30.1 Introduction

The axial skeleton not only functions as the central load-bearing column of the body but also assumes critical responsibility in the growth of the trunk and thorax and lung maturation. Disorders that affect the shape and growth potential of the vertebral column initiate a process of morbidity and potential mortality because the necessary environment for healthy lung maturation in the growing child cannot be established.

30.2 Spinal and Thoracic Growth

During the 13- to 15-year period between birth and adulthood, spinal growth does not proceed at a constant rate. Growth accelerates in two periods: the first 5 years of life, the first of which is the fastest, and the 1 to 2 years in the preadolescent period. ¹ Approximately 60% of sitting height comes from the vertebral column. During the period between birth and adulthood, the length of the spine will triple. A child of 5 years will have 62% of adult sitting height; at 10 years, they will have 78% of adult sitting height. In other words, if spinal growth ceases completely at age 10, the total loss of overall body height will be approximately 13%. However, in comparison, the thoracic volume matures at a much slower rate than the spine. The thoracic volume reaches 30% of adult volume at age 5 and approaches only 50% at age 10. Therefore, in the case of cessation of thoracic maturation in the longitudinal and circumferential planes at age 10, the loss of thoracic volume of the individual will be 50%. ¹ Lung development is directly related to the development of the thoracic cavity. During development, alveoli both multiply and enlarge. Although alveolar multiplication accelerates during the first 3 years to nearly reach maturation at age 8 (~300 million alveoli), the thoracic volume increases twofold between the ages of 10 and adulthood. Therefore, although the cessation of thoracic spinal growth has a negative effect of 20% on thoracic height, its effect on thoracic volume may depend upon corresponding growth of depth and width.

The elucidation of this direct relationship between thoracic and pulmonary growth and the demonstration of thoracic insufficiency with fusion procedures undertaken at young ages have increased the interest in the appropriate management of these deformities. ^{2 , 3} With the understanding that the thoracic volume grows at a much slower rate compared to the spine length, and that the spine must grow “normally” at least until age 10 to attain sufficient respiratory volume in adulthood, it has become necessary to redefine early-onset scoliosis (EOS). For this reason, surgeons have come together to extend the limits of EOS and to recommend consideration of growth potential when planning spinal deformity surgery in children younger than 10 years. ⁴

30.3 Surgical Alternatives

For spinal deformities in children younger than 10 years, growth-sparing instrumentation techniques are the norm, whereas for adolescent deformities, posterior instrumented fusion is generally accepted as the standard of care. However, in the gray area of children between 10 years of age and the adolescent period, the ideal approach to serious deformities that require surgical correction differs between institutions, surgeons, and individual patients.

30.4 Growth-Friendly Techniques

Vertical expandable prosthetic titanium rib (VEPTR), Shilla, and growing rods (GR) are common treatment alternatives utilized in EOS. Although examples of all three methods’ application to all kinds of EOS deformity can be found, growth guidance techniques tend to be preferred in children who are poor candidates for repetitive surgery due to serious comorbidities. VEPTR is most commonly used for congenital scoliosis in which the thoracic cage is also affected. GR are the most often used in children in whom the deformity mainly affects the spine, the thorax is unaffected, and the child is generally healthy. ⁵

There is no doubt that GR can control spinal deformities efficaciously while protecting spinal growth. Many studies performed in the past 10 years have shown that the correction obtained in the initial surgery can be maintained until the appropriate time for definitive fusion, and at that stage, if the deformity is beyond acceptable cosmetic ranges, posterior column osteotomies (PCOs) added to the definitive fusion and instrumentation procedure can achieve adequate correction. Although no formal fusion is attempted at the time of GR insertion (other than at the ends of the construct), repetitive distractions and spanning of a large portion of the spine with GR result in immobilization of these segments, leading to unavoidable spontaneous ankylosis or even arthrodesis. However, this situation does not appear to interfere with spinal growth for many years, and near-normal spine growth has been demonstrated. The technique is also successful in the restoration of sagittal plane alignment. Maintaining spinal height within normal parameters allows the thorax to develop appropriately and prevents the development of iatrogenic thoracic insufficiency. ^{5 , 6} A recent study indicated that patients with GR who have completed their lengthening procedures have no significant difference in pulmonary reserve compared to healthy adolescent idiopathic scoliosis (AIS) patients after instrumented fusion. ⁷

Despite these positive developments, consigning traditional GR patients to repetitive operations remains a daunting decision. It is possible to classify the problems associated with multiple surgeries into three main categories:

• 
  

  Likelihood of complications.

  
  
• 
  

  Psychosocial stress.

  
  
• 
  

  The lack of selective instrumentation and less correction compared to multilevel instrumentation and fusion.

Bess et al ⁸ have reported that the rate of complications increases with the increasing number of surgical sessions. However, this study included the retrospective analysis of multicenter data including a group of patients with diverse diagnoses operated on by different surgeons using different techniques. In the series where the standard GR technique is applied, where syndromic, congenital, neuromuscular, and especially hyperkyphotic patients with poor bone quality are excluded, complication rates are not as high as initially feared.

Another negative impact of repetitive surgical procedures is on the psychological health of these children. Previous studies comparing GR patients to their age-matched peers have revealed that these children have considerably more psychological symptoms. It is difficult to distinguish to what degree repetitive anesthesia sessions followed by painful procedures affect these scores versus the underlying disorder. However, the encouraging finding is that as time passes in the treatment process and children approach the end, they develop coping skills and form mechanisms that make it easier for them to deal with these symptoms. ⁹

As the GR is a distraction-based instrumentation and grasps the spine at two locations, the amount of correction initially obtained is directly proportionate to the flexibility of the deformity. This is an important problem especially for deformities that are severe and rigid. Again, the necessity to remain within the Harrington stable zone for instrumentation levels results in the need to include more spinal segments in the GR instrumentation construct compared to all-pedicle screw posterior fusion instrumentations where direct vertebral rotation maneuvers are employed. With GR constructs, selective instrumentation may be nearly impossible. As such, GR constructs often end up with longer ultimate fusions, but this is thought to be an acceptable trade-off depending on the additional spinal length gained with this strategy (Fig. 30‑1).

The use of magnetically controlled growing rods (MCGR) is one of the greatest advances in EOS management in the last 10 years. Its ability to achieve lengthening noninvasively at a rate more akin to normal biology without the need for anesthesia or an incision has generated great excitement. ⁵ Initial hopes were that with this technique, GR complication rates would be lowered, spinal growth would be better maintained, and all problems related to the GR would be eliminated. However, after a period of follow-up, this initial excitement was tempered by a continued need for unplanned surgery, increased difficulty in restoration of the sagittal plane, and its inability to show a meaningful improvement in health-related quality of life compared to the traditional GR. ¹⁰ Except for a decrease in surgical site infection from repetitive surgery, problems associated with the GR have not been completely solved despite this new technology.

Traditionally, GR patients were eventually converted to an instrumented fusion at some point when growth was complete or additional lengthening was not practical. These conversions to “final” fusion have been known to be difficult operations with limited ability to change the spinal alignment secondary to autofusion. A controversy now exists regarding whether such a conversion to fusion with reinstrumentation is actually required. In patients with acceptable alignment after their GR treatment, there is likely a role for a more reactive approach, only revising the instrumentation if a problem arises. The addition of MCGR to the field has complicated this decision to some extent as these implants are not Food and Drug Administration (FDA) approved for long-term implantation.

30.5 Instrumented Fusion

In patients who are older than 10 years but remain at Risser grade 0 with open triradiate cartilage, some experts recommend proceeding with a final fusion with maximum correction instead of initiating a difficult treatment process with the GR. This may result in less spinal length but often results in better ultimate deformity correction. ¹¹ A typical patient who will undergo this option has a surgical curve (>50 degrees) at the age of 8 or 9 but is being delayed until after age 10. Ideally, the patient is being treated with a brace to maintain some control over the curve. It is not uncommon for the curve to progress, especially if the patient hits his or her peak height velocity. ^{12 , 13} Therefore, early fusion is commonly performed in patients with severe and rigid deformities (Fig. 30‑2).

Modern segmental instrumentation techniques allow correction up to 80 to 90%, and this acute and remarkable correction results in a meaningful increase in height despite a loss of future spinal growth in the instrumented segments (~0.7 mm/y for thoracic levels and ~1.1 mm/y for lumbar levels fused). The data that 80% of spinal height has been achieved at age 10 are supportive of this option, especially if a significant number of levels of the spine can be spared fusion. The hope for a single operative intervention is attractive. The fact that the patient and family need not live for years in anxiety and at risk for developing a GR complication is also a positive point for both families and surgeons. Given the degree of deformity correction achievable with modern posterior instrumentation, the point at which surgical treatment might be recommended has also increased. Delaying surgery despite curve progression may allow a patient to complete “enough” growth to proceed with a standard posterior spinal fusion. The spinal length considered to be sufficient will vary based on the patient, family, and diagnosis. Based on T1–T12 height, a value of 20 to 22 cm is considered “sufficient.”

Besides inhibition of spinal growth, the potential for continued growth across the instrumented spine causing crankshaft is also a concern. Crankshaft results when continued anterior spinal growth is tethered by a posterior fusion, resulting in progressive spinal deformity especially in the axial plane. ¹⁴ To prevent this, early studies recommended the addition of an anterior fusion to any patient with significant growth potential. This added morbidity to the scoliosis treatment, especially if the procedure was performed open. Significant decreases in pulmonary function testing (PFT) is observed in patients treated with open anterior thoracic fusions when compared to posterior-only procedures. ^{15 , 16 , 17} Some authors have advocated a thoracoscopic approach to decrease some of the negative effects of an open approach. ^{18 , 19}

The use of segmental pedicle screws has made many surgeons question the need for an anterior fusion even in a very young patient. By engaging all three columns of the vertebra, the thought is that pedicle screws can suppress the growth that may occur from the vertebral body endplates. In addition, the improved axial correction may also minimize the risk of the apical vertebra spinning out that is typically seen with crankshaft.

When Sponseller et al ²⁰ first studied fusion in AIS patients with open triradiate cartilages, they found significant late increases in the deformity following posterior-only surgery. They did not see this deformity progression in patients treated with combined anterior and posterior spinal fusion. None of the patients, however, had all-pedicle screw instrumentation in this early study. A subsequent study investigating a similar group of patients treated by posterior-only spinal fusion with pedicle screws demonstrated a similar outcome with increased risk of curve progression compared to a combined approach. ²¹ In this study, the selection of the distal fusion level was also determined to be a significant risk factor for progression. Patients with open triradiate cartilage fused short of the stable vertebra progressed with a posterior-only fusion. Therefore, pedicle screws may not protect from late deformity progression in very young patients, and extra attention is needed when selecting the distal fusion level.

The concerns for late progression do not necessarily suggest that early fusion is bad for all patients. Especially for neuromuscular patients who have added risk in repetitive procedures due to comorbidities, and for whom three-planar fixation is critical due to poor bone stock, deformity characteristics, and rigidity, the definitive fusion option is a good alternative. ²² In this population, growth is slower compared to the normal population, the possibility of crankshaft is lower, and the possibility of strenuous physical activity in the future is low, so they do not require the same pulmonary capacity/reserve as a patient with idiopathic scoliosis. However, for patients with idiopathic scoliosis who are otherwise healthy with an expectation of normal height and activity level in life, controversy remains with many questions as yet unanswered regarding the efficacy, safety, and the point at which an “early” definitive fusion is better than a growth-sparing approach that delays the timing of the “final” fusion.

The sole study in the literature where these two methods are compared was published by Pawelek et al. ¹¹ Utilizing two large multicenter databases, a one-to-one patient match was performed comparing GR treatment to a cohort undergoing definitive fusion. The patients were matched based on age, major curve magnitude, and curve apex. As expected, the GR patients underwent significantly more surgical procedures, 54 compared to 13. A better correction was achieved in the posterior fusion group although there was no significant difference in total spinal growth achieved. There were no significant differences in the number of complications requiring unplanned surgery. The limitations of this study are (1) the low number of patients (11 in each group) and (2) that some patients had not reached skeletal maturity. Although the findings suggest that GR treatment does not benefit older patients with juvenile idiopathic scoliosis, the limitations make it difficult to draw hard conclusions regarding the safety and efficacy of this technique. In addition, this study does not account for the decreased surgical interventions seen with MCGR.