Socio-demographic characteristics.

Background
Interferon-gamma release assay and tuberculin skin test use is limited by costly sundries and cross-reactivity with non-tuberculous mycobacteria and Bacille Calmette-Guérin (BCG) vaccination respectively. We investigated the Monocyte to Lymphocyte ratio (MLR) as a biomarker to overcome these limitations and for use in monitoring response to tuberculosis preventive therapy (TPT).
Methods
We conducted a cross-sectional and nested prospective observational study among asymptomatic adults living with Human Immuno-deficiency Virus (HIV) in Kampala, Uganda. Complete blood count (CBC) and QuantiFERON-TB® Gold-plus were measured at baseline and CBC repeated at three months. Multivariable logistic regression was performed to identify factors associated with a high MLR and decline in MLR.
Results
We recruited 110 adults living with HIV and on antiretroviral therapy, of which 82.5% (85/110) had suppressed viral loads, 71.8% (79/110) were female, and 73.6% (81/110) had a BCG scar. The derived MLR diagnostic cut-off was 0.35, based on which the MLR sensitivity, specificity, positive predictive value, and negative predictive value were 12.8%, 91.6%, 45.5%, and 65.7% respectively. The average MLR declined from 0.212 (95% CI: 0.190–0.235) at baseline to 0.182 (95% CI: 0.166–0.198) after three months of TPT. A viral load of >50 copies/ml (aOR, 5.67 [1.12–28.60]) was associated with a high MLR while that of <50 copies/ml (aOR, 0.07 [0.007–0.832]) was associated with a decline in MLR.
Conclusion
MLR was highly specific in diagnosing latent TB and declined significantly following three months of TPT. Implications of a high MLR and decline in MLR after TPT need further evaluation in a larger cohort.