Resurgence of tuberculosis and the impact of HIV infection by ive16829


									Resurgence of tuberculosis and the impact of
HIV infection

Judith R Glynn
Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine,
London, UK

                               Tuberculosis is increasing in many countries. In some areas the major influences
                               on tuberculosis trends are the traditional ones: poverty, failures in the
                               treatment system, and immigration. In others, and increasingly, the HIV
                               epidemic is having a huge impact. HIV infection increases the risk of tuberculosis
                               approximately 7-fold, though this may vary with the stage of the HIV epidemic,
                               the prevalence of tuberculosis, and the age groups considered. Dually-infected
                               individuals develop tuberculous disease at a rate of 5-10% per year. HIV also
                               increases the risk of disease following recent infection, which makes a major
                               contribution to the tuberculosis burden in some settings. HIV-infected
                               individuals may transmit Mycobacterium tuberculosis less than do HIV-negative
                               individuals, but the extra cases will add to the transmission overall, and
                               evidence of HIV-attributable increases in the annual risk of infection is
                               beginning to be seen.

                               In 1993, the WHO took the unprecedented step of declaring
                               tuberculosis to be a 'Global Emergency'. This action was stimulated
                               both by the magnitude of the tuberculosis problem - an estimated 8
                               million cases and 3 million deaths each year1 - and by recent changes in
                               tuberculosis incidence. In many countries tuberculosis case rates, which
                               had been declining for decades, were reported to be increasing, or at
                               least to have stopped decreasing. That this was happening worldwide
                               was remarkable, not least because the underlying causes of the increase
                               varied from place to place.
                                 In the US, there was an estimated 13% increase in the case rate
                               between 1985 and 19922. In many countries of Western Europe case
                               rates stopped falling or started to rise in the mid to late 1980s3.
       Correspondence to
                               Dramatic increases have been reported from Eastern Europe since
Dr J R Glynn, Department       19894"6. Estonia recorded an 80% increase in incidence between 1982
of Infectious and Tropical     and 19865. Rises in childhood tuberculosis are particularly worrying6-7,
 Diseases, London School
                               as these are likely to reflect increased transmission of infection, not just
  of Hygiene and Tropical
 Medicine, Keppel Street
                               an increase in re-activation disease. In Russia, incidence in children
   London WC1E 7HT, UK         increased by almost 80% between 1989 and 19957. Most countries in

British Medical Bulletin 1998,54 (No 3) 579-593                                              CThe British Council 1998
Resurgent/emergent infectious diseases

                         Africa, for which sufficient data are available, show increases since the
                         mid 1980s, some of more than 10% per year8. In Asia, rates increased
                         from 1985 to 19909 and further increases are predicted1.

Increases in data or in disease?
                         Three sorts of data are used to estimate trends in tuberculosis in a
                         population: (i) incidence rates of all or selected types of disease; (ii)
                         mortality rates; and (iii) estimates of the annual risk of infection. Some
                         forms of tuberculosis are notoriously difficult to diagnose accurately,
                         and good diagnostic facilities are often lacking. There was no change in
                         case definition that would have accounted for the trends, but the com-
                         pleteness of case finding and notification may well have improved in
                         some countries. The proportion of cases found is very difficult to
                         estimate. Many tuberculosis programmes have used a 'rule', based on
                         estimates from survey data10, that for each 1% annual risk of infection
                         there should be about 50 smear positive tuberculosis cases per 100 000
                         population. Using this rule, case finding is often thought to be poor,
                         perhaps as low as 30% in some areas11-12, so that even the WHO goal
                         for case finding is only 70% n . This simple rule has been criticized and
                         is no longer valid13. The annual risk of infection is subject to con-
                         siderable error in measurement, particularly in the presence of neonatal
                         BCG vaccination and exposure to non-tuberculous mycobacteria13.
                         Furthermore, the presence of HTV infection, which increases the
                         progression from infection to disease, changes the relationship between
                         the annual risk of infection and disease incidence.
                           'Case rates' may refer to both new and relapse cases and are often crude
                         rates, not adjusted for the age and sex structure of the population.
                         Demographic changes may lead to changes in crude rates, although the
                         age-specific rates are stable14. Interactions with age are complex, older
                         individuals having a higher risk of disease following infection than do
                         younger individuals15. As the annual risk of infection falls, the average age
                         at infection rises, which could cause a paradoxical increase in disease.
                           Tuberculosis mortality rates reflect the efficiency of diagnosis and
                         treatment as well as the incidence of tuberculosis, so also need to be
                         interpreted cautiously. Many deaths from tuberculosis may be
                         misdiagnosed, particularly in HIV-positive patients16.
                           Although it is difficult to measure, much of the recorded increase
                         probably reflects real changes in the incidence of tuberculosis. The
                         increase is due both to 'traditional' factors which have been influencing
                         tuberculosis trends for decades (immigration, poverty and failures in the
                         treatment system), and to HTV.

580                                                                        British Medical Bulletin 1998,54 (No 3)
                                                                                        Tuberculosis and HIV

'Traditional' influences on tuberculosis trends

                              In some countries, much of the rise in tuberculosis cases can be
                              attributed to recent immigration. In the US, foreign-born cases
                              accounted for 60% of the increase in tuberculosis seen between 1986
                              and 19922. In The Netherlands, the incidence of tuberculosis in the
                              native-born population has continued to fall, though the total incidence
                              rate increased since 19873. The extent of spread of tuberculosis from the
                              immigrant communities to the general population has been studied in
                              The Netherlands using DNA fingerprinting techniques. Much of the
                              transmission occurred within each immigrant community, though the
                              extent of this varied by nationality. Among tuberculosis cases in Dutch
                              nationals, 17% (about half of those thought to be due to recent
                              infection) were attributed to transmission from non-Dutch nationals17.

Poverty and living conditions

                              Tuberculosis has traditionally been linked to poverty, though the relative
                              roles of the different aspects of poverty which may be responsible, for
                              example nutrition and overcrowding, are not well defined. Increases in
                              inner city poverty and homelessness have been blamed for some of the
                              increases in tuberculosis seen in high income countries18-19. Sometimes
                              this effect is direct, with transmission occurring in shelters for the
                              homeless19*20. Poverty is also a factor in the high rates of tuberculosis
                              seen in immigrants18. The displacement of populations due to war or
                              natural disasters leads to crowding and malnutrition which are likely to
                              foster tuberculosis21. High rates of tuberculosis, often in association
                              with HTV, have also been found in prisons14-22-23.

Collapse of treatment systems, and drug resistance

                              The duration of tuberculosis treatments ('short course' regimens are at
                              least 6 months long) bring particular problems to both health services
                              and patients. Treatment failure can have many causes: inappropriate
                              regimens may be prescribed; drugs may be unavailable; the patient may
                              have insufficient access to the health service; or may default even if drugs
                              are available. Sub-optimal prescribing is a particular problem where
                              tuberculosis services are not centralised, for example in populations
                              where many patients are treated privately. Failures in drug supply can
                              occur in any underfunded or neglected system. Failures of patients to

British Medical Bulletin 1998,54 (No 3)                                                                   581
Resurgent/emergent infectious diseases

                         collect or take drugs may be linked to costs (direct or indirect,
                         particularly if travel is involved) or lack of understanding of the
                         importance of continuing treatment beyond the acute phase of illness.
                         Drop out rates can be very high: of a group of patients in Harlem, New
                         York, many of whom were drug users or alcoholics, 89% failed to
                         complete treatment24. In a study in Madras, only 42% of smear positive
                         patients on short course therapy completed at least 80% of their
                         treatment and, of these, 22% remained sputum positive". The inevitable
                         consequence of these failures is a rise in drug resistance, first acquired
                         resistance, and then, with subsequent spread, primary resistance.
                         Published rates of drug resistance have recently been reviewed26.
                         Multidrug resistance was uncommon in new patients (less than 1% in
                         most studies), but was much more common (over 10% in some areas)
                         among patients with a history of previous treatment26. The impact of
                         drug resistance on tuberculosis is covered by Eltringham in this issue.

                         Whereas the factors discussed above have been influencing tuberculosis
                         trends for some time, and we can predict their impacts, even if not
                         always deal effectively with them, the full impact of HIV has yet to be
                         felt. There is no doubt that HTV infection increases the risk of
                         developing disease among individuals infected with M. tuberculosis.
                         HTV infection may not increase the risk of infection in those exposed23-27,
                         though this is difficult to study since delayed type hypersensitivity
                         reactions to tuberculin are lost as HTV infection progresses, and testing
                         for anergy can give inconsistent results28.

The association between HIV and tuberculosis

                         The relative risk of tuberculosis disease in individuals infected with HIV,
                         compared to those not infected, has frequently been studied. Cohort
                         studies, and case-control studies that have used community controls and
                         have adjusted at least for age and sex are shown in Figure 1. Many other
                         studies of the relationship between HTV and tuberculosis have been
                         reported, but have used inappropriate comparison groups, or have made
                         no adjustment for confounders. For case-control studies, it is difficult to
                         define suitable controls. Tuberculosis cases are usually identified in
                         hospital settings, but hospital controls4142 are rarely appropriate since
                         HTV is associated with many different diseases, and referral patterns
                         vary for different conditions. Blood donors have been used43"45, but are
                         unlikely to be representative of the general population as they will be

582                                                                       Bnhsh Medical Bulletin 1998,54 (No 3)
               Cohort studies ol women of chlldbearlna aae

                     Rwanda 88/93 Age, parity, urban (29)

                     Rwanda 88/90. Age, parity, urban (30)

                        Zaire 86/90. Age, parity, urban (31)

                        Zaire 86/90 Age, parity, urban (31)

                   Cohort studies in Intravenous drug users

                         USA 88/90, tuberculin positive (32)

                                     Netherlands 86/96 (33)

               Case-control studies with population controls
                                                                                                                        All TB
                           Tanzania 9 1 . Age, sex, area (34)
                                                                                                                        Pulmonary T 6 •
                        Guinea Bissau 92/93 Age, sex (35)                                                               (smear and/or.culture
                                                                                                                        positive) ; ;;;;
                           Kenya 89/90. Age, sex, area (36)

                          Malawi 88/89 Age, sex, area (37)

                          Malawi 90/95. Age, sex, area (38)
                                                            0.1                                 10                100               1,000
                                                                                           Relative risk

Fig. 1 Relative risk of TB in HIV-positive and HIV-negative individuals. Rate ratios are presented for the cohort studies and odds ratios for the case
control studies. In addition t o those shown, a cohort study among tuberculin positive drug users in New York found 7 cases of tuberculosis in 49        Q.
HIV-positives (7.9/100 person years at risk) and none in 62 HIV negatives 59 In Nairobi, a cohort study of commercial sex workers found tuberculosis     X
in 49/587 HIV positives (3.4/100 person years at risk) and in 0/132 HlV-negatives 40 .
Resurgent/emergent infectious diseases

                         healthy and may have been preselected, and in some countries are paid
                         to donate. Health care workers46 are also unrepresentative.
                           Even among the selected studies, there is a large variation in estimates
                         of the relative risk. Most are consistent with a relative risk of around 7,
                         but the relative risk may not be the same in different situations. The
                         relative risk should increase as the HTV epidemic progresses, as a higher
                         proportion of HTV-infected individuals develop more advanced levels of
                         immunosuppression. Several studies have found variations with age, with
                         higher relative risks in younger adults than in older adults34'41'43""*5'47. The
                         relative risk may vary with the incidence and prevalence of M.
                         tuberculosis infection in the population, since the relative risks of primary,
                         re-activation and re-infection disease may not be the same. The relative
                         importance of these different types of disease may explain any variations
                         found with age, since disease in a young adult is more likely to be due to
                         recent infection than is disease in an older adult.

Incidence of tuberculosis in HlV-infected individuals: re-activation or re-infection?

                         The absolute incidence of tuberculosis in HIV-positive individuals
                         depends on the prevalence of previous M. tuberculosis infection, the
                         current risks of infection, the extent of immunosuppression48 and the use
                         of antimicrobial (usually isoniazid) prophylaxis. In studies of incidence,
                         tuberculin testing, usually taking a cut-off of ^ 5 mm, has been used to
                         distinguish those thought to have previous infection. The extent of
                         immunosuppression and the current risk of infection depend on the
                         patient groups included and the setting of the study. Studies excluding
                         sicker patients and those with a history of previous tuberculosis, and
                         studies with stricter criteria for diagnosing tuberculosis, will tend to find
                         lower incidences of disease. Table 1 shows estimates of tuberculosis
                         incidence for 12 studies in which data were available for HIV-infected
                         patients with known previous positive tuberculin tests who had not
                         received isoniazid prophylaxis. The settings, exclusion and diagnostic
                         criteria are shown. The rates of tuberculous disease of 5-10% per year
                         in dually infected persons compare with an estimated lifetime risk of
                         10% in those without HIV infection58.
                           The appropriateness of the 5 mm induration cut-off as an indicator of
                         M tuberculosis infection will vary from setting to setting, its specificity
                         being lower in areas with higher rates of infection with environmental
                         mycobacteria, and those with widescale BCG use. This, and the multiple
                         exclusion criteria which will have resulted in a cohort of relatively newly
                         HTV infected patients, may explain the low incidence found in the recent
                         Ugandan study57 (Table 1).

584                                                                         British Medical Bulletin 1998,54 (No 3)
Table 1 Incidence of tuberculosis (TB) in HIV-positive individuals with tuberculin reactions of at least 5 mm who had not received isoniazid

 Ref Place               Year    Patient group                 Median CD4     Exclusion criteria                  All culture    Estimated length       Risk    Rate/100 person
                                                               count x10*/1                                       confirmed?    of follow-up (years)    n/n     years at risk (95 % Cl)

 49 Port-au-Prince       86-92   Symptom-free, newly                          Previous TB, abnormal                   No                24              6/25     10 0(3 6-21 4)
    Haiti                        diagnosed                                    chest X-ray, abnormal LFTs
 39 New York,            85-88   IVDUs in methadone                           ?None                                   Yes               1 8            7/36b     10 8b (4 3-22 2)
    USA                          maintenance programme*
 32 New York,            88-90   IVDUs in methadone                560        Documented previous TB                  No                1 7             4/25      9 7 (2 6-24 7)
    USA                          maintenance programme
 50 Hartford, USA        84-92   IVDUs seen as in or                          ? None                                  No                38              2/18      2 9(0 4-10 6)
 51 6 cities, USA       88-94    Mixed 23% IVDUs No or                        TB in last 12 months, acute             No                73 8            7/34      4 5 (1 6-9 7)
                                 < 6 months isoniazid                         pulmonary disease, AIDS
 52, Madrid,             85-94   Referred patients,                648        Previous or active TB                   No                28             24/84     10 4 c (6 7-15 5)
 53 Spain                        90% IVDUs                                                                                              49             43/92      9 4(6 8-12 7)
 54 Madrid,              89-92   Newly diagnosed, refused,        469         Previous or active TB                   No                1 7             7/26     16 2d (6 5-33 5)
    Barcelona,                   or discontinued isoniazid       (mean)
 48, Italy               90-93   In and outpatients,               380        TB in last 18 months                    Yes               1 4            15/197     5 4 (3 0-9 0)
 55                              73% IVDUs
 40 Nairobi,             89-92   Commercial sex workers                       Previous TB                             No                24             11/69      6 7 (3 3-11 9)
 56* Nairobi,            92-94   Commercial sex workers            385        Previous TB, possible current TB,       No                1 8            10/69      8 0(3 9-14 8)
        Kenya                    and clinic attenders                         abnormal LFTs, pregnant, life-
                                                                              threatening intercurrent illness
 57 Kampala,             92-95   Clinic attenders                             Previous TB, active TB, white           No                1 3            21/464     3 4 (2 1-5 2)
    Uganda                                                                    cell count < 3000/mm3, Hb < 80 g/l,
                                                                              abnormal LFTs, pregnancy,
                                                                              advanced HIV disease, major
                                                                              underlying illness
 f      Lusaka,          92-96   Clinic attenders, referrals                  Previous TB, possible current TB,       No                1 6             9/60      9 2 (4 2-17 4)
        Zambia                   from blood bank and                          abnormal LFTs, pregnancy,
                                 voluntary testing centres                    life-threatening illness

     IVDU = intravenous drug user, LFT = liver function test Hb = haemoglobin Cl = confidence interval,
     •Tuberculin reaction £10 mm, 'Without isoniazid Rate estimated assuming length of follow up same as for whole group,
     'Same rate for culture confirmed, dlncludes new converters,
     •And upbublished data, 'Unpublished data from Zambart Project (isoniazid prophylaxis tnal)
Resurgent/emergent infectious diseases

Table 2 Tuberculosis incidence in HIV-positive |patients who did not receive isoniazid, by initial tuberculin reaction
 Ref   Setting                                             Patient group                         Risk nln       Rate/100 person
                                                                                                                years at risk

 48    23 hospitals in Italy 1990-93 In and outpatients,   Tuberculin positive                   15/197         5 4 (3 0-9 0)
       73% IVDUs, 73% male, mean age 29                    Anergic                               62/1649        3 0 (2 3-3 9)
                                                           Tuberculin negative. non anergic      6/849          0 45(0 16-0 97)

 52    Madrid, Spain 1985-89 Patients referred             Tuberculin positive                   24/84          10 4(6 7-15 5)
       to one hospital (including referrals                Anergic                               20/112         12 4(7 6-19 2)
       from prison hospital), 78% IVDUs,                   Tuberculin negative, non anergic      19/151         5 4 (3 3-8 4)
        73% male, mean age 27

 In both studies, initial CD4 counts were similar in the tuberculin positive and tuberculin negative, non anergic groups, and much
 lower in the anergic patients

                               Tuberculin tests may be negative because the cut-off is too high,
                             because the individual is not infected with M. tuberculosis, or because
                             of a failure of the delayed type hypersensitivity response (anergy). Some
                             studies have used skin tests with additional antigens in order to
                             distinguish anergic from tuberculin negative, non anergic individuals.
                             Incidence rates of tuberculosis in HIV-positive individuals may be lower
                             in those who are anergic than in those who are tuberculin positive, since
                             only a proportion of the patients will be infected with M. tuberculosis,
                             or may be higher, since anergy is correlated with degree of
                             immunosuppression48'52. Among HIV-positive individuals in Italy, lower
                             rates of tuberculosis were found in anergic individuals than in those who
                             were tuberculin positive48; in Spain52 and New York32 (mainly among
                             intravenous drug users), and Uganda57, rates were similar in the two
                             groups, presumably reflecting higher background prevalence rates of M.
                             tuberculosis infection.
                               The incidence of tuberculosis in HTV-infected individuals who are
                             tuberculin negative but not anergic at the start of a study, compared to
                             that in those who are tuberculin positive, gives an indication of the
                             proportion of disease due to recent infection. Unfortunately this
                             information is rarely available (Table 2) and, given the apparent success
                             of isoniazid prophylaxis, further studies without prophylaxis are
                             unlikely. The higher rates in Spain than in Italy fit with the reported
                             prevalence of tuberculosis in the two countries. The data suggest that
                             most of the disease seen in Italy in those previously infected (tuberculin
                             positive) is a re-activation of the previous infection, whereas in Spain
                             perhaps half is due to a new infection. However, this interpretation
                             assumes that the tuberculin test is an accurate reflertion of previous
                             infection, and that the risk of becoming re-infected and developing
                             disease is the same as the risk of being infected and developing disease
                             for the first time, which is unlikely to be true15.

586                                                                                              British Medial BulleVn 1998,54 (No 3)
                                                                                          Tuberculosis and HIV

                                Incidence studies based on outbreaks with known index cases in
                              defined settings, such as HIV care facilities, suggest that infection and
                              disease rates in HIV-positive individuals can be very high and that the
                              incubation period can be very short59"61. These findings need to be
                              interpreted cautiously. Large, dramatic outbreaks will be investigated
                              and reported preferentially. Indeed 'outbreaks' where no secondary
                              cases are seen will be missed altogether although some transmission of
                              M. tuberculosis may have occurred. Occasional dramatic outbreaks
                              were also reported in the pre-HTV era6263.
                                An indirect method for estimating the proportion of disease due to
                              recent infection is provided by DNA fingerprinting studies. Groups of
                              patients with strains of M. tuberculosis with identical fingerprints, usually
                              defined by restriction fragment length polymorphism of the IS6110
                              insertion sequence, are said to be 'clustered' and the strains (or at least all
                              but one of a cluster) are assumed to have been recently transmitted. The
                              proportion of tuberculosis patients m a population who are in clusters is
                              used as an indication of the amount of recent transmission of tuberculosis.
                              Estimates of 30—40% clustering have been made in studies in the US and
                              South Africa64"68. However, the interpretation of proportion clustered as a
                              measure of recent transmission depends on the molecular clock of the
                              marker system, which is as yet not known, and on the sampling of the
                              population. Random incomplete inclusion of cases will underestimate
                              clustering, but non-random inclusion, for example contact tracing, will
                              tend to increase it69.
                                The proportion clustered has been found to be higher among HIV-
                              positive patients than among HFV-negative patients in some settings64'65,
                              and similar in others67-70. This is likely to reflect the relative exposure,
                              and perhaps susceptibility, to infection of HTV-positives and negatives,
                              but other considerations are also important. Relative clustering will be
                              influenced by the higher risk of progression to disease and shorter
                              incubation periods of HIV-positive individuals compared with HTV-
                              negatives, which apply, though not necessarily equally, to both recent
                              and past mfection. The amount of transmission of M. tuberculosis from
                              HTV-positives and HTV-negatives will also influence the clustering seen,
                              as this determines the chance of being the index case of a cluster.

The impact of HIV on tuberculosis: direct and indirect effects

                              It was initially feared that HTV-infected tuberculosis patients could be
                              more infectious than HTV-negative patients, since the lack of cell-
                              mediated immunity would encourage multiplication of the bacilli71'72.
                              On the other hand, more HTV-infected tuberculosis patients than HTV-
                              negative tuberculosis patients have extrapulmonary disease, and those

British Medical Bulletin 1998,54 (No 3)                                                                     587
Resurgent/emergent infectious diseases

                         with pulmonary disease are less likely to have cavities. Several studies
                         have measured tuberculous infection and disease in household contacts
                         of HIV-positive and negative pulmonary tuberculosis patients. In Nairobi
                         among HTV-negative contacts72, and in Kinshasa among all contacts73, the
                         patterns of tuberculin positivity in the contacts of HIV-positive and HTV-
                         negative index cases were very similar. In Lusaka, HTV-negative contacts
                         of HIV-positive patients were less likely to be tuberculin positive than
                         were similar contacts of HTV-negative patients, even after adjustment for
                         smear positivity74. In all three studies, background BCG coverage and
                         tuberculosis infection rates were high, making any difference between the
                         two sets of contacts hard to detect. No differences in disease rates among
                         contacts by HTV status of the index case were found in any of these
                         studies, but numbers were small. A study in Florida75, where BCG is not
                         routinely used and background infection rates are low, found lower risks
                         of tuberculin positivity in contacts of HTV-positive cases than of HTV-
                         negative cases, and this persisted after restricting the analysis to contacts
                         aged 5-14 years (who are unlikely to be HTV-infected) and adjusting for
                         smear positivity of the index case. A preliminary report from Rio de
                         Janeiro also found lower proportions tuberculin positive among contacts
                         of HIV-positive than of HTV-negative index cases76. Outside a study
                         setting, transmission from HIV-positive cases may be less than that from
                         HTV-negative cases, simply because undiagnosed and untreated HTV-
                         positive cases have a higher mortality rate than HTV-negative cases and so
                         have less time in which to transmit.
                             Most estimates of the impact of HIV on tuberculosis at the population
                         level have considered the direct effect77, for example in terms of the
                         population attributable fraction (PAF). PAFs of 30-40% have been
                         measured in various studies in sub-Saharan Africa38'4345'78 - sufficient to
                         account for all the observed increase in tuberculosis cases in these areas.
                         Worldwide, it is estimated that about 14% of tuberculosis cases will be
                         directly attributable to HIV by the year 2000 l .
                             Even if HlV-infected tuberculosis patients give rise to fewer secondary
                         cases than do HTV-negative cases, any increase in the number of infective
                         cases must increase M. tuberculosis infection above what it would have
                         been without them (an indirect effect). If transmission were equal, and
                         population mixing random, the annual risk of infection in the short-
                         term would be expected to increase by the same factor as the increase in
                         tuberculosis cases due to the direct effect of HTV. This factor is given by
                         the ratio of the risk of tuberculosis in the general population to the risk
                         of tuberculosis in the HTV-negative part of the population, or 1/(1-PAF)
                           '. Using this formula, PAFs of 30-40% translate into expected increases
                         in the annual risk of infection of 40-70% above what they would have
                         been in the absence of HTV. In the longer term, there would be a further
                         increase due to transmission from the extra second generation cases.

588                                                                        British Medical Bulletin 1998,54 (No 3)
                                                                                          Tuberculosis and HIV

                                 More complex models have been developed to try and estimate this
                              impact, some principally to estimate the effect of various interventions8'*"82.
                              All models are constrained by the validity of their assumptions, and these
                              seem to have (implicitly) assumed identical transmission probabilities from
                              HIV-positive and HIV-negative tuberculosis cases.
                                 It was initially unclear whether the impact of HTV would be sufficient
                              to reverse the long-term downward trend in the annual risk of infection.
                              It now appears that it may be. In the US, increases of 10% per year in
                              the risk of infection between 1987 and 1990 have been estimated from
                              the documented increase in tuberculosis cases in children under 5 years,
                              though not all of this can be attributed to HTV58. Preliminary results
                              from Kenya show that the annual risk of infection calculated from
                              tuberculin surveys in children has increased over the past decade in some
                              districts, and that this correlates with the HIV prevalence in those

                              There are multiple causes of the resurgence of tuberculosis, and the
                              predominant factors vary between different parts of the world. In sub-
                              Saharan Africa there is no doubt that HTV is of over-riding importance.
                              In Eastern Europe it is probably a combination of poverty and
                              deteriorating health care21. In parts of Western Europe immigration is
                              the key factor3. Immigration and HTV have probably contributed similar
                              amounts to the increase in the US58.
                                The HTV epidemic presents new challenges in tuberculosis control, and
                              the situation is likely to get worse. It is estimated that two-thirds of the
                              world's M. tuberculosts-iniected population live in Asia, many in areas
                              where the HTV epidemic is still in its early stages21. Apart from measures
                              to control HTV, the response depends on vaccination, prophylaxis, and
                              efficient case finding and treatment84. The only available vaccine, BCG,
                              is already widely used but is ineffective in many areas85. Prophylaxis
                              with isoniazid in dually-infected individuals has been shown to be
                              effective in most studies, at least in the short term49'57 and has been
                              recommended86, but has yet to be widely implemented. Safe use requires
                              exclusion of active tuberculosis in recipients, which is difficult on a large
                              scale; without this precaution the result will simply be an increase in
                              isoniazid resistance. Where much disease is due to recent infection, short
                              term prophylaxis in known infected individuals can have only limited
                                Currently the main emphasis in control programmes is on efficient
                              treatment following the WHO DOTS (directly observed therapy, short-
                              course) strategy. This aims to reduce transmission and hence the next

British Medical Bulletin 1998,54 (No 3)                                                                     589
Resurgent/emergent infectious diseases

                         generation of cases. While it is possible to reverse upward trends in
                         failing programmes14, and DOTS can be implemented on a mass scale12,
                         success becomes increasingly difficult in the face of high HIV

                         JRG is supported by the British Department for International
                         Development. I thank the Zambart Project for permission to include
                         unpublished data and Jo Haviland for providing details of the Nairobi
                         isoniazid prophylaxis trial.

                         1    Dolin PJ, Raviglione MC, Kochi A. Global tuberculosis incidence and mortality during
                              1990-2000. Bull World Health Organ 1994; 72 213-20
                         2   Cantwell MF, Snider Jr DE, Cauthen GM, Onorato IM. Epidemiology of tuberculosis in the
                             United States, 1985 through 1992 JAMA 1994, 272 535-9
                         3    Raviglione MC, Sudre P, Rieder HL, Spinaci S, Kochi A. Secular trends of tuberculosis in
                             Western Europe Bull World Health Organ 1993; 71: 297-306
                         4   Raviglione MC, Rieder HL, Styblo K, Khomenko AG, Esteves K, Kochi A. Tuberculosis trends
                             in Eastern Europe and the former USSR. Tubercle Lung Dis 1994; 75 400-16
                         5    Pehme L, Sillatsu H, Danilovits M, Kurve A Tuberculosis mortality in Estonia in 1996-1996
                             IntJ Tuberc Lung Dis 1997; 1 S94
                         6   Drobniewski F, Tayler E, Ignatenko N et al. Tuberculosis in Siberia 1. An epidemiological and
                             microbiological assessment. Tubercle Lung Dis \996; 77- 199-206
                         7   Aksenova V, Senkma T, Sevostianova T. Epidemiology of tuberculosis in children in the Russian
                             Federation in 1996. IntJ Tuberc Lung Dis 1997; 1 S152
                         8   Cantwell MF, Binkin NJ. Tuberculosis in sub-Saharan Africa- a regional assessment of the
                             impact of the human immunodeficiency virus and National Tuberculosis Control Program
                             quality. Tubercle Lung Dis 1996; 77: 220-5
                         9   Raviglione MC, Snider Jr DE, Kochi A. Global epidemiology of tuberculosis Morbidity and
                             mortality of a worldwide epidemic. JAMA 1995, 273 220—6
                        10   Murray CJL, Styblo K, Rouillon A. Tuberculosis in developing countries: burden, intervention
                             and cost. Bull Int Union Tuberc Lung Dis 1990; 65 6-24
                        11   Kochi A. The global tuberculosis situation and the new control strategy of the World Health
                             Organization Tubercle 1991; 72: 1-6
                        12   China Tuberculosis Control Collaboration. Results of directly observed short-course chemo-
                             therapy in 112,842 Chinese panents with smear-positive tuberculosis Lancet 1996; 347: 358-62
                        13   Rieder HL. Methodological issues in the estimation of the tuberculosis problem from tuberculin
                             surveys. Tubercle Lung Dis 1995; 76- 114-21
                        14   Kaye K, Fneden TR Tuberculosis control, the relevance of classic principles in an era of acquired
                             immune deficiency syndrome and mulndrug resistance. Epidemiol Rev 1996; 18 52-63
                        15   Vynnycky E, Fine PEM. The natural history of tuberculosis: the implications of age-dependent
                             risks of disease and the role of reinfection. Epidemiol Infect 1997; 119: 183-201
                        16   Lucas SB, De Cock KM, Hounnou A et al. Contribution of tuberculosis to slim disease in
                             Africa BMJ 1994; 308 1531-3
                        17   Borgdorff MW, Nagelkerke N, van Soohngen D, de Haas PEW, Veen J, van Embden JDA
                             Analysis of tuberculosis transmission between nationalities in The Netherlands in the period
                             1993-1995 using DNA fingerprinting Am J Epidemiol 1998; 147: 187-95

590                                                                                       British Medial Bulletin 1998.54 (No 3)
                                                                                                        Tuberculosis and HIV

                             18 Farmer P. Social inequalities and emerging infectious diseases Emerg Infect Dis 1996? 2:
                              19 Zolopa AR, Hahn JA, Gorter R et al. HTV and tuberculosis infection in San Francisco's homeless
                                 adults. Prevalence and risk factors in a representative sample. JAMA 1994, 272- 455-61
                             20 Nardell E, Mclnnis B, Thomas B, Weidhaas S. Exogenous reinfection with tuberculosis in a
                                 shelter for the homeless N Engl J Med 1986; 315: 1570-5
                             21 Kochi A. Tuberculosis: distribution, risk factors, mortality. Immunobiology 1994; 191: 325-36
                             22 Vail Mayans M, Maguire A, Mirret M, Alcaide J, Parron I, Casabona J. The spread of AIDS
                                 and the re-emergence of tuberculosis m Catalonia, Spam AIDS 1997; 11: 499-505
                             23 Ferreira MM, Ferrazoli L, Palaci M et al. Tuberculosis and HTV infection among female
                                 inmates in Sao Paulo, Brazil: a prospective cohort study. / Acquir Immune Defic Syndr Hum
                                 Retroutrol 1996, 13- 177-83
                             24 Brudney K, Dobkin J. Resurgent tuberculosis in New York City. Human immunodeficiency
                                 virus, homelessness, and the decline of tuberculosis control programs. Am Rev Resptr Dis
                                 1991; 144: 745-9
                             25 Datta M, Radhamani MP, Selvaraj R et al Critical assessment of smear-positive tuberculosis
                                 patients after chemotherapy under the district tuberculosis programme. Tubercle Lung Dis
                                 1993; 74. 180-6
                             26 Cohn DL, Busrreo F, Raviglione MC. Drug-resistant tuberculosis: review of the worldwide
                                 situation and the WHO/IUATLD global surveillance project. Cltn Infect Dis 1997; 24 S121-30
                             27 Gourevitch MN, Hartel D, Schoenbaum EE, Klein RS. Lack of association of induration size
                                 with HrV infection among drug users reacting to tuberculin Am] Respir Cnt Care Med 1996;
                                 154: 1029-33
                             28 Chin DP, Osmond D, Page-Shafer K et al. Reliability of anergy skin testing in persons with HIV
                                 infection. Am J Respir Cnt Care Med 1996; 153: 1982-4
                             29 Leroy V, Msellati P, Lepage P et al. Four years of natural history of HTV-1 infection in African
                                 women: a prospective cohort study in Kigali (Rwanda), 1988-1993. / Acquir Immune Defic
                                 Syndr Hum Retrovirol 1995; 9 415-21
                             30 Allen S, Batungwanayo J, Kerhkowske K et al. Two-year incidence of tuberculosis in cohorts of
                                 HIV-infected and umnfected urban Rwandan women. Am Rev Respir Dis 1992; 146' 1439—44
                             31 Braun MM, Badi N, Ryder RW et al. A retrospective cohort study of the risk of TB among
                                 women of childbearing age with HTV infection in Zaire. Am Rev Respir Dts 1991; 143: 501—4
                             32 Selwyn PA, Sckell B, Alcabes P, Fnedland GH, Klein RS, Schoenbaum EE. High risk of active
                                 tuberculosis in HIV-infected drug users with cutaneous anergy. JAMA 1992; 268: 504—9
                             33 Keizer ST, Langendam MW, van Deutekom H, Coutinho RA, van Ameijden EJC. Tuberculosis
                                 and HTV-infection m a cohort of drug users, Amsterdam 1986—1995. Int J Tuberc Lung Dis
                                 1997; 1: S157-8
                             34 Van den Brock J, Borgdorff MW, Pakker NG et al. HTV-1 infection as a risk factor for the
                                 development of tuberculosis: a case-control study m Tanzania. Int J Epidemiol 1993; 22:
                             35 Naucler A, Winqvist N, Dias F et al. Pulmonary tuberculosis in Guinea-Bissau: clinical and
                                 bacteriological findings, human immunodeficiency virus status and short term survival of
                                 hospitalized patients. Tubercle Lung Dis 1996; 77: 226-32
                             36 Orege PA, Fine PEM, Lucas SB et al. A case control study on human immunodeficiency virus-
                                 1 (HTV-1) infection as a risk factor for tuberculosis and leprosy in Western Kenya. Tubercle
                                 Lung Dis 1993, 74: 377-81
                             37 Ponnighaus JM, Mwanjasi LJ, Fine PEM et al. Is HTV infection a risk factor for leprosy' Int J
                                 Lepr 1991; 59:221-8
                             38 Glynn JR, Warndorff DK, Fine PEM, Msiska GK, Munthali MM, Ponnighaus JM. The impart
                                 of HTV on morbidity and mortality from tuberculosis in sub-Saharan Africa: a study in rural
                                 Malawi and review of the literature. Health Transition Rev 1997; 7(Suppl 2): 75-87
                             39 Selwyn PA, Hartel D, Lewis VA et al. A prospective study of the risk of tuberculosis among
                                 intravenous drug users with human immunodeficiency virus infection. N Engl J Med 1989;
                                 320: 545-50
                             40 Gilks CF, Godfrey-Faussett P, Batchelor BIF et al. Recent transmission of tuberculosis in a
                                 cohort of HIV-1-infected female sex workers in Nairobi, Kenya. AIDS 1997; 11 911-8

British Medical Bulletin 1998,54 (No 3)                                                                                      591
Resurgent/emergent infectious diseases

                        41 Long R, Scalcini M, Manfrcda J et al. Impact of HTV type I on tuberculosis in rural Haiti. Am
                           Rev Resptr Dis 1991; 143: 69-73
                        42 Gilks CF, Bnndle RJ, Otieno LS et al. Extrapulmonary and disseminated tuberculosis in HTV-1-
                           seropositive patients presenting to the acute medical services in Nairobi. AIDS 1990, 4- 981-5
                        43 De Cock KM, Gnaore E, Ad]orlolo G et al Risk of tuberculosis in patients with HTV-I and HTV-
                           II infections in Abid|an, Ivory Coast. BM] 1991, 302: 496-9
                        44 Chum HJ, O'Brien RJ, Chonde TM, Graf P, Rjeder HL An epidemiological study of
                           tuberculosis and HIV infection in Tanzania, 1991-1993. AIDS 1996; 10: 299-309
                        45 Van Cleef MRA, Chum HJ The proportion of tuberculosis cases in Tanzania attributable to
                           human immunodeficiency virus. Int ] Epidemiol 1995; 24: 637—42
                        46 Mighori GB, Borghesi A, Adnko C et al. Tuberculosis and HTV infection association in a rural
                           district of Northern Uganda: epidemiological and clinical considerations. Tubercle Lung Dts
                           1992; 73: 285-90
                        47 Houston S, Ray S, Mahan M et al. The association of tuberculosis and HTV infection in Harare,
                           Zimbabwe. Tubercle Lung Dis 1994; 75: 220-6
                        48 Antonucci G, Girardi E, Ravighone MC, Ippohto G. Risk factors for tuberculosis in HTV-
                           infected persons. A prospective cohort study. The Gruppo Itahano di Studio Tubercolosi e AIDS
                           (GISTA). JAMA 1995; 274: 143-8
                        49 Pape J, Jean SS, Ho JL, Hafner A, Johnson WD. Effect of isoniazid prophylaxis on incidence of
                           active tuberculosis and progression of HTV infection. Lancet 1993; 342 268-72
                        50 Rubinstien EM, Madden GM, Lyons RW Active tuberculosis in HTV-infected injecting drug
                           users from a low-rate tuberculosis area. / Acquir Immune Deftc Syndr Hum Retrovirol 1996;
                        51 Markowitz N, Hansen NI, Hopewell PC et al Incidence of tuberculosis in the United States
                           among HlV-infected persons. Ann Intern Med 1997; 126: 123-32
                        52 Moreno S, Baraia-Etxaburu J, Bouza E et al Risk for developing tuberculosis among anergic
                           patients infected with HTV. Ann Intern Med 1993; 119: 194-8
                        53 Moreno S, Miralles P, Diaz MD et al Isoniazid preventive therapy in human immunodeficiency
                           virus-infected persons Long term effect on development of tuberculosis and survival. Arch
                           Intern Med 1997, 157 1729-34
                        54 Guelar A, Gatell JM, Verdejo J et al. A prospective study of the risk of tuberculosis among HTV-
                           mfected patients. AIDS 1993; 7: 1345-9
                        55 Girardi E, Antonucci G, Ippohto G et al. Association of tuberculosis risk with the degree of
                           tuberculin reaction in HTV-infected patients. The Gruppo Itahano di Studio Tubercolosi e AIDS
                           Arch Intern Med 1997; 157 797-800
                        56 Hawken MP, Meme HK, Elliott LC et al. Isoniazid preventive, therapy for tuberculosis in HTV-
                           1-infected adults: results of a randomized controlled trial AIDS 1997, 11: 875-82
                        57 Whalen CC, Johnson JL, Okwera A et al. A trial of three regimens to prevent tuberculosis in
                           Ugandan adults infected with the human immunodeficiency virus. N Engl ] Med 1997; 337:
                        58 Bloom BR, Murray CJL Tuberculosis: commentary on a reemergent killer Science 1992; 257:
                        59 Couldwell DL, Dore GJ, Harkness JL et al. Nosocomial outbreak of tuberculosis in an
                           outpatient HTV treatment room AIDS 1996; 10: 521-5
                        60 Daley CL, Small PM, Schecter GF et al An outbreak of tuberculosis with accelerated
                           progression among persons infected with the human immunodeficiency virus. An analysis using
                           restriction-fragment-length polymorphisms N Engl ] Med 1993, 326 231—5
                        61 Di Pern G, Crucianf M, Danzi MC et al. Nosocomial epidemic of active tuberculosis among
                           HlV-infected patients Lancet 1989, 2: 1502^t
                        62 Hyge TV The efficacy of BCG-vaccination. Epidemic of tuberculosis in a State School with an
                           observation period of 12 years. Acta Tuberculosea 1955; 32: 89-107
                        63 Raffah J, Sepkowitz KA, Armstrong D. Community-based outbreaks of tuberculosis. Arch
                           Intern Med 1996; 156 1053-60
                        64 Small PM, Hopewell PC, Singh SP et al The epidemiology of tuberculosis in San Francisco A
                           population-based study using conventional and molecular methods. N Engl J Med 1994; 330:

592                                                                                    British Medical Bulletin 1998,54 (No 3)
                                                                                                       Tuberculosis and HIV

                               65 Alland D, Kalkut GE, Moss AR et al. Transmission of tuberculosis in New York City. An
                                  analysis by DNA fingerprinting and conventional epiderruologic methods N Engl ] Med 1994;
                                  330: 1710-6
                               66 Fneden TR, Woodlcy CL, Crawford JT, Lew D, Dooley SW. The molecular epidemiology of
                                  tuberculosis in New York City, the importance of nosocomial transmission and laboratory
                                  error. Tubercle Lung Dis 1996; 77: 407-13
                               67 Wilkinson D, Pillay M, Crump J, Lombard C, Davies GR, Sturm AW Molecular epidemiology
                                  and transmission dynamics of Mycobactertum tuberculosis in rural Africa. Trop Med Int
                                  Health 1997; 2. 747-53
                               68 Warren R, Hauman J, Beyers N et al. Unexpectedly high strain diversity of Mycobacterium
                                  tuberculosis in a high incidence community S Afr Med ] 1996, 86 45—9
                               69 Glynn JR, Vynnycky E, Fine PEM. The influence of sampling on estimates of recent
                                  transmission of M tuberculosis derived from DNA fingerprinting techniques Am J Epidemiol
                                  1998; In press
                               70 Samper S, Iglesias MJ, Rabanaque MJ et al Retrospective RFLP-population-based study of
                                  tuberculosis in Zaragoza (Spain). Combined meeting of the Concerted Actions on Tuberculosis,
                                  Cordoba 1997; A8
                               71 Di Pern G, Cadeo GP, Castelh F et al. Transmission of HTV-associated tuberculosis to health
                                  care workers Lancet 1992, 340 682
                               72 Nunn P, Mungai M, Nyamwaya J et al. The effect of human immunodeficiency virus type-1 on
                                  the infectiousness of tuberculosis. Tubercle Lung Dis 1994; 75: 25—32
                               73 Klausner JD, Ryder RW, Baende E et al Mycobacterium tuberculosis in household contacts of
                                  human immunodeficiency virus type 1-seropositive patients with active pulmonary tuberculosis
                                  in Kinshasa, Zaire. / Infect Dis 1993; 168: 106-11
                               74 Elliott AM, Hayes RJ, Halwundi B et al. The impact of HTV on uifectiousness of pulmonary
                                  tuberculosis: a community study in Zambia. AIDS 1993; 7: 981-7
                               75 Cauthen GM, Dooley SW, Onorato IM et al. Transmission of Mycobacterium tuberculosis from
                                  tuberculosis patients with HTV infection or AIDS. Am ] Epidemiol 1996; 144 69-77
                               76 Carvalho ACC, Figueira MM, Brito Z et al. Tuberculosis transmission from pulmonary
                                  tuberculosis patients (PTP) infected by HIV in Rio de Janeiro, Brazil. Int J Tuberc Lung Dis
                                  1997, 1: S159
                               77 Berme]o A, Veeken H, Berra A. Tuberculosis incidence in developing countries with high
                                  prevalence of HTV infection. AIDS 1992; 6: 1203-6
                               78 Richards SB, St Louis ME, Nieburg P et al. Impact of the HTV epidemic on trends in
                                  tuberculosis in Abidjan, Cote d'lvoire Tubercle Lung Dis 1995; 76 11—6
                               79 Lienhardt C, Rodngues LC. The estimation of the impact of HIV infection on tuberculosis:
                                  tuberculosis risks re-visited' Int J Tuberc Lung Dis 1997; 1: 196-204
                               80 Heymann SJ. Modelling the efficacy of prophylactic and curative therapies for preventing the
                                  spread of tuberculosis in Africa. Trans R Soc Trop Med Hyg 1993, 87 406-11
                               81 Schulzer M, Radhamani MP, Grzybowski S, Mak E, Fitzgerald JM. A mathematical model for
                                  the prediction of the impact of HTV infection on tuberculosis. Int J Epidemiol 1994; 23 400-7
                               82 Brewer TF, Heymann SJ, Colditz GA et al. Evaluation of tuberculosis control policies using
                                  computer simulation. JAMA 1996; 276 1898-903
                               83 Borgdorff MW, Kiambih FM, Odhiambo JA et al. Comparison between districts in the analysis
                                  of national tuberculin surveys: an example in Kenya. International Union Against Tuberculosis
                                  and Lung Disease, Conference Abstract 1997
                               84 Narain JP, Raviglione MC, Kochi A. HTV-associated tuberculosis in developing countries
                                  epidemiology and strategies for prevention Tubercle Lung Dis 1992, 73: 311-21
                               85 Fine PEM Variation in protection by BCG implications of and for heterologous immunity.
                                  Lancet 1995; 346 1339^*5
                               86 Tuberculosis preventive therapy in HTV-infected individuals. A joint statement of the
                                  International Union Against Tuberculosis and Lung Disease (IUATLD) and the Global
                                  Programme on AIDS and the Tuberculosis Programme of the World Health Organization
                                  (WHO). Tubercle Lung Dis 1994; 75. 96-8
                               87 Kornbluth RS, McCutchan JA Skin test responses as predictors of tuberculous infection and of
                                  progression in HIV-infected persons. Ann Intern Med 1993, 119: 241-3

British Medical Bulletin 1998,54 (No 3)                                                                                  593

To top