assessment of the head, entire spine, and extremities, including the skin; it should also encompass a detailed neurologic examination, including strength, tone, gait, coordination, sensation, and physiologic and pathologic reflexes. For example, neurofibromatosis may be suggested by the presence of café au lait spots or freckling, and patches of hair or skin dimpling can evidence underlying anomalies such as diplomyelia or lipomeningocele. Nonambulatory patients should be examined for evidence of decubitus ulceration, which may affect surgical planning. Posture should be assessed and may include sitting, standing, and walking.
The scoliometer or inclinometer is used to quantify the rib prominence and paralumbar prominence. A scoliometer reading greater than 5 degrees is associated with a scoliosis of at least 10 degrees ( Fig. 158-6 ).
Initial evaluation of suspected spinal deformity often includes full-length (36-inch) posteroanterior (PA) and lateral spinal radiographs to assess global and regional spinal alignment. PA images evaluate for scoliosis, with measures including Cobb angle and coronal balance. Coronal balance is typically measured as the distance between a vertical line from the center of the C7 vertebral body (C7 plumb line) and the central sacral vertical line (CSVL) ( Fig. 158-7 ). Lateral full-length spinal radiographs can be used to assess for regional kyphosis and lordosis and for global sagittal alignment. Global sagittal alignment is typically measured as the distance between a vertical line from the center of the C7 plumb line and the dorsal rostral corner of the S1 vertebral body. If the C7 plumb line is ventral to the dorsal rostral corner of the S1 vertebral body, the sagittal alignment measure is reflected as a positive value, and if dorsal, it is reflected as a negative value.
The flexibility of scoliotic curves can be assessed with side-bending PA films to the left and to the right either in standing or in supine position with the patient placed over a bolster or with the use of traction. Supine bending radiographs have been suggested to be more useful in assessing the flexibility than standing films. The flexibility of kyphotic deformities can be assessed with a bolster placed under the apex of the kyphosis, and the flexibility of lordotic deformities may be assessed with the spine and pelvis placed in flexion. Traction-assisted radiographs are more useful in the setting of larger curves (Cobb > 65 degrees) and in patients with neuromuscular disorders with contractures. This information can provide an estimate of surgical curve correction and can help in the assessment of whether the secondary curves need to be included in the fusion.
Computed tomography (CT) imaging provides greater detail of the bony anatomy and a three-dimensional view of complex deformities; such information may facilitate planning of surgical treatment. CT can better define congenital deformities with underlying anomalies, such as hemivertebra or unsegmented bars, that may be occult or not well defined on plain-film radiographs ( Fig. 158-8 ).
In the setting of spinal deformity, magnetic resonance imaging (MRI) can be used to evaluate for central canal and foraminal stenosis, as well as for underlying abnormalities, which may warrant treatment or alter surgical planning. Associated abnormalities, such as tethered cord, syringomyelia, and tumors, occur in up to 15% to 38% of congenital spinal deformity patients ( Fig. 158-9 ). Although MRI is often used as an adjunct in the imaging evaluation of pediatric spinal deformity, several specific indications necessitate this evaluation, including severe pain; neurologic findings, including motor weakness, muscle atrophy, and upper motor neuron signs; early-onset scoliosis with a Cobb angle greater than 20 degrees; atypical scoliosis curve patterns (e.g., left thoracic curves, sharp angular curves, congenital deformities, and curves that are > 70 degrees); scoliosis curves with a rapid progression (> 1 degree per month); neurofibromatosis; deformity associated with myelomeningocele; and lack of apical lordosis in idiopathic scoliosis ( Fig. 158-10 ).
Idiopathic Scoliosis: Early Onset and Adolescent Onset
Idiopathic scoliosis has a familial tendency and a bimodal frequency distribution. With the early-onset type, the majority of cases occur in infancy and a second major peak arises during adolescence. Idiopathic scoliosis is divided into two groups: early onset (< 5 years) and late onset (5 years to skeletal maturity). The most common type is adolescent idiopathic scoliosis (AIS).
The diagnosis of AIS is one of exclusion. In idiopathic scoliosis, deformity is the most common reason for referral. Occasionally, low-grade, activity-related back pain can result from lumbar curves or curves greater than 40 degrees. Atypical cases demand further evaluation to establish a diagnosis.
The prevalence of AIS with a curve of at least 10 degrees is 2% to 3% of the adolescent population. Curves in excess of 20 degrees have a prevalence of 0.3% to 0.5%, whereas the prevalence of curves in excess of 40 degrees is 0.1%.
Curves originating before the age of 5 years can exceed 100 degrees and affect both the cardiac and pulmonary systems. Thus, treatments are predicated by patient age and degree of deformity.
Adolescent Idiopathic Scoliosis
AIS occurs more frequently in girls than in boys. The typical curve pattern is a right thoracic curve with a compensatory left lumbar curve. Patients typically present during adolescence with physical deformity from body asymmetry, such as a trunk shift, unlevel shoulders, rib prominence, and breast or waist asymmetry. Though there is no clear evidence that screening has changed clinical outcomes, as a new trial sponsored by the National Institutes of Health (NIH) showed the effectiveness of bracing, it may be useful to screen and treat early by primary care physicians.
The natural history of AIS was studied extensively by Weinstein and colleagues, who followed a cohort of patients for more than 40 years, and by Collis and Ponseti. Factors that correlated with progression of idiopathic curves included physiologic younger age, female gender, curve magnitude, and double curves. During the adolescent years, curves typically progress an average of 1 degree per month and curves in excess of 50 degrees have a high risk of progression even after skeletal maturity. Finally, there is a significant inverse relationship between the pulmonary vital capacity and magnitude of thoracic scoliosis. Thoracic curves > 80 degrees are associated with an increased risk of shortness of breath. A retrospective cohort study showed that the risk of curve progression to ≥ 30 degrees is highest if the initial presentation curve is already > 25 degrees and the growth spurt is not yet complete.
Lenke and coworkers developed the currently most commonly applied classification for AIS. Based on both coronal and sagittal radiographs, it was developed to aid in determination of the appropriate vertebral levels to be included in an arthrodesis. The classification system includes six curve types (numbered 1 through 6), a lumbar spine modifier (A, B, or C) that is based on deviation of the apical lumbar vertebra, and a sagittal plane modifier (−, N, or +).
The structural curve types are main thoracic (type 1), double thoracic (type 2), double major (type 3), triple major (type 4), thoracolumbar/lumbar (type 5), and thoracolumbar/lumbar main thoracic (type 6). The left and right side-bending radiographs determine which curves are structural and nonstructural. The lumbar spine modifier is determined on the basis of the relationship of the CSVL to the lumbar curve on the coronal radiograph. The lumbar modifiers quantify the amount of lumbar curve as A (minimal), B (moderate), and C (large). The CSVL either (A) runs through the lumbar vertebra to the stable vertebrae, (B) runs between the medial border of the lumbar concave pedicle and the concave lateral margin of the apical vertebrae, or (C) falls completely medial to the entire concave lateral aspect of the apical vertebrae. The sagittal thoracic modifier describes overall thoracic kyphosis: hypokyphosis (−), a curve less than +10 degrees; normal (N), a curve +10 to +40 degrees; or hyperkyphosis (+), a curve more than +40 degrees.
For curves less than 20 degrees, close observation is recommended, and bracing is appropriate if the curve is > 20 degrees and < 40 degrees mainly if the patient has not reached skeletal maturity. Bracing is the traditional nonsurgical treatment for idiopathic scoliosis; however, the evidence supporting its benefit is primarily from uncontrolled observational studies, and patient compliance varies significantly. There is no standardization of indications for bracing, how it should be used, or the optimal type of brace. Bracing is mainly used to delay the progression or to stabilize the deformity until the patient completes the growth spurt at puberty. Typically it is used for a scoliotic curve with an apex at or below T7. Thoracic lumbar sacral orthosis (TLSO) or a cervicothoracolumbosacral orthosis may be used.
Bracing may be indicated in patients with AIS for curves greater than 20 degrees that have progressed more than 5 degrees. Adolescents with significant remaining growth potential who have curves between 20 degrees and 40 degrees may also warrant bracing, despite no documented progression. Bracing is typically stopped in adolescents if the curve progresses to surgical dimensions (45 to 50 degrees) or when skeletal maturity is reached.
Weinstein and colleagues, in their NIH-funded multicenter, randomized and preference cohort bracing in Adolescent Idiopathic Scoliosis Trail (BRAIST), showed that bracing significantly reduced the progression of high-risk scoliosis curves to the surgical threshold. Of the 242 enrolled patients, 146 were braced and 96 were observed with average follow-up of 24.2 and 21.3 months, respectively. Inclusion criteria were AIS patients 10 to 15 years old, with major curve Cobb angles between 20 to 40 and riser grade of 0, 1 or 2. Primary analysis showed 72% success in stopping the progression of curves with bracing compared to 48% in the observation group. The majority of braced patients were treated with Boston-type TLSO. Longer brace time was associated with better outcomes. Notably, this trial was stopped early due to the high success rate of bracing.
Surgery for AIS is often recommended in growing children once curves reach 45 degrees. Surgery is also recommended for skeletally mature teenagers for curves greater than 50 degrees, because curves of this severity have a high risk of progression in adult life. The goals of surgery include stopping curve progression, restoring alignment, balancing the spine in all anatomic planes, and minimizing the number of vertebral levels that are fused ( Figs. 158-11 to 158-13 ). Current dorsal instrumentation systems include hooks, wires, or pedicle screws that are typically connected with dual rods ( Figs. 158-14 to 158-16 ). Pedicle screw instrumentation of pediatric patients in experienced hands has been shown to be safe. Regardless of the instrumented system selected, the ultimate success of the procedure depends on achieving solid bony fusion.