Background

Spinal tuberculosis accounts for half the cases of osteo-articular tuberculosis. Vertebral inoculation occurs via the haematogenous route, and the process then spreads to the intervertebral disk and, in some cases, to the adjacent vertebra. Para-spinal abscesses may develop by direct spread from the vertebral lesion. The thoracic spine is predominantly involved [1]. The patients with spinal tuberculosis that were treated conservatively presented with kyphosis with an average of 15°, and eventually, 3–5% of them ended up with a kyphotic angle of more than 60° [2,3,4,5]. Furthermore, even if spinal tuberculosis heals during the period of growth and development of children, an imbalance in the spinal growth of the anterior and posterior columns could still aggravate kyphosis [6, 7]. Severe post-tubercular angular kyphosis generally affects children's self-confidence and appearance and could even lead to severe neurological impairment and cardiopulmonary dysfunction. Many surgical options have been described to manage post-tubercular angular kyphosis, but the standard approach has not yet been established.

Until now, surgical osteotomy has been the leading choice for correcting the deformity and realigning the spine. Posterior column osteotomy (PCO) is not indicated as a first-line treatment for post-tubercular angular kyphosis because of the difficulty in decompressing the spine canal anteriorly and the relatively small correction angle [8,9,10]. Conversely, pedicle subtraction osteotomy (PSO) is a wedge-shaped resection in one vertebra which includes the partial vertebral body and posterior column elements and can only acquire correction of 30°-40° [9,10,11]. Vertebral column resection (VCR) requires removing one or several whole vertebrae, concurrently including adjacent discs and a portion of the ribs in the thoracic region, which could significantly correct severe kyphoscoliosis. However, VCR has several disadvantages, such as neurological complications, blood loss, operation time, nonunion or pseudoarthrosis, and a significant increase in the incidence of postoperative complications [12,13,14,15]. Therefore, the present study was performed to evaluate the safety and efficacy of modified posterior wedge osteotomy within deformed complex vertebrae to treat severe thoracic post-tubercular angular kyphosis (> 70°) in children. This was a retrospective study analysing the outcome after grade 4/5 spinal osteotomies.

Materials and methods

Patient data

In this study, we retrospectively evaluate the clinical efficacy of 15 children (8 boys and 7 girls; age, 5–11 years; average age, 7.8 ± 2.0 years) that suffered from severe thoracic post-tubercular angular kyphosis and were treated in our department between 2010 to 2017. In all children, X-ray, CT, and MRI examinations revealed: deformed complex vertebrae involved with two vertebral bodies in 9 patients and three vertebral bodies in 6 patients; the lesions healed completely in all 15 cases without further destruction of the vertebral bodies or abscesses.

All cases presented with apparent kyphotic deformity and persistent back pain, and some of them incurred incomplete paralysis. C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) in all cases were normal. The Frankel scoring system was used to assess neurological function. Frankel's grade D was in 5 patients, and Grade E in 10 patients. The Ethics Committee of ** the growth potentials of the anterior column and to reduce surgical trauma and intraoperative blood loss. The anterior longitudinal ligament served as a hinge for the bone-to-bone closure of the gap. In this group of cases, the nerve roots were not ligated and slightly affected the visibility of the operative field and the freedom of osteotomy. However, there was no direct damage to the nerve roots, and this helped preserve nervous system functional integrity.

After DCVO, the postoperative mean kyphosis was reduced to 19.98° with a kyphosis correction of 63.41°. The correction rate was considerably better than that of PSO and VCR reported by some authors [9, 10, 12, 14, 17, 18]. Patients' body figure, sagittal alignment and balance (TK, LL, and SVA), pain (VAS), daily activities (ODI), and neurological function (Frankel's grade) all showed good improvement. However, the operation time, blood loss, difficulty of operation, incidence of complications, and neurological events were significantly lower than those of VCR reported in previous studies [12, 14, 17, 19, 20].

Although conservative treatment is effective with the use of antituberculosis medication, however, it is well known that progressive collapse of anterior spinal elements may still occur under gravity and compression. Moreover, due to eccentric loads and spinal growth imbalance of anterior and posterior columns in children, the wedge shape of the vertebral body and kyphotic deformity could be aggravated. Neurological deterioration, cardiopulmonary dysfunction, and severe persistent back pain are all possible side effects of post-tubercular kyphosis [2, 21, 22]. As a result, children with severe thoracic post-tubercular angular kyphosis frequently require surgery.

The primary goal of surgery is to treat spinal deformity and prevent further deformity, as well as to restore and reconstruct the physiological sagittal alignment and decompress the nerve to improve its function. To date, numerous techniques have been developed to treat kyphosis secondary to spinal tuberculosis, mainly including Smith-Petersen Osteotomy (SPO), PSO, and VCR [10, 14, 21, 23]. However, none of those mentioned above techniques was considered a golden criterion for attaining minor trauma and significant correction.

SPO could only achieve approximately 10.7° per segment of kyphosis correction. However, after opening the anterior column, it was determined that there was a risk of permanent neurological impairment and aortic rupture [10, 24]. In recent years, three-column osteotomy has been the primary technique for the treatment of severe angular kyphosis. Hu et al. [9] compared SPO and PSO in the treatment of rigid thoracolumbar kyphotic deformity, finding that PSO's kyphosis correction ranged from 31.7° to 48° with an average of 36.7°, while SPO's was 8.74° less. Furthermore, the incidence of biomechanical complications, including instrument breakage, anterior cortex fracture, pedicle screw loosening, pedicle fraction, vertebral body translation, and nonunion, in PSO was lower than in SPO [10,

Conclusions

In conclusion, DCVO was an individualized osteotomy for treating severe thoracic post-tubercular angular kyphosis in children. It could be safe and effective in reducing the incidence of complications and significantly improving kyphosis correction.