Tuberculosis

Tuberculosis is the leading infectious cause of death worldwide, being responsible for 3 million deaths annually. Among those aged over 5 years, tuberculosis kills more people than AIDS, malaria, diarrhoea, leprosy, and all other tropical diseases combined. The tragedy of this situation is that treating tuberculosis is one of the most effective and cost effective of all health interventions. The World Health Organisation has calculated that, unless urgent action is taken, the annual number of deaths could rise from 3 million to 4 million by the year 2004.¹ We urgently need improvements in the implementation of existing strategies for tuberculosis control, with particular emphasis on early diagnosis and delivery of effective treatments. In addition, basic research is needed for the development of simple and rapid diagnostic tests, more effective vaccines, and new drugs. This article focuses on new scientific approaches to tuberculosis control that could soon be incorporated into routine practice. However, the impact of new approaches will be negligible if the wealthy Western nations fail to address the gross global inequities in healthcare provision,² which account for the fact that 98% of deaths from tuberculosis occur in the poorer developing countries (fig 1).

Figure 1.

Estimated incidence and prevalence of, and number of deaths due to, tuberculosis in 1997 (data from World Health Organisation³)

So serious is the global threat of tuberculosis that, in 1993, the WHO took the unprecedented step of declaring this disease a global emergency.¹ The problem is fuelled by the pandemic of HIV infection and AIDS and the emergence of drug and multidrug resistance. HIV infection renders a person infected by Mycobacterium tuberculosis much more likely to develop overt tuberculosis, and the evolution of the disease is considerably accelerated. At present, about 8-10% of all cases of tuberculosis worldwide are related to HIV infection, but the association is much more common in many African countries, often 20% or more.⁴

Drug resistance is ubiquitous, although the incidence varies greatly from region to region. Globally, about 10% of cases of tuberculosis are resistant to one drug, and, although primary multidrug resistance is uncommon (about 0.2%), it occurs in a median of 4.4% of previously treated patients.⁵ These figures may be an underestimate, as countries with good facilities for testing for drug resistance also have good tuberculosis control programmes.

Potential futures

• Nucleic acid technology will provide rapid, specific, and sensitive diagnostic tests and rapid detection of drug resistance

• A new vaccine able to prevent the emergence of post-primary, infectious tuberculosis will be one of the principal means of controlling tuberculosis

• An immunotherapeutic agent used in conjunction with drug treatment will lead to a much lower failure rate, even in cases of drug resistant disease, and new “designer” drugs with specific anti-tuberculosis activity will be used to treat resistant cases

• Anti-HIV vaccines will reduce the burden of HIV related tuberculosis

• Adequate resources to implement the directly observed therapy short course (DOTS) strategy will be made available to developing countries by the Western nations

• The social factors responsible for the global emergency of tuberculosis—poverty, injustice, and conflict—will be more seriously considered and addressed

Priorities for research

The dilemma facing tuberculosis researchers and funding agencies is whether to give priority to operations research to determine the most effective ways of using the available control measures or to focus on basic research into new diagnostic tests, vaccines, and treatment regimens. Despite inadequate funding resources, much effort is being devoted to both approaches, involving a multidisciplinary approach from diverse disciplines such as molecular biology, social anthropology, and health economics.

Diagnostic tests

The traditional diagnostic tools, apart from a thorough clinical examination, are chest radiology, which is sensitive but non-specific, and sputum microscopy, which is specific but of limited sensitivity. Microscopy has two advantages: it detects open, infectious cases, and it does so rapidly. Culture is more sensitive, but several weeks elapse before the diagnosis is made and there is an even longer wait for the results of drug susceptibility tests. Automated systems for more rapidly detecting growth of M tuberculosis are available, but their cost and complexity restrict their use to major centres.

There is therefore a need for a simple, rapid, and specific test for tuberculosis. After an extensive but disappointing search for a serological test, attention has largely turned to nucleic acid technology, particularly the polymerase chain reaction (PCR) and related techniques.⁶ In these techniques, the replication of the DNA chain that occurs in dividing cells is induced in vitro by the addition of the requisite enzymes and building blocks (ribonucleotides) and short strands of DNA specific for M tuberculosis (primers). These primers bind to specific complementary parts of the DNA, and replication then proceeds from that point. The amplification cycle is repeated many times, and in a two hour period just a few DNA chains may be amplified to millions and are thus readily detectable. Use of such tests, some of which are available commercially,⁶ has met with several problems including cross contamination, lower than expected sensitivity, considerable variation between laboratories in collaborative trials, and false positive results.⁷ This is a rapidly developing subject, and, while these problems are being addressed, other modifications of the technique such as the ligase chain reaction (LCR) are being tried.

Nucleic acid technology has also been used to “fingerprint” strains of M tuberculosis for epidemiological purposes.⁸ The usual technique, restriction fragment length polymorphism (RFLP), is based on the ability of certain enzymes (endonucleases) to cut the DNA chains extracted from cultivated tubercle bacilli at certain specific points, resulting in a large number of fragments of varying sizes. These fragments are sorted by size on an electrophoretic gel, and the pattern of fragments produced by different strains are compared. It is also possible to do rapid fingerprinting on the DNA products of the polymerase chain reaction.⁶

There are also prospects for rapid detection of drug resistant strains based on detecting mutations in the genes determining drug susceptibility. Tests able to detect 95% of rifampicin resistant strains are available,⁹ and the recent successful sequencing of the entire genome of M tuberculosis should facilitate the development of rapid tests for other forms of drug resistance. Accordingly, it is already possible, in certain specialist centres, to make a diagnosis of tuberculosis within a few hours and, within a day or so, to establish possible sources of infection and to state with a high degree of confidence whether the strain is resistant to rifampicin. It is, however, unlikely that such technology will be available in the foreseeable future to the countries that shoulder most of the burden of tuberculosis.

Vaccines and vaccination strategies

The BCG vaccine has been used extensively since 1921, but its contribution to disease control has been limited for two reasons. Firstly, its protective efficacy varies considerably, from 0% to about 80%, in different regions.¹⁰ Secondly, when it is effective it affords a high degree of protection against the serious but usually non-infectious forms of primary tuberculosis but gives little or no protection against the post-primary forms of the disease, due to endogenous reactivation or exogenous reinfection, which are responsible for transmission of the disease.

Only those who are not infected by M tuberculosis should be vaccinated—that is, those who are tuberculin negative (Heaf grades 0 and 1 and Mantoux responses of 0-4 mm)—although infants up to 3 months old may be vaccinated without prior testing.¹¹ People with a history of BCG vaccination should be revaccinated only if they are tuberculin negative and have no apparent BCG scar. Being a living vaccine, BCG can cause complications in immunosuppressed people: it should not be given to people with symptomatic HIV infection.

Because of the disadvantages of BCG, attempts are being made to develop alternative vaccines, particularly those that can be given to people already infected to prevent post-primary disease. These attempts are facilitated by recent advances in the understanding of the immune response and the ability to clone specific antigens and to manipulate the genetic structure of mycobacteria. Approaches currently under investigation include modified ways of administering BCG, construction of genetically modified living vaccines, and the development of non-viable subunit vaccines, including naked DNA.¹²

In addition to developing new vaccines, we will need new and rapid ways of evaluating the efficacy of such vaccines in the community or there will be long delays before any such vaccines become available.

The immune response

In the absence of HIV infection or other cause of immunosuppression only about 10% of people infected by M tuberculosis develop overt tuberculosis, indicating that the immune response to this pathogen is usually good. There is evidence that the immune response in those who do develop tuberculosis is not weak but dysregulated. It is now known that helper T lymphocytes mature along two pathways, resulting in so called Th1 and Th2 cells that are distinguishable by the chemical messengers (cytokines) that they release. Regulation of these Th1-Th2 type responses may be influenced by glucocorticoids and dehydroepiandrosterone.¹³ The protective immune response in tuberculosis is mediated by Th1 cells, but a Th2 or a mixed Th1-Th2 response renders cells very sensitive to killing by the cytokine tumour necrosis factor (TNFα), thereby inducing the gross tissue destruction characteristic of progressive tuberculosis (fig 2).¹⁴

Figure 2.

Effect of maturation of T helper lymphocytes into Th1 and Th2 types on immune response in tuberculosis. Immunity requires a Th1 response, and tumour necrosis factor (TNFα) is needed as an additional, macrophage activating factor. With a mixed Th1-Th2 cytokine response, TNFα becomes toxic and mediates the gross tissue destruction characteristic of active, progressive tuberculosis. Stress and corticosteroids tend to drive newly recruited T cells towards Th2 responses

The factors determining the pattern of T cell maturation are not fully understood, but it seems that the immune system may be programmed to respond in a certain way by past experience, including exposure to the antigens of mycobacteria present in the environment. In addition, stress and glucocorticoids tend to drive newly recruited T cells towards Th2,¹³ while these steroids also directly inhibit the ability of macrophages to limit the growth of M tuberculosis.¹⁵ Vaccinating a person whose immune response is inappropriately programmed may therefore not lead to protection, and this may explain why BCG fails to afford protection in some regions and may even predispose people to the disease.¹⁰ Accordingly, the vaccine of the future may be one that contains not only specific antigens but also adjuvants able to modulate the immune response.¹⁶

Treatments

Modern short course treatment regimens based on rifampicin and isoniazid are highly effective. However, if resistance to these drugs develops, prolonged treatment with less effective, and often more toxic and expensive, drugs is required and the mortality is high, especially in patients infected with HIV. There are two main approaches towards developing new drugs: firstly, to screen many products randomly against M tuberculosis or a related rapidly growing mycobacterium or, secondly, to study metabolic pathways unique to this bacillus in the hope of producing “designer drugs.” The success of either strategy cannot be predicted, so it is important to prevent the emergence and transmission of drug resistance by effective use of available drug treatment and appropriate control measures.

An alternative approach to treatment is enhancing protective immune responses. Various agents have been and are being investigated, but, to date, evidence of efficacy is largely anecdotal.

Effective use of available control measures

The paradox of the global emergency of tuberculosis despite the availability of highly effective treatment reveals gross deficiencies in its deployment.² The principal cause of failed disease control is “non-compliance,” for which the patient is usually blamed but which is usually the fault of the healthcare provider. The WHO has advocated the directly observed therapy short course (DOTS) strategy,¹⁷ a strategy that will work only if attention is given to the many social, cultural, and ethnic factors that affect the use of tuberculosis services by the community.¹⁷

The conquest of this disease will not be achieved by medical advances alone. As the powerful tools for treatment and control that are currently available have made little impact, we cannot expect any new ones to do so unless there is global political willingness to address the gross inequities of wealth and healthcare provision in society.² While tuberculosis can affect anyone, the greatest burden of disease falls on the poor. In November 1997 the UK Department for International Development published its white paper Eliminating World Poverty: A Challenge for the 21st Century.¹⁸ This paper champions the cause of the world’s poor and sets precedence to the Western nations to target their aid packages for developing countries towards achieving reductions in world poverty. If this is successful, a parallel reduction in the incidence of tuberculosis can be anticipated. Indeed, the director of the WHO global tuberculosis programme has stated: “The growing tuberculosis epidemic is no longer an emergency only for those who care about health, but for those who care about justice.”¹