Extensively drug resistant tuberculosis

Tuberculosis outbreaks in the developed world are newsworthy.¹ However, in the developing world, where deaths from tuberculosis are common, it takes something exceptional for an outbreak to attract much attention. In response to a recent report at the 16th international AIDS conference² and to increasing South African media reports, the World Health Organization last week expressed concern about extensively drug resistant tuberculosis (also referred to as “XDR tuberculosis”).³

Among 536 culture confirmed cases of tuberculosis at a rural hospital in South Africa, 41% were multidrug resistant,² defined as resistance to rifampicin and isoniazid (two key first line drugs). This is cause enough for concern as multidrug resistant tuberculosis has a worse outcome and its management is very difficult even in high resource settings.⁴ Even more alarming was that 53 (24%) of the isolates from multidrug resistant tuberculosis fulfilled the definition of extensively drug resistant tuberculosis²—namely, multidrug resistant tuberculosis that is also resistant to at least three of the six classes of second line agents. Such tuberculosis is virtually untreatable.

All patients in this outbreak who were tested were HIV infected, and 52 of the 53 died after a median of just 25 days.² In 90% of the isolates the same genetic fingerprint was present, indicating extensive recent transmission. Fifty six per cent of patients had previously been admitted to hospital, raising the likelihood of nosocomial transmission.

Outbreaks of infectious diseases are always more newsworthy if their implications extend beyond the local context, which is the case with extensively drug resistant tuberculosis. For some years, such strains have been known to exist in Asia, North and South America, and Europe. In March this year, the Centers for Disease Control and Prevention and WHO reported a survey of over 17 000 tuberculosis isolates collected from around the world between 2000 and 2004.⁵ Overall, 2% of multidrug resistant strains were also extensively drug resistant, being most frequently found in eastern Europe, western Asia, and South Korea. Population based data from the United States, Latvia, and South Korea showed that 4%, 19%, and 15% respectively of multidrug resistant strains could be defined as extensively drug resistant.

The epidemiology and the limited genotypic data currently available²,⁶ indicate that this is not a single strain, but that extensively drug resistant strains are likely to have emerged in many different places and on multiple occasions. Paradoxically, this is both reassuring and alarming. It is reassuring in that the emergence of extensively drug resistant tuberculosis in more than one strain suggests that the mutations responsible are specific for drug resistance rather than reflecting a fundamental change in behaviour of the organism. This is nevertheless alarming because it also suggests that extensively drug resistant tuberculosis probably arises fairly regularly and is already disseminated.

Drug resistance to tuberculosis results largely from poorly managed care and control of the disease. Poor prescribing practices, low drug quality (or erratic supply), and suboptimal adherence can all contribute to this. Bacilli are subject to intense drug selection, and exposure to mono-therapy predisposes to an accumulation of mutations that confer resistance. Hence optimal treatment includes four drugs to which the organism is sensitive, and a single drug should never be added to a failing regimen. In much of the world, routine culture and sensitivity testing is not available. Thus, where multidrug resistant tuberculosis emerges, inappropriate treatment regimens may lead to serial acquisition of resistance mutations, with potential for emergence of extensively drug resistant tuberculosis. Widespread use of second line tuberculosis drugs (such as quinolones for respiratory tract infections) may also contribute to the development of resistance. Thus, the emergence of extensively drug resistant tuberculosis should come as no surprise—it was entirely predictable in the context of poor control practices.

The havoc that institutional transmission of multidrug resistant tuberculosis can wreak amongst HIV infected people was evident in the US in the early 1990s.⁷ The very modest actual rise in the incidence of tuberculosis that coincided with these outbreaks has now been reversed,⁸ albeit with extraordinary effort and cost. However, the huge potential for extensively drug resistant tuberculosis to further undermine control practices in communities in South Africa and elsewhere in the region is self evident and would be much more difficult to control. In some communities with an antenatal prevalence of HIV of 30%, annual notification rates for tuberculosis have already increased uncontrollably over the past 10 years, reaching 1500/100 000—a rate more than 250 times higher than rates in the US.⁹ Extensively drug resistant tuberculosis must now serve as a serious wake-up call. Although the potential consequences may be most grave in settings with a high prevalence of tuberculosis and HIV, extensively drug resistant tuberculosis is nevertheless already a very serious development in many other parts of the world too.⁵

What response is needed? The global scale and molecular epidemiology of extensively drug resistant tuberculosis require urgent assessment, and laboratory capacity needs to be greatly increased within a network of sentinel sites. Control practices must be rigorously and effectively implemented. Increasing cure rates for tuberculosis through directly observed treatment short course (DOTS) is crucial. Detection rates for cases of tuberculosis need to be improved, highlighting the need for a new diagnostic test. Technologies that can determine the presence of drug resistance at the point of care are needed, as are new drug treatments. The DOTS-Plus strategy¹⁰ for treatment of multidrug resistant tuberculosis needs to be further developed for areas where the disease is established. Nosocomial transmission of tuberculosis is probably commonplace in the developing world, and simple, effective strategies to reduce such transmission need to be urgently implemented. More fundamentally, the emergence of extensively drug resistant tuberculosis is a reminder that tuberculosis needs massive broader commitment: the incompletely funded Global Plan to Stop TB¹¹ demands political will and financial action.