BioEd Online

Researchers were surprised earlier this week when US health authorities declared that Atlanta lawyer Andrew Speaker, the American tourist recently diagnosed with a deadly strain of drug-resistant tuberculosis, doesn't have such an extreme strain after all. Despite a first test in March showing that he had a contracted an extensively drug resistant (XDR) strain before going on holiday, on 3 July the US Centers for Disease Control and Prevention (CDC) downgraded their assessment of his condition to multi-drug resistant (MDR).

The changed diagnosis is good news for the patient and his fellow travellers, who might have been exposed. About 70% of MDR infections are cured, versus up to 30% for XDR TB.

But for medics the news is simply confusing. "I think this whole case has been a mystery from the start, but this really adds a new twist," says Sarita Shah, a former CDC epidemiologist, now at Albert Einstein College of Medicine in New York.

The reclassification of the TB strain was based on new tests performed by a hospital in Denver, Colorado, where Speaker is in treatment, on samples taken between April and June. A CDC laboratory in Atlanta performed the one and only XDR-positive test in March.

The CDC stands by its first result. Mitch Cohen, head of its infectious disease division, said in press conference that Speaker may initially have been infected with several strains of TB. A disproportionate amount of XDR TB may have been in the sample used for their test, whereas in his body MDR strains predominated.

It's not uncommon for someone to carry multiple strains of TB at the same time. But experts say it would be unusual for a test to not reflect the true proportions of these strains in the patient. Some are calling for further scientific study to work out exactly what happened in this case.

Global challenge

Experts note that the odd case highlights just how hard it can be to diagnose TB. "The fact that we've had a reversal of the diagnosis in the US, which has the best diagnostic laboratories in the world, indicates how difficult it is to distinguish between MDR TB and XDR TB," says Christopher Dye, coordinator of TB monitoring and evaluation at the World Health Organization in Geneva.

Misdiagnoses can have a knock-on effect on treatment and on the evolution of resistant strains.

About 96% of tuberculosis infections can be effectively treated by a 'first-line' treatment of a four-drug cocktail. MDR strains are impervious to two of these. XDR strains are additionally untreatable by at least two 'second-line' drugs as well. Doctors treat these patients with what options are left, or untested antibiotics and surgery.

In the developing world, where MDR and XDR are taking hold, poorly equipped facilities, untrained staff and hard-to-replicate tests make it difficult to identify TB at all, experts note, let alone different strains.

So countries that have good access to second-line drugs tend to over-use them, to get the best chance of curing individual patients. This spurs antibiotic resistance to this 'last resort' treatment. Such "indiscriminate use" of second-line drugs is a major problem in the former Soviet Republics, China and India, Dye says.

Under-use of the drugs when they are really needed is equally devastating, says Karin Weyer, director of TB research at the South African Medical Research Council in Cape Town. "There is a very real possibility that patients are dying because the diagnostic capacity is absent," she says.

Until more accurate, simple and affordable tests of TB drug resistance are developed, doctors should decide on treatment on the basis of how each patient responds, Weyer says. "Treat the patient, not just the lab result."