New British test may improve TB diagnosis rate

A paper in today's Lancet reports a study in a British school which showed that a new test for exposure to tuberculosis is superior to the existing Heaf Tuberculin Skin Test (TST) used in the UK.

In 2001, a school student in Leicester was diagnosed with active pulmonary tuberculosis after coughing for nine months. All 1208 pupils at the school were offered testing and treatment for TB. Of the 1128 who were screened, 69 had active TB and another 254 were diagnosed with latent infection on the basis of TST skin tests. However, the only significant exposure in this population seems to have been from the first student (the 'index case'). This outbreak accounted for one third of all TB cases diagnosed in Leicester during the year.

This 'point source' outbreak provided an ideal opportunity to try out a new test, developed at Oxford University by a team led by Dr Ajit Lalvani and funded by the Wellcome Trust. The test is being evaluated internationally in a number of different populations, including people with HIV, and is one of several new diagnostic methods backed by the World Health Organization's TB programme. The developers plan to produce low-cost versions of the test for labs with minimal equipment, although it is still likely to cost more than current skin tests.

The new ELISPOT-based test looks in the blood for T-cells that respond to fragments of two proteins (ESAT-6 and CFP10) unique to the human form of Mycobacterium tuberculosis. It should therefore tell the difference between TB exposure and BCG vaccination or exposure to other relatively harmless mycobacteria. In contrast, TST and the similar Mantoux test can give false positive results, especially in populations such as schoolchildren, exposed to other mycobacteria.

Another key advantage of the new test is that it gives results the next day and does not require a delayed second clinical contact to evaluate the test result. (This may compensate for the increased cost.)

There is no agreed reference standard for diagnosing latent tuberculosis. Evidence that the new test is superior to TST therefore relied on showing (a) that it provided a better match than TST to exposure to the index case and (b) that it was not affected by BCG vaccination status or time spent living in countries where environmental levels of mycobacteria are higher.

Informed consent for this study was obtained from around half of the school students (and their parents or guardians) and their level of exposure to the index case was quantified by the number of hours during which they had been in the same room, using electronic school timetables. Thus, students in the same class had more exposure than others in the same year, and so on. BCG vaccination status was assessed, where possible, from medical records and the presence of vaccination scars. A substantial proportion of pupils were born outside the UK, in countries where BCG vaccination is routine, but vaccination records were often unavailable. The ELISPOT tests were done by scientists who had no information about the individual students.

535 11-15 year-old, predominantly Asian, students were included in the study, who were representative of the whole school population. 424 out of 467 BCG-vaccinated students had been vaccinated at birth. 38 children had a past history of household contact with TB and 76 were born in high prevalence countries in Africa and Asia.

In most cases the TST and ELISPOT results agreed, but where they did not, the positive ELISPOT results strongly predicted a high level of exposure to the index case but the positive TST results did not. Overall, there was a much stronger correlation between ELISPOT and exposure than between TST and exposure. TST but not ELISPOT was correlated with BCG vaccination at birth and with being born overseas (a marker for exposure to non-TB mycobacteria as well as exposure to TB).

Study findings also gave an estimate of the number of hours of exposure in a shared room needed to 'guarantee' infection: 130 hours.

The new test is a real advance but still requires further development before it can be made widely available. Treating people with TB before they can infect others is seen as one way to reverse the increasing global burden of this disease, driven in many countries by the HIV epidemic.

However, current treatments for latent TB are too slow-acting, vulnerable to drug-resistance, and are too much of a burden for those taking them, to provide a realistic chance of eradicating TB. Further progress will therefore depend not only on better diagnostics but also on finding and delivering effective new treatments for the 2,000 million people worldwide who have latent or active TB infection.

Further information on this website

Tuberculosis - overview