Study design; retrospective case- control study
Setting; This study was conducted in the Spine department of Xiangya hospital of central South University. Data were retrieved from the medical records. The study protocol was approved by the Ethics Committee of the Xiangya first Hospital of Central South University.
The study included 34 patients with lumbar TB who had undergone either mono-segment fixation or short-segment fixation in spine department of Xiangya first hospital of Central South University from January 2012 to December 2016.
Inclusion criteria (1) patients with the infection involving only a single motor unit. (2) patients with vertebral destruction not more than 2/3rd of the original vertebral body height as determined by three-dimensional CT. (3) patients with lumbar kyphosis between 10-60-degree Cobb. (4) patients with age above 18 years. Exclusion criteria; (1) patients with active pulmonary or extrapulmonary TB and cancer (2) Patients with osteoporosis. (3) patients with multisegmented infections. (4) patients with other spinal abnormalities.
Bias in this study were assessed in two levels; at study level and outcome level. Study level bias assessment involved assessing for selection biases and detection biases. On the outcome level, information bias and attrition bias were assessed.
At study level, selection biases were mitigated by two ways. Firstly, Cases were selected from Hospital's database for patients diagnosed with TB to ensure reliability. This is anticipated to minimize detection biases. Finally, all participants were selected from the same population of participants attending Xiangya Hospital, so that exposures will be more or less, similar.
At outcome level, two authors were involved in recording data into the spreadsheet. This is anticipated to reduce typing errors. To minimize attrition biases, only participants regularly visiting the Xiangya Hospital were eligible for inclusion. To minimize reporting biases, STROBE tool (Strengthening Reporting of Observational Epidemiological Studies) customized for cohort studies, were used in the write-up of this research work.
(1) Neurological deficit due to spinal cord or cauda equina compression. (2) Lumbar instability and kyphosis. (3) patients with either abscess, sequestrum, or long discharging sinus tract. (4) patients with severe pain caused by the diseased lesions.
In all patient’s complete debridement, bone grafting with titanium mesh cage, decompression, and deformity correction were done within the diseased segments through posterior only approach. Group A (cases) had a mono-segment fixation (n =18), in which posterior fixation was carried out within the diseased motor units. Group B （control) had the short-segment fixation (n = 16), in which posterior fixation included the diseased vertebrae together with one normal motor unit superior and inferior to the diseased vertebrae. Patients’ characteristics and preoperative variables are shown in Table 1.
On admission, patients were placed on complete bed rest. Preoperative anti-tuberculosis drugs HREZ, (isoniazid (5 mg/kg), rifampicin (10 mg/kg), ethambutol (15 mg/kg), and pyrazinamide (25 mg/kg) were administered 14 to 28 days before the operation. Coexisting diseases were routinely controlled. Blood pressure less than 140/90 mmHg, random blood sugar less than 11.0 mmol/L, good appetite, and weight gain were preferred. Antiosteoporosis drugs were administrated in older patients before surgery. Liver and Renal functions were monitored, and their damage was treated promptly. surgery was carried out after the patients’ condition had improved.
Surgical technique for group A
After administration of general anesthesia midline posterior approach was applied with patient placed in prone position. subperiosteal dissection was made along midline incision to display the posterior spinal elements. The laminae, facet joints, and transverse process were displayed. The exploration exposed one vertebra more above and below the diseased segment. Pedicle screws were placed with the help of C-arm guided fluoroscopy into the diseased vertebrae only. Routine pedicle screws were placed in the diseased vertebrae if the remained heights of the diseased vertebrae after debridement were more than two third. Intra-operative C-arm was used to determine vertebral height after debridement. laminectomy was done at both side on the diseased part and the cord was decompressed under clear visualization. A pre-bent temporary rod was applied on contralateral side of the focus and stabilized to avoid spinal cord injury caused by spinal instability during decompression and focal debridement. Laminectomy, facetectomy, transversectomy, excision of spinous processes and other posterior elements on the working side (opposite to the side with the pre-bent rod) were done to display the lateral wall of the pedicle. The nerve roots of the diseased lumbar level were moderately retracted to provide space for the surgical procedure of anterior decompression or debridement, as described by Rajasekaran et al (13). All diseased and necrotic disc and bony tissue were excised until the fresh bleeding bone was seen. Both paravertebral and epidural abscesses were evacuated, and debridement was accurately performed until all pus and infectious debris were finished. Debridement near the spinal cord was done with a great care, curettes and rongeurs were used very slowly. The same process was repeated on the other side of the lesion if necessarily needed. The rods were firmly secured, and kyphosis was gradually corrected with the aid of compression and stretching of internal instrumentation. A precontoured TMCs filled with autogenous bone (harvested from the spinous process and lamina) or allograft spongy bone were placed in the bone trough to reconstruct the spine. A cross linker was added to enhance stability in each patient as shown in figure 1 below.
Surgical technique for group B.
The surgical approach and all steps for group B were the same as those of group A. They only differed by the placement of the pedicle screws. Pedicle screws were placed on the diseased vertebra and on normal motor unit above and below the diseased vertebra, one on each side. A crosslinker was also used for all patients to enhance stability. Posterolateral bone grafting TMCs were the same as those of the mono-segmental fixation group. As shown in figure 2 below
Streptomycin 1.0 g and isoniazid 0.2 g were given locally on the diseased site before wound closure. A drain was inserted and well secured, incisions were closed. Excised specimens were sent for bacterial culture and other pathological diagnosis.
The drainage tube was removed after 3 days. Except when there was an extensive paraspinal abscess it was left fixed until the volume of drainage was less than 30 ml/24 h. Patients continued with oral anti-TB chemotherapy postoperatively. Pyrazinamide was withdrawn at 2 months. Patients kept on taking HRE chemotherapy for 9- to 12-months. Patients were encouraged to ambulate with a brace 6–8 weeks post-operative. Patients were limited to non-weight-bearing activities until there was radiographic evidence of bone fusion. For the first 6 months after surgery, patients attended our outpatient clinic each month. While, from 6 months to 3 years, the patients attended our outpatient clinic once a year. During visit, medical check-up and physical examination were taken. Ten-point Visual analog scale score VAS, Oswestry disability index score ODI and Frankel grade were recorded. We measured erythrocyte sedimentation rate ESR, C-reactive protein CRP, liver function test and renal function test. A decision to stop anti-TB drugs was made based on normal examination results and healing of the lesion (which was determined by X-rays, CT, magnet resonance imaging (MRI), and ultrasonography).
Cobb’s angle was used to judge the correction of the lumbar spine in the sagittal plane. Cobb’s angle was measured by drawing lines along the superior and inferior endplate of the diseased segment in the lateral radiograph. The radiolucent halo zone around the implant of more than 1 mm on lateral X-ray film or CT, was regarded as implant failure. Bone fusion was defined as no radiolucent lines covering more than half of either implant, a slight movement of 3mm or less, 50 or less on flexion-extension lateral radiographs, and continuous trabecular bone growth between the vertebral bodies. TB healing was identified by: no pain or tenderness on the diseased site, no cold abscess or discharging sinus, continuous normal range of both ESR and CRP, and complete bone fusion as seen from radiographic imaging.
Statistical analysis was done using IBM SPSS statistic 21 version software. Age, follow-up time, VAS, ODI, the volume of blood loss, operation time, hospital stay, CRP, ESR, Cobb’s angle and bone fusion time between the two groups were compared using the Student’s t-test. Gender, neurological function by Frankel and diseases location site were compared by chi-square (χ2) test. A P value of less than 0.05 was considered to be statistically significant.