Plasma kynurenine/tryptophan ratio as a sensitive blood-based diagnostic tool for tuberculosis: a diagnostic accuracy study among children living with HIV

Abstract

Background Pediatric tuberculosis (TB) is challenging to diagnose and confirm. The World Health Organization has prioritized development of non-sputum-based diagnostics. We investigated the role of plasma K/T ratio for TB diagnosis among children with and without HIV. Methodology Plasma tryptophan and kynurenine concentrations measured via enzyme-linked immunosorbent assay (ELISA) were used to calculate K/T ratios in a nested case-control design with an optimization cohort (the Lung Cohort Study (LCS)) and a validation cohort (the NIH Stool4TB. The discovery cohort included children with incident TB ( n = 23) and age- and sex-matched controls ( n = 75). The Stool4TB validation cohort included participants with and without HIV who had microbiologically confirmed TB ( n = 15) or clinically diagnosed TB ( n = 20), and controls ( n = 25). Results LCS participants with HIV-associated TB had higher plasma K/T ratio (median 0.511 [IQR 0.182–0.820]) than controls (median 0.053 [IQR 0.030–0.072]), p < 0.0001. Among participants with microbiologically confirmed TB living with HIV, plasma K/T ratio gave a sensitivity of 95% (CI 79–99), specificity of 85% (CI 0.75–0.91), PPV of 85%, NPV of 98%, and AUC of 0.96 (95% CI 0.93–0.99). In the validation cohort, among HIV uninfected children, applying a cut-off value of 0.090 gave a sensitivity of 83 % (CI 0.43–0.99), specificity of 96% (CI 0.80–0.99), PPV of 96%, NPV of 90%, and AUC of 0.92 (95% CI 0.82–1) for laboratory confirmed TB. In the LCS, participants who had received intermittent preventative therapy had lower K/T ratio compared to those who did not ( p = 0.0191). Conclusion Plasma K/T ratio is a sensitive non-sputum-based test for diagnosing TB in children and has promising application within pediatric TB diagnostic algorithms.