Abstract
Background
Novel diagnostic solutions using host-response-based gene signatures to differentiate between individuals infected with Mycobacterium tuberculosis presenting with active tuberculosis disease (ATB), individuals infected with the bacterium in the absence of active tuberculosis disease (TBI), and individuals with no evidence of ATB or TBI (no TB) have demonstrated limited utility. However, the impact of age-related immunological changes on diagnostic performance remains unclear.Methods
Using blood samples collected from 119 'adolescent/young adults' aged 10-24 years, 259 'adults' aged 25-55 years, and 79 'older adults' aged 56 years and above we explored differences in expression of eight gene targets using a novel research-use-only Cepheid Xpert TB/LTBI assay. We plotted relative expression for each gene by age stratified by TB status, then constructed two composite classifiers using a subset of gene targets. These composite classifiers were compared using AUCs derived from logistic regression models fitted within each age stratum.Findings
Expression of IL2, IFNG, and MIG appeared to best differentiate individuals who were classified as TBI or ATB from no TB and were used to construct an 'exposure' classifier which yielded an AUC of 0.89 (95% CI 0.83-0.96) among adolescent/young adults, 0.82 (95% CI 0.76-0.87) among adults, and 0.71 (95% CI 0.58-0.84) among older adults. Similarly, expression of SLPI, VEGFA, PLAU, and GBP5 appeared to best differentiate between individuals who were classified as ATB from TBI or no TB. The corresponding composite 'disease' classifier yielded an AUC of 0.88 (95% CI 0.80-0.96) among adolescent/young adults, 0.74 (95% CI 0.66-0.82) among adults, and 0.56 (95% CI 0.40-0.73) among older adults.Interpretation
This study demonstrates the importance of considering age in the development and evaluation of host-response-based diagnostics for ATB and TBI and suggests that host-response-based diagnostics may prove most useful in differentiation of ATB, TBI, and no TB among adolescent/young adults.Fundings
This study was supported by the US Department of Defence (award number W81XWH-18-1-0253) and National Institutes of Health (award number R01AI137681). Salary support was received from the National Institutes of Health (training grant number T32AI007384).