Advertisement

Surgical Management of Kyphosis and Hyperlordosis in Children with Cerebral Palsy

  • Kirk W. DabneyEmail author
Living reference work entry

Abstract

Sagittal plane spinal deformities, excessive kyphosis or lordosis may occur in the child with cerebral palsy (CP) and may be seen either with or without scoliosis. These spinal deformities may cause difficulty with seating and/or pain, especially when the deformity is greater than 70 degrees. And hyperlordosis has also been reported to cause superior mesenteric artery syndrome. Pain and difficulty with seating are the most common indications for correcting sagittal plane spinal deformities in the child with CP. Mild and some moderate sagittal plane deformities can be treated by wheelchair modifications and bracing. Symptomatic moderate and severe deformity may require surgical treatment. More flexible kyphosis and hyperlordosis can be corrected by posterior spinal fusion and segmental instrumentation alone while rigid deformity may require posterior osteotomies (for kyphosis) or anterior discectomies (for hyperlordosis). Instrumentation and correction techniques vary from screw/rod constructs using distraction/compression correction to wire or screw/rod constructs using cantilever correction. Overall, natural history and surgical outcome studies focused solely on sagittal plane spinal deformities in the patient with CP are limited. Those authors that do measure functions report improvements in pain, sitting balance, head and neck control, breathing, and hand use. Patients with kyphosis undergoing spinal fusion with instrumentation are at risk for loss of proximal and/or distal fixation. Patients with hyperlordosis appear to be at greatest risk for postoperative complications.

Keywords

Cerebral palsy Neuromuscular Kyphosis Hyperlordosis Sagittal plane Spinal deformity Spinal fusion 

Introduction

Sagittal plane spinal deformities, excessive kyphosis or lordosis may occur in the child with cerebral palsy (CP) and may be seen either with or without scoliosis. These spinal deformities may cause difficulty with seating and/or pain, especially when the deformity is greater than 70 degrees. And hyperlordosis has also been reported to cause superior mesenteric artery syndrome (Lipton et al. 2003; Karampalis and Tsirikos 2014). Pain and difficulty with seating are the most common indications for correcting sagittal plane spinal deformities. Separate surgical strategies are necessary to correct each deformity (Dabney et al. 2004). Similar to scoliosis, the surgical outcomes of corrective spine surgery in the CP child may be adversely affected by associated comorbidities in the CP child. Optimum medical management is therefore important to minimize adverse postoperative outcomes.

Etiology/Pathogenesis/Natural History

Similar to scoliosis in cerebral palsy, sagittal plane deformities are a consequence of muscle imbalance. Iliopsoas contracture can cause lumbar hyperlordosis, severe anterior pelvic tilting, and a horizontal sacrum. On the other hand, McCarthy et al. showed an association between loss of lumbar lordosis or even frank lumbar kyphosis with hamstring contracture that may cause a posteriorly tilted pelvis and a prominent vertically oriented sacrum (McCarthy and Betz 2000; Dabney 2018). Lumbar kyphosis can result in a greater weight-bearing load shifted on to the sacrum resulting in a sacral pressure sore. Truncal hypotonia and poor head control can initially result in postural thoracic kyphosis eventually resulting in a more rigid kyphosis over time. Currently, there is insufficient research to ascertain whether or not the severity of motor involvement is directly proportional to the severity of sagittal plane deformity in CP, although our clinical experience tells us that this is likely the case. Understanding whether or not quality of life is impacted by the sagittal plane spinal deformity should be the main determinant and rationale for treatment.

No current natural history studies exist for either kyphosis or hyperlordosis. On the other hand, the CP child’s life expectancy is largely dependent on the number of comorbidities. Accordingly, the surgeon must recognize if the sagittal plane spinal deformity is having a significant impact on the child’s quality of life or not. In addition, the number and severity of comorbidities may impact the surgical timing and perioperative risks associated with spinal surgery.

Patient Assessment and Preoperative Considerations

After a shared decision-making process between the medical care provider and the family to determine whether to proceed with surgery, all children with CP should have a detailed preoperative medical evaluation similar to that of scoliosis in the child with CP. The surgeon and interdisciplinary medical team must ensure that all associated comorbid conditions are optimized medically. Common comorbidities that exist include gastroesophageal reflux, aspiration pneumonia and reactive airway disease, poor nutrition, seizure disorders, and low bone mineral density and may be risk factors for postoperative complications (Lipton et al. 1999; Samdani et al. 2016; Nishnianidze et al. 2016). Risk factors should be identified preoperatively and comanaged medically (Rappaport et al. 2013a, b; Rappaport and Pressel 2008). The child with CP should also have a detailed neurological examination, including sensory and motor testing, the assessment of upper and lower extremity reflexes as well as abdominal reflexes. This helps to establish the child’s baseline neurological function and serves to assess for any undiagnosed intraspinal pathology such as tumor, tethered cord, or syringomyelia which may also occasionally occur in the child with CP.

The orthopedic preoperative assessment should evaluate sitting and/or standing ability, as well as coronal, sagittal, and rotational deformity of the pelvis. Also, flexibility of the spinal and pelvic deformity is very important to evaluate. The flexibility of kyphosis can be assessed by either physical and/or radiographic examination by placing the child in the supine position and placing a bolster directly under the apex of the kyphosis to determine how well the kyphosis corrects. The flexibility of a hyperlordosis is more difficult to assess, which, however, can be assessed by gently flexing both hips and thighs to the chest with the child in the supine position. Radiographs can also be taken in this position to evaluate flexibility. For better accuracy, the surgeon should position the patient for these radiographs in the x-ray suite instead of allocating them to the x-ray technician. The surgical treatment of more rigid versus stiffer sagittal plane deformities will be discussed later in the chapter.

The assessment for the coexistence of sagittal spinal deformity, hip flexion contracture, and/or adduction contracture is also important. This can be done by performing the Thomas Test, done by stabilizing the pelvis in a neutral position by flexing the opposite hip and assessing for hip flexion contracture on the opposite side. Alternatively, assessing for the presence of hip adduction contracture can be achieved by measuring the amount of abduction achieved with the pelvis in neutral obliquity. If these contractures are present, the parents should be advised that muscle releases may be needed 4–6 months after corrective spine surgery in order to balance infra-pelvic deformity. Also, assessment with an AP radiograph for the presence of hip subluxation should always be done in patients with spinal deformity.

The surgeon should also assess if the pelvis is part of the kyphosis (posterior pelvic tilt) or lordosis, anterior pelvic tilt in order to determine whether instrumentation and fusion to the pelvis will be necessary. In the nonambulatory patient with cerebral palsy (GMFCS VI and V), fusion to the pelvis is almost always necessary to prevent distal extension of the deformity especially when lumbar kyphosis or hyperlordosis are present. Also, patients with poor head control and/or thoracic kyphosis should be considered for instrumentation and fusion up to T1 or T2 to prevent a junctional kyphosis at the cervical-thoracic junction. Patients with very proximal kyphotic deformities may even require instrumentation and fusion into the lower cervical spine.

Another special high-risk group is children who develop severe lumbar lordosis following posterior dorsal rhizotomy for spasticity management. The lordosis component is often very stiff combined with a dense posterior surgical scar. Often the intrathecal nerve roots are also scarred. Correction with strong posterior stretching forces a center of rotation of deformity correction to be in the anterior aspect of the lumbar vertebra requiring large excursion of the nerve roots which will lead to nerve root nerve pain. It is strongly advised in the situation to first focus on anterior spinal shortening with large disk vertebral end plate wedge resections or vertebral body resection. This spinal shortening will allow lordosis correction without stretching the intrathecal neural elements.

Nonoperative Treatment

Postural (flexible) kyphosis treatment in CP may be accomplished initially by using wheelchair modifications such as a tilt-in-space wheelchair with a chest harness to stop the trunk from leaning forward (Fig. 1a). In addition, a wheelchair tray table (Fig. 1b) may help prevent forward lean, while a Hensinger or soft cervical collar may assist with head control. As thoracic kyphosis becomes more rigid, a clamshell orthosis that is high in the front of the trunk and lower in the back, usually below the scapulae (Fig. 1c), may provide some assistance with upright seating. Larger patients and/or those with stiff kyphotic deformities are generally not amenable to bracing (Dabney 2018). Alternatively, these patients can be treated nonoperatively with a custom-molded seat back if the kyphosis is not causing pain, recognizing that this only accommodates the kyphotic deformity. Conversely, hyperlordosis is not amenable to bracing. Spinal orthotics may not be tolerated in the child with CP due to discomfort, excessive sweating in warm weather, pressure sores, restriction of the child’s breathing, and abdominal constraint when feeding the child. The latter can be alleviated by removing the brace during and an hour after feedings.
Fig. 1

Nonoperative methods for flexible kyphosis: (a) Tilt-in-space chair can be used to reduce a flexible neuromuscular kyphosis by tilting the chair back and (b) in children with flexible kyphosis, the tray table prevents the child from leaning too far forward. (c) Clamshell brace used for kyphosis with high anterior extension

Surgical Treatment

The only treatment that can make a permanent impact on the correction of CP sagittal plane deformity is spinal instrumentation and fusion. The standard surgical procedure is a posterior spinal fusion with segmental instrumentation from T1 or T2 down to the sacrum if the pelvis is part of the deformity. The pelvis and lumbar spine is almost always involved in hyperlordosis. Even if the pelvis is not part of the deformity, in the nonambulatory patient (GMFCS level IV and V) or ambulatory patient with poor balance (GMFCS level III), the surgeon should strongly consider fusion to the pelvis to prevent late pelvic deformity (Dabney 2018).

In the past, the gold standard to instrument the correction of neuromuscular scoliosis was Luque rod instrumentation and sublaminar wires with Galveston instrumentation to the pelvis (Ferguson and Allen 1988). This was later improved upon with the Unit Rod (Surgical Treatment of Scoliosis due to Cerebral Palsy) or the cross-linkage of separate rods to prevent rod shift and rotation (Bell et al. 1989; Rinsky 1990; Dias et al. 1996; Sponseller et al. 2009; Peelle et al. 2006; Sink et al. 2003) However, this surgical approach has had mixed outcomes in managing hyperkyphosis and hyperlordosis (Lipton et al. 2003; Karampalis and Tsirikos 2014; Sink et al. 2003).

Medical/Anesthesia Considerations (Anesthesia for Cerebral Palsy Spine Fusion Surgery)

The general medical status of the CP child should be evaluated using a multidisciplinary approach prior to spinal fusion with instrumentation. These considerations are described in more detail in the scoliosis chapter (Medical Evaluation for Preoperative Surgical Planning in the Child with Cerebral Palsy). Increased perioperative morbidity, including increased blood loss, and postoperative infection are associated with increasing lumbar lordosis (Karampalis and Tsirikos 2014; Sponseller et al. 2010a; Sponseller et al. 2013; Jain et al. 2012). Our experience has also shown increased blood loss in patients with hyperlordosis.

Guidelines to prevent infection are described in the scoliosis section and should also be followed for the correction of sagittal plane deformity (Vitale et al. 2013) (Surgical Treatment of Scoliosis due to Cerebral Palsy). In addition, the utilization of tranexamic Acid (TXA) should be followed to reduce blood loss, especially in hyperlordotic deformity (Dhawale et al. 2012). Even with the use of TXA, the surgeon should take additional measures to prepare excessive blood loss when correcting hyperlordosis which include the availability of: typed and cross-matched blood (up to twice the patient’s blood volume), fresh-frozen plasma, and platelets (Brenn et al. 2004). The use of cell-saver blood should be considered. Vascular access through the use of a central line should also be considered in patients with poor peripheral access (Anesthesia for Cerebral Palsy Spine Fusion Surgery). In patients with poor nutrition, a central venous catheter can also be used for postoperative hyperalimentation. We place a central venous catheter in high-risk patients with severe sagittal plane deformity. The use of pedicle screws instead of sublaminar wires has also been reported to have lower blood loss in the correction of scoliosis in CP, and this may also be true in sagittal plane deformity as well (Fuhrhop et al. 2013).

Another concern in the treatment of sagittal plane spinal deformity is the neurologic risk to the spinal cord. While spinal cord injury is a risk during the correction of spinal deformity in all children with CP, rigid thoracic lordosis and kyphosis may present greater risk. Hyperlordosis may cause the spinal cord to shift more posteriorly over the apex of the lordosis which causes greater neurologic risk for posterior spinal cord injury during the corrective process and also greater risk during the passing of sublaminar wires (Fig. 2), especially in thoracic hyperlordosis where the space for the cord may be narrowed. In rigid hyperlordotic deformity, anterior release may be helpful in reducing this risk by shortening the spine and decreasing the amount of tension on the spinal cord at the apex of the lordosis. Alternatively, a rigid thoracic kyphosis may result in stretching the anterior thoracic spine over the apex of the kyphosis during correction. In severe rigid kyphosis, posterior multilevel vertebral osteotomies which shortens the spinal column can help lessen neurologic risk of spinal cord stretch. In severe rigid combined anterior release, posterior osteotomies or vertebral resection may also be helpful to lessen tension on the spinal cord in very rigid higher magnitude deformity. Many children with mild to moderate cerebral palsy (those with less severe motor cortex involvement) can have successful spinal cord monitoring using a combination of somatosensory and motor-evoked potentials (DiCindio et al. 2003). DiCindio et al. showed approximately 60% of children with cerebral palsy to be monitorable with only severe quadriplegic cerebral palsy children with poor motor function unable to be monitored (DiCindio et al. 2003) (Surgical Spinal Cord Monitoring in Cerebral Palsy). As a general rule, somatosensory and motor evoked potential monitoring should be attempted for the child with CP who has ambulatory function and the capability to assist with standing transfers. There may also be some efficacy in monitoring neuromuscular patients with intact sensation and bowel and bladder control. Related to this, the child with CP and a neurogenic bladder preoperatively should be carefully evaluated for urinary tract infection, and, if present, should be treated prophylactically to clear the urine prior to surgery.
Fig. 2

This patient with spastic quadriplegic cerebral palsy (GMFCS 5) and severe thoracolumbar hyperlordosis has narrowing of the spinal canal between the lamina and the vertebral bodies especially in the thoracic spine. This places increased neurologic risk during wire passage. Using current modular fixation, curve correction may be more safely achieved using pedicle screws as fixation

The next important consideration in the correction of sagittal plane deformity in the child with CP is whether the child has low bone mineral density, especially prevalent in the child with CP with: greater motor involvement (GMFCS IV and V), patients on anticonvulsant medications, and patients who are nutritionally deprived (Sees et al. 2016). Adequate bone density is especially important during the cantilever correction of the sagittal plane deformity. These are highest at the apex in lordosis posteriorly and very high posteriorly at the distal- and proximal-most ends of the spine in kyphosis. In both the deformities, wire pull-out when wires are used or screw plowing when pedicle screws are used may occur when low bone mineral density is present. Any nonambulatory child with low-impact long bone fractures should be checked for low bone density using dual energy x-ray absorptiometry (DEXA scan). Intravenous pamidronate is recommended for the child with CP with bone density two or more Z-scores below the mean and with frequent fractures (Sees et al. 2016). Children on seizure medication should have preoperative calcium, phosphorous, and vitamin D levels measured (Managing Bone Fragility in the Child with CP).

Operative Principles

The principles of spinal deformity correction for sagittal plane spinal deformity in CP are to: (1) correct coronal, sagittal, and transverse plane pelvic deformity with the sitting or standing surface as a reference plane. (2) Restore coronal and sagittal truncal balance in order to center the head over the trunk and pelvis and correct anatomic sagittal alignment of the trunk and pelvis (average sacral slope of approximately 40°, pelvic tilt 13°, and lumbar lordosis of 40–60°) (SW1 et al. 2013). Nonambulatory children with CP should be corrected to a slightly greater than anatomic lumbar lordosis to balance the child’s body weight over the posterior thigh muscle mass. This helps to prevent sacral decubitus ulcers from occurring. Ambulatory children with CP should have relatively equal lumbar lordosis and thoracic kyphosis in order to optimize standing balance. Suh and colleagues showed a significant difference between sagittal spinopelvic parameters in the CP child compared to normal control children and that these abnormal parameters may be related to the symptoms seen in CP children (SW1 et al. 2013). The correction of these parameters during spine surgery is therefore critically important in the child with primary sagittal plane deformity. (3) Maximize segmental fixation in the face of what is often osteoporotic bone and (4) minimize operative time since children with CP often have multiple comorbidities, excessive bleeding, and a greater risk for wound infection (Sponseller et al. 2010a; Sponseller et al. 2013; Jain et al. 2012).

Preoperative Planning

When planning preoperatively for surgery, three technical questions deserve careful thought: (1) Should fusion include the pelvis? (2) Is there a rotational component to the spinal deformity that is affecting sitting or standing balance that will definitively require pedicle screw fixation over sublaminar wires? (3) Is there poor flexibility of the sagittal deformity that will warrant: preoperative or intraoperative traction, anterior release, concurrent posterior-only osteotomies, or total vertebral resection?

Current Preferred Surgical Treatment Methods (Spinal Procediure Atlas for Cerebral Palsy Deformities)

Intraoperative Positioning

The patient is positioned prone with the abdominal area left free in order to minimize abdominal pressure and therefore bleeding. Patients with lumbar kyphosis should have their hips and knees flexed to create maximum lordosis, while patients with hyperlordosis should have their legs left to hang freely, minimizing excessive lordosis (Fig. 3). This position also helps to minimize the stress on the wire or screw/bone interface during the correction maneuver for hyperlordosis. All bony prominences should be well padded and minimal tension/pressure should be placed on contracted extremities. Urinary catheters should be free flowing, especially in children with neurogenic bladder, vesicostomy, and/or other bladder reconstruction.
Fig. 3

Positioning of the patient should allow the abdomen to be free and in cases of hyperlordosis, the hips are flexed to 90 degrees with the trunk and the legs hang freely to allow as much passive correction of the hyperlordosis as possible

Instrumentation

The Unit Rod was previously shown to be effective in the correction of both hyperlordosis and kyphosis; however, the pelvic limbs are difficult to insert, especially with hyperlordosis, making the risk of pelvic limb penetration through the pelvic wall high (Lipton et al. 2003; Dabney et al. 2004). Newer methods of instrumentation allow modularization of the Unit Rod concept and cantilever correction by combining wires or pedicle screws with two pelvic screws placed independently into the pelvis and connected to two rods pre-contoured in the sagittal plane to restore sagittal alignment and a proximal connector (Fig. 4). The pelvic screws have varying diameter of 7–10 mm and length between 65 mm and 100 mm that can be selected according to the pelvic size of the child. Because the pelvic screws can be positioned separately into the pelvis, there is less risk of penetration through the inner pelvic wall as did the Unit Rod, especially in hyperlordotic spinal deformity (Fig. 5). Pelvic screws also provide better fixation into the pelvis, decreasing distal screw pull-out (Erickson et al. 2004). Achieving proper sagittal balance is critical as described earlier which may require recontouring the lumbar bend to have greater or less lordosis than the manufactured contour. Either sublaminar wires or pedicle screw fixation can be used; however, many surgeons prefer to use pedicle screws given their greater rigidity instead of wires for segmental fixation, especially if there is a severe kyphotic or lordotic deformity. On the other hand, sublaminar wires/tapes may be just as efficient with correcting a hyperlordotic lumbar spine into a more anatomic alignment (Dabney et al. 2004). In hyperlordotic spines, great caution should be taken to prevent pedicle screw pull-out (“plowing”) if using screws, and laminar fracture if using wires, especially with osteopenic bone. If significant sagittal plane stiffness is present on either physical or radiographic examination, preoperative halo-femoral/intraoperative traction, posterior only osteotomies, anterior discectomies, and/or total vertebral resection should be considered. The choice of these procedures is determined by the magnitude and stiffness of the deformity and the type of sagittal plane deformity and will be discussed later in this chapter.
Fig. 4

This modular system consists of (a) two rods with a sagittal contour which are (b) connected by a closed connector proximally and a cross-link at the thoracolumbar junction. The pelvic screws are anchored into the pelvis separately which allows easier pelvic placement than the Unit Rod

Fig. 5

Placement of the pelvic limbs of the unit rod is difficult in hyperlordosis due to the far anterior start point required for the pelvic limbs to enter the pelvis. The drill hole and rod limbs (the latter of which must be crossed in order to enter the pelvis properly) must aim just in front of the sciatic notch and aim distal and posterior. Failure to do so may cause the rod limb to penetrate the inner pelvic table as shown. Placing separate pelvic screws attached to pre-contoured rods is technically easier

Fusion to the Pelvis

The extension of fixation and fusion to the pelvis should be considered in every patient with CP with a sagittal plane spinal deformity that extends to the pelvis. Similar to the correction of pelvic obliquity in scoliosis, cantilever correction is an excellent method to correct both anterior and posterior pelvic tilt in the sagittal plane using pelvic screws connected to dual pre-contoured rods connected to a proximal connector as described with less complication than the Unit Rod construct (Fig. 6). Pelvic screws can be inserted into the pelvis at the traditional posterior superior iliac spine entrance or can be placed using an S2 iliac approach. A pedicle probe is used for either approach to enter the pelvis. In the former, the pedicle probe enters at the posterior superior iliac spine and aims just above the sciatic notch using intraoperative fluoroscopy. This is the region where the pelvis is most dense for pelvic screw fixation (Miller et al. 1990). By not fully exposing the sciatic notch as previously done with the Unit Rod procedure, blood loss is minimal. Intraoperative AP and “tear drop” fluoroscopic views are taken to confirm the placement of the probe to make sure that there is no penetration through the inner or outer pelvic table, or into the sciatic notch (Fig. 7). A pelvic screw with the largest diameter possible (usually 7–10 mm) is placed in this trajectory and should be an adequate length to pass the sciatic notch by at least 1 cm. The author prefers to use a closed polyaxial screw head which maximizes the rigidity of the final rod-pelvic screw construct. Usually, only two pelvic screws alone are used for pelvic fixation; however, if additional fixation is needed to improve the correction of a rigid pelvic deformity, S1 screws can be added to the construct.
Fig. 6

This patient with spastic quadriplegic cerebral palsy and severe lumbar hyperlordosis underwent posterior spinal fusion with Unit Rod instrumentation. Due to the inability to achieve the proper trajectory of the Unit Rod pelvic limbs into the pelvis, the correction of anterior pelvic tilt was insufficient. The introduction of separated pelvic screws and pedicle screws in the lumbar spine would have obviated the failure to correct this severe anterior pelvic tilt

Fig. 7

Intraoperative AP (a, b) and oblique (c, d) views showing proper placement of pelvic screw. Note the AP view shows the trajectory of the pedicle probe from the PSIS to just superior and adjacent to the sciatic notch and the final screw position at least 1 cm lateral to the notch. The oblique view is taken parallel with the probe and shows the probe and the final screw position between the inner and outer cortex just superior to the sciatic notch which appears as a “teardrop”

Alternatively, pelvic screws can be placed using the medial portal (S2-iliac approach) as described by Chang et al. and Sponsellar (Chang et al. 2009; Sponseller et al. 2010b). Advocates for this method claim less exposure time, less bleeding, and that the pelvic screw head is less prominent and more in line with the rod, making the need for a separate connection to the rod unnecessary. While we have not found bleeding or exposure time to be less in our hands, the screw is less prominent using this approach and lines up more directly with the pre-contoured rods, avoiding the need for lateral rod connectors. If the traditional PSIS start point is used, notching the ilium at the entrance point with a rongeur and countersinking the screw prevents screw head prominence. A fixed lateral rodded connector, 10–20 mm in length) is used to connect each pelvic screw to a pre-contoured rod. Critical to the correction is to attach and secure each of the pre-contoured rods to the iliac screws with the fixed lateral connectors so that each of the rods are perfectly perpendicular to the horizontal axis of the pelvis and that the sagittal contour of the rods are parallel with one another and aligned with the sacrum (Figs. 4 and 8) (Dabney 2018). The sagittal bend of each rod should be identical from proximal to distal. In addition, the sagittal contours for lumbar lordosis and thoracic kyphosis of the rod should match the length of the lumbar and thoracic spine, respectively. If these steps are not meticulously followed, the sagittal alignment of the pelvis and thoraco-lumbar spine may not be optimally corrected with the cantilever maneuver. Once the correction is obtained, the set screws on both the pelvic screws can be tightened and torqued down onto the rod. A proximal connector is added at the top of the construct which strengthens the proximal construct. A cross connector should be added at the thoracolumbar junction to augment the stability of the construct.
Fig. 8

Pelvic fixation is performed first with any distal to proximal cantilever correction. (a) The pelvic screws are placed as shown in this anteroposterior radiograph. (b) The construct is then assembled from distal to proximal securing the rods to the pelvic screws using the rodded connectors shown if using a traditional PSIS entrance into the pelvis. A proximal closed connector and crosslink at the thoraco-lumbar junction connect the two pre-contoured rods which should be parallel to one another

Only if the patient preoperatively has a level pelvis, a correct sagittal pelvic position, and adequate balance, should the surgeon consider ending the fusion and fixation more proximally at the L4 or L5 vertebrae. If fixation to the pelvis is not done, distal pedicle screw fixation in the lumbar spine at a minimum of four levels is recommended. In severe lumbar lordosis, pedicle screws should be considered at each level. Cantilever correction with fixation using pedicle screws or sublaminar wires to correct the remainder of the sagittal alignment can then be done to complete the thoracic spine correction. Hyperlordosis correction requires another corrective technique that will be described.

Kyphosis Correction

Lumbar and Thoracolumbar Kyphosis

Cantilever correction is very effective in correcting both lumbar and thoracolumbar kyphosis in the child with CP. Each is effectively corrected utilizing a distal-to-proximal cantilever correction with the modular dual contoured rod construct described, beginning with fixation to the pelvis similar to the cantilever correction described with the Unit Rod (Dabney et al. 2004). The surgeon places the pelvic screws, attaches the pre-contoured rods, and then begins cantilever correction (Fig. 9). Next, the surgeon progressively pushes the rod down to each vertebra at a time, securing each vertebral level with sublaminar wires or screws. The surgeon should not use the fixation to pull the rod to the spine, as this may cause loss of fixation (either wires cutting through the laminae or pedicle screw pull-out). The process of securing the rod to the fixation at each level begins at the L5 vertebral level and progresses gradually up to the T2 or T1 vertebral level. The placement of pedicle screws with reduction posts at the proximal ends of the spine is helpful to capture the proximal rod ends as the kyphosis is gradually being corrected. Pedicle screws along the entire spine may also be helpful to perform compression of the vertebrae posteriorly to further correct the kyphosis. The pelvis which is typically posteriorly tilted should also be corrected as a part of the cantilever correction.
Fig. 9

In lumbar and thoracolumbar kyphosis, a distal to proximal cantilever correction is performed, first fixing the rod to distal vertebrae and then pushing down (anterior) on the rod after the rod is anchored to the apical vertebrae. The sagittal placement of the rod should initially be parallel to the pre-corrected sagittal alignment of the sacrum which is tilted posterior along with the pelvis in kyphosis. As the rod is moved to the spine using cantilever correction, the sagittal alignment of the pelvis and spine will correct to the sagittal contour of the rod

Thoracic Kyphosis

Thoracic kyphosis is difficult to correct using a distal to proximal cantilever correction technique because the lever arm remains too short above the apex of the kyphosis to provide an adequate cantilever correction. Accordingly, thoracic kyphosis is difficult to correct with the Unit Rod since it requires distal fixation into the pelvis first. With this type of curvature, a proximal to distal cantilever correction is preferred when using the more modular system (Fig. 10). After exposing the spine and pelvis, the pre-contoured rods are connected using a proximal closed rod connector at the T1 level and a cross connector in the lumbar spine. The rods should be parallel from proximal to distal with respect to their contour. Next, pelvic screws and sublaminar wires are placed as previously described. The top of the rod construct is then secured to the spine using sublaminar wires or pedicle screws from T1 down to the apex of the kyphosis. In thoracic kyphosis, great care should be taken to preserve the spinous process ligaments in order to prevent a junctional kyphosis (Fig. 11). After the apical vertebrae is secured to the rod, cantilever correction can be performed by gradually pushing the rod down to the next more distal vertebrae, tightening the sublaminar wire or securing it to the pedicle screw, performing the same maneuver progressively down the spine until the rod is secured to each of the two pelvic screws. Pelvic screws with reduction posts are helpful to capture the distal-most end of the rods. Similar to lumbar or thoracolumbar kyphosis, pedicle screws placed along the kyphosis can be helpful to compress and further correct the kyphosis. The fixed rodded lateral connectors are then utilized to connect the rod to the pelvic screws or directly to the pelvic screw if the S2 iliac technique is utilized. Ambulatory patients without pelvic deformity can be fused short of the pelvis and secured to pedicle screws at the L4 or L5 vertebrae. In thoracic kyphosis, it is critical that fixation be completed up to at least the T1 vertebral level and occasionally the C7 level to prevent “drop-off” at the cervicothoracic junction. Firm fixation at the proximal-most end with two wires, hooks, or screws is recommended (Dabney 2018).
Fig. 10

It is difficult to cantilever thoracic kyphosis using the Unit Rod due to insufficient lever arm. This diagram shows a proximal to distal cantilever technique that can be used for thoracic kyphosis. The rod is preassembled and secured proximally and then delivered into lumbar pedicle screws if there is no pelvic sagittal misalignment. Preoperative and postoperative radiographs are shown

Fig. 11

(a) This patient with quadriplegic CP (GMFCS 5) and thoracic kyphosis with poor head control (b) underwent posterior spinal fusion with Unit Rod instrumentation. (c) By 3 months post-op, he developed a significant junctional kyphosis at the cervical-thoracic junction. (d) This eventually necessitated extension of the fusion up to the occiput

Hyperlordosis

Isolated neuromuscular lumbar hyperlordosis does occur but is more frequently seen in combination with scoliosis or thoracic kyphosis. Pelvic screws are placed first followed by pedicle screw fixation with reduction posts in the vertebrae within the hyperlordosis (usually in the lumbar spine) (Fig. 12) (Dabney et al. 2004). After securing the pre-contoured rods to the pelvic screws, the rods are pushed down into the reduction posts and secured set using screws. Reduction of the hyperlordosis cannot be solely achieved using cantilever correction alone but requires incrementally screwing down the set screws, gradually increasing the load share over each of the screws, a small amount at a time. Great care is taken to notice any evidence of posterior plowing of the pedicle screws. Maximizing the diameter of the screws may help with improved pedicle fixation. In addition, the supplementation of sublaminar wires can be used at the same level as the screw for additional fixation if screw pull-out is beginning to occur. After the hyperlordosis is corrected, the rest of the spinal instrumentation can be completed using cantilever correction.
Fig. 12

(a) Correction of hyperlordosis can be achieved using pedicle screws with reduction posts in the hyperlordotic region of the deformity after pelvic fixation of the rod is done. Screw tightening should be gradual and incremental to share the load across all screws (b) Preoperative and postoperative photographs are shown

Rigid Kyphotic and Hyperlordotic Deformities

Rigid thoracic and thoracolumbar kyphosis may be difficult to correct using posterior spinal fusion with instrumentation alone. Some authors have shown that multiple level posterior-only (Ponte, vertebral, or Smith-Petersen) osteotomies with or without anterior discectomies as a first stage followed by posterior instrumentation is successful to correct severe kyphotic deformity. Corrective osteotomies may also decrease the excessive corrective forces required by shortening of the vertebral column (Diab et al. 2011; Dorward and Lenke 2010; Geck et al. 2007; Auerbach et al. 2009). It may also lessen neurologic risk to the cord during correction. Preoperative or intraoperative traction has been recommended by some as an alternative to anterior release for rigid spinal deformities, specifically scoliotic deformity (Takeshita et al. 2006; Keeler et al. 2010; Jackson et al. 2018). Little is written about halo-femoral traction and its use in sagittal plane deformity. Rigid hyperlordotic deformity may require staged anterior release (multiple anterior discectomies) at the rigid apex of the lordosis followed by posterior spinal fusion with instrumentation (Lipton et al. 2003; Dabney et al. 2004; Geck et al. 2007). In severely rigid hyperlordotic and kyphotic deformity, vertebral column resection can produce excellent curve correction and restoration of sitting balance (Helenius et al. 2012; Sponseller et al. 2012; Modi et al. 2011). This can be achieved as a staged anterior/posterior vertebral resection or posterior only vertebral resection (Modi et al. 2011) (Fig. 13).
Fig. 13

(a) Severe lordoscoliosis (with primary lordosis) which underwent (b) a staged anterior vertebrectomy followed by (c) posterior completion to a total vertebrectomy with posterior instrumentation and fusion

Evidence-Based Outcomes

Lipton et al. was the first to describe a series of 24 children with cerebral palsy with isolated sagittal plane spinal deformity (8 with hyperlordotic deformity, 14 with kyphotic deformity, and 2 with both) (Lipton et al. 2003). Each sagittal plane deformity underwent posterior spinal fusion and cantilever correction using Unit Rod instrumentation. The indications for surgery included back pain, seating problems despite wheelchair modifications, and two cases of superior mesenteric artery syndrome refractory to conservative treatment in children with hyperlordosis. In children with kyphotic deformity, the mean preoperative kyphosis of 93.8° was corrected to a mean postoperative kyphosis of 35.8°, while the mean preoperative hyperlordosis of 91.8° was corrected to a mean postoperative lordosis of 43.6° in children with hyperlordosis. Postoperatively, caregivers reported improvements in: sitting balance, head control, pain relief, and physical appearance. Both cases of superior mesenteric artery syndrome resolved after spinal deformity correction.

Karampalis and Tsirikos reported on 13 patients with lumbar hyperlordosis and lordoscoliosis who underwent posterior spinal fusion with instrumentation (Karampalis and Tsirikos 2014). The mean lumbar lordosis was corrected from 108° to 62° postoperatively. Sacral slope (horizontal sacral inclination) improved from 79° to 50°. Sagittal imbalance was improved from a mean of −8 cm to −1.8 cm. Preoperative lumbar lordosis and sacral slope had an increased risk of perioperative morbidity. Reduced lumbar lordosis and increased thoracic kyphosis were associated with improved sagittal balance at follow-up. Postoperative questionnaires at the final follow-up showed relief of severe preoperative back pain and improvements in physical appearance and function. There were also improvements in head control, breathing, and hand use (Dabney 2018).

Sink et al. looked at a retrospective case series of 24 patients with patients had preoperative kyphotic deformities (Sink et al. 2003). Preoperative thoracic, thoracolumbar, and lumbar kyphosis were risk factors for loss of proximal and distal sagittal fixation and therefore correction. The authors stated that increased forces at the proximal- and distal-most end (Galveston fixation) of the instrumentation during kyphosis correction resulted in the greatest potential for failure. They recommended reinforcing these ends with stronger fixation. We prefer to use the largest diameter pelvic screw fixation which in our experience is less likely to pull-out compared to the Unit Rod or Galveston fixation. Proximal loss of correction occurred in 11 patients who developed a junctional kyphosis. Securing fixation proximally with two wires, screws, or hooks provide a more secure proximal fixation.

Summary

Sagittal plane spinal deformities (kyphosis and hyperlordosis) are uncommon by themselves in cerebral palsy; however, when present can interfere with proper sitting and standing balance. Sagittal plane spinal deformity in conjunction with scoliosis is more common and must be treated surgically as a component of the scoliosis. Mild and some moderate sagittal plane deformities can be treated by wheelchair modifications and bracing. Symptomatic moderate and severe deformity may require surgical treatment. More flexible kyphosis and hyperlordosis can be corrected by posterior spinal fusion and segmental instrumentation alone while rigid deformity usually requires posterior osteotomies (for kyphosis) and/or anterior discectomies (for hyperlordosis). Instrumentation and correction techniques vary from screw/rod constructs using distraction/compression correction to wire or screw/rod constructs using cantilever correction. Overall, natural history and surgical outcome studies focused solely on sagittal plane spinal deformities in CP are limited. Those authors that do measure functions report improvements in pain, sitting balance, head and neck control, breathing, and hand use. Patients with kyphosis undergoing spinal fusion with instrumentation are at risk for loss of proximal and/or distal fixation. Patients with hyperlordosis appear to be at greatest risk for postoperative complications.

Cross-References

References

  1. Bell DF, Moseley CF, Koreska J (1989) Unit rod segmental spinal instrumentation in the management of patients with progressive neuromuscular spinal deformity. Spine (Phila Pa 1976) 14(12):1301–1307CrossRefGoogle Scholar
  2. Brenn BR, Theroux MC, Dabney KW, Miller F (2004) Clotting parameters and thromboelastography in children with neuromuscular and idiopathic scoliosis undergoing posterior spinal fusion. Spine (Phila Pa 1976) 29(15):E310–E314CrossRefGoogle Scholar
  3. Chang TL, Sponseller PD, Kebaish KM, Fishman EK (2009) Low profile pelvic fixation: anatomic parameters for sacral alar-iliac fixation versus traditional iliac fixation. Spine (Phila Pa 1976) 34(5):436–440CrossRefGoogle Scholar
  4. Dabney K (2018) Sagittal plane spinal deformity in patients with neuromuscular disease. In: Samdani AF et al (eds) Neuromuscular spine deformity.  https://doi.org/10.1055/b-0038-162475, Part II: Disease SpecificCrossRefGoogle Scholar
  5. Dabney KW, Miller F, Lipton GE, Letonoff EJ, McCarthy HC (2004) Correction of sagittal plane spinal deformities with unit rod instrumentation in children with cerebral palsy. J Bone Joint Surg Am 86-A(Suppl 1(Pt 2)):156–168CrossRefGoogle Scholar
  6. Dhawale AA, Shah SA, Sponseller PD et al (2012) Are antifibrinolytics helpful in decreasing blood loss and transfusions during spinal fusion surgery in children with cerebral palsy scoliosis? Spine (Phila Pa 1976) 37(9):E549–E555CrossRefGoogle Scholar
  7. Diab MG, Franzone JM, Vitale MG (2011) The role of posterior spinal osteotomies in pediatric spinal deformity surgery: indications and operative technique. J Pediatr Orthop 31(1 Suppl):S88–S98CrossRefGoogle Scholar
  8. Dias RC, Miller F, Dabney K, Lipton G, Temple T (1996) Surgical correction of spinal deformity using a unit rod in children with cerebral palsy. J Pediatr Orthop 16(6):734–740CrossRefGoogle Scholar
  9. DiCindio S, Theroux M, Shah S, Miller F, Dabney K, Brislin RP, Schwartz D (2003) Multimodality monitoring of transcranial electric motor and somatosensory-evoked potentials during surgical correction of spinal deformity in patients with cerebral palsy and other neuromuscular disorders. Spine (Phila Pa 1976) 28(16):1851–1855CrossRefGoogle Scholar
  10. Dorward IG, Lenke LG (2010) Osteotomies in the posterior-only treatment of complex adult spinal deformity: a comparative review. Neurosurg Focus 28(3):E4CrossRefGoogle Scholar
  11. Erickson MA, Oliver T, Baldini T et al (2004) Biomechanical assessment of conventional unit rod fixation versus a unit rod pedicle screw construct: a human cadaver study. Spine (Phila Pa 1976) 29:1314–1319CrossRefGoogle Scholar
  12. Ferguson RL, Allen BL Jr (1988) Considerations in the treatment of cerebral palsy patients with spinal deformities. Orthop Clin North Am 19(2):419–425Google Scholar
  13. Fuhrhop SK, Keeler KA, Oto M, Miller F, Dabney KW, Bridwell KH, Lenke LG, Luhmann SJ (2013) Surgical treatment of scoliosis in non-ambulatory spastic quadriplegic cerebral palsy patients: a matched cohort comparison of unit rod technique and all-pedicle screw constructs. Spine Deform 1(5):389–394CrossRefGoogle Scholar
  14. Geck MJ, Macagno A, Ponte A, Shufflebarger HL (2007) The Ponte procedure: posterior only treatment of Scheuermann's kyphosis using segmental posterior shortening and pedicle screw instrumentation. J Spinal Disord Tech 20(8):586–593CrossRefGoogle Scholar
  15. Helenius I, Serlo J, Pajulo O (2012) The incidence and outcomes of vertebral column resection in paediatric patients: a population-based, multicentre, follow-up study. J Bone Joint Surg Br 94(7):950–955CrossRefGoogle Scholar
  16. Jackson TJ, Yaszay B, Pahys JM, Singla A, Miyanji F, Shah SA, Sponseller PD, Newton PO, Flynn JM, Cahill PJ, Harms Study Group (2018) Intraoperative traction may be a viable alternative to anterior surgery in cerebral palsy scoliosis ≥100 degrees. J Pediatr Orthop 38(5):e278–e284CrossRefGoogle Scholar
  17. Jain A, Njoku DB, Sponseller PD (2012) Does patient diagnosis predict blood loss during posterior spinal fusion in children? Spine (Phila Pa 1976) 37(19):1683–1687CrossRefGoogle Scholar
  18. Auerbach JD, Spiegel DA, Zgonis MH, Reddy SC, Drummond DS, Dormans JP, Flynn JM (2009) The correction of pelvic obliquity in patients with cerebral palsy and neuromuscular scoliosis: is there a benefit of anterior release prior to posterior spinal arthrodesis? Spine (Phila Pa 1976) 34(21):E766CrossRefGoogle Scholar
  19. Karampalis C, Tsirikos AI (2014) The surgical treatment of lordoscoliosis and hyperlordosis in patients with quadriplegic cerebral palsy. Bone Joint J 96-B(6):800–806CrossRefGoogle Scholar
  20. Keeler KA, Lenke LG, Good CR, Bridwell KH, Sides B, Luhmann SJ (2010) Spinal fusion for spastic neuromuscular scoliosis: is anterior releasing necessary when intraoperative halo-femoral traction is used? Spine (Phila Pa 1976) 35(10):E427–E433CrossRefGoogle Scholar
  21. Lipton GE, Miller F, Dabney KW, Altiok H, Bachrach SJ (1999) Factors predicting postoperative complications following spinal fusions in children with cerebral palsy. J Spinal Disord 12(3):197–205Google Scholar
  22. Lipton GE, Letonoff EJ, Dabney KW, Miller F, McCarthy HC (2003) Correction of sagittal plane spinal deformities with unit rod instrumentation in children with cerebral palsy. J Bone Joint Surg Am 85-A(12):2349–2357CrossRefGoogle Scholar
  23. McCarthy JJ, Betz RR (2000) The relationship between tight hamstrings and lumbar hypolordosis in children with cerebral palsy. Spine (Phila Pa 1976) 15:211–213CrossRefGoogle Scholar
  24. Miller F, Moseley C, Koreska J (1990) Pelvic anatomy relative to lumbosacral instrumentation. J Spinal Disord 3:169–173CrossRefGoogle Scholar
  25. Modi HN, Suh SW, Hong JY, Yang JH (2011) Posterior multilevel vertebral osteotomy for severe and rigid idiopathic and nonidiopathic kyphoscoliosis: a further experience with minimum two-year follow-up. Spine (Phila Pa 1976) 36(14):1146–1153CrossRefGoogle Scholar
  26. Nishnianidze T, Bayhan IA, Abousamra O, Sees J, Rogers KJ, Dabney KW, Miller F (2016) Factors predicting postoperative complications following spinal fusions in children with cerebral palsy scoliosis. Eur Spine J 25(2):627–634CrossRefGoogle Scholar
  27. Peelle MW, Lenke LG, Bridwell KH et al (2006) Comparison of pelvic fxation techniques in neuromuscular spinal deformity correction: Galveston rod versus iliac and lumbosacral screws. Spine (Phila Pa 1976) 31:2392–2398CrossRefGoogle Scholar
  28. Rappaport DI, Pressel DM (2008) Pediatric hospitalist comanagement of surgical patients: challenges and opportunities. Clin Pediatr (Phila) 47(2):114–121CrossRefGoogle Scholar
  29. Rappaport DI, Cerra S, Hossain J, Sharif I, Pressel DM (2013a) Pediatric hospitalist preoperative evaluation of children with neuromuscular scoliosis. J Hosp Med 8(12):684–688CrossRefGoogle Scholar
  30. Rappaport DI, Adelizzi-Delany J, Rogers KJ, Jones CE, Petrini ME, Chaplinski K, Ostasewski P, Sharif I, Pressel DM (2013b) Outcomes and costs associated with hospitalist comanagement of medically complex children undergoing spinal fusion surgery. Hosp Pediatr 3(3):233–241CrossRefGoogle Scholar
  31. Rinsky LA (1990) Surgery of spinal deformity in cerebral palsy. Twelve years in the evolution of scoliosis management. Clin Orthop Relat Res 25(3):100–109Google Scholar
  32. Samdani AF, Belin EJ, Bennett JT, Miyanji F, Pahys JM, Shah SA, Newton PO, Betz RR, Cahill PJ, Sponseller PD (2016) Major perioperative complications after spine surgery in patients with cerebral palsy: assessment of risk factors. Eur Spine J 25(3):795–800CrossRefGoogle Scholar
  33. Sees JP, Sitoula P, Dabney K, Holmes L Jr, Rogers KJ, Kecskemethy HH, Bachrach S, Miller F (2016) Pamidronate treatment to prevent reoccurring fractures in children with cerebral palsy. J Pediatr Orthop 36(2):193–197CrossRefGoogle Scholar
  34. Sink EL, Newton PO, Mubarak SJ, Wenger DR (2003) Maintenance of sagittal plane alignment after surgical correction of spinal deformity in patients with cerebral palsy. Spine (Phila Pa 1976) 28(13):1396–1403Google Scholar
  35. Sponseller PD, Shah SA, Abel MF et al (2009) Scoliosis surgery in cerebral palsy: differences between unit rod and custom rods. Spine (Phila Pa 1976) 34:840–844CrossRefGoogle Scholar
  36. Sponseller PD, Shah SA, Abel MF, Newton PO, Letko L, Marks M (2010a) Infection rate after spine surgery in cerebral palsy is high and impairs results: multicenter analysis of risk factors and treatment. Clin Orthop Relat Res 468(3):711–716CrossRefGoogle Scholar
  37. Sponseller PD, Zimmerman RM, Ko PS et al (2010b) Low profile pelvic fixation with the sacral alar iliac technique in the pediatric population improves results at two-year minimum follow-up. Spine (Phila Pa 1976) 35(20):1887–1892CrossRefGoogle Scholar
  38. Sponseller PD, Jain A, Lenke LG et al (2012) Vertebral column resection in children with neuromuscular spine deformity. Spine (Phila Pa 1976) 37(11):E655–E661CrossRefGoogle Scholar
  39. Sponseller PD, Jain A, Shah SA et al (2013) Deep wound infections after spinal fusion in children with cerebral palsy: a prospective cohort study. Spine (Phila Pa 1976) 38(23):2023–2027CrossRefGoogle Scholar
  40. Suh SW, Suh DH, Kim JW, Park JH, Hong JY (2013) Analysis of sagittal spinopelvic parameters in cerebral palsy. Spine J 13(8):882–888CrossRefGoogle Scholar
  41. Takeshita K, Lenke LG, Bridwell KH, Kim YJ, Sides B, Hensley M (2006) Analysis of patients with nonambulatory neuromuscular scoliosis surgically treated to the pelvis with intraoperative halo-femoral traction. Spine (Phila Pa 1976) 31(20):2381–2385CrossRefGoogle Scholar
  42. Vitale MG, Riedel MD, Glotzbecker MP et al (2013) Building consensus: development of a Best Practice Guideline (BPG) for surgical site infection (SSI) prevention in high-risk pediatric spine surgery. J Pediatr Orthop 33(5):471–478CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Department of OrthopedicsNemours/AI DuPont Hospital for ChildrenWilmingtonUSA

Section editors and affiliations

  • Freeman Miller
    • 1
  1. 1.AI DuPont Hospital for ChildrenWilmingtonUSA

Personalised recommendations