Published online 28 November 2007 | Nature | doi:10.1038/news.2007.306


The evolution of a killer

Could better surveillance have staved off a deadly tuberculosis outbreak?

Mycobacterium tuberculosis can develop a deadly resistance to antibiotics.CAMR / A.B. DOWSETT / SCIENCE PHOTO LIBRARY

Failure to test patients with tuberculosis for drug-resistant strains may have spurred the development of a deadly antibiotic-resistant outbreak, researchers have found.

The work suggests that if better tuberculosis surveillance programmes had been in place during the past decade in developing countries, antibiotic treatments could have been better tailored to patients, and the emergence of extensively drug-resistant strains of tuberculosis (XDR-TB) could have been delayed.

“The main message to any public health programme anywhere in the world is that you cannot control infectious disease without a surveillance programme,” says Willem Sturm, a doctor and microbiologist at the University of KwaZulu-Natal in South Africa and an author on the study.

XDR-TB is a lethal descendant of the multidrug-resistant strains of tuberculosis that emerged in the 1990s. Whereas multidrug-resistant strains are resistant to the two antibiotics that have traditionally been used as a first line of defence, XDR-TB is resistant to both the first-line antibiotics and at least three of the six remaining second-line drugs. The resulting infection is incredibly hard to treat.

XDR-TB was first seen in patients with HIV in KwaZulu-Natal in 2005 (see Extreme TB strain threatens HIV victims worldwide). Since then, the strains have been found across the globe, in patients with and without HIV. Almost 10% of cases of multidrug-resistant tuberculosis are now believed to be XDR-TB1. Models have predicted that without improved control of multidrug-resistant tuberculosis, the XDR-TB outbreak will become an uncontrollable epidemic2.

“Between HIV and TB — not only in Africa but in parts of India and China — there are going to be so many deaths that it will change the shape of those societies,” says Michael Iseman, a doctor at the University of Colorado School of Medicine in Denver, who was not affiliated with the study.


Doctors have tried to suppress the emergence of drug-resistant tuberculosis by treating patients with multiple antibiotics. The use of several drugs is expected to drastically slow the development of resistant strains because simultaneously developing resistance to more than one drug should be a very rare event. But this only works if the infection in question is susceptible to a number of drugs in the treatment cocktail.

Now, a survey of drug resistance in tuberculosis cultures taken from infected patients between 1994 and 2005 shows that strains found more than a decade ago in the KwaZulu-Natal province were already resistant to several of the antibiotics being used to treat the disease3. “People were treated with four drugs, but two to three of them didn’t work because the organism was already resistant,” says Sturm.

If South African officials had known that, they could have switched the treatment of such patients to a set of four drugs that did work against the dominant strains, says Sturm. Instead, they effectively treated the infections with a single drug, so the local tuberculosis became resistant to that too, generating XDR-TB.

The spread of HIV in the region provided an increasing population of particularly susceptible patients, worsening the problem.

Culture test

Many countries in the developing world cannot afford to routinely test tuberculosis cultures, and diagnosis is often based solely on microscopic examination of sputum taken from patients. “They have no laboratory services,” says Iseman. “If the smear is positive they treat the patient.”

South Africa is rich enough and has the labs, says Sturm. “It was not a matter of resources in South Africa. It was a matter of priorities.”


Sturm says that when he was in charge of laboratory services for KwaZulu-Natal in the early 1990s, he instituted a surveillance programme in two regions of the province. But in 1995, this programme stopped, he says. “It was considered a waste of money,” says Sturm. “At that time, we stopped doing any susceptibility testing. We were treating patients with a blindfold on.”

With the emergence of XDR-TB, doctors in KwaZulu-Natal now need to perform drug susceptibility tests on all of their patients. “So we are now financially much worse off then we would have been if we had implemented surveillance in 1995,” says Sturm.

The results have implications for all infectious disease programmes, from malaria to HIV, says Sturm. "The main point is that we must learn to prioritize surveillance." 

  • References

    1. Shah, N. S. et al. Emerg. Infect. Dis. 13, 380-387 (2007). | PubMed | ChemPort |
    2. Blower, S. & Supervie, V. Lancet Infect. Dis. 7, 443 (2007). | Article | PubMed |
    3. Pillay, M. & Sturm, A. W. Clin. Infect. Dis. 45, 1409-1414 (2007). | PubMed |
Commenting is now closed.