Comparison of single posterior debridement, bone grafting, and instrumentation with single anterior debridement, bone grafting, and instrumentation in treatment of thoracic spinal tuberculosis

Objective To explore the clinical effect of single posterior debridement, bone grafting, and instrumentation and single anterior debridement, bone grafting and instrumentation in the treatment of thoracic spinal tuberculosis of adult patients. Methods: A retrospective analysis was made by involving 38 adult patients with thoracic spinal tuberculosis from between June 2013 and December 2015. Of the 38 patients, 21 patients were categorized in single anterior approach group and underwent single posterior debridement, bone grafting, and instrumentation (Group A). The remaining 17 patients were classified in single posterior approaches group (Group B), which received single posterior debridement, bone grafting and instrumentation. Clinical manifestations, laboratory and imaging results of the two groups were analyzed subsequently. Results: All patients were followed up for 23.9 ± 3.8 m (range, 19–36 m). Bony fusion was achieved in all bone grafts. The operation time and intraoperative blood loss in group B were significantly less than those in group A (P<0.05). The VAS scores, ESR, and CRP levels 6 weeks after operation and at the final follow-up were significantly lower than the preoperative levels (P<0.05). At the last follow-up, ASIA improvement no significant difference between groups (P>0.05). Furthermore, the postoperative and final-follow-up kyphosis angles in group B were both significantly smaller than those in group A (P<0.05). Group A had a postoperative angle correction rate smaller than group B, and its postoperative angle loss was greater than group B’s (P<0.05). Conclusion: Single posterior debridement, bone grafting, and instrumentation can achieve similar curative effect as single anterior debridement, bone grafting and instrumentation in the treatment of thoracic spinal tuberculosis, but also accompanied by additional advantages of shorter operation time and less bleeding .


Introduction
The incidence of spinal tuberculosis (TB) is increasing in developing countries, and every year, two to three million deaths worldwide are related to spinal TB [1][2][3]. STB is the most frequent skeletal tuberculosis, accounting for almost 50% of all skeletal tuberculosis cases [4]. Spinal infections can cause destruction and collapse of the vertebral body and further lead to kyphosis and neurological impairment [5]. When the blood supply to the thoracic spinal cord is inadequate because of thoracic spinal canal stenosis, patients suffering from severe destruction of bone or spinal instability are more prone to neurological impairment [6,7].
Although chemotherapy plays a crucial role in spinal TB treatment, it is often necessary to correct kyphosis and improve neurological function through surgical intervention. Surgical treatment for spinal TB mainly includes debridement, kyphosis correction, and reconstruction of spinal stability [8]. The optimal surgical treatment for spinal TB remains controversial. The single anterior approach allows for direct debridement, bone grafting, and instrumentation; yet, outcomes of this approach regarding kyphosis correction and maintenance are far from satisfactory [8,9]. The single posterior approach is a proven STB treatment of favorable outcomes [10,11], in spite of which it fails to achieve thorough debridement of lesions in the anterior spine.
In this study, a comparative analysis was performed on the clinical outcomes of single anterior and posterior debridement, bone grafting, and instrumentation in treating adult patients with thoracic spinal tuberculosis (TSTB).

Basic information
This retrospective study included 38 active pulmonary tuberculosis-free patients who were

Preoperative therapy
All patients diagnosed with TSTB were treated by HREZ (300 mg/d isoniazid, 450 mg/d rifampicin, 750 mg/d ethambutol and 1.5 g/d pyrazinamide) for at least two to four weeks before operations.

Operative technique
All patients were intubated under general anesthesia. In group A, the patients were placed in a lateral position, and incisions were placed using the extra pleural or extra peritoneal anterolateral approach. From the severely affected side of the lesion, purulent fluid, dead bone, the diseased vertebra and intervertebral disk tissue, and caseous necrotic tissue were removed radically. In a less severe case, pedicle screws could be inserted in the short-segment fixation and correction of kyphosis deformity. An autogenous iliac bone or rib graft was closely embedded into the intervertebral bone grafting groove (Fig. 1).
Patients in group B were placed in a prone position, and a posteromedial incision was made to expose the vertebral plate, facet joints, and costovertebral joints. After the location of the lesion by C-arm fluoroscopy, pedicle screws were inserted in the normal vertebral body adjacent to the upper and lower affected vertebrae, followed by kyphosis correction. From the costovertebral joints or pedicles of the vertebral arch, the lesion in the anterior vertebral body was removed, so as the dead bone, necrotic intervertebral disc, and caseous necrotic tissue. Then, an autogenous iliac bone graft was embedded into the focal zone (Fig. 2). After the operation, a drainage tube was placed, and then the incision was closed. Following the operation, conventional bacterial cultivation and pathological diagnosis were performed.

Postoperative treatment
Prophylactic antibiotics were used for 72 h, and oral administration of HREZ continued after the operation. Six months later, Pyrazinamide was withdrawn while daily isoniazid, rifampicin, and ethambutol (HRE) were administered for another 10 to 12 months. The drainage tube was removed when 24-h drainage volume was less than 50 ml. Following that, patients were allowed to walk with the support of a mobility aid. At first, non-weightbearing activity was recommended only. Normal weight-bearing activity was not allowed until confirmation of intervertebral fusion with x-ray and CT examinations.

Evaluation Standard
Visual analog scale (VAS) scores were used for pain assessment. The Frankel grading system was employed in the assessment of pre-and postoperative spinal cord injuries. 6 ESR and CRP levels were assessed to monitor the disease activity. Pre-and postoperative kyphosis angles were measured using the method proposed by a previous study [12]. Xray and CT examinations were conducted to assess bone graft fusion, loss of correction angle, and internal fixation failure [13]. An analysis was performed on the operation time, intraoperative blood loss, and length of hospital stay in the two groups.

Statistical analysis
The software SPSS 19.0 (SPSS, Inc., Chicago, IL, USA) was used for data analysis.
Wilcoxon's signed-rank test was used for assessing the differences between pre-and postoperative American Spinal Injury Association (ASIA) impairment scale scores. The paired t-test was performed to compare pre-and postoperative degrees of kyphosis deformity and ESR and CRP levels. The independent-samples t-test was used to compare the laboratory and physical parameters of the two groups. The Wilcoxon rank-sum test was employed in the analysis of discrepancy in normal distributions. P < 0.05 was considered indicative of a significant difference.

Results
Definite diagnosis was established through bacterial cultivation and/or pathological diagnosis in all the 38 cases, and mycobacterium tuberculosis was found in 20 patients.
Out of the 38 patients, 21 [10 male and 11 female; mean age 43.5 ± 8.4 years (range, 24-69)] were included in group A while the other 17 patients [8 male and 9 female; mean age 41.4 ± 8.9 years (range, 21-71)] were included in group B. The mean operation time in group A was 222 ± 38 min, significantly longer than that in group B (174 ± 41 min) (P < 0.05). The mean intraoperative blood loss was greater in group A than in group B (600 ± 100 mL vs. 490 ± 110 mL, P < 0.05). The mean length of hospital stay was 22 ± 4 d in group A and 22 ± 5 d in group B, with the difference indicating no statistical significance (P > 0.05).

7
The mean time of bone graft fusion was 6.4 ± 0.2 m in group A and 6.5 ± 0.2 m in group B, and the difference was not statistically significant (P > 0.05). Varying degrees of paraspinal abscess around the psoas muscle was observed in 19 patients of group A and 14 of group B. It took group A 7.4 ± 0.7 m (range, 6-11 m) and group B 7.2 ± 0.6 m (range, 6-12 m) for the abscess to disappear, and the difference showed no statistical significance (P < 0.05). Table 1 lists the sites of lesions and lumbar fusion in the two groups.
Follow-ups were performed 100% in both groups, with the mean duration of follow-up in the entire study population as 23.9 ± 3.8 m (range, 19-36 m). Disruption of wound occurred in one patient in group A, which healed 23 d after redressing and suture; in group B, there was a case of postoperative sinus track formation, and the sinus tract was treated properly and closed four weeks after the operation. A case of breakage of internal fixation was found in group A during a follow-up eight months after the operation. The internal fixation rod was removed after confirmation of bone graft infusion. During the follow-up period, no TB recurrence was reported in either group. Table 2 shows the pre-and postoperative kyphosis angles, angle correction, angle loss, and angle correction rates of the two groups. Differences between the two groups in the preoperative kyphosis angles lacked statistical significance (P > 0.05). The postoperative and final-follow-up kyphosis angles in group B were both significantly smaller than those in group A (P < 0.05). Group A had a postoperative angle correction rate smaller than group B, and its postoperative angle loss was greater than group B's. These differences were statistically significant (P < 0.05).
All patients had neurological impairment before operations (Frankel Grade C or D, Table   3). At the most recent follow-up, 18 patients in group A and 15 patients in group B had their neurofunctional parameters being brought to the normal range. Between-group 8 differences in the neurofunctional parameters at the final follow-up showed no statistical significance. Yet, significant differences were found between the neurofunctional parameters before operations and at the final follow-up in each group (P < 0.05). Table 4 summarizes the changes in the VAS scores, ESR and CRP levels six weeks after operations and at the most recent follow-up. The VAS scores, ESR, and CRP levels six weeks after operations and at the final follow-up were significantly lower than the preoperative levels (P < 0.05). Beyond that, the VAS scores, ESR, and CRP levels were significantly lower than those recorded six weeks after operations (P < 0.05).

Discussion
Existing STB treatment mainly includes anti-TB drugs and surgical intervention. Anti-TB drug therapy plays an essential role in treating STB and provides a basis for surgical treatment. The goals of surgical treatment are to eradicate the lesion(s), relieve spinal cord compression, correct kyphosis deformity, and restabilize the spine [8]. Surgery alone, without regular anti-TB treatment, is extremely dangerous and ineffective. Effective surgical treatment is only possible in combination with effective anti-TB drug therapy. It is difficult to relieve spinal cord compression, improve nerve dysfunction and prevent progressive spinal deformity using conservative treatment alone. In contrast, surgery is an effective solution to this problem [14].
It remains a controversial issue which surgical approach to use when accessing the lesion sites. Single-stage anterior debridement, bone grafting, and instrumentation has been widely accepted as the "gold standard" for STB treatment [15] because it allows for thorough removal of lesions, decompression of the anterior structure of the spine, easy bone grafting, kyphosis correction, and reconstruction of spinal stability under direct vision, with the lesions being fully exposed [16]. However, it shows inadequate fixation rigidity and orthopedic strength [8,9] and imposes a high risk of vascular injury because 9 of poor access to the lesion site [17]. Considering the high exposure, the large wound, and a great risk of hemopneumothorax, this method is not a good choice for patients with multisegmental TSTB [18]. In the present study, the patients who underwent single anterior debridement, bone grafting, and instrumentation in combination with regular anti-TB drug therapy pre-and postoperatively experienced significant improvement in the VAS scores, ESR, and CRP levels as compared to the preoperative levels. Moreover, no serious complications were caused by this approach. Therefore, the anterior approach is a safe and effective option for TSTB treatment under specific circumstances. Despite its benefits, this study found the anterior approach associated with postoperative kyphosis angle loss, which has been agreed upon by other authors [17,18]. However, the single posterior approach also has its own limitations. To be specific, it is difficult to remove TB lesions radically using the posterior approach; TB bacteria may invade the normal tissue in the posterior structure of the spine; the surgery may bring damage to the normal posterior spine, compromise the spinal stability, interfere the spinal cord and increase postoperative intraspinal scar adhesion. Beyond that, this approach only gives limited exposure of the anterior structure of the spine, and thus it is not an option when there is a large paraspinal abscess. It was stated in our previous study that the single posterior approach had shorter operation time and a smaller volume of intraoperative blood loss compared to the anterior approach [8], and these findings were proved by this study. Through analysis, it is believed that these advantages of the posterior approach are associated with the development of the pedicle screw insertion technique and the posterior approach itself through day-to-day applications, as well as the use of the approach in a large number of short-segment fixation (no more than three segments) cases. Also, it was found that the posterior approach had an angle loss rate smaller than the anterior approach in terms of kyphosis correction, which was reported in another relevant study [21]. The rationale for the use of the posterior approach lies in the removal of the lesion and the sclerotic bone surrounding the lesion to clear the path for anti-TB drugs. As to the residual TB-like lesion and purulent fluid, long-term, standard anti-TB chemotherapy is an effective postoperative solution. This study has some limitations, including the small sample size and the short follow-up. A prospective multicenter randomized comparative study having a larger sample size is needed to 11 investigate the long-term efficacy of the two approaches.

Availability of data and materials
The datasets supporting the conclusions of this article are included within the article. The raw data can be requested from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.   * One-way analysis of variance , compared pre-operative kyphosis angle between two groups, P >0.05 Δ One-way analysis of variance, compared kyphosis angle with pre-operative in two groups, P A <0.05, P B <0.05 † One-way analysis of variance, compared angle correction between two groups, P<0.05 ▲ One-way analysis of variance, compared correction rate between two groups, P<0.05 One-way analysis of variance, compared kyphosis angle between two groups, P<0.05 □ One-way analysis of variance, compared angle lost between two groups, P<0.05 One-way analysis of variance, compared angle lost rate between two groups, P <0.05