Indications for surgery in spinal tuberculosis (TB) have been widely debated over the years [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15]. The indications for urgent surgical intervention in spinal TB vary from neurologic deficit due to spinal cord compression (clinical or radiological) [16] to spinal instability. In developing countries, there is a gross mismatch between the disease burden and the surgical facilities that are available. With the literature supporting good neurological outcomes following treatment (medical or surgical) in patients of spinal TB, some of these patients can be considered for non-operative treatment, especially the ones who do well with supervised anti-Koch's therapy (AKT). There is very little in the literature that discusses systematic non-surgical treatment in the situation of cord compression. We present a series of 50 patients with tuberculous cord compression who were treated with a methodical non-surgical treatment.

We reviewed 50 consecutive adult patients (who fit the inclusion criteria) treated in our clinic for spinal cord compression due to tuberculous spondylitis, between May 1993 and July 2002. This was a prospective study, with all enrolled patients being given a fixed non-surgical management protocol. There was no 'control' group, as all patients who suited the inclusion criteria were offered the same protocol of management.

The results of our pre-set, methodically applied, conservative treatment under very close observation were as follows. At the time of presentation, 10 patients had a motor deficit of Frankel grade D, 18 patients had clinically detectable hyper-reflexia, while the other 22 patients had normal neurology (Frankel grade E). Seven patients had been advised to have surgery elsewhere and they were admitted for the same. Four patients required more than 4-drug (first line) AKT based on the clinico-radiological criteria and response to first line AKT. Of these, one of the patients was a culture/biopsy proven drug-resistant TB patient.

On MRI scan, all the patients presented with an epidural abscess compressing the spinal cord. Forty-seven out of 50 patients responded completely to non-operative treatment and healed with no residual neurological (Frankel grade E) deficit. Three patients worsened neurologically to Frankel grade C while on AKT, and hence, they were operated upon with excellent eventual recovery, progressing to normal neurology (Frankel grade E) in the immediate postoperative period, as well as at the final follow-up.

With the description of the anterior approach in the 1970s, and subsequently, with the technical revolution in spinal reconstruction and instrumentation, spinal TB was being treated increasingly with surgery till the mid 1990s [17]. The surgical options ranged from simple drainage of pus to radical resection of the tuberculous infected focus/affected vertebra, followed by reconstruction using bone grafts, cages and stabilization with implants.

However, the treatment of spinal TB seems to have come full circle today, with a better understanding of the microbiology of the disease, as well as the better availability of modern culture and sensitivity methods. Spinal TB today is considered a medical disease in the absence of neurological deficit and gross deformity [18,19,20].

However, there are no clear cut guidelines in the literature regarding how much neurology is significant to warrant surgery. It is obviously difficult to set a cut off point for neurologic deficits, for the movement from nonsurgical to surgical treatment. In the literature, guidelines vary from offering surgery to patients with any grade of neurologic deficit to performing surgery only for patients with advanced motor deficits. Patients with significant neurologic deficits, i.e., TB paraplegia with epidural pus and granulation tissue compressing the spinal cord (on an MRI scan), have straight forward indications for surgery [21], as the prognosis of recovery is better when compared to other pathological compressive lesions. But in the TB spine, most of the times the disease is evolving, i.e., the development of a neurologic deficit is a gradual process with a set pattern. Hence, there remain a large subgroup of spinal TB patients with radiologic epidural cord compression but with only early neurological signs, i.e., with clumsy gait, hyper-reflexia, clonus and early motor deficits (Frankel grade D and E). It is in this group of patients that a trial of conservative treatment can be considered. In our current study, 47 out of 50 such patients responded well to non-surgical treatment and did not end up with progressive neurological worsening, clearly showing that radiological evidence of cord compression and early neurologic signs need not be an absolute indication for surgery in the management of spinal TB.

There always remains a chance of failure, as well as the fear of a worsening neurologic deficit while on conservative treatment, for which close monitoring of the patient's neurological status is extremely important. For this we did (and do advocate) initial monitoring of these patients as inpatients, with twice daily complete neurological examinations to check for the slightest possible deterioration in motor power. The same regimen is continued for two weeks once the patient's condition stabilizes. Patient education and counseling about the red flags and the possibility of emergent surgery while on treatment remain of paramount importance.

Any worsening in motor power or persistence of the preexisting neurologic status despite the patient being on AKT and the above mentioned supportive treatment calls for a decision to be made regarding surgical treatment. A worsening neurologic deficit or the development of a major neurologic deficit (≤Frankel grade C) or progressive paraplegia while on treatment under supervision becomes an absolute indication for surgical treatment. Epidural soft tissue swelling consisting of pus, granulation tissue and edema may subside and resolve completely with antitubercular medicines alone as shown in Figs. 1,2,3. Forty-seven out of fifty (94%) patients in our series showed complete resorption and healing of the compressing epidural tissue with chemotherapy alone. In three patients (6%), further worsening of the neurologic deficit (≤than Frankel grade C) despite chemotherapy made us consider surgical decompression (Fig. 4). In these operated patients, the neurologic status improved completely to normal in the immediate postoperative period, which persisted even after stopping the AKT, at the final follow-up.

In the long term follow-up, conservatively treated TB spine patients developing a kyphotic deformity and its sequelae, i.e., instability (due to concertina collapse in central body lesions) remain a possible entity [22,23], but as substantially evident now, this possibility becomes relevant only in the growing spine [24,25], and the advantages of conservative treatment abundantly outweigh surgical treatment as shown in the literature [26]. In this series, we had six patients with vertebral collapse due to a central body lesion, of which four patients presented with instability complaints at the start of treatment. These complaints resolved completely with time, after treatment with proper AKT.

The issue of drug resistance is very important while treating spinal TB patients [27], as the incidence of drug-resistant TB is on rise [28]. Hence, tissue diagnosis and rapid in vitro culture and sensitivity testing methods such as the automated BACTEC Mycobacterial Growth Indicator Tube 960 TB system gain importance [29]. CT guided biopsies increase the accuracy of needle placement, and hence, also the results [30]. In the event of equivocal results of these tests, one can treat spinal TB patients based on the clinico-radiological grounds.

Of importance is patient education and active participation. The patient's views which may include a preference for non-surgical treatment should be given due consideration in the presence of spinal cord compression, which otherwise is thought as a surgical problem. The intricate details of the non-operative treatment should be explained to the patient as well as the pros and cons of both types of treatment.

Radiological evidence of cord compression and early neurological signs need not be an emergency surgical indication in the management of spinal TB. Methodically applied non-surgical treatment can certainly give excellent results. However, patient education and participation remain crucial and should be given due importance when implementing non-operative treatment, especially in the presence of spinal cord compression.