Biography
James Seddon is a Professor of Global Child Health at Imperial College London, a Professor at Stellenbosch University, and a Consultant in Paediatric Infectious Diseases at St. Mary’s Hospital in London. His research focuses on children with TB, including TB diagnostics, TB therapeutics, advanced imaging for child TB, drug-resistant TB, TB meningitis, adolescent TB, TB burden modelling and the evaluation of TB biomarkers. James studied medicine at Cambridge University and Imperial College London and has trained in paediatric infectious diseases at centres of excellence in the United Kingdom, with extensive clinical experience in low resource settings.
Key Impacts
Development of a biomarker to monitor children treated for non-severe tuberculosis
SHINE was a randomised phase 3 multi-country trial which found that 4-months of standard treatment was non-inferior to 6-months for non-severe tuberculosis (TB) drug-susceptible in children, allowing for global treatment shortening recommends. There are however no biomarkers of treatment response that allow for tailoring treatment in individual children. We aimed to identify a transcriptomic signature of treatment response in children and assess its viability as a biomarker of treatment response.Methods: Of children enrolled into SHINE, 198 children enrolled in Cape Town were retrospectively classified into having confirmed, unconfirmed, and unlikely TB by an independent endpoint committee (n=771 samples). PAXgene whole blood RNA samples were collected at enrolment, 2, 8 weeks, and at the end of treatment (week 16 or 24), and expression of 85 TB-associated genes measured by qPCR, compared across treatment timepoints.Results: Using the enrolment and end -of-treatment samples in the confirmed TB group (n=37), we identified a new 4-gene monitoring biomarker-based disease risk score (area under the receiver operating characteristic curve, AUROC: 0.89, 95% CI: 0.81-0.97). Performance did not significantly differ between the 4- and 6-month arms. This signature was also validated in an independent, publicly available RNA-Seq dataset of children from India with TB (n=16, PMID: 31796955, AUROC: 0.81, 95% CI: 0.64-0.98).We adapted this 4-gene signature into a monitoring score, with equal weighting assigned per gene. A score of 1 was assigned per gene if its expression value was within 1.5 standard deviations of each groups averaged end of treatment value, subsequently summed (Figure 1).Figure.1 Monitoring score proof of concept – higher scores associated with a return to “baseline” expression levelsConclusions: We have demonstrated proof of principle that a novel transcriptomic treatment monitoring biomarker could be used to measure antituberculosis treatment responses in children allowing for future tailored of treatment in children.
Source: Conference 2024