Non-respiratory biomarkers to diagnose and monitor response in pediatric TB

University Of Virginia, Charlottesville VA

Investigators

Abstract

? DESCRIPTION (provided by applicant): Background: Tuberculosis (TB) is the leading cause of death from a bacteria worldwide and infants and young children are disproportionately affected. However, there is currently no single accurate test to diagnose or monitor children with TB disease. Traditional diagnostic methods that rely upon detecting the actual mycobacterial organisms from sputum are insensitive in children. Biomarkers, like the Antibodies in Lymphocyte Supernatant (ALS) assay and the trans-renal TB DNA assay, are available in adults and hold promise for use in children. ALS uses blood (and not serum) to measure TB-specific antibodies which are actively being secreted from immature plasma cells found in circulation among children who are suffering from TB. Concurrently, the immune-mediated breakdown of Mycobacterium tuberculosis (Mtb) from pulmonary TB results in the release of small, cell-free nucleic acids into the plasma which are filtered through the kidney and can be detected as fragments of Mtb DNA in urine. Objectives: Our overall goal is to use non-respiratory specimens to evaluate biomarkers for pulmonary TB in children. We propose to evaluate the ALS and trans-renal TB DNA assays as diagnostic biomarkers of disease activity. The changes in monthly ALS measurements will also be evaluated as a response-predictive biomarker. Methods: This will be a prospective study among 408 young children (<5 years) from Dhaka, Bangladesh who are hospitalized with signs and symptoms of pneumonia are suspected of having pulmonary TB. Children will be followed over 6 months. We will determine the diagnostic accuracy of each biomarker in comparison to a gold standard reference of GeneXpert and/or TB culture-positive disease. Secondary analysis will evaluate the diagnostic performance of each biomarker using an NIH-consensus clinical case definition, since many children do not have microbiologically confirmed TB. Logistic regression models will be used to determine whether or not either biomarker contributes additional information to predicting TB in the presence of clinical predictors and confirmatory laboratory tests. Lastly, the kinetics of monthly ALS values will be correlated with clinical response and TB treatment outcome. Impact: Successful completion of these aims will provide much-needed minimally-invasive diagnostic and monitoring strategies for children suspected of having TB, thereby limiting misdiagnosis.