Fingerstick blood test shows promise for TB screening

Researchers have been exploring alternatives such as testing for TB in blood, exhaled breath, stool and urine.

Last week, researchers presented promising interim results showing that a new fingerstick blood test meets target criteria that the WHO set for such triage tests. Presenting the research at the 52nd Union World Conference on Lung Health, Dr Jayne Sutherland, head of the TB research group at the Medical Research Council Unit in Gambia, said the triage test can quickly rule out the majority of patients who do not suffer from TB. Confirmatory tests will still be needed — to confirm infection and establish whether the TB is drug resistant.

How it works

Developed by medical diagnostics company Cepheid, the new fingerstick test — the Xpert-MTB-Host Response (HR)-Prototype — can produce results within an hour, provided the right equipment is available. In some cases, samples might still have to be sent to faraway labs. Cepheid also produces the sputum-based Xpert MTB-RIF test used for most TB diagnoses in South Africa.

As with other fingerstick tests, a person’s finger is pricked to obtain a drop of blood, which is put into a cartridge that is slotted into a machine to run the analysis. Sutherland said the machine looks for three different genes (technically it looks at gene expression) in the blood and then applies an algorithm to come up with a “TB score” that indicates the likelihood of having TB.

“Because this test is based on a small amount of blood that can be collected from virtually anyone by fingerstick, it has particular utility for diagnosis in persons who cannot produce a sputum sample, which is required for the traditional diagnostic tests for tuberculosis,” says Professor Thomas Scriba, deputy director of immunology at the South African Tuberculosis Vaccine Initiative. Scriba explains that currently TB is diagnosed using one of three methods — smear microscopy, sputum culture, or a nucleic acid amplification test (NAAT) — which depend on a sputum sample. The MTB-RIF test mentioned earlier is a type of NAAT. Counting against culture is that it can take two to six weeks, while microscopy suffers from poor sensitivity.

“The first limitation is thus that not all people with TB can produce a sputum sample — this is especially an issue in children, people with HIV infection, and those who have extrapulmonary TB. Such a blood-based test may be particularly useful for healthcare professionals to aid with these difficult-to-diagnose cases,” he says.

Scriba adds that it is unlikely that the blood test will replace existing diagnostic tests, especially NAATs in people who can produce sputum, since detecting the bacterium will always be a more definitive diagnostic method than detecting changes in gene expression in blood cells, which is indicative of an inflammatory response to the TB bacillus but can also occur in people who have certain viral infections.

Interim results

While the study is ongoing, Sutherland presented results based on the first 195 participants. The full study will include 1,200 adults and adolescents and 600 children. It is being conducted in South Africa, Uganda, Vietnam and Gambia.

In the study, people with persistent cough and at least one other TB symptom are being given the fingerstick test. They are also tested with the gold-standard sputum-based Xpert MTB-RIF and/or culture tests — testing positive on either of the latter constitutes a positive TB diagnosis. The key question the researchers asked was how well the fingerstick result predicted the TB diagnosis.

Of the 195 people, 75 had a TB diagnosis using Xpert MTB-RIF and 120 did not. Measured against MTB-RIF, the fingerstick had an estimated sensitivity of 87% and a specificity of 94%. (Sensitivity is a measure of how well a test does in picking up a disease in people who have the disease, while specificity is a measure of how well a test does at not picking up anything if someone does not have the disease being tested for.)

“It is not accurate enough for diagnosis alone,” says Sutherland, “but use of blood rather than sputum reduces biohazard risk and hopefully will allow screening in populations that find it difficult to produce sputum such as children and people living with HIV. Only those with a high likelihood of TB will require further TB testing.”

She says about 30% of patients presenting with respiratory symptoms have TB, the rest have other respiratory diseases. Getting a result within an hour will allow healthcare workers to streamline follow-up testing.

Critically for the South African context, the tests seem to work as well in people living with HIV as they do in people who aren’t. “HIV patients often have a low bacterial load or can’t produce sputum, so detection of the bacteria is not possible. This test instead detects host markers in blood that give a likelihood of TB so it’s not reliant on detection of the pathogen,” explains Sutherland.

Scriba adds that the proportion of people who can produce sputum for testing and those among them with detectable bacteria in their sputum, is lower in the HIV-positive population than in the HIV-uninfected population. “HIV-positive people are also more likely to have extrapulmonary TB and therefore do not produce sputum. However, there are important additional considerations. For example, in people who are not on antiretroviral treatment and who do not have suppressed HIV viral load, the blood test appears to have a lower specificity,” he says.

‘A gratifying development’

Professor and pulmonologist Keertan Dheda at Groote Schuur Hospital. (Photo: Supplied / Spotlight)