Objectives: Tuberculosis (TB) is the leading infectious cause of death in the world. Cheaper and more accessible TB treatment monitoring methods are needed. Here, we evaluated white blood cell (WBC) absolute counts, lymphocyte, and monocyte proportions during TB treatment, and characterized their association with treatment failure. Methods: This multicentered prospective cohort study was based in Bangladesh, Georgia, Lebanon, Madagascar, and Paraguay. Adult, non-immunocompromised patients with culture-confirmed pulmonary TB were included and followed up after two months of treatment and at the end of therapy. Blood counts were compared to treatment outcome using descriptive statistics, logistic regression, and Receiver Operating Characteristic (ROC) analyses. Results: Between December 2017 and August 2020, 198 participants were enrolled, and 152 completed treatment, including 28 (18.5%) drug-resistant patients. The rate of cure at the end of treatment was 90.8% (138/152). WBC absolute counts decreased, and lymphocyte proportions increased throughout treatment. In multivariate analyses, baseline high WBC counts and low lymphocyte proportions were associated with positive sputum culture results at the end of treatment (WBC > 11,450 cells/mm 3 : p = 0.048; lymphocytes <16.0%: p = 0.039; WBC > 11,450 cells/mm 3 and lymphocytes <16.0%: p = 0.024). Conclusion: High WBC counts and low lymphocyte proportions at baseline are significantly associated with the risk of TB treatment failure.

There is a crucial need for non-sputum-based TB tests. Here, we evaluate the performance of RISK6, a human-blood transcriptomic signature, for TB screening, triage and treatment monitoring. RISK6 performance was also compared to that of two IGRAs: one based on RD1 antigens (QuantiFERON-TB Gold Plus, QFT-P, Qiagen) and one on recombinant M. tuberculosis HBHA expressed in Mycobacterium smegmatis (IGRA-rmsHBHA). In this multicenter prospective nested case–control study conducted in Bangladesh, Georgia, Lebanon and Madagascar, adult non-immunocompromised patients with bacteriologically confirmed active pulmonary TB (ATB), latent TB infection (LTBI) and healthy donors (HD) were enrolled. ATB patients were followed-up during and after treatment. Blood RISK6 scores were assessed using quantitative real-time PCR and evaluated by area under the receiver-operating characteristic curve (ROC AUC). RISK6 performance to discriminate ATB from HD reached an AUC of 0.94 (95% CI 0.89–0.99), with 90.9% sensitivity and 87.8% specificity, thus achieving the minimal WHO target product profile for a non-sputum-based TB screening test. Besides, RISK6 yielded an AUC of 0.93 (95% CI 0.85–1) with 90.9% sensitivity and 88.5% specificity for discriminating ATB from LTBI. Moreover, RISK6 showed higher performance (AUC 0.90, 95% CI 0.85–0.94) than IGRA-rmsHBHA (AUC 0.75, 95% CI 0.69–0.82) to differentiate TB infection stages. Finally, RISK6 signature scores significantly decreased after 2 months of TB treatment and continued to decrease gradually until the end of treatment reaching scores obtained in HD. We confirmed the performance of RISK6 signature as a triage TB test and its utility for treatment monitoring.

No abstract

BackgroundTuberculosis (TB) is a leading infectious cause of death. To improve treatment efficacy, quicker monitoring methods are needed. The objective of this study was to monitor the response to a heparin-binding hemagglutinin (HBHA) interferon-γ (IFN-γ) release assay (IGRA) and QuantiFERON-TB Gold Plus (QFT-P) and to analyze plasma IFN-γ levels according to sputum culture conversion and immune cell counts during treatment.MethodsThis multicentered cohort study was based in Bangladesh, Georgia, Lebanon, Madagascar, and Paraguay. Adult, non-immunocompromised patients with culture-confirmed pulmonary TB were included. Patients were followed up at baseline (T0), after two months of treatment (T1), and at the end of therapy (T2). Clinical data and blood samples were collected at each timepoint. Whole blood samples were stimulated with QFT-P antigens or recombinant methylated Mycobacterium tuberculosis HBHA (produced in Mycobacterium smegmatis; rmsHBHA). Plasma IFN-γ levels were then assessed by ELISA.FindingsBetween December 2017 and September 2020, 132 participants completed treatment, including 28 (21.2%) drug-resistant patients. rmsHBHA IFN-γ increased significantly throughout treatment (0.086 IU/ml at T0 vs. 1.03 IU/ml at T2, p &lt; 0.001) while QFT-P IFN-γ remained constant (TB1: 0.53 IU/ml at T0 vs. 0.63 IU/ml at T2, p = 0.13). Patients with low lymphocyte percentages (&lt;14%) or high neutrophil percentages (&gt;79%) at baseline had significantly lower IFN-γ responses to QFT-P and rmsHBHA at T0 and T1. In a small group of slow converters (patients with positive cultures at T1; n = 16), we observed a consistent clinical pattern at baseline (high neutrophil percentages, low lymphocyte percentages and BMI, low TB1, TB2, and MIT IFN-γ responses) and low rmsHBHA IFN-γ at T1 and T2. However, the accuracy of the QFT-P and rmsHBHA IGRAs compared to culture throughout treatment was low (40 and 65% respectively). Combining both tests improved their sensitivity and accuracy (70–80%) but not their specificity (&lt;30%).ConclusionWe showed that QFT-P and rmsHBHA IFN-γ responses were associated with rates of sputum culture conversion. Our results support a growing body of evidence suggesting that rmsHBHA IFN-γ discriminates between the different stages of TB, from active disease to controlled infection. However, further work is needed to confirm the specificity of QFT-P and rmsHBHA IGRAs for treatment monitoring.