Mycobacterium tuberculosis associated immune activation, inhibition and exhaustion profiling to predict active tuberculosis disease among presumptive tuberculosis participants and monitor anti-tuberculosis treatment response.

Abstract

Globally, tuberculosis (TB) is the leading cause of mortality and morbidity especially among resource-limited populations including Uganda. This statistic is greatly fueled by the delay in diagnosis of the disease and initiation of treatment, treatment failure, disseminated TB forms, and the burden of coinfections such as HIV. Presumptive pulmonary TB (PTB) subjects who present to treatment centers with symptoms of PTB offer a pool of probable TB that can continue to spread in the community given the delays in seeking medical attention and challenges in diagnosis. Currently, the available World Health Organisation (WHO)-approved tests are sputum based which have shortcomings such as long turnaround time, and infrastructure requirements on top of the challenge of acquiring a good quality sputum samples, especially among children, the very ill who are unable to cough, and the HIV infected person who usually present with smear-negative or paucibacillary pulmonary TB. The WHO's high target product profiles target the need for a non-sputum-based test for triaging, screening and diagnosing PTB given the challenges with sputum-based tests. These should use easily available samples such as blood, saliva, urine and breath condesate. These tests should be field-friendly and applicable in all patient populations, health care settings and for monitoring treatment response. We therefore set out to evaluate blood biomarkers for predicting active PTB disease and monitoring treatment response. This doctoral study utilized archived blood samples collected from presumptive PTB participants who self-presented at a TB treatment centre in Kampala, Uganda. Participants were later classified as having PTB or ORD using standard microbiological confirmatory tests. Patients with ORD were further sub-classified as having LTBI or no LTBI using the QuantiFERON Gold-Plus test. The concentrations of soluble markers of immune activation and exhaustion markers, the protein expression, expression of associated genes as well as 27 host cytokine biomarkers were evaluated in patient sera using ELISA, RT-PCR, microfluidic technology, and Luminex platform respectively. We evaluated the abilities of the evaluated markers to discriminate between PTB, ORD_LTBI, and ORD_NoLTBI in serum, QFT pellets, and PAXgene blood samples respectively. We found increased expression of PD1 in active PTB, and increased HLADR, FOXp3 and CD80 in the serum of ORDs at baseline. All markers remained high during anti-PTB treatment with only FOXp3 having significant differences between anti-TB treatment visits. Host biomarkers including IP10, IL6, IL2, IL1β, TNFα, IFNγ, and IL12p70, were significantly higher in PTB patients (n=55) than ORDs (n=106), and between PTB and the two ORD sub-groups. A bio-signature comprising IP10+IL6+TNFα+IL1β+IL1ra+IL12p70 best diagnosed PTB disease, with an area under the ROC curve of 90%. When clinical and microbiological parameters were included, a biosignature comprising TTP+TNFα+PDGF-BB+IL9+GCSF best predicted sputum culture conversion at month 2- post initiation of anti-TB therapy with a sensitivity and specificity of 82%. There was increased gene expression of PD1 in active PTB, and higher HLA-DR, and CTLA4 among ORDs in QFT pellets. In addition, there was higher gene expression of NOD2, STAT1, BAFT2, CIQA, and GBP2 in blood PAXgene samples of active PTB patients. We identified blood host biomarkers including cytokines, chemokines and genes that were differentially expressed among the different presumptive PTB participant groups. The identified biomarkers also had the potential for predicting active PTB disease and for monitoring anti-PTB treatment response among confirmed active PTB patients. In addition, we identified that QuantiFERON pellets show potential for use in gene expression analysis among presumptive PTB patients. These biomarkers may be considered as additional candidate markers for future active PTB diagnostic and treatment monitoring point-of-care tests.