Multiple three-column osteotomies successfully correcting cervicothoracic kyphosis in the setting of ankylosing spondylitis: illustrative case

Abstract

BACKGROUND

Ankylosing spondylitis (AS) is an autoimmune spondylarthritis often associated with rigid kyphoscoliosis. The authors describe a surgical approach that employs multilevel three-column osteotomies for the restoration of normal global alignment.

OBSERVATIONS

A 48-year-old male with a past medical history of AS presented to the clinic with a stooped-over posture: his chin-brow vertical angle (CBVA) was 58.0°; T1 slope (T1S), 97.8°; thoracic kyphosis (TK; T1–12), 94.2°; proximal TK (T1–5), 50.8°; distal TK (T5–12), 43.5°; and sagittal vertical axis (SVA), 22.6 cm. A two-stage procedure was planned. During stage 1, instrumentation was placed from C5 to T10, followed by a T3 vertebral column resection. During stage 2, bilateral pedicle screws were placed from T11 to the pelvis. An L3 pedicle subtraction osteotomy (PSO) was completed and was followed by a T7 PSO. Postoperatively, the patient had significant postural improvement: CBVA was 29.3°; T1S, 57.8°; TK, 77.3°; proximal TK, 33.5°; distal TK, 43.8°; and SVA, 15 cm. At 6 years postoperatively, the patient continued to do well and was without evidence of construct breakdown.

LESSONS

The authors propose that multilevel three-column osteotomies, if optimally located, successfully correct spinal malalignment associated with AS.

Ankylosing spondylitis (AS) is an irregular bone homeostasis-induced spondylarthritis, often leading to secondary kyphoscoliosis.^1,2 AS-induced fixed kyphoscoliosis can lead to limited vital capacity from mechanical restrictions on pulmonary function, an increased risk of falling due to impaired balance from a lack of mobility, and an overall decreased quality of life and limited daily activities, which compound and result in a higher risk of premature death because higher mortality rates correlate with an increased kyphotic angle.^3,4 There is, however, no current guideline for surgical correction.⁵ In cases of extreme kyphoscoliosis (>60° in the sagittal plane), double pedicle subtraction osteotomy (PSO) conducted at the lower thoracic and L2–3 vertebrae has been found to be effective in correcting spine misalignment.^6,7 Unfortunately, PSO requires a longer operative time than lower-grade osteotomies, carrying greater associated blood loss and risk of neurological injury.⁸ Despite these risks, PSOs are tolerated well, can provide significant correction in the setting of a rigid spine, and are often necessary in the setting of focal deformity compared with lower-grade osteotomies such as Smith-Peterson osteotomies.^8,9 We describe a novel approach using three-stage, noncontiguous PSO at the T4, T7, and L3 vertebra levels for the correction of the fixed kyphoscoliosis and restoration of global spinal alignment.

Illustrative Case

Presentation

A 48-year-old male presented to the clinic with neck pain and an inability to stand up straight. Being upright and trying to straighten his neck caused significant pain and fatigue. His past medical history was significant for AS, which had been diagnosed 15 years prior. He also had a diagnosis of osteoporosis with a T-score of 2.7 and was being treated with Voltaren and Fosamax. His chin-brow angle was 0°. Neurologically, the patient was intact and without balance issues. Preoperative radiographs (Fig. 1) and radiographic measurements were as follows: pelvis and shoulders were level; pelvic incidence (PI) was 59.3°; pelvic tilt (PT), 38.2°, lumbar lordosis (LL), 5.0°; and sacral slope (SS), 20.7°. Chin-brow vertical angle (CBVA) was 58.0°; T1 slope (T1S), 97.8°; thoracic kyphosis (TK; T1–12), 94.2°; proximal TK (T1–5), 50.8°; distal TK (T5–12), 43.5°; and sagittal vertical axis (SVA), 22.6 cm. Surgimap (Nemaris Inc.) was used for planning the surgical intervention.

FIG. 1.

Preoperative imaging. A: Sagittal 36-inch standing scoliosis radiograph. B: Coronal 36-inch standing scoliosis radiograph. C: Sagittal cervical spine T2-weighted magnetic resonance imaging (MRI). D: Sagittal thoracic spine T2-weighted MRI. E: Sagittal lumbar spine T2-weighted MRI.

Operation

The patient was intubated and positioned prone. After exposure, instrumentation was placed from C5 to T10, including C5–6 lateral mass screws and pedicle screws from T1 to T10. Laminectomies were completed from T2 to T4. A vertebral column resection (VCR) was done at T3 with insertion of an interbody cage. Using bivector traction and a compressor, we placed pressure across the osteotomy site, and rods were placed. Of note and secondary to poor bone quality, T1 and T2 screws pulled out, necessitating relaxation of some of the correction. Drains were placed, and the patient was closed. Total estimated blood loss for stage 1 was 500 ml. Both cell saver and tranexamic acid were used for intraoperative blood loss control. The patient was discharged home on postoperative day 6. Intraoperative images are provided in Fig. 2.

FIG. 2.

Intraoperative imaging of stage 1. A: Sagittal fluoroscopy demonstrated pedicle screw placement within the thoracic spine. B: Sagittal fluoroscopy demonstrated cages placement at T3 after completion of the VCR. C: Anteroposterior fluoroscopy demonstrating cage placement and temporary rod placement after the T3 VCR.

After stage I, the patient reported improvement in his posture and appetite. He started Forteo in anticipation of stage 2. One and one-half years later, after completing courses of Forteo and vitamin D, the patient elected to undergo stage 2 of the surgery for final correction.

The patient was brought to the operating room, intubated, and positioned prone with neuromonitoring. After exposure, laminectomies were done from T6 to T8 and from L2 to L4. After decompression, bilateral pedicle screws were placed at T11–pelvis, although L3 was skipped. Bilateral S2 alar-iliac screws were placed. Intraoperative fluoroscopy was used for confirmation of placement. Next, an L3 PSO was completed using temporary rod pedicle subtraction spoons. A cage was placed ventrally after the bone was removed. After creating the wedge osteotomy, a cantilever maneuver was used to close the osteotomy. Next, a PSO was completed at T7 in the same fashion, although no cage was placed. After completion of both PSOs at T7 and L3, permanent rods were placed spanning T4 to the sacrum. Accessory rods were placed, which spanned the T3 and T7 osteotomy sites. An additional accessory rod was secured across the L3 PSO on the right. For fusion, decortication was complete, and an off-label-use bone morphogenetic protein was placed. Total estimated blood loss for the lumbar three-column osteotomy (3CO) was 1 L and for the thoracic was 800 ml. The same intraoperative protocol was used for blood loss control. A drain was positioned, and the patient was closed. Intraoperative fluoroscopic images are presented in Fig. 3.

FIG. 3.

Intraoperative imaging of stage 2. A: Anteroposterior fluoroscopy demonstrating placement of bilateral S2 alar-iliac screws. B: Sagittal fluoroscopy of the final L3 PSO with cage placement. C: Sagittal fluoroscopy of T7 PSO without cage placement.

The hospital stay was uneventful, and the patient was discharged home on postoperative day 8. Postoperative measurements were as follows: pelvis and shoulders remained level; PI, 59.5°; PT, 28.8°; LL, 38°; SS, 28.0°; T1S, 57.8°; TK, 77.3°; proximal TK, 33.5°; distal TK, 43.8°; CBVA, 29.3°; and SVA, 15 cm. The patient was seen in follow-up for many years. Figure 4 shows clinical images of the patient before and after surgery. The most recent follow-up was 6 years after the first surgery. He was neurologically intact and pain free. Unfortunately, no long-term pulmonary data were available. Figure 5 shows radiographs obtained at that time.

FIG. 4.

Lateral photographs obtained in clinic. Left: Preoperative photograph demonstrating alignment. Right: Postoperative photograph after the correction.

FIG. 5.

Six-year postoperative 36-inch scoliosis radiographs. Left: Coronal standing radiograph. Right: Sagittal standing radiograph.

Patient Informed Consent

The necessary patient informed consent was obtained in this study.

Discussion

Observations

Rigid kyphotic deformity presents a challenge from a surgical perspective. Our patient had significant kyphoscoliosis with TK of 85.9°, which resulted in a CBVA of 58° and very low T1S. As a result of his AS, he also had no LL. Although severe kyphosis is generally thought of as >70°, modern pedicle screw fixation techniques have permitted correction without anterior column release in pediatric populations.¹⁰ Our patient presented in his 40s. Depending on the rate and degree to which AS has caused autofusion, surgery should be undertaken when global alignment has degraded and there is a significant impact on quality of life. In our case, the patient presented to the clinic in extremis, given the extent of disease progression that had already occurred.

It has been shown that global alignment is worse in patients with AS, given their inability to compensate.² AS removes any natural compensation so that a normal spine would have to adopt a focal deformity, which results in fixed malalignment with even small changes throughout the thoracolumbar spine. The very nature of the fixed deformity necessitates multiple, noncontiguous 3COs to achieve global alignment. This is true even if minimal correction is needed, given the rigidity of the spine and the inability for any compensation.^7,11 Low-grade osteotomies and gradual correction are not possible.¹² Fortunately, this pathology does typically have a focal deformity and a relatively low deformity angular ratio, making it favorable for a 3CO because of its associated lower neurological complication rate.¹³

Historically, 3COs have proved efficacious for the correction of focal deformity in both adult and pediatric populations.^14,15 When examining AS, 3CO and specifically pedicle subtraction osteotomies are effective for overall correction.¹⁶ Ha et al.⁶ demonstrated that the use of two noncontiguous VCRs is helpful in the restoration of alignment in the context of AS. This technical report differs from ours in that it employed only two 3COs, and they were full VCRs. Our case involved an apical VCR and two noncontiguous PSOs to achieve correction. These 3COs permitted gradual global correction and the establishment of a harmonious physiological curve. Looking to the future, employing preoperative three-dimensional printed templates for a planned PSO, surgical corrections to severe kyphosis secondary to AS through staged osteotomy in lateral position result in a greatly improved SVA.^17,18 Here, multiple 3COs were useful to improve global alignment as manifested by a decrease in the SVA and CBVA and an increase in T1S.

Osteotomy site selection was deliberate. The apical vertebra was T4, which resulted in horizontalization of the cervical spine and loss of LL. This made T4 the appropriate initial site of correction, offering the additional benefit of avoiding nerve root motor injuries associated with 3COs performed at the cervicothoracic junction and the paresthesia occasionally seen with a T1 PSO.^19,20 Using preoperative surgical planning software, we estimated that a T4 3CO would result in 20° of apical correction, altering global alignment and establishing T7 as the new apex. Thus, T7 was chosen as the second 3CO site. L3 was the final site because it would permit restoration of the most physiological LL. The resulting global alignment and proportion score for our case was 10, secondary to AS.²¹ However, at 5-year follow-up, the patient was without signs of pseudoarthrosis or proximal junctional kyphosis.

Lessons

Surgical correction of kyphoscoliosis in the setting of AS remains a challenging undertaking. Given the ridged nature of this pathology, the restoration of physiological alignment necessitates the employment of multiple strategically placed 3COs. Planning the location and grade of each osteotomy is essential to obtain correct alignment. We demonstrate that multilevel PSO represents a successful approach for primary sagittal correction in AS.

Author Contributions

Conception and design: Mugge, Brewer, McHugh. Acquisition of data: Mugge, Brewer, McHugh. Analysis and interpretation of data: Mugge, Gorka, Brewer. Drafting the article: Mugge, Gorka. Critically revising the article: Mugge, Gorka, McHugh. Reviewed submitted version of manuscript: Mugge, Gorka, McHugh. Approved the final version of the manuscript on behalf of all authors: Mugge. Statistical analysis: Mugge. Administrative/technical/material support: Mugge, Brewer. Study supervision: Mugge.