RESUMO
Background: Diagnostic specimens for spinal tuberculosis (STB) are mostly collected via open surgery. Percutaneous computed tomography (CT)-guided biopsies are used in times of limited surgical availability. However, poor diagnostic accuracy of Mycobacterium tuberculosis (Mtb) culture has been reported with this method, due to limited sample volume and the paucibacillary nature of STB. We evaluated Xpert MTB/RIF Ultra on open and CT-guided biopsies as compared with the gold standard Mtb culture and histopathology. Methods: We conducted a prospective diagnostic accuracy study of Xpert Ultra, as compared with tuberculosis culture and histopathology, in adults with signs and symptoms of STB at a tertiary academic hospital in South Africa from November 2020 to December 2021. Diagnostic testing was performed on 31 patients with available samples. Results: Xpert Ultra had a sensitivity of 94.7% (95% CI, 75.3%-99.7%) and specificity of 100% (95% CI, 75.7%-100.0%) against a reference standard of Mtb culture and histopathology. Xpert Ultra had high diagnostic accuracy in open and CT-guided biopsy samples with sensitivity and specificity of 100% and 100% (open) and 89% and 100% (CT), respectively. Mtb culture had limited specificity for CT-guided biopsies (43%; 95% CI, 15.8%-74.9%). HIV-1 coinfection did not affect Mtb abundance measures by Xpert Ultra or culture. Xpert Ultra was also superior to culture for STB diagnosis in patients concurrently treated for pulmonary tuberculosis. Conclusions: Xpert Ultra detected more STB cases than culture for CT-guided biopsy samples. There was also no difference in sensitivity for open biopsies, irrespective of HIV-1 status, making it an important tool for rapid diagnosis, especially during times or in locations where open surgery is not possible or concurrent pulmonary tuberculosis treatment is initiated.
