Tuberculosis of the spine in children - does drug resistance affect surgical outcomes?

Background Context: The emergence of drug resistance has complicated the management of spinal tuberculosis (TB). While it is well known that the medical management of drug-resistant spinal TB is more difficult, the surgical outcomes of the same have not been studied sufficiently, particularly in children.

Purpose: To analyze the surgical outcomes in a cohort of children treated for spinal TB, and to thus assess whether drug resistant (DR) disease is associated with poorer surgical outcomes.

Study Design/setting: Retrospective observational study.

Patient Sample: All children diagnosed and treated for tuberculous spondylodiscitis at a single center between January 2014 and June 2017.

Outcome Measures: Surgical outcomes in terms of neurological status and kyphosis angle at final follow-up, and complication rates.

Methods: Radiographic and clinical data of children treated for spinal TB with minimum two-year follow-up were retrospectively analyzed. Data gathered included age, gender, level of spine affected, number of vertebrae involved, neurology (Frankel grade), microbiological reports, duration and type of anti-tuberculous therapy (ATT), details of Orthopaedic management and complications during treatment. In DR cases, the time from presentation to starting of second-line ATT was also assessed. Radiographs were reviewed to note the pre- and post-operative degree of kyphosis as well as the angle at final follow-up. Patients that developed major complications were compared statistically with those that did not.

Results: Forty-one consecutive children (mean age 8.5 ± 4.2 years, 20 boys, 21 girls) were treated for spinal TB with a mean follow-up of 31.2 ± 6.4 months. Fifteen were managed conservatively, of which only one had DR-TB. Of the 26 managed surgically, 13 were managed with first-line ATT and 13 required second-line ATT. Of this latter group, eight had microbiologically proven drug resistance, whereas five were switched to second-line therapy presumptively because of failure to show an adequate response to first-line regimen. At last follow-up, all children had completed the prescribed course of ATT and had been declared cured. Neurological improvement was seen in all but one patient; and at last follow-up, 18 children were Frankel E, seven were Frankel D, and one was Frankel B. The immediate post-operative Kyphosis angle averaged 24.38° ± 15.21°. However, six children showed a subsequent worsening of kyphosis, and the Kyphosis angle at last follow-up averaged 30.96° ± 23.92°. Five children had major complications requiring revision surgery; complications included wound dehiscence, vertebral collapse, screw pull-out and implant breakage. Significantly higher number of patients in the group with complications had required second-line ATT (p < .05).

Conclusions: In a cohort of children treated surgically for spinal tuberculosis, a higher complication rate, and thus poor surgical outcomes, were found to be associated with drug resistant disease.