One-stage Posterior Transforaminal Thoracic Debridement and Interbody Fusion with Instrumentation for Multi-segment Tuberculosis of the Thoracic Spine in Adults: A Single-center Mid-to-long-term Follow-up Study


 Background Multi-segment tuberculosis of the thoracic spine in adults poses a therapeutic challenge due to the complexity of infection and the need for stabilization. Experience with a mid-to-long-term follow-up for management that uses a single-stage debridement through the posterior transforaminal approach coupled with stabilization that uses fusion and instrumentation is presented here.Methods From January 2007 to October 2015, a total of 42 adult patients were diagnosed with tuberculosis of the thoracic spine that involved 3 or more levels; they underwent a single-stage procedure with posterior transforaminal debridement. The instrumentation, in addition, to interbody fusion was performed for all patients. All patients received the appropriate medical management for tuberculosis. The treatment response was recorded as an improvement on serial biochemical markers (ESR, CRP levels), pain scores, and the measurement of kyphotic angles and spine scores (ASIA, SF-36 scale).Results At a mean follow-up of 73.5 + 9.6 months, all 42 patients were eligible for final evaluation. The reduction at 3 months and a return to physiological levels of biochemical markers as well as pain scores were noted in all patients at end of treatment. VAS and SF-36 scales were significantly improved compared with preoperative values. In total, 30 patients with a preoperative neurological deficit underwent the procedure and improvement in neurology was noted in all. The mean kyphotic angle loss of 1.7±1.1° was recorded at final the follow-up and bone fusion was observed at a mean of 10.6±2.1 months with no failure of instrumentation. There were no failures of graft lysis or implant breakages noted in this study.Conclusions Our mid-to-long-term experience with the management of multilevel thoracic tuberculosis in adults demonstrated the ability to offer a single staged procedure for posterior debridement as well as fusion with instrumentation and this approach has shown promising results in this study.


Background
Skeletal tuberculosis (TB) remains a signi cant disease burden, in particular, in the developing world [1]. Signi cant challenges in spinal TB arise from the resulting bone destruction that might lead to deformity, neurological de cit, and even paraplegia [2]. However, anti-TB chemotherapy remains the mainstay of treatment for spinal TB [5], to date, various therapeutic options have been suggested to manage spinal TB, which includes drug therapy alone, surgery followed by medical therapy, or concomitant use of these [3,4]. A conservative approach with medical management cannot fully prevent the potential progression of a kyphotic deformity, which often results in chronic back pain, or a residual deformity on completion of therapy, or both [6]. Although the incidence is relatively low, multi-segment thoracic TB can provoke acute tuberculosis symptoms and a high disability rate. This makes treatment di cult, and more cases require surgical treatments compared with the single-segment thoracic TB [7,8]. A lack of consensus on the management protocol for these infections has been noted in the literature. The central segment of the vertebral body is often characterized by extensive damage with sequestrum fragments and massive cold abscesses, which can easily invade the spinal canal [9]. Vital organs are located in the vicinity of the thoracic spine, which brings challenges to multi-segment thoracic TB surgery. The adjacent organs of the thoracic spine should be protected from damage and attention should be paid to the reconstruction of the thoracic spine stability after the lesion is cleared.
With the development of advanced treatments and surgical options, stabilization through posterior instrumentation remains the mainstay for correction of angular deformities and provides stabilization; however, fusion across the spinal segments has been noted when the anterior approach is advocated for easier access to the diseased body segment of the vertebras [10,11]. The transforaminal thoracic interbody fusion aimed to avoid retraction of the neural elements that are necessary when using the anterior approach [12]. Due to a lack of evidence and guidelines for the optimal treatment and management strategies, the treatment of multi-segment thoracic spine TB remains controversial. Our experience with a mid-to-long-term follow-up for management that uses a single-stage debridement through the posterior transforaminal approach coupled with stabilization using fusion and instrumentation is presented.

Patient diagnosis, inclusion, and exclusion criteria
Ethical approval was sought by the institutional review board at Xiangya Hospital, Central South University and all investigations and research performed conformed with the ethical principles of research. From January 2007 to October 2015, 178 patients with multi-segment (3 or more levels) thoracic spine TB were evaluated at our center. In total, 76 patients did not undergo surgical management and 42 patients were not suitable for a single-stage posterior approach procedure due to patient or disease characteristics. In total, 60 patients underwent a single-stage posterior approach debridement through the transforaminal corridor and stabilization using interbody fusion as well as instrumentation. Patients with limited vertebral body, or disc destruction, or both, a large paravertebral abscess, or sequestrum formation, or both in the regions anterior to the vertebrae, and bone destruction without spinal instability and spinal cord compression, were excluded. A total of 42 patients were eligible for nal evaluation at a minimum 5-year follow-up and were included in this study (Table 1) and 18 patients were lost to follow-up before completion of treatment or the study endpoints were achieved. Patients that had the following criteria for surgery were included: (1) vertebral lesions that caused spinal instability, which led to persistent back pain; (2) progressive neurological dysfunction caused by spinal cord compression; (3) unpreventable progressive kyphosis; and (4) weak e cacy of conservative treatment. Among the patients, 25 were male and 17 were female. The average age at the time of surgery was 45.8 ± 13.8 years. All the patients presented with at least one constitutional symptom, such as back pain, weakness, malaise, night sweats, low fever, and weight loss.
Disease-speci c presentation was noted in 22 patients (52.4%) with intercostal neuralgia, 30 patients (71.4%) showed incomplete paraplegia, and 1 patient (2.4%) suffered from complete paraplegia. The initial assessment and suspicion of tuberculosis were made based on biochemical and radiological parameters. The con rmatory diagnosis was based on histopathological ndings or a positive Mycobacterium tuberculosis culture of the tissue samples. Preoperative procedures All patients received standard anti-TB chemotherapy as advised by the institutional protocol and medication composed of isoniazid, rifampicin, ethambutol, and pyrazinamide. Doses were adjusted to patient weight and considerations for maximum daily dose were taken before the medical prescription. Surgery was scheduled after recovery from anemia and hypoproteinemia as well as a reduction in in ammatory markers levels were noted following medical therapy. Upfront surgery was advocated only if a patient experienced acute paralysis or progressive neurological impairment when on medical therapy.

Operative procedures
The patient was positioned prone after the administration of general anesthesia. The diseased segment was localized under an image intensi er and a posterior midline approach was centered over the affected levels. Exposure of the posterior elements was followed by decompression at the disease levels on either side to expose the costotransverse joints. This allowed complete debridement through posterior access after a unilateral facetectomy and laminectomy. Exposure was increased as appropriate by sacri cing the thoracic nerve roots on the worse affected side to allow a more thorough debridement when offering a > 270 decompression. Before destabilizing the vertebral column, temporary rod stabilization was utilized on the side with milder destruction from the infection. This allowed disease clearance and avoided spinal cord injury that results from instability. The nal stabilization was planned for two levels above and below the diseased segment. The affected segment was sometimes included in the xation if there was adequate bone stock for screw purchase. The extent of the interbody fusion was determined and the body endplate was debrided to achieve bone surfaces to facilitate fusion. An autograft from excised posterior elements with or without allografts was used for the fusion. Local installation of streptomycin and isoniazid (1.0 g and 0.2 g, respectively) was utilized before closure over a suction drain. All diseased tissue was collected for pathology and microbial culture assessment.

Postoperative procedures
Postoperative orthogonal radiographs were taken to document the decompression and instrumentation. Patients were advised in-bed physical therapy when maintaining a supine position for 4 to 6 weeks. Out of bed mobilization was initiated gradually, and the use of orthosis was suggested as required. Anti-TB therapy was continued for 12 to 18 months following surgery as recommended by the institutional protocol.

Follow-up
Regular outpatient reexaminations were required for disease assessment as well as spinal parameters every 3 months in the rst year postoperatively and every 6 months thereafter.

Statistical analyses
Using SPSS version 18.0 (IBM Corp, Armonk, NY, USA), statistical analysis was performed for descriptive data as well as comparative analysis using a paired t-test, which was utilized to assess the difference in study parameters (preoperative versus postoperative) with a two-tailed p-value signi cant at p < 0.05.

Results
In total, 42 patients with a mean follow-up of 73.5 ± 9.6 months were analyzed at the nal follow-up. All patients had a pathology diagnosis of tuberculosis that was demonstrated by either tuberculous granulomas or caseous necrosis. Culture positivity for Mycobacterium tuberculosis was seen in 11 patients.
The average intraoperative bleeding volume and surgery duration were 842.9 ± 182.7 mL and 182.2 ± 26.0 min, respectively. The average pretreatment ESR and CRP were 60.6 ± 18.6 mm/h and 43.9 ± 11.2 mg/L, which signi cantly decreased to 11.4 ± 2.7 mm/h and 4.4 ± 1.2 mg/L 3 months after surgery, respectively. Furthermore, at the nal follow-up, the ESR and CRP values respectively returned to 4.2 ± 1.5 mm/h and 1.8 ± 0.7 mg/L. Statistical analyses indicated that the preoperative ESR and CRP values were signi cantly lower (p < 0.05) from those for 3 months postoperative and the nal observation period.
The mean pain (VAS) score was 7.0 ± 1.2 preoperatively and decreased to 0.8 ± 0.7 at the nal follow-up (Table 2). Patients' post-operation life quality was signi cantly improved compared with that before operation according to the SF-36 scale score (Table 3).   ostoperatively. At the nal follow-up, the mean deformity angle was 23.7 ± 2.1°, with only a 1.7 ± 1.1° correction loss (p < 0.05). Spontaneous intervertebral bone fusion was achieved 10.6 ± 2.1 months after surgery ( Table 2). No nonunion, pseudoarthrosis, loosening, or fracture of instruments was observed at the nal follow-up ( Figs. 1 and 2).
There were no wound-related complications in the series, except for the one case that experienced delayed healing of the incision, which was successfully managed with conservative treatment. Postoperative intercostal neuralgia required nonsteroidal anti-in ammatory drugs in 5 patients.

Discussion
Spinal TB, which is the most common presentation of skeletal TB, continues to increase in prevalence globally [13][14][15]. Multi-segment thoracic TB is a relatively rare but very serious disease. Although medical therapy with anti-TB drugs is the mainstay in the management of tuberculous spondylitis, active and early surgical intervention and stabilization might be advocated to limit the risk of neurologic de cit or progressive instability and deformity [16,17]. The reduction of bacterial load, and therefore, a better response to medical therapy as well as the ability to avoid prolonged immobilization when on medical therapy alone appear to favor of early surgical intervention [18].
The primary aim of surgery would be to offer stability when decompressing the spinal cord through debridement at the same time. An anterior surgical approach has been popular due to the predominantly anterior body involvement by tuberculous infection, which allows debridement through direct access to the focus point [19][20][21]. However, for multi-segment thoracic TB, the anterior approach alone is limited in exposure and ability to offer stabilization across a long disease segment. Instability is associated with the failure of a graft to heal or implant breakages that lead to spinal cord injury. Due to this limitation, a combined anterior and posterior approach has been adopted by surgeons to offer disease control, which allows for stabilization and posterior instrumentation [22,23]. However, a staged procedure requires the patient to undergo surgical procedure twice and this adds to stress during recovery in an already nutritionally and metabolically challenged state in which TB thrives [24,25].
With these considerations, the single posterior only approach has gained popularity for the treatment of multilevel spinal TB in the thoracic spine [26][27][28]. The posterior approach avoids the disturbance of the physiological functions of organs in the thoracic cavity, which causes reduced surgical trauma. Besides, the three-column xation of pedicle screws could provide strong biological xation for a short time after surgery, the spine; therefore, could be reconstructed immediately, and a favorable kyphosis correction could be obtained. Finally, the intervertebral defect is adequately implanted with size-matched bone block to reconstruct the anterior column after lesion removal. The long-term stability of the spine could be achieved after bone graft fusion. In addition, favorable outcomes after posterior debridement and internal xation in elderly patients with multi-segment tuberculous spondylitis have been reported [7,29]. The comparative assessments in this study supported this nding based on a quantitative analysis of the correction of kyphotic deformities, The VAS score re ecting pain intensity dropped over 80% over the mid-to-long-term follow-up after surgical treatment. Signi cant increases in the SF-36 scores suggested favorable overall health status during follow-up. These assessments demonstrated that the patients' quality of life was improved, as expected.
From this series and experience with this posterior only approach, it was noted that the posterior instrumentation provided improved biomechanical support due to its three-column xation. this type of xation can correct kyphosis, reduce the angle loss of deformity correction and relieve pains due to spinal instability. Although Mycobacterium tuberculosis does not adhere to the internal xation, other bacteria can adhere and even form a bio lm. The diagnosis of spinal tuberculosis was often postoperative and di cult to distinguish from other bacterial infections. Because tuberculous lesions are commonly involved in the anterior column, a posterior approach allows internal xation away from bacteria. In addition, as a chronic disease, many patients with spinal TB are in a poor nutritional state. The posterior approach reduces blood loss and anesthesia time, thereby reducing the risk of intraoperative and postoperative complications. Finally, where there is a need to mobilize the complex anatomical structures through the anterior approach, the posterior has a direct approach; therefore, the complications to the great vessels or viscera are minimized.
This surgical treatment of multi-segment thoracic TB can achieve posterior decompression, unilateral anterior decompression (total 270-360° decompression), and reconstruction of anterior load support by interbody fusion is also achieved by the thoracic reconstruction technique [30]. A thorough removal of TB lesions is the key to surgical treatment of spinal tuberculosis. Since the surgery cannot achieve full sterility of the lesion, effective anti-TB drug treatment and improvement of the patient's general condition are important aspects of the treatment of spinal TB. The debridement of the surgical focus is to promote the quiescence and healing of the tuberculosis focus, destroy the environment favored by the survival of tuberculosis bacilli, and promote the anti-TB drugs to penetrate the focus area. Remove pus, caseous necrotic tissue, dead bone, granulation tissue and necrotic intervertebral discs in the lesion area, and do not emphasize the expansion of lesion removal [31]. For the tubercular lesion wall that has not sclerosed in multi-segment spinal TB, the curettage of the lesion wall should be cautious due to the osteoporosis around the lesion, and imaging data should be referred during the operation to prevent the loss of healthy bone.
Complete removal of the central lesion was also feasible for spinal canal decompression, xation and bone grafting to reconstruct the stability. For satellite lesions, paravertebral pus, caseous necrotic tissue and tuberculous granulation tissue were removed and the lesions were scratched to the normal bone surface with a curette. Paravertebral abscesses can be eliminated by catheter lavage, negative pressure suction and postoperative postural drainage.
The authors agree that the posterior only approach might be radical and requires careful patient selection.
They advocated that these considerations should be taken into account when managing multilevel thoracic spine TB: (1) whenever possible and in the presence of adequate bone stock, the affected vertebrae should be incorporated into the instrumentation system; (2) enough graft should be impacted into the defect, and the graft between vertebrae should be xed to promote fusion; (3) temporary rod stabilization for the spinal cord during transforaminal thoracic debridement should be provided; (4) careful patient selection is required for method, which is mainly used in spinal TB with limited bone destructions in the middle column; (5) careful monitoring to ensure that the spinal cord is not affected (stretched or distracted) when debriding the lesions and spinal monitoring could be added for safety during these procedures. During the whole followup process, all patients achieved bone fusion, and no bone nonunion or recurrence caused by insu cient lesion removal was observed.
This study is limited by its small size; however, multilevel spinal TB remains rare, and therefore, these numbers might not be disregarded. Despite the favorable outcomes of this study, the authors would recommend a larger multicenter study with the ability to compare results between the various surgical approaches before advocating only a single-stage posterior approach for all patients.

Conclusion
The mid-to-long-term experience with the management of multilevel thoracic TB in adults demonstrated the ability to offer a single staged procedure for posterior debridement as well as fusion with instrumentation and this approach has shown promising results in this study.

Consent for publication
Written informed consent was obtained from all the patients for publishing their personal details in this article. A copy of the patient's written consent is available for review by the Editor-in-Chief of this journal.

Availability of data and materials
The datasets and materials generated or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This work was supported by the National Natural Science Foundation of China (No. 81672191). No bene t in any form has been or will be received from a commercial party related directly or indirectly to the subject of this manuscript.
Authors' contributions ZC Xu and Z Zhang participated in the design of this study, performed the statistical analysis, and drafted the manuscript. XY Wang collected the clinical data and follow-up details of the study and conducted the study. YQ Wu directed the study design and manuscript preparation. All authors read and approved the nal manuscript.

Figure 1
A 49-year-old female demonstrating lesions received a one-stage posterior transforaminal thoracic debridement, interbody fusion, and instrumentation (a-f). The preoperative images from radiographs, CT scans, and MRI scans show tuberculosis of T5-T8 vertebral bone destruction with deformity (kyphosis angle was 36°), and paravertebral abscess formation. Postoperative radiographs (g-h) that demonstrate the correction of the deformity (kyphosis angle was 23°). Postoperative MRI (i) that shows that the tuberculosis lesions has cleared, and the compression of the spinal cord was relieved. CT (j) that shows satisfactory bone fusion at 12 months. Radiographs that display good internal xation position (k-l) and solid bone fusion, with a correction loss of 1° after 81 months follow-up.  Postoperative radiographs (f-g) that demonstrate that instrumentation was achieved in a good position.
CT (h) that shows satisfactory bone fusion was achieved at 12 months. (i-j) Radiographs (i-j) that display good internal xation position and solid bone fusion, with no loss of correction after 63 months follow-up.