Background: Cervical spinal tuberculosis is relatively common in some developing countries. It erodes vertebrae and discs, which sometimes results in cervical kyphosis and myelopathy. However, to our knowledge, no studies have evaluated improvements to patient-reported outcomes among patients who undergo surgical cervical sagittal realignment after kyphotic cervical spinal tuberculosis has been treated by débridement and reconstruction.
Questions/purposes: (1) Can a spine with kyphotic cervical spinal tuberculosis be returned to normal alignment and fused successfully? (2) Will patient-reported outcomes be improved with this intervention? (3) Are patient-reported outcomes correlated with realignment?
Methods: Forty-six patients with kyphotic cervical spinal tuberculosis were evaluated in this retrospective study. We generally performed surgery on patients with this condition when patients with cervical spinal tuberculosis presented with cervical kyphosis with or without neurologic deficits. Patients who did not meet these criteria were treated with other surgical procedures during the study period. Study patients were evaluated with cervical imaging, patient-reported outcomes questionnaires (Neck Disability Index [NDI], and the Japanese Orthopaedic Association [JOA] score), and physical examinations. Scores were collected by fellows preoperatively and at followup. No patient died during the followup. The mean followup was 26.8 months (range, 20-35 months). Preoperative and 2-year followup radiologic parameters were measured, including C0-2 Cobb angle, C2-7 Cobb angle, C2-7 sagittal vertical axis, center of gravity (CG) to C7 sagittal vertical axis (CG-C7 sagittal vertical axis), thoracic inlet angle, T1 slope, and neck tilt. The correlations between cervical alignment and the NDI and JOA score were analyzed. Factors correlated with the NDI and JOA score improvements were identified by multiple stepwise regression analysis. CT was used to assess bone fusion after surgery.
Results: All 46 patients showed bone fusion on CT scans. The preoperative C0-2 Cobb angle improved after surgery (mean difference, 5.0°; 95% CI, 2.3°-7.7°; p = 0.0068), as did C2-7 Cobb angle (mean difference, -33°; 95% CI, -35° to -31°; p = 0.0074), C2-7 sagittal vertical axis (mean difference, -28 mm; 95% CI, -30 mm to -26 mm; p = 0.0036), CG-7 sagittal vertical axis (mean difference, -26 mm; 95% CI, -28 mm to -24 mm; p = 0.0049), T1 slope (mean difference, 6.0°; 95% CI, 3.7°-8.3°; p = 0.0053) and the thoracic inlet angle (mean difference, 8.0°; 95% CI, 3.7°-12°; p = 0.0072). With the numbers available, the neck tilt angle did not improve (mean difference, -0.2°; 95% CI, -1.0° to 0.6°; p = 0.079). The preoperative NDI of 34 ± 5.1 decreased to 17 ± 4.6 (p = 0.0096) at followup. Improvements in NDI were correlated with the magnitude of correction of the cervical deformities, including C0-2 Cobb angle (r = -0.357, p = 0.007), C2-7 Cobb angle (r = 0.410, p = 0.002), T1 slope (r = -0.366, p = 0.006, thoracic inlet angle (r = -0.376, p = 0.005), C2-7 sagittal vertical axis (r = 0.450, p = 0.001), and CG-C7 sagittal vertical axis (r = 0.361, p = 0.007). The JOA score improved to 13 ± 2.6 from 7.2 ± 1.9, which did not correlate with postoperative cervical realignment. After controlling for potential confounding variables like Cobb angles and T1 slope, we found C2-7 sagittal vertical axis was the most influential factor correlated with NDI improvement (r = 0.450, p = 0.002).
Conclusion: When treating kyphotic cervical spinal tuberculosis by débridement, decompression, and reconstruction, more attention should be drawn to realigning the cervical spine, in particular to restoring the C2-7 sagittal vertical axis. However, how best to restore the C2-7 sagittal vertical axis and cervical alignment in a kyphotic cervical spine needs further study.
Level of evidence: Level III, therapeutic study.