Using neutrophil-specific biomarkers as a measure of Mycobacterium tuberculosis burden, lung damage and treatment response kinetics.

Abstract

The spectrum of TB disease remains poorly characterized impacting case identification and linkage to care. Furthermore, the absence of simple, sensitive and easy to measure biomarkers of disease severity and treatment response makes it difficult to distinguish patients at risk of treatment failure from those who would benefit from shorter treatment regimens. Ultimately, these challenges contribute to perpetuating the global TB epidemic. This thesis explores the potential of blood neutrophils as biomarkers for delineating TB disease, characterizing disease severity and monitoring TB treatment response. Neutrophils are short lived and rapidly responsive phagocytes that have been shown to have a strong TB associated signature. Using longitudinal samples, collected during a 2-year observational TB treatment response study, I analyzed the neutrophil response over the course of standard TB drug therapy. This included the novel assessment of a panel of neutrophil surface markers by flow cytometry. Overall, I show that an active TB infection is associated with elevated blood neutrophil counts that resolved over the course of TB treatment and is associated with disease severity, shown by direct correlation with bacterial load and extent of lung involvement. Importantly, these associations were not impacted by the high background of HIV infection present in this cohort. In addition, I identified the expression level of surface CD15 as a novel, rapid, robust and highly sensitive marker of TB disease. This correlated strongly with clinical characteristics and, uniquely, baseline CD15 expression predicted TB treatment response, as shown by solid culture conversion at 2 months. To extend these findings I measured a panel of inflammatory cytokines and neutrophil specific soluble factors by Luminex and ELISA. As shown in other studies, many plasma cytokines are elevated in subjects with active TB and reduced in response to treatment. Baseline levels of the neutrophil associated markers, S100A8/9 and TREM-1 associated most strongly with disease severity, again supporting the hypothesis that neutrophils are a good sentinel immune cell in TB. Hierarchical clustering indicated a coordinated inflammatory response to TB infection, which resolved over the course of treatment. However, these did not associate with disease severity or predict treatment outcome. Finally, I tested the potential of simple neutrophil characteristics (abundance and neutrophil:lymphocyte ratio) to identify clinical and subclinical TB identified in the community, using blood smears obtained during a recent community

wide TB screen conducted in rural Kwa-Zulu Natal. Despite reasonable performance in the clinic, however, these metrics failed to distinguish symptomatic TB from sub-clinical TB, and either of these from TB uninfected community controls. Unfortunately, it was not possible to evaluate neutrophil phenotype, particularly CD15 expression level, in this setting. Taken together, these data suggest that detail measuring of the neutrophil response to TB in the clinic may provide useful information about disease severity, offer a sensitive measure of treatment response and, potentially, identify patients at baseline who may be at risk of poor treatment outcomes.