AddiCAD

The challenge

An estimated 10 million people develop tuberculosis disease each year, yet around a quarter are never formally diagnosed. This means they do not receive the appropriate treatment and can act as a source of additional household and community infections. 

This gap reflects difficulties in TB diagnosis. Respiratory symptoms are non-specific, culture methods take weeks to perform, and molecular platforms are expensive and not available in all settings. As a result, new approaches to TB detection are highly prioritised, such as triage strategies to identify those most likely to have TB disease for intensive assessment, and confirmatory testing methods that are more accessible, quicker, and easier to use.

Progress is being made in the search for alternative approaches. For example, computer-aided diagnosis (CAD) using chest X-rays can improve diagnosis rates. There is also great interest in using blood biomarkers as the basis of tests to identify patients with active TB.

The project

The AddiCAD project is based on an innovative approach that combines chest X-ray CAD and blood biomarker technologies to identify tuberculosis cases. 

The project team has developed a blood test that detects three protein markers associated with TB disease: SAA-1, CRP, and IP-10. A lateral flow test has been developed, known as 3P-TB, based on these markers. In addition, the team has assessed the additional value of combining 3P-TB with an existing chest X-ray CAD product, CAD4TB (developed by Delft Imaging).  

In an initial study of 600 people with possible TB disease, the combined approach (AddiCAD) showed very good performance. When tuned for triage, AddiCAD achieved a significant increase in specificity without loss of sensitivity (fewer patients would be sent for confirmatory testing, with no increased risk of cases being missed). When tuned for diagnosis, AddiCAB achieved the best performance yet seen for non-sputum-based tools, nearly meeting WHO benchmarks.

In the AddiCAD project, the team is building on these promising results and advancing the tool towards implementation in routine practice. In a first step, the 3P-TB test is being transferred into a new technology platform, based on ‘aptamers’ – specially designed nucleic acids that bind specifically to the three targeted biomarkers. This will enable the markers to be detected at lower concentrations, using smaller volumes of blood (so fingerprick samples can be used) and with results available sooner (almost immediately). Importantly, the improved detection technology will not be more expensive. 

The team will also develop a simple-to-use app to integrate CAD and 3P-TB results. The new tool will then be evaluated in a clinical study in The Gambia, Namibia and South Africa. Around 1,000 patients with signs of possible TB will be recruited, and the results obtained with AddiCAD will be compared with those from culture and molecular testing technologies. The study will examine the use of AddiCAD for both triage and confirmatory testing. 

The project will assess performance in key patient subgroups, including those with a low microbial load and people living with HIV. The device will also be ‘stress tested’ to ensure it is robust enough for use in challenging primary care settings in the region.

The project team will also engage with key stakeholders to map out pathways for intellectual property management, regulatory approval and commercial sustainability.

Impact

The AddiCAD project could have a major impact on tuberculosis detection. If successful, it will:

• Show whether the promising performance data seen to date are confirmed in a larger study.
• Significantly enhance triage, reducing the number of patients referred for confirmatory testing, thereby saving costs.
• Provide an on-the-spot test for TB disease, allowing treatment to begin immediately.
• Create commercial opportunities for African-based product developers.

If shown to perform as expected, AddiCAD could significantly reduce the number of missed cases of TB, ensuring patients start treatment more rapidly and reducing opportunities for community spread of TB.