Hopewell: I'm struck by how the situation in Ravensmead and Uitsig compares with that in Alaska in the late 1950s and early 1960s, where there was a prevalence of tuberculosis infection in five-year olds of nearly 100%. Mass campaigns promoting chemotherapy for those with disease and preventive therapy for those with tuberculous infection but not the disease resulted in a rapid decrease in prevalence, and by the late 1960s the prevalence of infection was comparable to the low prevalence in the USA. It seems to me that now, although tuberculosis control programmes have remained in place, they are not of the same intensity as those used to bring tuberculosis under control. This suggests that social and biological factors can be overcome by a decent tuberculosis control programme, and that once they are overcome they can be maintained without the same intensity of effort. I realize that the magnitude is much greater in the areas you have studied, but have you considered that they can be used as demonstration areas to test this approach?

'Donald: The areas in question have now been identified by the local government as being demonstration areas, and the World Health Organization DOTS (directly observed therapy short-course) approach is being implemented there precisely as a demonstration. Preventive therapy is not a prominent part ofit because we haven't been ensuring that the adults who are excreting organisms are actually completing their treatment, but I would like to see a more aggressive chemoprophylactic programme targeted at certain groups. In the Canadian study there was a relatively small group amongst the Inuit who could be identified and intensively treated, but for a whole population chemotherapy would almost be like sprinkling salt over a salad.

'Hopewell: You could administer it in drinking water.

Donald: That is probably not a realistic option, but certainly a more explorative approach to chemoprophylaxis may work; for example, a combination of isoniazid and the sterilizing drugs rifampicin and pyrazinamide given to close contacts of sputum smear-positive patients for a period of two to three months may deliver considerable benefits. However, it is unlikely that in our current political climate tuberculosis control programme administration will be considered such an option.

Anderson: The age-specific incidence curves are extremely unusual for any infectious agent that I'm familiar with. Careful interpretation of these curves, which are observed in all regions of the world and are moderately independent of the intensity of transmission, is essential. What are the current hypotheses that explain these curves?

Donald: They may indicate that hormones, and variations of them in different populations, are also involved in modulating the immune response.

Fine: The striking age patterns are identical to those reported in many other populations and over a long period of time; and they have little or nothing to do with local genetics. They reflect how human populations respond to Mycobacterium tuberculosis.

van Helden: But I believe that genetics can be used to explain these age-specific curves. Our work with the mannose-binding protein, for example, shows that the risk of tuberculosis meningitis in those with mutant alleles is higher than for pulmonary tuberculosis in adults. I would like to ask Albert Beyers to comment on this with respect to the differences between children and adults.

Beyers: I will discuss our preliminary work on this in my presentation. It is not only the incidence of tuberculosis that changes with age, but also the type of disease because childhood tuberculosis and adulthood tuberculosis differ in terms of pathology. The immunopathology of adulthood pulmonary tuberculosis has previously been ascribed to the Koch phenomenon. Childhood tuberculosis occurs in a person who has not been exposed to the organism before, and adulthood tuberculosis with severe immunopathology is the result of reactivation or reinfection. We do not exclude that the Koch phenomenon contributes to the differences in pathology, but we propose that age-related factors also play a role. There are clear examples of children in high incidence communities who have second bouts of tuberculosis even though they are HIV negative and not immunosuppressed, as far as we can tell. During the second presentation with tuberculosis, children again have the childhood form of the disease. Age-related changes in the type and incidence of disease may be due to alterations in sex steroids, although this is currently purely speculative.

'Donald: I have a comment about other infectious diseases that show the same age-related variation in clinical phenomena. For example, when the Epstein—Barr virus is contracted during childhood nothing much happens, but infectious mononucleosis appears during adolescence. The same is true for chickenpox, which amounts to a few vesicles during childhood, but contracted as an adult it becomes a serious disease. It suggests that the immune system undergoes a fundamental change during the process of maturation, which perhaps makes it more aggressive towards infecting agents or intruding antigens.

Fourie: I was struck by the base rate of 3000 tuberculosis cases per 100 000 population in some of the subdistricts, and I thought when has there ever been a rate that high? I recall a report on tuberculosis at the start of the 19 th Century in the USA when the mortality rate was around 1600 per 100 000 population in New England (Stead & Dutt 1988). This means that the incidence rate would have been at least twice as high. It might be of interest to note that the highest mortality rate ever recorded was 9000 per 100 000 population in North American Indians in 1886 (Ferguson 1955, see also Daniel et al 1994). Under a good chemotherapy programme, mortality rates drop quickly, thus creating the impression that the tuberculosis epidemic had been brought under control, although transmission is probably still occurring in the survivors. We must surely be seeing a similar pattern in the Western Cape, but the question is whether the exceptionally high incidence rates are true or false. Cummins (1920)

reported the tuberculosis rates in South Africa during World War I, and showed that people of mixed race had high rates of tuberculosis morbidity and mortality. This has also been reported in the literature earlier than the 1920s. Therefore, one has to acknowledge that there is something different about this group of people. I believe that the focus should not so much be on the organism, suspecting there might be differences in virulence between strains in different geographical areas, but rather on human genetics.

In addition, one strange phenomenon in the tuberculin skin test response of Coloured groups is that they have two modes above 10 mm following Mantoux testing with 2 tuberculin units of RT23. We would normally expect in areas of high prevalence a mode at 14 or 15 mm. However, in the Western Cape Coloureds, there is a second mode situated between 20 and 22 mm (Fourie 1983). It might be worthwhile investigating this because it suggests that there might be two subgroups, raising the question of differences in susceptibility for disease.

Why do I think the rates in Ravensmead are false? Because the degree of variation is unusual for such a small area. Also, after the introduction of the new tuberculosis registration system by standardized case definition in South Africa at the end of 1995, the number of reported tuberculosis cases in the Western Cape reduced by 30%. We need to look at the statistics of the epidemic in the areas you have studied in terms of the new case definition, and then make statements on the epidemiology.

I would also like to ask, why haven't we seen any impact on the epidemic in that area despite the intensive efforts of a dedicated team over many years? Is it because cases that are being fed back into the community are propagating the transmission of the disease? Is it that the diagnosis is good and the initial treatment phase is appropriate, but the cases are not seen through to the end?

van Melden: If it is true that the majority of cases are not due to recent transmission, then if there is a high infection rate the majority of cases will reactivate. There will be a huge case load despite the treatment of active people, and this will persist for some time because the disease is so endemic in that population. Also, in multiple-case households about half of the multiple cases are infected outside the home, so in other words these people are spending a large amount of exposure time outside the home. We don't know where that exposure time occurs; it could be in the neighbouring suburb because the suburb is continuous with another suburb on at least one side. While a criticism of the work done in this suburb is that it is an isolated component of the whole, the fact of the matter is that we cannot look at a larger area.

Bateman: I would like to comment briefly about the Cape epidemic and the question of over-reporting. My first comment relates to a study of the criteria for diagnosis, used in cases notified to the State authorities, and comparing diagnoses made at a major hospital in the Western Cape with those made in equivalent hospitals in other regions. This study confirmed that the diagnosis was based on sound criteria, usually bacteriological, in the Western Cape, whereas in other regions a large proportion was diagnosed according to X-ray abnormalities alone. This suggests that the high notification figures for the Western Cape are correct. Differences in reporting procedures at local clinics may also play a role. For example, in the Ravensmead/Uitsig area, compared to other regions of the Western Cape, there is a higher paediatric notification rate. This is probably the result of an enthusiastic team in that area that has a special interest in the paediatric population. The ratio ofsmear-positive to smear-negative diagnosed active disease in this area is considerably higher than it is elsewhere.

My second comment is that we must not expect all the answers concerning differing incidences in different population groups to come from genetics. There has been a dramatic upturn in infection rates in African Blacks in this region. One reason for this is the relatively recent influx of Africans to the region. However, these recent arrivals are displaying the same increase in tuberculosis as is found in the mixed race group that has traditionally populated this region. This suggests that transmission and environment are key factors.

'Donald: We can dispute the accuracy of a certain percentage of diagnoses, but the most important points from a genetic point of view are those concerning the varying sex ratios and the age influence upon disease.

Tine: It appears that the incidence of tuberculosis is as high in the Ravensmead/ Uitsig area as it is anywhere in the world today, but it does not seem to be as high as the incidence in Europe 100 years ago. Peter Donald talked of mortality rates of 1% per population, and we have been told that the incidence of disease is up to 3000 per 100 000 population — but what is the denominator?

Donald: It is about 900 per enumerator subdistrict.

Tine: The standard error based on that denominator would be rather large. If you look at the national statistics anywhere, for example in the UK, over the last century, you will find some areas that have higher and some areas that have lower disease incidences than the national average. There is always a danger of focusing on the area with the highest disease incidence, particularly if it implies a small denominator.

'Bellamy: Worldwide, only 10% of people who are infected with M. tuberculosis will develop disease. I'm sure that many of us here have been exposed to the bacterium and are probably infected, but have not developed tuberculosis. In contrast, at least in Ravensmead and Uitsig, the risk of developing disease following infection is probably more like 30% or even 50%. This extremely high rate suggests that there is something different in this region, either a particularly virulent strain of M. tuberculosis, an unusually genetically susceptible population or some environmental factor. If it does involve the environment, then it must be specific environmental factors rather than just generally poor living conditions, because there are many populations in the world that live in equally bad conditions but have a much lower rate of tuberculosis.

Fine: I'm surprised by the way the phrase 'the tuberculosis epidemic' is being widely used in the field of tuberculosis nowadays, given that many infectious diseases have declined over the last 200 years in developed countries and we don't talk about 'the polio epidemic', 'the diphtheria epidemic' or 'the typhoid epidemic'. We understand the dynamics of acute infectious disease epidemics, and how they decline because of exhaustion of susceptibles and increase when the susceptibles are replenished. The phrase is simplistic, and I'm convinced that the situation is much more complicated for tuberculosis. It can probably be traced back to Grigg (1958), who implied that when tuberculosis was introduced into populations its incidence first increased and then declined in an almost predictable manner, and he attributed the decline largely to genetic selection. Many people have denied that genetic selection explains the incredible declines of tuberculosis observed all over the world, and there is much evidence that they have been associated with improvements in socioeconomic conditions and treatment, etc. I wonder whether a population geneticist has ever looked at the rates of tuberculosis decline in Europe, for example, and asked 'what would I have to assume of a gene to explain that?'

Donald: I tried to illustrate that there is a debate on this, and I quoted Grigg (1958) as one of the people who has influenced our thinking on this. In his defence, tuberculosis is a different disease from the other infectious diseases in which wave patterns are clearly seen. With tuberculosis one has to stand back much further from the disease than in the case of measles, for example, in which the wave phenomenon can be seen within a matter of months. Youmans (1979) demonstrated mathematically that the elimination of susceptibles cannot be the only factor, and he suggested that suburbanization plays a role.

Bellamy: Whether or not changes in the incidence of tuberculosis can be affected by the weeding out of tuberculosis susceptibility genes may depend on the population being studied. One of the classic examples where genes were used as an excuse to justify appalling social conditions was in South Africa during the time when the Blacks worked in mines, when there were high incidences of tuberculosis. Cummins (1929), for example, said that this was due to their racial susceptibility as opposed to the dire conditions they were working in. However, there is evidence to support his virgin soil hypothesis. For example, if we look at the Qu'Appelle Indians when tuberculosis was introduced in 1890, initially tuberculosis caused an annual mortality rate of 10% of the population, and within 40 years the mortality rate had decreased to 0.2% of the population. Half of the families were wiped out, and it is believed that this dramatic fall in tuberculosis death rates was due to a weaning out of tuberculosis susceptibility genes (Motulsky 1960).


Cummins SL 1920 Tuberculosis in primitive tribes and its bearing on tuberculosis in civilised communities. Int J Publ Health 1:10-171 Cummins SL1929 Virgin 'soil' — and after. A working conception of tuberculosis in children, adolescents and aborigines. Br Med J 2:39-41 Daniel TM Bates JH, Downes KA 1994 History of tuberculosis. In: Bloom BR (ed)

Tuberculosis: pathogenesis, protection and control. ASM Press, Washington DC, p 13-24 Ferguson RG 1955 Studies in tuberculosis. University of Toronto Press, Toronto Fourie PB 1983 Patterns of tuberculin hypersensitivity in South Africa. Tubercle 64:167-179 Grigg ERN 1958 The arcana of tuberculosis. Am Rev Tuberc Pulm Dis 78:151-172, 426-453, 583-596

Motulsky AG 1960 Metabolic polymorphisms and the role of infectious diseases in human evolution. Hum Biol 32:28-62 Stead WW, Dutt AK 1988 Epidemiology and host factors. In: Schlossberg D (ed) Tuberculosis.

3rd Edn. Springer Verlag, New York, p 1-15 Youmans GP 1979 Tuberculosis. WB Saunders, Philadelphia, p 356-369

Novartis 217: Genetics and Tuberculosis. Copyright © 1998 John Wiley & Sons Ltd Print ISBN 0-471-98261-X elSBN 0-470-84652-6

Using conventional and molecular epidemiological analyses to target tuberculosis control interventions in a low incidence area

Philip C. Hopewell

'Division of Pulmonary and Critical Care Medicine, San Francisco General Hospital, 1001 Potrero Avenue, Room 5K1, San Francisco, CA 94110, USA

Abstract. To consolidate the gains made in controlling tuberculosis in industrialized countries in the past five years, we must analyse carefully the epidemiology of the disease and the effectiveness of various control interventions. In San Francisco we have performed conventional and molecular epidemiological analyses that have shown that there are in essence two parallel epidemiological patterns, one in the foreign-born population and the second among US-born persons, with little interaction between them. Most tuberculosis in the foreign-born population is a result of endogenous reactivation of latent infection, whereas recent infection with rapid progression to illness is a more frequent course in US-born cases. Among the US-born cases specific risk factors — homelessness, HIV infection, drug abuse—are highly prevalent. Although there has been a progressive reduction in the number of cases and in the proportion resulting from recent infection in San Francisco in the past five years, there continues to be a high proportion of cases that result from recent infection among US-born persons. These findings suggest that existing control interventions should be tailored to specific target groups and that new interventions are needed to provide for increased efficacy.

1998 Genetics and tuberculosis. Wiley, Chichester (Novartis Foundation Symposium 217) p42-56

The increase in tuberculosis cases and case rates in the US in the late 1980s and early 1990s was a phenomenon previously unseen in an industrialized country (Fig. 1A). With the exception of minor deviations and statistical artefacts, tuberculosis has been decreasing consistently during this century (Fig. 1B), and, between 1953 and 1984, the annual rate of decline averaged at 5—6% (Centers for Disease Control and Prevention 1997). Why should there have been an increase in the incidence of a disease for which both curative and preventive measures are well known and widely available? Moreover, why should the resurgence occur at all

FIG. 1. (A) Reported annual tuberculosis cases and case rates in the US 1953—1996. Data from Centers for Disease Control and Prevention (1997). (B) Reported annual tuberculosis case and death rates in the US 1900—1980. (Modified from Health Education and Welfare Publication No. Centres of Disease Control 78—8360.)

FIG. 1. (A) Reported annual tuberculosis cases and case rates in the US 1953—1996. Data from Centers for Disease Control and Prevention (1997). (B) Reported annual tuberculosis case and death rates in the US 1900—1980. (Modified from Health Education and Welfare Publication No. Centres of Disease Control 78—8360.)

in a country that in 1990 spent more than US$2700 per person on health care (compared with, for example, Tanzania which spent US$4 per person in 1990) (Ad Hoc Committee on Health Research Relating to Future Intervention Options 1996)?

The answers to these questions are not simple and require understanding of a complex interplay of social, biological and political factors that came into unfortunate alignment, most notably in New York City but also in other urban centres, such as San Francisco (Brudney & Dobkin 1991, Cantwell et al 1994). However, the coming together of these factors does not represent a unique occurrence, and the resulting damage to tuberculosis control could occur again in the US or in other countries. Many of the conditions that prevailed in the US in the 1980s and early 1990s are equally prevalent today. The socioeconomic status of the most at-risk segment of the population has not improved, HIV infection is still frequent, and many persons are still in shelters, correctional facilities and hospitals. What has changed is the attention paid to tuberculosis control, and this attention has been translated to substantially increased funding for control interventions (Fig. 2).

Having now re-established control of tuberculosis, it is essential that we evaluate as carefully and precisely as possible the epidemiology of tuberculosis in the US in order to be most efficient in further control interventions, to use the current resources to identify targeted strategies, to indicate where new approaches are needed and to develop new tools to incorporate into the control effort.

How To Bolster Your Immune System

How To Bolster Your Immune System

All Natural Immune Boosters Proven To Fight Infection, Disease And More. Discover A Natural, Safe Effective Way To Boost Your Immune System Using Ingredients From Your Kitchen Cupboard. The only common sense, no holds barred guide to hit the market today no gimmicks, no pills, just old fashioned common sense remedies to cure colds, influenza, viral infections and more.

Get My Free Audio Book

Post a comment