Innovations in drug susceptibility testing for tuberculosis: toward decentralized, accessible diagnostics

Tuberculosis (TB) remains the world's leading infectious disease killer. Progress in its control remains below expectations: the global incidence rate has declined only by 8.7% since 2015, far short of WHO's 50% target set for 2025. Historically, the ‘diagnostic gap’ has been a fundamental challenge in tuberculosis, with as of 2022 only around 60% of estimated cases being diagnosed, and about the same proportion of diagnosed cases being bacteriologically confirmed.
Given the ~450.000 cases of drug-resistant TB yearly, also drug-susceptibility testing (DST) is central to effective TB control, yet access remains severely limited. For instance, for bedaquiline, the most important recent addition to the TB treatment arsenal, resistance is already emerging within a decade after its introduction, often in settings where DST is unavailable. Without robust, accessible bedaquiline DST, this drug risks being compromised before its potential is fully realized. Current reference methods, such as whole-genome sequencing (WGS), targeted sequencing, and phenotypic broth microdilution (BMD), require advanced laboratory infrastructure, restricting their use to well-resourced settings.
This PhD focuses on technological innovations to expand access to DST in low-resource settings. Specific objectives are:
•Development and evaluation of low-cost imaging techniques for BMD-based phenotypic DST
•AI-guided interpretation tools to simplify interpretation of DST assays by non-expert operators
•Investigation of novel imaging modalities for alternative phenotypic DST assays such as thin-layer agar (TLA)
•Identification of resistance biomarkers usable in a point-of-care phenotypic DST assay through AI-driven biological network modeling