Surgical treatment of mono-segmental spinal tuberculosis using a combination of debridement and extreme lateral interbody fusion (XLIF): A retrospective study

Background This study investigated the clinical ecacy and safety of a combination of debridement and extreme lateral internal fusion (XLIF) for mono-segmental spinal tuberculosis (TB). The medical records of 9 patients (aged 21–70 years; 6 males) with mono-segmental spinal TB treated at our hospital between January 2014 and November 2016 were retrospectively reviewed. The involved vertebral bodies included 2 cases each at the T6/7, L2/3 or L3/4 level and 1 case each at the T7/8, T8/9, or T9/10 level. All patients were successfully treated with a combination of debridement and XLIF. The mean length of the operation was 97.3 ± 20.6 min (range, 65–126 min), and the mean blood loss during surgery was 151.1 ± 25.7 ml (range, 105–185 ml). Two patients experienced sensory disturbance over the left thigh and iliopsoas muscle weakness after surgery. After a mean follow-up of 12.3 ± 3.7 months (range, 6–17 months), the mean Cobb angle decreased from 28.7 ± 6.7 degrees to 12.1 ± 3.7 degrees. The mean VAS pain score decreased from 7.2 ± 1.1 preoperatively to 2.3 ± 0.9 postoperatively. The mean bone graft fusion time was 5.6 months (range, 4.7–7.2 months). At the nal follow-up, spinal cord injuries were evaluated as ASIA Grade D in 5 patients and ASIA Grade E in 4 patients.


Page 3/11
has rarely been used for spinal TB or spondylodiscitis. In this study, we retrospectively reviewed the medical records of patients with mono-segmental spinal TB treated with a combination of debridement and XLIF in a one-stage operation.

Study subjects
The study protocol was approved by the Institutional Review Board of our Hospital. Patients with mono-segmental spinal TB treated at our hospital between January 2014 and November 2016 were eligible for this study. Inclusion criteria were: 1) imaging ndings (X-ray, CT, and MRI) showing typical signs of spinal TB; 2) TB involving ≤ 2 spinal vertebrae; 3) lesions involving the anterior and middle column of the spine; 4) presence of abscesses, bone destruction, and spinal instability; and 5) TB con rmed on postoperative pathology. Exclusion criteria were: 1) contraindications to surgery; 2) TB involving > 2 vertebrae; or 3) severe osteoporosis.
Preoperative procedure All patients received standard HREZ chemotherapy (300 mg/day isoniazid, 450 mg/day rifampicin, 750 mg/day ethambutol, 750 mg/day pyrazinamide) and nutritional support for 2-3 weeks before the procedure. Surgery was performed when patients had recovered from anemia, hypoproteinemia, and neurological symptoms and their erythrocyte sedimentation rate (ESR) and body temperature were normal.

Operative procedure
After administration of anesthesia with standard single lumen endotracheal intubation, patients were placed in a 90 degree lateral position and the lesion segment of the vertebral body and the skin incision were marked using uoroscopy. An oblique incision (about 4-6 cm) was made in the median axillary line above the rib to avoid injuring the neurovascular bundle. Under realtime C-arm uoroscopy, the lung was pushed away using nger dissection, and a dilator was placed directly against the vertebral column at the surgical level (Figure 1). At the lumbar level, the oblique externus abdominis and obliquus internus abdominis were bluntly separated to provide access to the retroperitoneal space, the peritoneum was pushed forward in front of the psoas major muscle with nger dissection, and sequential dilation was performed with progressively larger dilators ( Figure 2). A guide wire was inserted into the middle or anterior part of the diseased intervertebral space to avoid injuring the spinal cord or nerve root. Carm uoroscopy was used to con rm the position of the dilator. One-step lesion debridement was used to drain the prevertebral abscess and remove diseased bone. The length of the bone grafting groove was measured. The same length of iliac bone was implanted into the bone groove and xed with an anterior plate and screw. Streptomycin (1g) and isoniazid (0.5g) were administered around the bone graft. A drainage tube was placed and the incision was sutured. The debrided material underwent histopathologic examination.

Postoperative care
Antibiotics were administered for 72 h after surgery. The drainage tube was removed when drainage volume was <50 ml/day. Standard anti-TB drugs were administered for 12-18 months, and liver and kidney function and ESR were monitored at regular intervals. After one week of bed rest, patients were permitted to perform ambulatory activities wearing a brace, which was worn for at least 6 months after surgery.

Curative e cacy evaluation
Plain radiographs and CT of the spine were obtained one week after surgery and every three months thereafter. Patients were evaluated based on American Spinal Injury Association (ASIA) grading, the visual analogue scale (VAS) pain score and Cobb angle, which were recorded at one month after surgery and the nal follow up. Suk's criteria were used to assess bone graft fusion [10]. 41.1 years (range 21-70 years), and mean body mass index (BMI) was 22.1 kg/m 2 . All patients experienced the following symptoms: fever, night sweats, back pain, and weight loss. Spinal cord injuries were evaluated as ASIA Grade B in 2 patients, ASIA Grade C in 3 patients, ASIA Grade D in 3 patients, and ASIA Grade E in 1 patient. The involved vertebral bodies included 2 cases each at the T6/7, L2/3 or L3/4 level and 1 case each at the T7/8, T8/9, or T9/10 level (Table 1).
All surgeries were successful. Mean operative time was 97.3 ± 20.6 min (range, 65-126 min), and mean blood loss during surgery was 151.1 ± 25.7 ml (range, 105-185 ml). There were no complications, such as incision infection, recurrence of TB, pleural effusions, rib pain, or delayed pneumothorax, and there were no fatalities. Two patients experienced sensory disturbance over the left thigh and iliopsoas muscle weakness after surgery. They were provided symptomatic treatment and recovered within 3 months.
Mean bone graft fusion time was 5.6 months (range, 4.7-7.2 months) (Figs. 3 and 4). At nal follow-up, spinal cord injuries were evaluated as ASIA Grade D in 5 patients and ASIA Grade E in 4 patients (Table 3).

Discussion
Our data demonstrates that spinal TB can be effectively treated using debridement and XLIF during a single operation. The technique made it possible to retain the spine's stability, while minimizing the length of the operation and blood loss, with few complications. As detailed above, patients had improvements in both their VAS scores and postoperative neurologic recovery.
Compared with laparoscopic or thoracoscopy-assisted approaches for the treatment of mono-segmental spinal TB, the XLIF approach has several advantages. First, XLIF eliminates the need for a general surgical assistant and the steep learning curve associated with laparoscopic or thoracoscopic approaches [8,11]. Second, the mean length of the operation and blood loss in our patients were signi cantly minimized compare with previous series of video-assisted thoracoscopic surgery for spinal TB described by Lv et al. and Kapoor et al. [11,12]. Third, a long period of lung de ation can be avoided, which makes XLIF ideal for use in elderly patients or those with respiratory compromise [13,14]. Finally, the technique allows direct visualization, which is particularly helpful for removing the infected tissue [15].
Perhaps most importantly, the stability of the spine was retained. Although there was a decrease in the Cobb angle, the average correction rate of the kyphotic angle was 58.0%. Biomechanical studies [16,17] indicated that transpedicular xation conferred better stabilization than lateral xation, but both methods can provide adequate stability. The major disadvantage of transpedicular xation combined with XLIF is that it prolongs the operation. Timothy et al. [15] reported that posterior instrumentation was not necessary, especially in the setting of infection, because of the high bone fusion rate associated with using XLIF. In addition, mono-segmental xation can retain the spinal motion units and slow the degeneration of adjacent vertebral bodies. The average bone fusion time was 5.6 ± 0.8 months, which was similar to other studies [5-7, 13, 14]. Moreover, the quality of life improved signi cantly in accordance with the better VAS score and neurological improvement.
Notably, surgeons should be aware of the possible neurological complications associated with XLIF. In the present study, two patients experienced sensory disturbance over the left thigh and iliopsoas muscle weakness after surgery. The patients recovered following three months of physical therapy and oral methylcobalamin. A nationwide survey conducted between March 2013 and April 2015 in Japan reported that the most common complications associated with XLIF were sensory nerve injury (5.1%) and psoas weakness (4.3%) [18]. Tohmeh et al. [19] proposed the use of dynamically-evoked electromyography (EMG) to detect and prevent neural injury during XLIF. We suggest that surgeons should have accurate knowledge of the spinal nerve rather than relying on EMG to avoid potential neurological complications [20,21]. segmental spinal TB without a large abscess or severe bone defects. The operating area and the guide needle puncture should be in front of the psoas muscle and vertebra, respectively. Standardized anti-TB treatment and regular follow up are very important. Patients should be made aware of potential complications, including delayed pneumothorax or hemothorax, vascular injury, and non-pulmonary complications, so that they can be carefully monitored postoperatively [13,14].
This study has several potential limitations. First, the results are from a relatively small sample size and there was no control group. Second, the follow-up was relatively short. Third, the indications for the procedure were based on our experience.
Additional multicenter prospective studies with a longer follow-up are necessary to better characterize the e cacy and safety of XLIF for the treatment of spinal TB. The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was approved by the institutional review board and the committee of our hospital.

Consent for publication
Informed consent to use their data was obtained from all patients before surgery.

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

Competing interests
The authors have no competing interests to disclose.   Figure 1 Illustration of the patient's posture and the XLIF approach in the thoracic spine Illustration of the patient's posture and the XLIF approach in the lumbar spine