Tuberculosis Severity Predictive Model Using Mtb Variants and Serum Biomarkers in a Colombian Cohort of APTB Patients

Abstract

Currently, tuberculosis (TB) is a bacterial infection caused by Mycobacterium tuberculosis (Mtb) that primarily affects the lungs. The severity of active pulmonary TB (APTB) is an important determinant of transmission, morbidity, mortality, disease experience, and treatment outcomes. Several publications have shown a high prevalence of disabling complications in individuals who have had severe APTB. Furthermore, certain strains of Mtb were associated with more severe disease outcomes. The use of biomarkers to predict severe APTB patients who are candidates for host-directed therapies, due to the high risk of developing post-tuberculous lung disease (PTLD), has not yet been implemented in the management of TB patients. We followed 108 individuals with APTB for 6 months using clinical tools, flow cytometry, and whole-genome sequencing (WGS). The median age of the study population was 26.5 years, and the frequency of women was 53.7%. In this study, we aimed to identify biomarkers that could help us to recognize individuals with APTB and improve our understanding of the immunopathology in these individuals. In this study, we conducted a follow-up on the treatment progress of 121 cases of APTB. The follow-up process commenced at the time of diagnosis (T0), continued with a control visit at 2 months (T2), and culminated in an exit appointment at 6 months following the completion of medical treatment (T6). People classified with severe APTB showed significantly higher levels of IL-6 (14.7 pg/mL; p < 0.05) compared to those with mild APTB (7.7 pg/mL) at T0. The AUCs for the ROC curves and the Matthews correlation coefficient values (MCC) demonstrate correlations ranging from moderate to very strong. We conducted WGS on 88 clinical isolates of Mtb, and our analysis revealed a total of 325 genes with insertions and deletions (Indels) within their coding regions when compared to the Mtb H37Rv reference genome. The pattern of association was found between serum levels of CHIT1 and the presence of Indels in Mtb isolates from patients with severe APTB. A key finding in our study was the high levels of CHIT1 in severe APTB patients. We identified a biomarker profile (IL-6, IFN-γ, IL-33, and CHIT1) that allows us to identify individuals with severe APTB, as well as the identification of a panel of polymorphisms (125) in clinical isolates of Mtb from individuals with severe APTB. Integrating these findings into a predictive model of severity would show promise for the management of APTB patients in the future, to guide host-directed therapy and reduce the prevalence of PTLD.