The anomalous pattern of segmental vascularity that often accompanies these deformities can make adequate hemostasis difficult to achieve and often is responsible for the substantial blood loss that is encountered during the procedure. Gaining control of a bleeding vessel at the deep limit of the surgical field often requires a combination of adequate aspiration, bipolar cautery, thrombin-soaked gel foam, direct pressure, and time. Once the hemivertebra has been removed, the adjoining end plates are denuded to expose cancellous bleeding bone. A wedge that approximates the desired amount of correction is created in the anterior column, with care being taken to preserve the anterior longitudinal ligament so that it serves as a rotational hinge and stabilizing restraint to translation during correction. The resected hemivertebra is morselized for bone graft and is placed loosely in the anterior aspect of the gap left by the resection. For patients with acquired kyphoscoliosis without an associated hemivertebra, a decancellation osteotomy14 of the apical vertebra is performed under direct visualization.
Following resection of the hemivertebra, posterior spinal instrumentation is placed for correction of the deformity. The pattern of fixation that is used depends on the type of the deformity and the preference of the surgeon. Our current preference is to use pedicle screws at least two vertebrae caudad to the level of the resection and a combination of pedicle screws and hooks cephalad to the level of the resection, depending on the specific deformity and the ease of pedicle navigation. Given the relative ease of placing pedicle screws inferior to the deformity, we typically begin by inserting pedicle screws at the inferior limit of the instrumentation. We then proceed superiorly with pedicle screw fixation until the superior limit of the instrumentation is reached or the difficulties of pedicle navigation warrant a switch to pedicle, lamina, or transverse process hooks, as appropriate. Fluoroscopy is used to confirm satisfactory placement of the fixation points. Two rods are contoured to approximate the desired amount of correction and are attached to the inferior points of fixation. A combination of cantilever bending and translation forces are then applied, and the deformity is slowly corrected to approximate the normal sagittal and coronal balance of the spine. During this process, the dura and its contents can be directly inspected for signs of compression by visualizing and palpating the anterior aspect of the thecal sac at the level of the hemivertebra and the apex of correction. If compression of the thecal sac or neural elements does occur with correction of the overall deformity, additional bone can be resected at the point of compression if it can be seen or palpated. If there are changes in somatosensory evoked potentials during correction and there is not an obvious source of compression, the etiology of the compromise is thought to be vascular and the rods are recontoured to reduce the amount of correction. With a combination of these maneuvers, the deformity can be corrected without compromising the contents of the spinal canal.
A standard posterior spinal arthrodesis is then completed (Figs. 1-A and 1-B). Our preference is to use local autogenous bone and supplemental morselized allograft bone. A surgical drain with a closed self-suctioning reservoir is placed, a chest tube is placed into each hemithorax that has been violated by a pleural opening, and the wounds are closed in a routine fashion. The patient is mobilized with the assistance of a physical therapist on the first or second postoperative day, depending on the comfort level. An orthosis was not used postoperatively for any patient in the present series. Specific activity restrictions that avoid bending, twisting, heavy lifting, and sporting activities are maintained for six months.
All sixteen patients were managed with vertebral resection through a simultaneous anterior and posterior approach to the spine that involved a single posterior midline incision and a costotransversectomy. All but one of the patients had an anterior and posterior spinal fusion and posterior segmental spinal instrumentation; the remaining patient was too small for spinal instrumentation at the time of vertebral resection. The mean correction of the major kyphotic deformity was 31° (range, 0° to 82°), and the mean correction of the major scoliotic deformity was 25° (range, 0° to 68°). The mean duration of follow-up was 5.0 years (range, 2.0 to twelve years).
Four patients had a major complication. One patient had a failure of posterior instrumentation that necessitated revision surgery, one patient had development of persistent lower extremity dysesthesias of uncertain etiology, and two patients had radiographic evidence of late progression of pelvic obliquity caudad to the fusion. There were no substantial neurological injuries affecting bowel or bladder function or muscle strength in the lower extremities.
Thirteen patients had a satisfactory outcome; that is, both the patient and the family were satisfied with the results of the operation, there were no obvious and clinically important postoperative complications, the fusion appeared solid, and there was no evidence of hardware failure or progression of deformity during the follow-up period. Two patients had a fair outcome; that is, they had a substantial postoperative complication but did not require additional surgery. One patient had a poor outcome because additional surgery was needed for the treatment of a postoperative complication during the follow-up period. Specifically, a catastrophic failure of the initial posterior instrumentation necessitated revision anterior and posterior surgery for stabilization. Despite this additional surgery, the patient believed that the overall outcome was satisfactory; nevertheless, the result was rated as poor because of the need for a second major surgical procedure.
The natural history of congenital kyphosis has been well described in the literature1-3,15,16. Although less common than congenital scoliosis, congenital kyphosis is associated with a greater risk of anterior cord compression and neurologic compromise with growth and progression of the deformity if left untreated16. Early intervention to prevent progression of the deformity and to allow for some correction with growth is currently recommended3,16. In a review of ninety-four patients in whom progressive congenital kyphosis had been treated after the age of five years, Winter et al.3 recommended a posterior arthrodesis for curves of ≤50° and a combined anterior and posterior arthrodesis for curves of >50°. McMaster and Singh16 also recommended that all children with Type-I or III congenital kyphosis should be managed with posterior arthrodesis without instrumentation before the age of five years and before the kyphosis exceeds 50°. In this group of younger patients, the authors reported that posterior arthrodesis alone resulted in a mean correction of 15° with growth, from a mean preoperative deformity of 43°. Among patients who had been managed with posterior arthrodesis without instrumentation alone, with and without cast application, at the age of six years or more, there was limited correction of only 9° from a mean preoperative deformity of 70°, and there was a substantial rate of pseudarthrosis or kyphosis progression during the follow-up period (mean, 6.6 years). The authors recommended anterior release, strut-grafting, and posterior arthrodesis with instrumentation (if possible) for these older patients and for patients in whom the curve exceeds 60°.
Kim et al.15 noted that correction of kyphosis may occur with continued growth in patients with Type-I and III deformities, especially when a posterior arthrodesis is performed when the patient is two years old or less. In their study of twenty-six cases of surgically treated congenital kyphosis and kyphoscoliosis, a variety of surgical techniques were used. Five patients with a mean age of sixteen months underwent posterior arthrodesis without instrumentation alone, with improvement of the mean kyphotic deformity from 49° to 26°. Pseudarthrosis developed in two patients, necessitating subsequent posterior augmentation or anterior arthrodesis. Five patients with a mean age of 13.6 years underwent posterior arthrodesis with instrumentation, with improvement of the kyphotic deformity from 59° to 29°. Seven patients with a mean age of sixteen months underwent anterior release or vertebral resection followed by posterior arthrodesis, with improvement of the kyphotic deformity from 48° to 22°, whereas nine patients with a mean age of 11.5 years demonstrated improvement from 77° to 37° after the same procedure. The authors did not report the amount of intraoperative blood loss. They did not comment that spinal instrumentation reduced the need for subsequent augmentation of the fusion and was associated with a low prevalence of pseudarthrosis. Two of the twenty-six patients in that study had development of a postoperative neurological deficit, and the authors identified a number of risk factors for neurologic injury: an older age at the time of correction, combined anterior and posterior arthrodesis procedures, more severe deformity, and preexisting spinal cord compromise.
The studies by McMaster and Singh16, Winter et al.17, and Kim et al.15 suggest that whereas early posterior arthrodesis is effective in the younger child, it is not sufficient for the older child who presents with congenital kyphoscoliosis and a substantial deformity. The deformities that arise in older children as a consequence of an abnormal hemivertebra present a difficult surgical challenge. Some degree of vertebral resection often is required in order to achieve satisfactory correction of the deformity, and this is not possible through a posterior-only exposure and fusion. A combined anterior and posterior arthrodesis for the treatment of kyphoscoliosis has inherent limitations. With use of separate approaches, it is not possible to expose and resect the hemivertebra from the anterior side, to perform spinal instrumentation, and to observe the effects of posterior correction on the contents of the spinal canal during surgery. The adequacy of the anterior decompression must be estimated prior to closing the anterior incision and proceeding to the posterior approach for instrumentation and correction of the deformity. This limitation can be overcome by performing a surgical resection of the hemivertebra and correction of the deformity through the same posterior approach. While a posterior-only approach for resection of the hemivertebra seems to be technically challenging and fraught with risk to the neural elements, several published reports have described acceptable clinical outcomes in association with this technique12,13. Increasing familiarity with anterior and posterior column surgery through a posterior approach has resulted in a tendency to fuse both columns at a younger age. For example, Ruf and Harms18 recently reported on twenty-eight consecutive cases of congenital scoliosis in very young children who underwent hemivertebra resection through a posterior-only approach at a mean age of 3.3 years. We concur with their conclusions that this procedure permits excellent correction in the frontal and sagittal planes and produces a short-segment fusion that allows for normal growth in the unaffected parts of the spine, and we add that it also avoids the uncertainties associated with unbalanced growth of the anterior column that may lead to progression, recurrence, or crankshaft-type deformities.
The location of a typical hemivertebra makes it amenable to resection from the posterior approach. The anomalous vertebral body is located at the apex of the kyphosis, and several authors have reported excellent visualization of the hemivertebra during resection from a posterior approach. In addition, the surgical dissection required to remove the hemivertebra exposes the neural elements that are at risk of compression during correction of the deformity. It should be noted that substantial blood loss should be anticipated during this procedure. We speculate that large losses of blood are a result of the combination of an anomalous vascular supply to the hemivertebra, exposure of the cancellous surfaces of end plates of the adjacent vertebral bodies, and the technical difficulty of obtaining hemostasis in the depths of the operative field.
Shono et al.12 recently reported the results of one-stage posterior hemivertebra resection and posterior segmental spinal instrumentation in a study of twelve patients with congenital kyphoscoliosis who were between eight and twenty-four years of age. The procedure was associated with satisfactory correction of both the scoliosis (from 49° to 18°) and the kyphosis (from 40° to 17°). The authors reported no postoperative neurologic complications, a 100% union rate, and a mean blood loss of 600 mL. They commented that visualization of the pathologic hemivertebra was enhanced by removal of the adjoining rib. This approach was thought to be safe and effective for adolescents with congenital kyphoscoliosis, and the authors concluded that the single-stage posterior approach was effective for resection of an isolated hemivertebra in the thoracic and lumbar spine. Ruf and Harms13 described a similar approach for hemivertebra resection in a study of twenty patients ranging from less than two years of age to fourteen years of age. The approach, which emphasized limited segmental fixation with use of pedicle screws, was associated with correction of the scoliotic deformity from 41° to 14° and correction of the kyphotic deformity from 24° to 11°. Again, the authors were able to achieve excellent visualization of the hemivertebra as well as of the effects of correction of the deformity on the adjacent neural structures. The mean blood loss was 635 mL. In a subsequent report on patients who were five years of age or less18, twenty-eight children with a mean age of 3.3 years were managed with a posterior-only approach with transpedicular instrumentation. The mean scoliotic deformity was corrected from 45° to 13°, and the mean kyphotic deformity was corrected from 22° to 10°. The authors reported no neurologic complications, one infection, three cases of implant failure, and a mean blood loss of 496 mL.
Two patients in the present series were treated for acquired kyphoscoliosis. Both had had extensive, repeated laminectomies for the treatment of tethered-cord syndrome and subsequently had had development of progressive, painful kyphoscoliosis at the thoracolumbar junction. Neither patient had undergone instrumentation and fusion at the time of the laminectomies. The use of the costotransversectomy approach following the failure of previous surgery allows for excellent visualization of the spinal cord during correction of the deformity.
In conclusion, we believe that congenital kyphosis and acquired kyphoscoliosis can be safely corrected with use of a one-stage posterior approach to the spine. The addition of a costotransversectomy approach to the anterior part of the spine allows excellent visualization during resection of a hemivertebra and also allows access to the anterior column for fusion, but substantial blood loss should be anticipated. We recommend arthrodesis of the entire deformity and stabilization with use of pedicle fixation where possible. ▪
The authors did not receive grants or outside funding in support of their research or preparation of this manuscript. They did not receive payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.
Investigation performed at Primary Children's Medical Center, Salt Lake City, Utah
1. , Singh H. Natural history of congenital kyphosis and kyphoscoliosis. A study of one hundred and twelve patients. J Bone Joint Surg Am. 1999;81: 1367-83.
2. . Kyphoscoliosis. J Bone Joint Surg Br. 1955;37: 414-26.
3. , Moe JH, Wang JF. Congenital kyphosis. Its natural history and treatment as observed in a study of one hundred and thirty patients. J Bone Joint Surg Am. 1973;55: 223-56.
4. , Herring J. Congenital vertebral displacement. J Bone Joint Surg Am. 1993;75: 656-62.
5. . Etude pratique sur le mal du Pott. Paris: Masson et Cie, 1900. In: Edmonson AS, Crenshaw AH, editors. Campbell's operative orthopedics. 6th ed. St. Louis: C.V. Mosby; 1980. p 2091.
6. . In: Platt H, editor. Modern trends in orthopedics, second series. London: Butterworth; 1956.
7. , Herkowitz HN. A modified posterolateral approach to the thoracic spine. J Spinal Disord. 1995;8: 69-75.
8. , Tompkins J, Clark SB. Transcostovertebral approach for thoracic disc herniations. J Neurosurg. 2001;94(1 Suppl): 38-44.
9. , Matsuda H, Konishi S, Yamano Y. Single-stage excision of hemivertebrae via the posterior approach alone for congenital spine deformity: follow-up period longer than ten years. Spine. 2002;27: 110-5.
10. , Uede T, Igarashi K, Tatewaki K, Morimoto S. Thoracic dumbbell-shaped neurinoma treated by unilateral partial costotransversectomy—case report. Neurol Med Chir (Tokyo). 1997;37: 354-7.
11. . Modification of costotransversectomy to approach ventrally located intraspinal lesions. Preliminary report. Acta Neurochir (Wien). 1995;136: 12-5.
12. , Abumi K, Kaneda K. One-stage posterior hemivertebra resection and correction using segmental posterior instrumentation. Spine. 2001;26: 752-7.
13. , Harms J. Hemivertebra resection by a posterior approach: innovative operative technique and first results. Spine. 2002;27: 1116-23.
14. , Lewis SJ, Rinella A, Lenke LG, Baldus C, Blanke K. Pedicle subtraction osteotomy for the treatment of fixed sagittal imbalance. Surgical technique. J Bone Joint Surg Am. 2004;86 Suppl 1: 44-50.
15. , Otsuka NY, Flynn JM, Hall JE, Emans JB, Hresko MT. Surgical treatment of congenital kyphosis. Spine. 2001;26: 2251-7.
16. , Singh H. Surgical management of congenital kyphosis and kyphoscoliosis. Spine. 2001;26: 2146-55.
17. , Moe JH, Lonstein JE. The surgical treatment of congenital kyphosis. A review of 94 patients age 5 years or older, with 2 years or more follow-up in 77 patients. Spine. 1985;10: 224-31.
Copyright 2005 by The Journal of Bone and Joint Surgery, Incorporated
18. , Harms J. Posterior hemivertebra resection with transpedicular instrumentation: early correction in children aged 1 to 6 years. Spine. 2003;28: 2132-8.