Advertisement

New automated diagnostic test can identify TB infections in 2 hours

Share

A new automated test to detect tuberculosis infections and the presence of an antibiotic-resistant TB strain can shave days to weeks off the time it takes to identify new infections, allowing treatment to be started immediately to prevent further spread of the bacterium. The new test, which can be performed by technicians with only limited training, could prove highly valuable not only in the developing world, but also in the inner cities of the developed world where outbreaks of the deadly disease are becoming more common. The World Health Organization is expected to decide by the end of this month whether to recommend use of the new test by its member countries. “These results suggest that it has the potential to revolutonize TB care, and WHO will treat it as a top priority,” the agency said in a statement.

TB is one of the most common and deadly infectious diseases in the world, striking an estimated 9.4 million people each year and killing 1.8 million. It is especially deadly in people who are already infected with HIV. Moreover, researchers are growing concerned about increasingly common drug-resistant strains of the bacterium, which last year struck about 440,000 people and killed 150,000.

The most commonly used test for an infection, developed 125 years ago, is called the smear test and involves looking at a “smear” of sputum from a patient under a microscope to check for the presence of the TB bacterium. But that test can take days to perform because of backlogs of specimens and the need for trained personnel and is only 50% accurate. Moreover, it doesn’t work in patients who are HIV-positive. A more accurate way to identify an infection, also old, involves culturing the specimen in a test tube to confirm that the bacterium is present. Different antibiotics can then be tested to determine if the bug is resistant. That process can take six weeks or longer and also requires skilled technicians. During that period, a person who is not given the expensive drugs required to treat antibiotic-resistant strains can infect 12 to 15 people, on average.

The new test, developed by Cepheid Inc. of Sunnyvale, Calif., in contrast, requires placing the sample in a small cartridge that is inserted into an automated processor called GeneXpert, which performs the test. The test itself uses what is known as PCR technology to make many copies of any bacterial DNA in the specimen for analysis. It also amplifies the DNA for the gene that confers resistance to the antibiotic rifampin, one of the most common treatments. Interestingly, the technology was originally created for the U.S. Post Office to help postal workers check for anthrax contamination in mail, and has been adapted to diagnose TB.

In the new study reported online Wednesday in the New England Journal of Medicine, a team headed by Dr. Catherina C. Boehmer of the Foundation for Innovative New Diagnostics in Geneva, studied specimens from 1,730 patients from Peru, Azerbaijan, South Africa and India who were suspected of having tuberculosis. The samples were also subjected to conventional analysis for comparison. The test accurately identified 98% of all confirmed smear-positive TB cases and 72% of those that were negative in the smear test. Putting the latter specimens through two additional cycles of the machine raised the success rate to 90%. The test also accurately identified 98% of those samples resistant to rifampin.

There are some limitations to the study. The tests were actually performed in labs in developed countries, and need to be verified in local laboratories, where conditions are often much dustier--which might interfere with the test--and where interruptions of electrical power are not uncommon. Cost could also be an issue. The GeneXpert machine costs about $30,000 and each test costs about $63. Cepheid said it will sell the machine in developing countries for $20,000 and tests for about $32, but even those rates may be beyond the budgets of the impoverished countries.

But as tuberculosis specialist Richard Chaissons of the Johns Hopkins Center for Tuberculosis Research in Baltimore told the journal Nature, “We can say that smear is cheap, but given that it misses most cases, who cares how cheap it is?”

Funding for the study was provided primarily by the Bill and Melinda Gates Foundaton and the National Institute of Allergy and Infectious Diseases.

Thomas H. Maugh II / Los Angeles Times

Advertisement