Tuberculosis in a changing world

The reversal in the decline of tuberculosis during the past few years has been widely publicised. The World Health Organisation has declared a global emergency in the light of the rising incidence of the disease in developing countries, where the interaction between tuberculosis and AIDS can be devastating. In Zambia, for example, two thirds of patients who are diagnosed as having tuberculosis are HIV positive.1 But the problem is not confined to developing countries. In some American cities rates have risen alarmingly and strains resistant to multiple antibiotics have emerged.2 Notifications of tuberculosis have also risen in Britain since 1987, with drug resistant isolates increasing from 8% in 1987 to 14% in 1991.3 How should we respond to this re-emerging threat? The answer will differ among countries, but an understanding of the factors affecting transmission of the pathogen is crucial.

The recent development of molecular typing for Mycobacterium tuberculosis allows us to track the spread of individual strains. The species is genetically relatively homogeneous, but the chromosomes of most strains of the bacterium contain a variable number of copies (up to 20) of a small DNA element called IS6110.4,5 This is a transposable element, which means that it has the capacity to make a copy of itself that can “jump” into another part of the chromosome. The number and location of copies of IS6110 can be shown by Southern blotting, a standard method in molecular biology in which DNA is extracted from the organisms, cut into small pieces with specific enzymes, and separated by size in an agarose gel. The DNA is then transferred on to a filter and incubated with a specific DNA probe, and the locations of fragments containing IS6110 appear as a series of bands in an x ray film. Most strains isolated from patients in Britain result from reactivation of tuberculosis acquired many years before in different parts of the world; since they are only distantly related they produce different DNA fingerprints. The rate at which the pattern changes, however, is so slow that multiple isolates from one person, even after a long dormant period, or isolates taken from someone else infected by this person usually produce identical patterns.

How can this technology help us? Some investigations show that transmission has occurred but merely confirm what is already known or suspected from epidemiological analysis.6 The technology may, however, provide important information that could not be obtained in other ways and may generate data that challenge and affect policy. One useful application of tuberculosis fingerprinting is in determining whether cultures that are unexpectedly positive are the result of microbiological contamination.7 A second application is to identify outbreaks. A short report in a recent issue of the BMJ described how in London a small cluster of four identical strains of tuberculosis were identified among patients who were HIV positive.8 Further investigation showed that three of the four were epidemiologically related and suggested that transmission could have occurred in a centre caring for patients infected with HIV. That such cases can occur is known,*RF 9-11* but this is a valuable method of surveillance in countries with a low incidence of tuberculosis.

Two recent papers in the New England Journal of Medicine have surveyed transmission of tuberculosis in San Francisco and New York, comparing both fingerprinting and conventional epidemiological methods. The results have important implications for the control of tuberculosis.12,13 In both reports fingerprinting showed that recent transmission, as opposed to reactivation of pre-existing disease, was much more common than had been suspected (30% and 40% respectively). In the San Francisco study conventional tracing of contacts missed 90% of the cases shown to be related by fingerprinting analysis. Another startling finding was that one person was responsible for 6% of all cases studied.

The usefulness of tuberculosis fingerprinting as a research and surveillance tool has been greatly helped by the rapid standardisation of methods.14 This allows laboratories to exchange information and provides a yardstick against which new techniques can be assessed. The European Union has recently funded a concerted action programme on tuberculosis fingerprinting, which takes this further, acting as a forum for researchers from interested laboratories to exchange results and for scientists and epidemiologists to discuss priorities. One aim of the programme is to enhance computerisation of the data analysis and to examine the feasibility of a European database.

The experiences of New York and San Francisco show that the threats posed by infectious diseases can never be forgotten and that we abandon surveillance and control measures at our peril. At a time when finances in the public sector are under fierce scrutiny it is important not to make false economies. The treatment and control of tuberculosis remain among the most cost effective health strategies available today.