The monocyte-to-lymphocyte ratio: Sex-specific differences in the tuberculosis disease spectrum, diagnostic indices and defining normal ranges.

BACKGROUND: The monocyte-to-lymphocyte ratio (MLR) has been advocated as a biomarker in tuberculosis. Our objective was to evaluate its clinical value and associations. METHODS: Blood counts, inflammatory markers and clinical parameters were measured in patients with and those screened for tuberculosis. Complete blood counts (CBCs) from a multi-ethnic population aged 16 to 65 years were evaluated; a sub-group with normal hematological indices was used to define the range of MLRs. RESULTS: Multivariate analysis in proven tuberculosis (n = 264) indicated MLR associated with low serum albumin, high white cell counts and a positive culture; values were higher in sputum smear-positive pulmonary tuberculosis (S+PTB). Analysis in S+PTB (n = 296) showed higher MLRs in males and those with high neutrophil counts, low serum albumin and high C-reactive protein. The diagnostic value of MLRs was assessed by comparing notified patients with TB (n = 264) with denotified cases (n = 50), active case-finding in non-contacts (TB n = 111 and LTBI n = 373) and contacts of S+PTB (n = 149) with S+PTB found at screening (n = 75). Sensitivities and specificities ranged from 58.0-62.5% and 50.0-70.0% respectively for optimal cut-off values, defined by ROC curves. In CBCs obtained over one month, ratios correlated with neutrophil counts (ρ = 0.48, P<0.00001, n = 14,573; MLR = 0.45 at 8-8.9 x 109/L) and were higher in males than females (P<0.0001). The MLR range (mean ± 2SD) in those with normal hematological indices (n = 3921: females 0.122-0.474; males 0.136-0.505) paralleled LTBI MLRs. Ratios did not predict death (n = 29) nor response to treatment (n = 178 S+PTB with follow-up CBCs). Ratios were higher in males than female in the 16-45 years age group, where immune differences due to sex hormones are likely greatest. CONCLUSIONS: Severe tuberculosis and male sex associated with high MLRs; the same variables likely affect the performance of other biomarkers. The ratio performed poorly as a clinical aid.