ABSTRACT
PURPOSE: Considering the current recommendation of the World Health Organization to replace sputum smear microscopy with Xpert MTB/RIF as an initial diagnostic test for tuberculosis (TB), and that culture takes time to provide results, the cycle threshold (CT) of the Xpert test may be the only way to assess bacillary load. The objective of this study is to evaluate the association of bacillary load, measured by the Xpert CT, with the TB treatment outcomes. METHODS: In cohort study, Xpert CT values were evaluated in cured and non-cured (failure and death) patients. Multivariate analysis was performed to evaluate if CT is independently associated with TB treatment outcomes. RESULTS: During this study period, 155 patients (84 cured and 71 non-cured) met the inclusion and were included in the analysis. The mean CT value for Xpert MTB/RIF test was 20.7 ± 5.6 in cured patients and 17.1 ± 5.6 in non-cured patients (p < 0.0001). Previous TB was more frequent in non-cured (28.2%) than in cured patients (7.1%) (p < 0.0001). Non-cured patients were younger than cured ones (37.1 ± 13.3 vs 43.6 ± 16.2; p = 0.006). HIV was more frequent in non-cured (28.2%) than in cured patients (15.5%), although this difference was not statistically significant (p = 0.054). In multivariate analysis, CT values, age, previous TB, and HIV were independently associated with non-cure. CONCLUSIONS: Lower Xpert MTB/RIF CT values were independently associated with worse treatment outcomes. The information from even a single test performed before starting treatment proved to be a relatively good predictor of TB treatment outcome.
