What Should a New TB Test Look Like?

The tests used today to detect tuberculosis (TB) are not good enough ? but what should a new TB test look like exactly?

On March 17-18, 2009, Doctors Without Borders/Médecins Sans Frontières (MSF) brought together a number of doctors, lab workers, community activists and test developers to answer this question. MSF Access Campaign Director Dr. Tido von Schoen-Angerer talks us through what is needed in a new TB test and explains why we can’t settle for anything less.

What was the point of this meeting?

The TB tests we have today are woefully insufficient – for children, for HIV positive people, for people with certain forms of TB: the list of people missed by TB tools is long. There is wide agreement that the current microscopy tests aren’t anything like good enough to detect tuberculosis – they only detect the disease in less than half the patients who have TB.

So the idea for this meeting was to ask experts about the minimum criteria that the new test needs to meet. The reason we need minimum criteria is to guide test developers – so that we avoid a situation where tests that have lesser performance are promoted as solution.

What did the participants identify as the key requirements for a new TB test?

What we agreed is that the test needs to detect a much higher rate of pulmonary TB – the form of TB that is in the lungs – regardless of whether the patient has HIV or not. The current tools fare very badly at detecting TB in HIV positive people so a new test needs to change this.

It also needs to be much better at detecting TB in children. And the new test cannot be based on examining a sputum sample from children – this is crucial because it is so difficult to get a sputum sample from children that many of them are being missed.

It’s also vital that a test works in the places the patients really are, at the first point of contact of a person who is unwell with health services. It’s at this level that most TB patients are seen today and so a real point-of-care test is needed.

And the results need to be quick. To make this test worthwhile at the point-of-care level, patients need a result in one day. And it must be simple: any healthcare worker should be able to perform it with only the smallest training.

It is a huge scientific challenge to come up with this test– but we all agreed that we cannot settle for anything less.

Is the TB research community moving in this direction – and if not, why not?

The current efforts to develop a new TB test that can be implemented at peripheral level are completely insufficient – for one there are clearly not enough actors involved and more people need to work on TB diagnostics.

But there are also some practical obstacles that need to be resolved. For example, test developers need to have better specimen banks so that they can see whether their new tests work or not. More work needs to be done on what we call biomarkers – these are the different ways you can detect the disease – and we need urgent work to see if existing biomarkers are good enough to be developed into a test.

There is also a desperate need for much more open access to information on the current progress in TB diagnostics research. In the meeting we discussed setting up a clearing house so that information can be exchanged freely and for current progress to be regularly and critically reviewed. Today, people are working on their own in their own corner – of course that limits progress.

And then obviously there is a big funding gap. The funding that goes into TB research & development is already limited, but only seven percent of that small amount goes to diagnostics research and development. So much more money is needed – and that money needs to be provided differently, in new funding mechanisms – we need more grants to researchers, but also new ways to stimulate the research such as prize funds to bring in more actors into this neglected field.