Surgical outcomes of posterior trans-facetal decompression and stabilisation in tuberculous spondylodiscitis with neuro-deficit

Abstract

Study design

Retrospective study.

Purpose

To evaluate the functional, neurological and radiological outcomes of posterior trans-facetal decompression and stabilisation in tuberculous spondylodiscitis patients with neuro-deficit.

Overview of literature

Spinal tuberculosis mainly involves anterior column and for that anterior approach has been the most frequently used surgical technique in the past as it allows direct access to the infected tissue providing a good decompression. However, anterior surgery is associated with higher morbidity which can be reduced by posterior trans-facetal approach.

Materials and methods

The study included 100 Tuberculous Spondylodiscitis patients with neuro-deficit who underwent posterior trans-facetal decompression and stabilisation from 2009 to 2014. Demographic data, clinical parameters (back pain score-VAS, ODI), neurological status (Frankel’s grade), radiological parameters (Kyphosis angle) and complications were evaluated.

Results

Out of the total 100 patients there were 58 males and 42 females. 84 patients had thoracic and 16 had thoracolumbar region involvement. The mean age of the patients was 34.7 years. The extent of fixation was 2 segments in 52 patients and >2 segments in 48 patients. Postoperatively significant improvement in VAS (pre-op 6.5 ± 0.65 to post-op 1.73 ± 0.64) and ODI (pre-op 76.54 ± 6.96 to post-op 30.5 ± 6.56) were noted. The mean kyphosis angle was corrected from 22.33° ± 5.59° to 5.14° ± 1.32°. 86 patients showed at least 1 grade of improvement in neurology (Frankel’s grading) and there was no deterioration in any patient. 3 patients developed superficial infection and 2 had an intra-operative dural tear. 94 patients showed bony fusion at 2 years follow-up.

Conclusion

Posterior trans-facetal decompression and stabilisation is an effective procedure in the management of thoracic & thoracolumbar tuberculous spondylodiscitis patients with neuro-deficit. It offers circumferential decompression with stabilisation and also maintains kyphosis correction.

1. Introduction

Spinal tuberculosis is the most debilitating form of extra-pulmonary TB. Sir Percivall Pott first described tuberculous spondylitis in 1779 since then it is known as Pott’s spine.¹ Spinal TB accounts for 2% of all the cases of TB, 15% of extra-pulmonary and 50% of the skeletal TB.^2,3 In spine, thoracic and thoracolumbar regions are the common sites of involvement.⁴ It is mainly a medical problem and antitubercular drug therapy remains the mainstay of treatment. Surgical treatment is reserved for patients with cold abscess, neurological involvement, spinal instability, kyphosis, and/or failure of conservative treatment. The incidence of neurological complications varies from 10 to 43%.¹ The surgical treatment consists of debridement, segmental instability reconstruction, neural decompression and kyphotic deformity correction. For anterior and middle column of the spine, anterior approach has been the most frequently used surgical technique as it allows direct access to the infected tissue providing a good decompression. However, anterior surgery is associated with high risk of structural damage (such as injury to lungs, liver, heart, kidneys, bowel, and ureter), graft failure & has limited application in mechanical instability and deformity correction giving more advantage to the posterior approach allowing circumferential decompression and three-column fixation.

Our study aims to evaluate the functional, neurological and radiological outcomes of posterior trans-facetal decompression and stabilisation in tuberculous spondylodiscitis patients with neuro-deficit.

2. Materials and methods

The study included 100 Tuberculous Spondylodiscitis (thoracic/thoracolumbar) patients from 2009 to 2014 after institutional ethics committee approval, who underwent posterior trans-facetal decompression and stabilisation. Patients with Neuro-deficit and having a minimum 4 years follow-up were included in our study. Diagnosis was based on MRI scans and histopathology reports. All the patients were started on antitubercular drug therapy (ATT) based on the drug sensitivity pattern.Demographic data (age, gender, duration of illness, comorbidities), Pre-operative and Post-operative clinical parameters (back pain score-VAS, ODI), Neurological status (Frankel’s grade), Radiological parameters (Kyphosis angle), Peri-operative parameters (operative time, blood loss, hospital stay, Dural leak), Post-operative complications (neurology worsening, implant failure) and fusion rate were evaluated.

2.1. Surgical procedure

Under general anaesthesia in prone position the posterior midline approach was used in all the cases. Spinous processes, lamina, facet joints and transverse processes were exposed (subperiosteal dissection). Pedicle screws were placed above and below the affected level by free hand technique. One side rod was fixed to pedicle screws temporarily to avoid spinal cord injury during debridement and decompression. The inferior facet of the superiorly affected vertebrae along with the adjacent part of lamina was resected exposing the affected disc space. Debridement was done with the help of rongeurs and curettes working in a lateral to medial direction. (Fig. 1). Local bone graft was used to fill the void. No nerve root was sacrificed in any case. In some cases where the destruction was more and anterior column reconstruction was needed in those cases along with the facet, pedicle was excised and mesh/expandable cages were used. Rods were placed on both sides and compressed to correct the kyphotic deformity if any. The Infected material was sent for culture and sensitivity, AFB staining (acid-fast bacilli), grams stain, fungal smear, polymerase chain reaction/Gene expert test and tissue for histopathologic examination in all the cases.

Case 1

T3-T4 Spondylodiscitis with cord compression (Fig. 2a and b). Posterior stabilisation with trans-facetal decompression done (Fig. 2c).

Case 2

T7-T8 spondylodiscitis (Fig. 3a and b) treated with trans-facetal decompression and fixation (Fig. 3c). Postoperative MRI imaging showing well decompressed cord (Fig. 3d and e).

Fig. 1.

Trans-facetal approach.

Fig. 2.

2a,2b: T3-T4 Spondylodiscitis with cord compression. 2c: Posterior stabilisation with trans-facetal decompression.

Fig. 3.

3a,3b: T7-T8 spondylodiscitis treated with trans-facetal decompression and fixation (3c).

3d, 3e: Postoperative MRI imaging showing well decompressed cord.

2.2. Post-operative protocol

Log roll, side turning, and pelvic lift exercises were started on postoperative day one and out of bed mobilisation was started as early as possible. The mean hospital stay was 8 days (5–11 days) and suture removal was done on postoperative day 14 in all except in 3 cases with superficial infection (postoperative day 21). Patients were regularly followed-up and neurological assessment was done. Radiographs were taken immediate post-operative period, at 6 weeks, and at 3, 6, 12, 24, 36 and 48 months to ensure maintenance of spinal column alignment [kyphosis angles by Konstam and Blesovsky method (Fig. 4)], bony fusion of affected levels (Bridwell criteria), loosening of implants, and loss of correction of the kyphosis angle. Blood tests (ESR, CRP) were done monthly for the assessment of progression/regression of disease.

Fig. 4.

Kyphosis angle measurement by Konstam and Blesovsky method.

3. Results

The study was started with 112 patients, 12 lost to follow-up and 100 patients were included in the study with 58 males and 42 females. 84 patients had thoracic and 16 had thoracolumbar region involvement. The mean age of the patients was 34.7 years (16–70 years) and the mean follow-up period was 5.8 years (4–9 years). The extent of fixation was 2 segments in 52 patients when the disease lesions involved a single vertebra and more than 2 segments in 48 patients when the disease lesions involved two or more vertebra. All the patients were started on antitubercular therapy according to the drug sensitivity pattern. AKT in all patients was started by infectious disease expert. Routinely it was given for 9–12 months. 8 out of 100 (0.8%) showed multidrug resistant TB (MDR-TB) infection for which the drug therapy was extended for 18–24 months. Al the patients underwent MRI after the full course of AKT, and AKT was stopped on the basis of clinical, radiological and blood parameters. The mean surgical time was 154.96 ± 17.15 min and the mean blood loss was 155 ± 33.32 ml. Post-operatively significant improvement in VAS (pre-op 6.5 ± 0.65 to post-op 1.73 ± 0.64) and ODI (pre-op 76.54 ± 6.96 to post-op 30.5 ± 6.56) were noted (Fig. 5, Fig. 6). Preoperatively, the mean kyphosis angle was 22.33° ± 5.59° and was corrected postoperatively to 5.14° ± 1.32°. The mean loss of correction at 1 year follow up period was 1.93⁰ and at 4 year follow up was 2.67⁰ (Fig. 7).According to the Frankel’s grading (Grade A: completely paralyzed; Grade B: preserved sensation only and no motor function; Grade C: preserved sensory function and non-functional motor function; Grade D: preserved sensory function and functional motor function; Grade E: normal motor and sensory function), there were 14 patients classified as grade A, 32 as grade B, 36 as grade C, 18 as grade D. 86 patients showed at least 1 grade of improvement in neurology and there was no deterioration in any patient.Frankel’s grading of neurology at final follow-up is shown in (Table 1). ESR and CRP values became normal within 6 months post-operatively.

Fig. 5.

Graph comparing visual analogue scores (pre-op and post-op).

Fig. 6.

Graph comparing Oswestry Disability Index scores (Pre-op and Post-op).

Fig. 7.

Graph comparing Kyphosis angles (Pre-op, Post-op, 1 year and 4 years follow-up).

Table 1.

Changes in Frankel’s grade before surgery and at final follow-up.

Frankel grade	Before surgery	At final follow-up
		A	B	C	D	E
A	14	1		4	9
B	32		6	10	11	5
C	36			5	20	11
D	18				4	14

3 patients developed superficial infection which resolved with injectable antibiotic therapy for 3 weeks. 2 patients had an intra-operative dural tear which was managed by facial graft intraoperatively.94 patients showed bony fusion & 6 patients showed non-union at 2 years follow-up out of which 3 patients showed implant loosening for which revision surgery was performed however, 3 patients of non-union showed no signs of instability and were managed conservatively.

4. Discussion

Thoracic spinal TB accounts for the largest proportion (30.3–55.8%) of spinal TB.⁵ Usually associated with kyphosis, neurological deficit, and even paraplegia.^6,7 Thoraco-lumbar lesions have the highest propensity to develop morphological changes and likely to predispose to deformity progression.⁸ TB is a medical disease where the antitubercular drug therapy is the mainstay of treatment. Although it effectively inactivates TB, vertebral collapse may continue. Kyphosis is a common complication of spinal TB requiring extensive reconstruction.⁹

Rajasekaran et al. suggest that surgery is indicated in patients without neurologic deficits if their disease is progressive or recurrent, with the failure of chemotherapy to decrease symptoms in 6–8 weeks and to prevent progression of kyphosis.¹⁰

Spinal TB is traditionally treated surgically by anterior instrumented fusion, combined anterior-posterior procedure, and extra-pleural anterolateral procedure or posterior procedure only.^11,12 Historically, an anterior approach to the spine was popularized by Hodgson et al., and since then many anterior-based surgeries were performed including anterior debridement and fusion with rib or iliac crest bone graft alone.¹³ This method can provide direct access to the disease focus, effective debridement of the lesion, and decompression of the spinal cord. However, these approaches were often associated with a high rate of morbidities and mortality.¹⁴

These were followed by procedures involving two-staged anterior and posterior arthrodesis with local bone achieving 360 fusion with a reduced rate of pseudo-arthrosis. Certain studies showed that the combined approach of anterior debridement with posterior instrumentation had a significant result in achieving both spinal decompression and kyphosis correction in tuberculous spondylodiscitis.^15,16 However, anteriorly based procedures had issues regarding the stability of implants placed anteriorly in the infected bone.¹⁷

Benli et al. observed that anterior instrumentation increases the rate of correction of kyphotic deformity (79.7 ± 20.1%) and was effective in maintaining it, with an average loss of 1.1 ± 1.7°.¹¹

Kim et al. who operated on 21 patients by anterior instrumentation, mentioned that although a 67.7% correction (11.3°) was achieved initially, a correction of 83% (9.4°) was lost at the last follow-up.¹⁸

Lee et al. Reported the correction of the kyphotic angle and loss of correction was statistically significant (p < 0.05) in both the anterior instrumentation and fusion groups and the single-stage trans-pedicular decompression and the posterior instrumentation groups.¹⁹

Recently, posterior-only procedures with or without anterior reconstruction are state of the art for management of spinal TB. The versatility of transpedicular approach in achieving adequate debridement/decompression of the abscess and granulation tissues (both anteriorly and laterally) by is well documented in the literature.²⁰ Posterior instrumentation also has the advantage of obtaining better correction of the kyphotic deformity as compared to anterior surgery and offers appropriate access to the spinal canal for efficient decompression of the neural elements, particularly in cases of epidural suppuration, and renders excellent clinical results.²¹ Bezer et al. Reported transpedicular drainage and posterior instrumentation as a less demanding single-stage procedure in patients with lumbosacral tuberculosis.²² Wang et al. Studied 115 patients with Mono-segmental disease and concluded that combined anterior with posterior may not be the first choice for treating patients with spinal TB and a single-stage instrumented posterior debridement with bone grafting produces excellent clinical results.¹⁵ In our study we have used trans-facetal approach for anterior decompression which was first described by bransford et al. for thoracic disc herniation.²³ The approach is different from the transpedicular approach in several aspects. First, in Trans-facetal the pathology is approached by resecting inferior facet of the superior affected vertebrae and disc/paradiscal area which is the primary involved site is reached through foraminal route same as in TLIF. Second, It is a bone preserving approach since pedicle is not removed and Even the contralateral lamina was not removed which provides additional stability and helps in fusion. In some cases in our series where the destruction was more and anterior column reconstruction was needed in those cases along with the facet, pedicle was excised. Third a screw can be inserted in the vertebrae which helps to provide an extra anchor for the construct and reduces the construct length.

We believe that fixation of the unstable vertebrae reduces the pain and inflammation and surgery per se decreases the infection load which in collaboration with the anti-tubercular drug therapy was responsible for the improvement. Pedicle screws crossing the vertebral body pedicle, the strongest part of the vertebral body, providing three-dimensional correction and stability, which was much stronger than anterior instrumentation. Our study showed 86 patients had at least 1 grade of improvement in neurology (Frankel’s grading). This improvement might be because our study included mainly young patients with acute onset of neurology worsening (wet lesions). However, there was no deterioration in neurology in any patient at a 4year follow-up and associated with significant pain relief. The results were consistent with the other studies.^24,25

There was a significant correction of kyphotic angle post-operatively in our study (pre-op 22.33° ± 5.59° to post-op 5.14° ± 1.32°) with less significant loss of correction at final follow-up which also does not lead to clinical worsening similar to previous studies. Our study included patients mainly with less kyphotic angle preoperatively. We believe that the intact lamina, which acts as a posterior tether preventing kyphosis and also acts as a good bed for fusion. This negates the need for a cage or bone graft (anterior column reconstruction), particularly in single-level spondylodiscitis. Jain et al. Found that there was a greater correction of kyphosis in lumbar and lumbosacral as compared to thoracolumbar studies.²⁶ This may be attributed to a more rigid thoracic vertebrae with rib attachments. Overall, the loss of kyphosis is lesser in posterior plus anterior surgery than anterior alone that has been documented in many studies.

The mean duration of achieving fusion in our study was 6.2 months which was similar to previous studies with a 94% fusion rate.²⁷ Again, the fusion rates in our study were similar to other studies but better as compared to Sahoo et al.²⁸

5. Limitations

Posterior debridement bears the potential risk of TB spreading to the posterior healthy regions, resulting in infection and fistulas. Fortunately, these complications have not been observed in our study. Secondly, the normal posterior column of spine was destroyed to achieve complete debridement and decompression in this procedure, which would affect the stability of the spine theoretically but we had removed a part of lamina only which ensures the stability of the spine in long term follow-up.

6. Conclusion

Posterior trans-facetal decompression and stabilisation is an effective procedure in the management of thoracic and thoracolumbar tuberculous spondylodiscitis patients with neuro-deficit. It provides adequate exposure for the drainage of para-spinal abscesses and debridement of diseased bone. It offers circumferential decompression with clearance of foci and also maintains kyphosis correction.conclusion.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Declaration of competing interest

The authors have no conflicts of interests to disclose.