A new approach to point of assessment diagnosis of tuberculosis

Tuberculosis remains one of the most significant causes of death, particularly in individuals
co-infected with HIV-AIDS; co-infection can lead to a life expectancy of a few weeks and has,
for example, reduced life expectancy in South Africa by around 10 years. Although there are
many methods that are supposed to detect the disease, in practice none of these can give a
diagnosis within a single visit and many do not distinguish latent from active TB - a major
problem given that the World Health Organisation estimates that a third of the world
population has latent disease. The normal protocol for diagnosis uses a combination of several
biochemical assays as well as physical observation. The WHO has identified a clear need for a
reliable point of care assay that will provide a result within minutes and which can be
distributed in high-burden TB populations for less than $ 10. This project seeks to validate just
such a device, based on the recognition by antibodies in the serum of patients with active TB
of novel synthetic lipid antigens identical to components of TB cells. The assay (patents
applied for) has been shown to provide good levels of sensitivity (correct detection of TB
positive serum compared to clinical diagnosis) and specificity (correct detection of TB
negative serum) in a blind trial on 250 samples from high burden TB populations, and to do so
within 30 minutes without the need for laboratory equipment. This project will seek to
improve further the performance of the device and provide proof of concept that it works to
the required standard with serum from a range of different types of co-infected patients,
including those co-infected with HIV/AIDS and those with non-pulmonanry TB. It will put in
place all the necessary data and quality control systems to provide a design lock for the device
and to take it to the next stage of external validation and then to commercial production.