A Common TB Test Is Inaccurate, WHO Warns Countries

This week the World Health Organization (WHO) blasted the use of unreliable blood tests commonly used to diagnose active TB, as well as the aggressive marketing used to promote them. In its first "negative" policy recommendation, the agency urged countries to ban these serological tests. "They are a waste of time, and they are a waste of money, and very importantly, they put at risk those people who suffer from TB," Mario Raviglione, director of WHO's Stop TB Department, said at a press conference Wednesday.

Instead, WHO recommends that countries use standard microscopy or a new molecular testing system called GeneXpert.

The blood tests, used on at least 2 million people a year, are wrong 50% of the time, says WHO, yielding either false positives or negatives, both of which are dangerous.

WHO started investigating the blood tests in 2005 after getting questions from governments and doctors who were concerned about their increasing prevalence. The tests are mostly manufactured in Europe and North America, but they are primarily used in developing nations where regulations are more lax but the TB burden is greatest.

"It's a moneymaking business. It's a multimillion-dollar business, which is centered on selling substandard tests with unreliable results, and we feel this needs to change," Karin Weyer, coordinator of TB Diagnostics and Laboratory Strengthening for WHO's Stop TB Department, said at the press conference.

In a 2007 evaluation, WHO, along with UNICEF, the United Nations Development Programme, and the World Bank, concluded that the 19 commercially available blood tests were highly unspecific and insensitive. But since then, the use of serological testing for TB has increased, prompting WHO to do an updated review of the evidence. The final review, performed by WHO and then assessed by a group of outside experts, was delayed until yesterday, Weyer says, because after the 2007 report, several of the companies simply changed the names of their products and remarketed them, making them difficult to track. In addition, Weyer says, given the hard-hitting nature of their recommendation, "we had to make sure that the science behind that policy could be defended at any challenge."

The consequences of such inaccurate diagnoses are significant. When TB goes untreated, individuals are not only at risk of death but also unknowingly transmit the disease to others. And when individuals are incorrectly diagnosed with TB, the underlying cause of their illness goes untreated, and the unnecessary use of TB medicines contributes to drug resistance.

These blood tests are inaccurate because they rely on antibody detection of antigens produced by Mycobacterium tuberculosis (MTB), the organism that causes tuberculosis. Antibody response is a poor indicator, however, because the various types of mycobacteria share antigens, and other infections raise similar antibody responses. As a result, blood tests may be detecting the presence of other mycobacterial infections; they may also produce false positives when antibody levels remain elevated after cases of TB have been treated.

Despite these shortcomings, manufacturers make claims of very high accuracy, Weyer says, citing one drug package insert that says the test is 100% accurate, although the study included only 10 subjects. She also mentioned reports of doctors receiving economic incentives from the tests' manufacturers.

Since December 2010, WHO has endorsed the Xpert MTB/RIF test, a new molecular diagnostic manufactured by Cepheid, a company based in Sunnyvale, California. The test runs on the GeneXpert system, which uses real-time polymerase chain reaction to rapidly detect MTB and rifampicin resistance, identifying patients infected with drug-resistant strains.

"The advantage of GeneXpert is that you can immediately diagnose TB and [determine] whether that TB is drug resistant," Weyer says. It provides results in 2 hours compared with the weeks required to complete a traditional TB culture diagnosis. Unlike microscopy, the coffee machine-sized instrument doesn't require a laboratory or trained personnel, she says. The downside is that it costs $20,000, expensive compared with the $1000 needed to purchase a microscope, although the costs of individual tests are similar. GeneXpert also requires a constant and uninterrupted flow of electricity, lacking in many developing nations, although Weyer says studies are under way to attempt to power the machines with batteries and solar energy.

Several studies published in PLoS Medicine this week conclude that the MTB/RIF test is very effective at diagnosing TB in HIV-positive patients, which is particularly difficult because they often have smear-negative results because their TB is disseminated or extrapulmonary. One study found a 45 % increase in case detection for HIV-infected patients using Xpert MTB/RIF assay compared with smear microscopy. Last week, in
a paper in The Lancet, researchers reported that GeneXpert is also effective for diagnosing TB in children, another population at high risk of dying from TB.="#article_upsell">

WHO is still hoping for a "point of care" diagnosis that will deliver results at patients' bedsides, but for now, Weyer says "there is a lot of basic science still missing from the picture."