Identifying individuals at risk of progression to active tuberculosis

点击翻译按钮获取中文摘要

Project Summary Almost 2 billion people are infected with Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis (TB). Approximately 10% of these individuals will progress to active TB disease over their lifetimes, but there is currently no clinical test to distinguish those that will progress to active TB disease, from those that will not. If we are to realize the World Health Organization's (WHO) goal of a world free of TB by 2035, the massive reservoir of TB infection must be addressed with a cost-effective, ethical therapy for preventing progression, based on treating only those most likely to progress. A diagnostic test that can accurately predict the risk of progression is critical for treating these high-risk individuals and the eradication of TB. Our goal is to develop such an assay. Our central hypothesis is that five independent host immune biomarkers, combined into a single multimetric signature will predict progression from latent to active TB with at least 90% sensitivity and specificity. We will test this hypothesis and achieve our goal by implementing the following specific aims: Aim 1: Compile a comprehensive dataset of biomarkers in a prospective cohort of individuals who are at risk of progressing to active TB. Working with the Moldova Ministry of Health's National TB Program, we will enroll 3,685 close contacts of active TB cases. All participants will be followed for two years to determine who progresses to active TB. We expect to identify ≥ 140 progressors. We will assess three previously established blood-based predictors of active TB progression, and two novel assays. We will verify the performance of previously published biomarkers in this population to discriminate progressors from non-progressors and identify new candidate biomarkers using RNA-Seq of antigen stimulated PBMC and detection of Mtb-peptides by NanoDisk MS. Aim 2: Use a discovery set of samples to develop predictive models of progression to active TB. Using data from 140 progressors and 140 non-progressors from Aim 1 we will (1) Verify the performance of existing biomarkers, (2) Use a cross-validation to identify new candidate biomarkers, and (3) derive predictive models using logistic regression and machine learning methods to identify optimal biomarker signatures that best predict progression to active TB within 12 months. Aim 3: Verify the ability of the model to predict progression to active TB disease. Using the same approach as Aim 1, we will enroll a new set of 1,340 household contacts of active TB and identify at least 60 progressors and 60 matched non-progressors and verify clinically the sensitivity/specificity of our models and biosignatures (Aim 2) to predict progression to active disease. A combined host biomarker signature that can predict TB progression from a small blood volume will have significant impact on the WHO End TB Program.