If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Posterior only surgery has been widely performed in the treatment of thoracic and lumbar spinal tuberculosis. Surgical options include debridement with posterior instrumentation only or combined with anterior reconstruction. The aim of this study is to investigate and compare the clinical, functional and radiological outcomes using a single-stage posterior only surgery in thoracolumbar spinal tuberculosis by three different surgical techniques.
Patients undergoing posterior only surgery for thoracic and lumbar spinal tuberculosis were followed up prospectively and included. Three different procedures, Group-A: Posterior instrumentation with anterior cage reconstruction (n = 49), Group-B: Posterior instrumentation and anterior autologous bone-grafting (n = 21) and Group-C: Posterior column shortening without anterior-reconstruction (n = 52) were compared for kyphosis correction achieved, kyphosis at final follow-up and degree of correction lost. Neurological assessment was done using ASIA impairment Scale(AIS) grades. Functional assessment was done using Visual analogue score (VAS), Modified McNab criteria and NASS satisfaction score.
A total of 122 patients were included in the study, Group-A (49), Group-B (21) and Group-C (52). Radiological correction of kyphotic deformity in anterior reconstruction, Group-A (20.17 ± 9.25⁰) was higher than 13.97⁰ ± 6.06⁰ and 14.27⁰ ± 6.47⁰ achieved in Groups B and C respectively. There was no significant difference in correction lost amongst the three groups (p-value, 0.76). Surgical duration, blood loss and hospital stay were significantly higher in the anterior reconstruction group (p-value, 0.001). Similarly, no significant difference was noted between the three groups in neurological and functional outcomes at 2 years.
Posterior only approach is eminently satisfactory for treating Thoracolumbar Spinal Tuberculosis (STB). All three groups had similar functional and neurological outcomes. However there was a better correction of deformity in patients with anterior cage reconstruction.
Over the past few decades, there has been a major shift from the traditional radical anterior debridement and reconstruction surgeries to posterior approaches due to lesser morbidity by avoiding thoracotomy related complications, achieving better deformity correction with lesser blood loss, surgical complexity, duration of surgery and hospital stays.
Using a single posterior approach; stabilization of the spine, decompression of the neural elements and if necessary, reconstruction of the anterior column can be performed. In active stages of STB, while debridement followed by posterior column shortening has shown good outcomes, global reconstruction by posterior approach has been documented to achieve better deformity correction.
However, there is no literature on whether anterior reconstruction achieves superior functional, radiological, and neurological outcomes when compared to posterior column shortening alone. The purpose of this study is to investigate the outcomes of three different surgical procedures using posterior approach in the treatment of active thoracolumbar STB in adults.
This was a single center prospective observational study conducted at a tertiary level hospital from January 2016 to January 2019 with approval from the Institutional Ethical Committee and Institutional Review Board (IRB approval no: 2016/02/04). During this period, 143 patients were operated for thoracic and lumbar spine tuberculosis of which 122 patients had the complete data and a minimum follow up of 2 years and formed the study population. Inclusion criteria were: 1) age more than 18 years, 2) thoracic, thoracolumbar, or lumbar spine tuberculosis with neurodeficit, 3) patients with significant pain but without neurological deficit, 4) conservatively treated patients not responding to ATT following 6 weeks of chemotherapy. Patients of age less than 18 years, or those with cervicothoracic or lumbosacral junctional or posterior column involvement and those with active pulmonary TB at presentation were excluded from the study.
All the patients were operated by three senior surgeons (SR, APS, RMK) in the institute and all patients underwent single stage posterior only approach. All the patients were planned for pedicle screw instrumentation, debridement, and posterior column shortening. The author's decision to perform additional anterior reconstruction following debridement was based on two factors, one being significant vertebral body loss visualized in preoperative images and the other being intraoperative decision, where the author felt inadequate bone-on-bone contact following shortening of posterior column. Anterior reconstruction was performed using autologous grafts in first scenario and using titanium cages in case of latter. Depending on the procedure done they were divided into following three groups as depicted in Fig. 1, Group-A: Posterior instrumentation with anterior cage reconstruction, Group-B: Posterior instrumentation and anterior autologous bone grafting, Group-C: Posterior column shortening and pedicle screw instrumentation only.
The patient's demographic details, clinical presentation, radiological parameters like level of affection, number of vertebrae involved, vertebral body loss, preoperative kyphosis, postoperative kyphosis, correction achieved, kyphosis at final follow-up and degree of correction lost were all recorded from the PACS and analyzed. Kyphotic angles were calculated from lateral x-rays by the method described by Konstam and Blesovsky.
Loss of correction was calculated as difference between the deformity angle measured at 2nd year of follow up and that of immediate postoperative value. The initial vertebral body loss (VBL) was measured in the lateral radiograph using the method described by Rajasekaran and Shanmugasundaram
(Fig. 2). Segmental VBL ratio was calculated by dividing the total VBL by the number of involved segments. Neurological assessment was done using ASIA Impairment Scale(AIS) grading. Functional assessment was done using Visual analogue score (VAS), Modified McNab criteria and NASS satisfaction score.
2.1 Surgical procedure
All surgeries were performed under general anesthesia in prone position. Prone position with flexed knees aided reduction of the initial kyphotic deformity. Using standard posterior midline incision, paravertebral muscles were stripped by subperiosteal dissection. The pedicle screws were inserted at the desired vertebrae above and below the involved level and screw positioning was monitored using intraoperative fluoroscopy. Screws were also applied in affected levels whenever possible. After temporarily stabilization with a contoured connecting rod, all patients underwent laminectomy and debridement was done by either of the trans-pedicular/trans-facetal/costotransversectomy or transforaminal approaches based on the level involved and location of the diseased tissue. Post debridement, intraoperative decision was made either to reconstruct the anterior column with titanium cage, bone grafting or posterior column shortening and stabilization with contoured rod. Sequentially, rods were tightened, and the kyphosis was slowly and carefully corrected by compression over posterior column. In Group-A, the inserted anterior cage was used as a pivot for posterior column shortening. Pre and paravertebral abscess were treated by drainage and antituberculous chemotherapy. Negative pressure drainage was used, and the debrided tissue samples were sent for GeneXpert, bacterial culture, AFB culture and histopathological studies in all patients. Postoperatively, drainage tube was removed when the volume was less than 50 ml over 24 h. Patients without neurological deficit were allowed to mobilize the next day with TLSO brace. Bracing was continued for 3 months postoperatively. Patients with deficits were started on a rehabilitation program. Once the laboratory results were confirmatory of tuberculosis, anti-tuberculosis regimen was started in all patients according to the standard WHO guidelines.
2.2 Follow-up protocol
Patients were reviewed at 1 month, 2 months, 3, 6 and 9 months and 2 years. Following initial evaluation of liver function tests, ESR and CRP were monitored every visit. Radiographs were taken at each follow-up to assess position of implants, alignment of spinal column, angle of kyphosis, remineralization, sclerosis at the end of 2 contiguous involved vertebral bodies, resolution of abscess shadows, fusion, and overall healing (Fig. 3). MRI was done at 9th month to access signs of healing including resolution of abscess, fatty infiltration of the marrow. ATT was discontinued after the clinical, laboratory and radiological evidence of healing.
2.3 Statistical analysis
Data were analyzed using SPSS v21 software. Results are presented as Mean ± SD for continuous measurements and frequencies as number and percentages. Continuous variables were analyzed using unpaired t-test and categorical variables were analyzed using the Chi-square test. P value of 0.05 or less was considered statistical significant.
A total of 122 patients who completed a minimum of 2 year follow up were included in the study. There was no significant difference in sex distribution amongst the groups, Group-A: 49 patients (29 males, 20 females), Group-B: 21 patients (9 males, 12 females) and Group-C: 52 patients (22 males, 30 females). Back pain was the main presenting symptom in the all the patients in the study population, along with weakness of lower limbs and walking difficulty in 17.21% and 13.93% respectively. Only 1 patient presented with complaints of deformity over the back.
3.1 Demographic and surgical parameters
The mean age in Group-A, B and C was 46.92 ± 12.24(range; 22–76) than 45.85 ± 14.03(range; 19–75) and 46.36 ± 16.33 (range; 20–80) respectively. The mean duration of symptoms in months was similar in Group-A (3.23), Group-B (5.12) and Group-C (3.63) (p > 0.05). 14% of patients had signs of myelopathy in Group A, compared to 23.81% and 17.31% in groups B and C respectively. In group A, B and C, 44.9%, 52.38% and 61.54% of the patients had normal neurology at presentation respectively. Of the 122 cases, thoracic vertebral level involvement was seen in 77 patients (Group A:31, B:11 and C:35). Thoracolumbar and lumbar in 39 patients (Group A:15, B:7, C:17) and 6 patients (Group A:3, B:3) respectively. 9 patients (7.37%) had noncontiguous involvement of the vertebrae, of which only 1 patient required surgery at both the levels.
The average number of levels of instrumentation were similar among the three groups - 5.86(range; 2–9), 5.86(range; 4–8) and 6.19(range; 4–10). However, duration of surgery was significantly higher (p < 0.05) in Group-A (202 ± 19.61 min) compared to Group-B (165.71 ± 18.05 min) and C (156.54 ± 21.32 min). Mean intraoperative blood loss was also higher in Group A: 760.82 ± 135.03 ml (range; 550–1200 ml) than group B: 607.14 ± 60.76 ml (range; 500–750 ml) and C: 606.15 ± 62.12 ml (range; 450–750 ml) (p-value, 0.001). The duration of hospital stay was again higher in Group-A (p < 0.05) when compared to B and C (Table 1).
Table 1Preoperative Clinical data and Surgical variables of the three groups treated by posterior surgery.
Preoperative Vertebral Body Loss (VBL) was 0.73 ± 0.33 in Group-A, followed by 0.61 ± 0.26 and 0.55 ± 0.33 in Group-B and Group-C respectively. Preoperative kyphotic deformity was higher in Group-A (29.42⁰ ± 12.18⁰) (p-value, 0.046) compared to Group-B (24.40⁰ ±7.10⁰) and Group-C (24.90⁰ ± 8.85⁰). Significant deformity correction was achieved following surgery in all three groups with better correction in Group-A (20.17⁰ ± 9.25⁰) compared to 14⁰ ± 6.06⁰ and 14.27⁰ ± 6.47⁰ achieved in Groups B and C respectively. There was no significant difference between the loss of correction (3.4⁰ vs 3.2⁰ vs 3.1⁰) amongst the three groups (p-value, 0.755) at 2 years of follow up and the final deformity angles were similar (p-value, 0.076), 12.62⁰ ± 8.22⁰, 13.66⁰ ± 5.76⁰, and 13.51⁰ ± 5.53⁰ in Groups- A, B and C respectively (Table 2). Radiological fusion at 3 months was observed in 6.1% of Group-A comparable to 9.5% in Group-B and 3.8% in Group-C. Radiological fusion at 9 months was similar and achieved in 96% of Group-A, 95.2% of Group-B and 94.2% of Group-C respectively. All our patients showed good remineralization of the involved end plates, sclerosis with resolution of abscess at final follow up of 2 years.
Table 2Preoperative and Postoperative radiological parameters amongst the three groups.
Preoperative VAS scores were similar across all the groups (p-value, 0.47) with significant improvement achieved in all following surgery. VAS scores were comparable between the three groups at 2-year follow up (p-value,0.134). There was no significant difference in the NASS satisfaction score (p-value,0.522) and Modified McNab outcome criteria (p-value,0.377) in the study population (Table 3). 3 patients had immediate post-operative neurological worsening (one in each group), but all of them improved subsequently.
Table 3Functional and neurological outcomes amongst the three groups.
A total of 3 wound related complications included 2 wound dehiscence (treated by 1: debridement and secondary suturing, 2: Vacuum assisted closure) and 1 superficial infection (treated by intravenous antibiotics). 1 patient had developed hepatitis due to ATT drugs and was then treated with modified ATT regimen. Screw loosening, seen as translucent areas around the implant was seen in postoperative follow-up radiographs in 2 patients. However, they were asymptomatic and did not require any treatment. 1 patient died at 2 and half years following surgery due to a cardiac event.
Tuberculosis continues to be one of the common infectious diseases affecting 10 million people annually and claiming at least 4000 lives per day. Spinal Tuberculosis (STB) is the most common form of skeletal tuberculosis, and it mainly affects anterior structures causing vertebral body loss, pathological collapse, and kyphotic deformity. Traditionally radical debridement and fusion by anterior approaches as described by Hodgson was the treatment of choice. With advancements in MRI enabling earlier diagnosis, and multidrug chemotherapy showing excellent healing response, STB is mainly a medical disease nowadays. However, surgical intervention in active disease is still required in certain scenarios such as incapacitating pain due to instability, neurological worsening, or progressive deformity. However radical debridement by anterior approach has been replaced by posterior approaches using pedicle screws due to familiarity amongst surgeons and more importantly lesser morbidity by avoiding complications related to anterior approach.
The goals of surgical intervention include debridement of the affected area, decompression of neural structures, correction of deformity, and stabilization of spine, all of these can be achieved by a single-stage posterior surgery.
However, it is debatable as to whether in addition to this anterior reconstruction is required in active cases of STB. The current study aimed to investigate and compare the clinical, functional, and radiological outcomes of anterior reconstruction using titanium cages filled with autografts (Group-A) and autografts (Group-B) compared to posterior debridement and stabilization alone (Group-C).
The results of the present study demonstrate that irrespective of whether anterior reconstruction was used or not, all the patients treated by posterior instrumented debridement and decompression combined with chemotherapy achieved the status of complete healing without any evidence of recurrence until their last follow-up. All patients showed neurological improvement and only four complications were noted in this study, 3 related to surgical wound and one related to chemotherapy-related hepatitis which were managed appropriately. There was no recurrence or need for revision surgery in any of these cases despite screw loosening evidenced by translucency surrounding screws in two patients.
4.1 Surgical morbidity
Posterior only approach has been well documented to have lesser operative time, blood loss, hospitalization period, and complication rate in addition to adequate deformity correction when compared to anterior or combined approaches.
In the current study, anterior reconstruction in addition to posterior decompression and stabilization had increased the operative time, blood loss, and hospital stay (p-value; 0.001), but complication rates were similar and minimal.
4.2 Deformity correction
Sahoo et al., in a series of 18 patients treated by posterior decompression and stabilization alone observed a deformity correction from 17.7⁰± 5.8⁰ to 9.4⁰± 4.6⁰.
In our study, we could achieve a similar deformity correction from 24.9⁰± 9⁰ to 10.4⁰± 4.7⁰ in Group-C. While additional anterior reconstruction using autografts (Group-B) as in Fig. 4, did not differ in fusion rates or deformity correction, better deformity correction was observed in Group-A (20⁰± 9.3⁰) compared to 14⁰± 6.1⁰ and 14.3⁰± 6.5⁰ achieved in Groups B and C respectively. Also, this correction achieved in anterior column reconstruction using titanium cages (Fig. 5) in Group-A is better than the 8⁰ and 12.5⁰ of correction achieved in other studies.
In a study comparing anterior column reconstruction with titanium mesh cages filled with graft harvested from surgical site compared to reconstruction using iliac crest autografts, no significant difference was observed, unlike our study.
This could probably be due to surgical technique, where we used the anterior cage as a pivot to achieve better deformity correction resulting in a radiographic correction of 20⁰± 9.3⁰ compared to only 4.4⁰± 8.3⁰ observed by Gao et al.
achieved a deformity correction of 21.8⁰ ± 5.4⁰ using a similar technique used by Gao et al.
Adequate deformity correction was achieved in all the posterior approach surgeries in this study, and this is in line with meta-analysis comparing anterior and posterior approaches where Cobb angle correction, loss of deformity correction at final follow-up in posterior approach were similar to that of anterior approach.
In a study of STB with extensive abscess formation, reconstruction using anterior approach achieved better deformity correction when compared to posterior approach but did not differ in correction lost during follow-up.
Similar to this observation, anterior reconstruction using posterior approach in this study achieved better deformity correction but did not differ in correction lost during follow-up. It was observed by the author during the surgery that, when deformity approached 30°, posterior column shortening alone resulted in poor bone to bone contact and also kinking of cord in some, and therefore anterior reconstruction was performed and when anterior cages were used as a pivot, addition deformity correction was achieved.
4.3 Time to fusion and fusion rates
Gorensek et al. observed solid bony fusion in 88% of patients in cases of pyogenic osteomyelitis at 6 months follow up following single-stage posterior stabilization and anterior column reconstruction.
In the current study in Group-A, where patients underwent anterior column reconstruction using titanium cages, 57.1% achieved radiographic fusion at 6 months and 96% at 9 months. This is similar to many studies where the time taken to fusion has been reported between 8 and 10 months in STB.
Further fusion rates were observed to be similar in all the three groups in this study at the 9th month of follow up, where 47 patients in Group-A (96%), 20 patients in Group-B (95.2%), and 49 patients in Group-C (94.2%) achieved radiographic fusion. The remaining patients also achieved fusion in subsequent follow-ups.
4.4 Functional and neurological outcomes
In uncomplicated STB, no significant difference has been observed between conservative and surgical treatment.
Further to date, there is no study comparing neurological outcomes amongst different surgical techniques in posterior only approach. In the current study, in addition to posterior decompression alone (Fig. 6), anterior reconstruction did not provide any added benefit in terms of neurological recovery, but a significant difference was observed when comparing pre-operative and postoperative neurological status in all three groups. Further, no difference was observed in final VAS scores or patient satisfaction scores measured by North American Spine Society and McNab's patient satisfaction indices amongst three posterior surgical techniques.
Anterior reconstruction using autografts in posterior approach did not bring any advantage but when titanium cages were used as a pivot, better deformity correction was achieved when compared to stand-alone decompression and posterior column shortening. No difference was observed in complication rates, neurological recovery, and functional outcomes amongst the three posterior surgical techniques but reconstruction using metallic cages increased the operating time, blood loss, and hospital stay. Posterior decompression, debridement, and stabilization alone would be sufficient to treat thoracolumbar STB, but additional anterior reconstruction using metallic cages as pivot could be used as a potential tool to achieve better deformity correction.
The study was supported by GOREF (Ganga Orthopaedic Research and Education Foundation) .
Authors' contributions declarations
All authors contributed equally towards the preparation of this manuscript
The study was conducted after the approval of the Institutional ethics committee.
Consent to participate
Appropriate informed consent was obtained before the procedure.
Consent for publication
The participants have given consent to publish data without revealing identity.
Declaration of competing interest
All authors declare that there are no conflicts of interests.
All authors contributed equally towards the preparation of this manuscript.