Perspective

For reprint orders, please contact:

reprints@future-drugs.com

Antimicrobial resistance
in tuberculosis:
an international perspective
Paul D van Helden†, Peter R Donald, Thomas C Victor, H Simon Schaaf,
Eileen G Hoal, Gerhard Walzl and Robin M Warren
Drug-resistant tuberculosis is a threat to tuberculosis control programs and community
health. This growing problem mirrors the increasing incidence of tuberculosis in general.
Public health problems include the absence of early diagnosis and effective treatment.
The real need is to identify tuberculosis patients far earlier, particularly those with
CONTENTS drug-resistant strains, and to begin appropriate therapy, which is of the shortest possible
duration with minimal risk of acquiring further drug resistance or permitting further
Disease pathology,
epidemiology & diagnostics transmission. This article will address the epidemic of drug resistance and discuss some of
the inherent difficulties in the treatment of drug-resistant tuberculosis. We highlight some of
Optimal therapy
the controversies and new findings in this area, as well as future perspectives requiring
Factors affecting choice
more active interventions, in addition to new technology and developments.
of therapy
Drug-resistant TB in children Expert Rev. Anti Infect. Ther. 4(5), 759–766 (2006)
Expert commentary Disease pathology, epidemiology methods to detect drug-susceptible and -resist-
Five-year view & diagnostics ant bacteria, each one of which has its own
Key issues Tuberculosis (TB) caused by Mycobacterium advantages and drawbacks. Sputum smear
tuberculosis in humans is largely a pulmonary positivity on acid-fast staining is regarded
References
disease characterized by large lesions in the currently as the basis of all TB control
Affiliations lung and the destruction of lung tissue. How- programs [101] since it is simple, cheap and
ever, other forms of disease may occur, particu- identifies those individuals most likely to
larly in children and immunocompromised spread infection. However, a major problem is
individuals. It is estimated that approximately that, by the time that sputum diagnosis based
one third of the global population is infected on positive microscopy can be made, the
by M. tuberculosis but, in the absence of patient will have had a high bacterial load for
immunosuppression, perhaps only 10% of some time and extensive transmission could
†
Author for correspondence infected individuals will progress to active have occurred already. Although pathology can
Molecular Biology and Human
disease in their lifetime. The diagnosis of TB is occur in many locations, for example, internal
Genetics, MRC Centre for
Molecular and Cellular Biology, generally made through a combination of organs or bones, where it is more difficult to
DST/NRF Centre of Excellence for techniques, usually including clinical symp- diagnose, transmission is unlikely from those
Biomedical TB Research, toms, such as night sweats, fever, coughing, individuals. Therefore, from the point of view
Faculty of Health Sciences, lost of weight, a positive skin test reaction to of continuing the epidemic, such individuals
Stellenbosch University,
purified protein derivative (PPD) and chest (including children) pose a negligible risk.
PO Box 19063, Tygerberg 7505,
South Africa x-ray and the identification of acid fast bacteria As with all chemotherapeutic treatments,
Tel.: +27 219 389 401 after sputum microscopy. Culture of live drug resistance can occur. Resistance to an
Fax: +27 219 389 476 mycobacteria from the sputum is a more defin- anti-TB agent was detected during the first
pvh@sun.ac.za itive and positive identification and, although randomized clinical trial evaluating strepto-
KEYWORDS: slow, it is commonly regarded as the gold mycin [1]. It is now appreciated that any rela-
drug resistance, drug trials, standard for diagnosis. More recently, other tively large population of M. tuberculosis will
ethambutol, first line, isoniazid,
rifampicin, second line, faster technologies have been used, ranging inevitably contain or develop drug-resistant
tuberculosis from serology to PCR-based or phage-based mutants and that these emerge at a predictable

www.future-drugs.com 10.1586/14787210.4.5.759 © 2006 Future Drugs Ltd ISSN 1478-7210 759

van Helden, Donald, Victor et al.

rate as a result of chromosomal mutations [2–6]. Using a single and thus constitute a ‘fitness cost’ [13]. However, laboratory
antibiotic agent gives a resistant mutant a selective advantage, studies show that the relative fitness of mutants may vary [14]
such that it will soon emerge as the dominant strain in any and some resistance mutations appear to confer no cost (<1%
situation, be it a test tube, an experimental animal or a human reduction in fitness). In a community setting, we have found
TB lesion. Acquired multidrug-resistant tuberculosis (MDR- that the same strain family can contain larger clusters of drug-
TB) results from the gradual selection of resistant bacteria resistant strains than susceptible strains [15], although time of
within the patient – ‘acquired resistance’ – who can in turn exposure to MDR strains may overcome relative loss of fitness
spread the resistant strain to other individuals, who will develop to boost numbers.
primary MDR-TB.
The appreciation of the need for multidrug regimens and the Optimal therapy
contribution of pyrazinamide (PZA) to preventing relapses led Generally, WHO recommendations are followed, which are
to the development of modern short-course chemotherapy based on the premise that the aims of treatment of TB are [101]:
(SCC) of 6-month duration with reported relapse rates of less • Cure the patient
than 5% in controlled clinical trials (Hong Kong Chest • Prevent death
Service/British Medical Research Council, 1987). The critical
• Prevent recurrence
role of isoniazid (INH) and rifampicin (RMP) in these regimens
means that the development of INH and RMP resistance is con- • Decrease transmission
sidered as a therapeutic disaster that seriously threatens patient • Prevent the development of (acquired) drug resistance
management and TB control activities worldwide. Conse- The recommended strategy characterizes patients into one of
quently, INH and RMP resistance has been labeled as multidrug four basic categories and treatment is based on those categories.
resistance. It should be noted that, in the presence of non- By far, the largest number of TB patients fall into Category
compliance, even fixed-dose combination preparations may not (Cat) I, new smear-positive patients. (The Cat III procedure is
be sufficient to prevent the development of drug resistance [7]. for new smear-negative pulmonary TB and less severe forms of
The chromosomal loci responsible for the resistance to various extrapulmonary TB. The treatment regimen is as for
drugs are not linked, so the risk of a double spontaneous muta- Cat I patients).
tion is extremely low: 9 × 10-14 for both INH and RMP [8]. To For such individuals the recommended treatment regimen is
date, approximately 13 genes are known to be linked to resist- 2 months of intensive phase therapy with INH, RMP, PZA and
ance to a variety of antibiotics and many functional mutations ethambutol (EMB), followed by a continuation phase of
have been described in these genes, whereas drug-susceptible 4 months with INH and RMP only [101].
isolates lack these mutations. Therefore, MDR-TB is due to an After the initial 2-month intensive phase of treatment, the
accumulation of mutations. patient’s sputum is checked for acid-fast bacilli. If the sputum
In some regions, MDR-TB rates exceed 14% of total cases [101]. has converted, the patient is placed onto the continuation
The consequences of MDR-TB are that treatment success is phase of therapy for a further 4 months. If the sputum is still
reduced from 80% for drug-susceptible TB to only 50% for smear positive, further follow-up is carried out and a decision
MDR-TB and the mortality rate rises. tree followed.
By definition, drug-resistant cases are refractory to first-line If the patient has treatment failure or recurrence, or returns
drugs used in directly observed treatment short-course after default, then the Cat II procedure is recommended. The
(DOTS). This has led to considerable controversy since many standard retreatment regimen consists of five drugs in the
have argued that the primary objective in TB control is to diag- initial phase and three drugs in the continuation phase. In this
nose and treat the majority of susceptible cases and, without case, streptomycin is added to the intensive phase and the con-
achieving this, one would not be in a position to tackle drug- tinuation phase is extended to 5 months and consists of INH,
resistant cases. Some have even argued that SCC would cure RMP and EMB. Drug-resistance testing is recommended.
drug-resistant TB or at least remove the problem [9], despite Clearly, this differs from Cat I by one drug in each phase.
evidence to the contrary [10]. Supporting evidence for this claim Should patients remain sputum-positive for a considerable
is that, over time, the relative proportions of drug-resistant TB period, they would be classified as Cat IV with suspected
can decline when usage of first-line drugs is protected [11,12]. MDR-TB and the appropriate diagnosis and treatment
Unfortunately, this logic is based on managerial principles and followed. It is important to note that EMB is being used both
not necessarily on scientific evaluation. Furthermore, while the in the intensive phase of Cat I and the intensive and follow-up
relative proportion of drug-resistant cases may decline, the total phase of Cat II patients. We will return to this problem later.
case load will have increased considerably, as has happened in Should a patient be diagnosed with drug-resistant TB, there are
South Africa. three options available:
The common assumption that drug-resistant strains are less • Standardized treatment regimen
likely to cause epidemics than drug-susceptible strains is based
on the observation that mutations that confer resistance often • Empirical treatment
alter gene products that are important to the pathogen’s survival • Individualized treatment

760 Expert Rev. Anti Infect. Ther. 4(5), (2006)

 Drug resistance in TB

The following basic principals are involved in the treatment probably the first child reported with drug-resistant TB [20].
regimen design for any suspect or proven MDR-TB case: Children usually have paucibacillary disease and therefore
• They should be based on the history of drugs taken by the rarely develop acquired resistance owing to previous anti-TB
person treatment. In New York (USA), both INH and RMP
• Commonly used drugs should be taken into account and resistance in children followed the increase in resistance to
the prevalence of resistance to first- and second-line drugs these drugs in adults [21,22]. Schaaf and colleagues demon-
be considered strated a higher infection rate among children in contact with
adults with MDR pulmonary TB compared with adults with
• Regimens should consist of at least four drugs with certain drug-susceptible pulmonary TB [23]. Although the initial
effectiveness disease rate was significantly lower in MDR-TB contacts, on
• Treatment of at least 6 days per week is recommended for 30-month follow-up, many more children developed active
MDR-TB and should continue for a minimum of TB [24]. The diagnosis of MDR-TB should be considered in
18 months beyond conversion children in the following circumstances, where cultures are
The drugs that can be used are listed on the WHO website [101]. not obtained:
• A known adult source case with infectious MDR-TB
Factors affecting choice of therapy
• A child whose disease deteriorates despite adequate and
Steps to prevent and manage MDR-TB include: adherent anti-TB treatment
• Adherence to the principles of DOTS • A child in contact with an adult source case with unknown
• Prevention of the development of an extended pattern of DST result who is a treatment failure, a retreatment case or a
resistance in those already INH resistant chronic TB case
• The use of standardized second-line regimens for chronic • A high prevalence of MDR-TB in the community where the
patients and MDR patients child resides
• Individualized treatment regimens where sufficient resources Children with MDR-TB should be aggressively managed, as
are available to supply the necessary drugs and monitor failure to do so could lead to increased morbidity and mortality.
their use It is most important not just to add a single drug to a failing
regimen but to add a minimum of three new drugs to which
Adherence to the principles of DOTS the strain is susceptible or the source case is naïve. The addition
Several studies have demonstrated that the use of a standardized of second line drugs, such as the fluoroquinolones, which are
retreatment regimen in countries with a well functioning DOTS not recommended for use in children generally, and amino-
program may lead to the development of MDR in patients with glycosides (amikacin or kanamycin), are essential in the
monoresistance or a pattern of resistance not usually classified as management of childhood MDR-TB cases.
MDR and may create resistance to additional drugs in patients
who can already be classified as MDR patients [16]. However, it Expert commentary
has been claimed that, even in a setting with moderate MDR Ideally, full drug-susceptibility testing on all first-line drugs (at
rates, DOTS can reduce the transmission and incidence of both least) for all presenting TB cases should be carried out. It is a
drug-susceptible and resistant TB [9,17]. Drug-susceptibility test- bitter irony that this probably happens only in developed
ing (DST) and either standardized or individualized regimens countries with a low case load and little drug resistance. The
are, however, required to reduce the mortality of MDR-TB. countries with a high burden of disease frequently consider
themselves to have inadequate resources to address this prob-
Standardized second-line drug treatment for chronic TB lem. It is precisely at this point that the problem arises. Given
Against the background of a well functioning DOTS program, it that the case load of drug-susceptible and -resistant TB is
becomes possible to introduce the use of second-line drugs for the increasing in many, if not most, regions [101], a careful analysis
management of chronic TB patients [18] or for those identified as of this problem is required in order to arrive at some solutions.
suffering from MDR-TB [19]. With both approaches, a standard- In this section, some of the problematical and controversial
ized regimen of kanamycin, ethionamide (or prothionamide), issues will be discussed.
EMB, PZA and a quinalone is used. In some situations, para-
amino-salicylic acid and cycloserine may also be used. These Recurrence: relapse/reinfection
approaches can provide an acceptable alternative to individualized The design of the current SCC was based on the concept of
treatment programs in resource-limited settings. ‘relapse’. It is perhaps better to use the word ‘recurrence’ to deal
with repeat episodes and to redefine recurrence into two
Drug-resistant TB in children categories: true relapse and reinfection. In high incidence
The diagnosis of TB in children is complex and the diagnosis of settings, reinfection may considerably exceed true relapse [25].
MDR-TB is even more so. An 11-week-old infant who The automatic and often unnecessary use of Cat II or IV treat-
presented with streptomycin-resistant miliary TB in 1948 was ments with additional antibiotics would probably increase the

www.future-drugs.com 761
van Helden, Donald, Victor et al.

risk of development of further acquired drug resistance since Some of the failures may simply be owing to drug-resistant
first-line drugs may be perfectly adequate in reinfection cases. TB. These cases pose a serious risk to the community as they
Furthermore, patients with a prior episode of TB have an remain infectious and are probably responsible for epidemics
enhanced likelihood of progression to disease after infection [26]. and outbreaks of drug-resistant TB. Thus, failure to convert is a
Therefore, recurrent cases are not necessarily at high risk of drug signal for intensive follow-up and susceptibility testing and not
resistance, assuming compliance. necessarily blind adherence to a program.

Shorten SCC further? Risk of classification of patients in Cat I–IV based on

Given that the rate of true relapse may be very low, there is a WHO definitions
case to be made for further trials to evaluate a reduction of While there are obvious benefits to using this classification
standard SCC from 6 months to 4–5 months at different doses scheme, there are also risks based on assumption. In this recom-
of sterilizing drugs (e.g., RMP). Under standard conditions, the mendation, it is assumed that a new case is first-line drug suscep-
majority of patients convert to sputum-smear negativity at tible. While this is probably true in most settings, in an Eastern
2 months but up to 25% of initially smear-positive patients European prison setting, for example, it may very well not be.
may not [OWN RESULTS, UNPUBLISHED DATA]. Therefore, SCC Similarly, classification into Cat II and treatment accordingly
could be tailored according to the outcome of the 2-month may be quite unnecessary, leading to unnecessary costs for the
response. A shortening of SCC to 4-months should considera- program, increased work load for the healthcare staff and
bly alleviate the burden on the program, giving it more scope to decreased patient compliance owing to the unpleasantness of
deal with MDR-TB cases. extra medication and longer therapy. Adherence to standard
protocols without further DST may enable a drug-resistant case
Risk associated with use of second-line antibiotics to spread to further contacts and lead potentially to the
The choice facing the healthcare professional in the case of a acquisition of more resistance owing to the monotherapy effect.
Cat II patient or a slow responder is complex and risky. If the
patient defaulted before cure without acquiring resistance, Technical problems with diagnosis of drug resistance in TB
standard therapy should still be effective and should, therefore, The technical problems are numerous and dependent on the
be continued. However, the recommendation is to increase the methods and experience of the laboratory doing the work, as
number of drugs and treatment time. The risk with this philo- well as the interpretation of the results. The first problem is the
sophy is that, should the bacteria indeed have acquired resist- choice of method to be used. Culture-based susceptibility test-
ance to some drugs, the addition of only one new antibiotic ing is perhaps the simplest method but the major disadvantages
may effectively be akin to monotherapy and may generate are the time taken to result, the relative instability of some
strains that accumulate resistance to new drugs as they are drugs in solution and the fact that some drugs may need activa-
added. This practice therefore needs to be re-examined. tion, prefer a microaerophilic or acidic environment or do not
The issues that should be addressed are diagnosis and classifi- work satisfactorily in vitro (e.g., PZA). Culture-based methods
cation of TB patients, particularly the so-called ‘problem can be accelerated by the use of phage-based assays (Albert
patients’. Such patients may be defaulters, noncompliant 2002), but these have not been adopted widely. A new, simple,
patients, those who do not sputum convert and those who have culture-based direct microscopy method known as microscopic
recurrent episodes. observation broth-drug susceptibility assay (MODS) has been
developed recently, which takes 8 days, is inexpensive and
Failure to convert performs well but, unfortunately, not for all antibiotics [29].
Failure to convert after 2–3 months of SCC treatment (but Molecular methods for the identification of mutations in
particularly after 6 months) could be an indicator for full resistance-causing genes may offer a means to rapidly screen
DST. The simplest explanation for delayed conversion is that M. tuberculosis isolates for antibiotic resistance. Mutation
the initial bacterial burden and the general health condition screening methods are fast and include methods, such as DNA
is such that the patient’s immunity is unable to reduce the sequencing, probe-based hybridization methods, PCR-
bacterial load adequately at 2 months. However, a further restriction fragment length polymorphism (RFLP), single-
explanation and the rationale behind the relatively long treat- strand conformation polymorphism, heteroduplex analysis,
ment period for SCC is that different populations of bacteria molecular beacons and amplification refractory mutation
exist. The bacterial populations are either actively growing system (ARMS)-PCR [30]. However, these methods require
and metabolizing and are killed quickly by antibiotics or more skill and can only be used if the mutations are known.
existing in a dormant state refractory to SCC [27,28]. The use of any single method alone is not recommended at
Dormant bacilli can reawaken to become actively replicating present and validation of the testing method should be
(recurrence) or can be the explanation for conversion failure. performed. For example, we showed recently that routine culture
Antibiotics may not be effective against dormant bacteria based methods in a high-throughput routine laboratory detected
and are therefore graded according to their bactericidal and EMB resistance in less than 2% of INH- and RMP-resistant iso-
sterilizing abilities. lates, but molecular methods detected EMB resistance in 25% of

762 Expert Rev. Anti Infect. Ther. 4(5), (2006)

 Drug resistance in TB

these [31]. Similarly, PZA is important in various regimens for its (USA). However, estimates of costs must be viewed in the
sterilizing ability but testing for PZA resistance is not usually context of the country where the treatment is being carried out.
performed owing to culture-based problems [32]. Cost estimates for South Africa have been in the order of
US$5000 and, in Peru, perhaps US$600 [9]. The question should
Identify the source of the problem (primary or not be, can we afford it, but rather, we must afford it so how do
acquired resistance) we keep the costs as low as possible? If we cannot control drug-
The use of surveillance data gathered by the WHO shows that, resistant TB cases, a rising incidence is inevitable as can be
in most environments, there are more secondary than primary predicted by the consultation of any compound interest table.
cases of resistance [101]. This is interpreted traditionally as
suggesting that acquisition of resistance is the driving force Five-year view
behind the epidemic, allegedly demonstrating the inability of Diagnosis of MDR-TB
the TB control program to ensure compliance. New diagnostics are set to come forward and these are sorely
However, in high-incidence communities, most drug-resistant needed. The ideal would be a point-of-care diagnostic, with
isolates can be clustered and are therefore transmitted [4,33]. The susceptibility testing. This is not that unrealistic, given that there
traditional clinical interpretation of acquired resistance was are simple portable FACS instruments for the assessment of T-cell
incorrect in this scenario since most cases were secondary. In one counts for HIV positive individuals and the US military has
sense, this is good news, since it suggests that, in most cases, developed a number of mobile molecular biology-based point-of-
therapy works well and acquisition of resistance is rare. However action diagnostics for other infectious agents. Chip/strip technol-
it also shows that we need to adopt alternative strategies to ogy with multiplex PCR should be developed for TB [35] as this is
combat the dissemination and amplification of antibiotic- rapid and requires no culture, thus reducing the biosafety risk.
resistant TB. In many environments, outbreaks of resistant Enhanced efforts are necessary to understand the molecular
strains have been found, which suggest that there may be a few mechanisms of resistance in second-line anti-TB drugs in clinical
bacterial strains only responsible for the majority of cases; these isolates. The next generation of molecular methods for the pre-
could be known as ‘superspreaders’ [34] and are responsible for diction of drug resistance in M. tuberculosis will possibly consist
epidemics. This information suggests that we need to disrupt the of matrix hybridization formats, such as DNA oligonucleotide
chains of transmission to prevent further spread of resistant TB arrays on slides or silicon chips [36], particularly if these systems
since acquisition of resistance in a reasonable program is a minor can be fully automated and reused. This may be particularly
component of the total burden of drug-resistant TB cases. useful for mutations in the rpoB gene, which can serve as markers
Furthermore, the relative proportion of drug-resistant cases can for MDR-TB [37] and also for the multiple loci involved in INH
decline in a good program [11,12]. This may require a rethink of resistance. Selection of a limited number of target mutations that
the control programs to include introduction of extensive contact enable the detection of most drug resistance [38] would be useful
tracing and rapid drug-resistance testing, otherwise we are in this strategy. It is essential that developers of new techniques
unlikely to prevent further dissemination of those strains already recognize that the majority of drug-resistant cases occur in
in circulation. In particular, we need to identify the resource-poor countries [39] and, therefore, that the
superspreaders and devote our efforts to combating these. methodologies must not only be cheap but also robust.
The application of rapid methods to break the chains of
Model the effect of intervention ongoing transmission of drug-resistant TB will become increas-
During the delay period from infectiousness to diagnosis and ingly important as recent mathematical modeling indicates that
treatment, the patient can act as a source case for ongoing the burden of MDR-TB cannot be contained in the absence of
transmission. A rapid test for susceptibility at diagnosis is there- specific efforts to limit transmission [40,41], which may include
fore desirable. Modeling to determine the cost-effectiveness of rapid detection of drug resistance by molecular methods.
interventions, such as widespread DST and contact tracing,
needs to be carried out in different settings, as it is usually New drugs & clinical trials
argued that it is not affordable in most settings. Clearly, this There are a number of new candidate drugs at various stages of
should include the projected cost of continuing transmission development and testing but the very long clinical trial period for
and generation of future disease burden (opportunity cost). It TB suggests that there will be few, if any, new drugs available for
may be that the most cost-effective route (long term) would be programmatic use within the next 5 years. However, there is no
active case finding and standard diagnosis and therapy. doubt that trials will be carried out.
Given that so few new antibiotics have been developed for
Cost of programs TB in the last 40 years, it would be a tragedy if new candidates
Opponents of a strategy of widespread DST and use of second- were missed because of poor understanding of trial results.
line antibiotics have argued that many poorer nations cannot Therefore, it is important that it is realized that a new drug may
afford to diagnose or treat MDR cases. This assertion was based be highly bactericidal and perform well in early bactericidal
on the popularization of statements that suggested that the cost activity (EBA) and therefore pass through to the next trial
of treating one MDR-TB case may be as high as US$250,000 phase but that some may be potent sterilizers and perform

www.future-drugs.com 763
van Helden, Donald, Victor et al.

poorly in EBA trials. It would therefore be wrong to condemn a parameters, such as the standard delay from infection to active
potential candidate based only on EBA. Furthermore, trials disease and costs of various activities in such settings, as well as
supported by new technologies must be carried out to assess data to estimate the case-finding ability. All indications are that,
whether or not we can reduce the time for SCC and what the in most high-burden environments, the case-finding ability is
best doses are for SCC, for both existing and new drugs. There low. In that case, no strategy will reduce incidence. Therefore,
is a great need to better inform both pharmaceutical companies we need to ask: can we carry out active case finding? Can we
and regulatory agencies concerning the dynamics of this disease afford not to?
and the nature of the bacterium. They need to appreciate the
urgency involved in approving deserving trials with the shortest Other issues
delay and the real meaning of trial outcomes. One issue relates to the fact that isolation will stop transmis-
New formulations for existing TB drugs could also be sion. How can we isolate patients in the modern environment?
developed. These could include inhaled antibiotics, which may If a MDR case refuses therapy, should we enforce isolation?
even be supported by nanotechnology (microencapsulation). This could be carried out by incarceration or therapy. We need
to consider the ethical issues: when do the rights of the individ-
Points of intervention ual exceed those of the community, or vice versa? In this case, it
We need to develop mathematical models to assess where the may be argued that health care is for the common good and
most effective intervention can be performed. More accurate therefore the rights of the community should take precedence
data from a variety of settings are therefore required, specifically as TB is a notifiable disease.

Key issues

• Diagnosis of tuberculosis (TB), particularly antibiotic-resistant TB, is too slow. Rapid diagnostics for antibiotic-resistant TB must be
introduced widely.
• Current WHO recommendations form an excellent basis for TB control but are a minimum standard and are insufficient.
• Revision of the classification of TB patient categories should be considered, based on better diagnostics. Treatment of TB cases must
be based on actual susceptibility, not assumption.
• New drug trials for improved short-course chemotherapy are required.
• Active case findings must be considered.
• Reliable epidemiological data from different settings must be gathered to plan better control programs and interventions.

References the Western Cape Province of South 9 Farmer P, Walton D. The social impact of
Papers of special note have been highlighted as: Africa. J. Clin. Microbiol. 42(2), 891–894 multi-drug-resistant tuberculosis: Haiti
• of interest (2004). and Peru. In: The Return of the White
•• of considerable interest 5 Victor TC, Lee H, Cho SN et al. Plague. Gandy M, Zumla A (Eds). Verso,
Molecular detection of early appearance of London, UK 163–177 (2003).
1 Mitchison DA. Development of
streptomycin resistant strains of tubercle drug resistance during Mycobacterium 10 Espinal MA, Kim SJ, Suarez PG et al.
bacilli in pulmonary tuberculosis; results of tuberculosis infection. Clin. Chem. Lab. Standard short-course chemotherapy for
simultaneous sensitivity tests in liquid and Med. 40(9), 876–881 (2002). drug-resistant tuberculosis: treatment
on solid media. Thorax 5(2), 144–161 6 van Rie A, Warren R, Mshanga I et al. outcomes in 6 countries. JAMA 283(19),
(1950). Analysis for a limited number of gene 2537–2545 (2000).
2 David HL. Probability distribution of codons can predict drug resistance of 11 Weyer K, Kleeberg HH. Primary and
drug-resistant mutants in unselected Mycobacterium tuberculosis in a high- acquired drug resistance in adult black
populations of Mycobacterium tuberculosis. incidence community. J. Clin. Microbiol. patients with tuberculosis in South Africa:
Appl. Microbiol. 20(5), 810–814 39(2), 636–641 (2001). results of a continuous national drug
(1970). 7 Mitchison DA. How drug resistance resistance surveillance programme
emerges as a result of poor compliance involvement. Tuber. Lung Dis. 73(2),
3 Watterson SA, Wilson SM, Yates MD,
during short course chemotherapy for 106–112 (1992).
Drobniewski FA. Comparison of three
molecular assays for rapid detection of tuberculosis. Int. J. Tuberc. Lung Dis. 2(1), 12 Kim SJ, Bai GH, Hong YP. Drug-resistant
rifampin resistance in Mycobacterium 10–15 (1998). tuberculosis in Korea, 1994. Int.
tuberculosis. J. Clin. Microbiol. 36(7), 8 Dooley SW, Simone PM. The extent and J. Tuberc. Lung Dis. 1(4), 302–308
1969–1973 (1998). management of drug-resistant (1997).
4 Streicher EM, Warren RM, Kewley C et al. tuberculosis: the American experience. 13 Andersson DI, Levin BR. The biological
Genotypic and phenotypic characterization Clinical tuberculosis Chapman & Hall, cost of antibiotic resistance. Curr. Opin.
of drug-resistant Mycobacterium London, UK 171–189 (1994). Microbiol. 2(5), 489–493 (1999).
tuberculosis isolates from rural districts of

764 Expert Rev. Anti Infect. Ther. 4(5), (2006)

 Drug resistance in TB

14 Mariam DH, Mengistu Y, Hoffner SE, 25 van Rie A, Warren R, Richardson M et al. 35 Park H, Song EJ, Song ES et al. Comparison
Andersson DI. Effect of rpoB mutations Exogenous reinfection as a cause of of a conventional antimicrobial susceptibility
conferring rifampin resistance on fitness of recurrent tuberculosis after curative assay to an oligonucleotide chip system for
Mycobacterium tuberculosis. Antimicrob. treatment. N. Engl. J. Med. 341(16), detection of drug resistance in
Agents Chemother. 48(4), 1289–1294 (2004). 1174–1179 (1999). Mycobacterium tuberculosis isolates. J. Clin.
15 Streicher EM, Warren RM, Kewley C et al. •• Paradigm shift paper that demonstrates Microbiol. 44(5), 1619–1624 (2006).
Genotypic and phenotypic characterization the importance of distinguishing between 36 Vernet G, Jay C, Rodrigue M, Troesch A.
of drug-resistant Mycobacterium tuberculosis reinfection and relapse in recurrent cases. Species differentiation and antibiotic
isolates from rural districts of the Western 26 Verver S, Warren RM, Beyers N et al. Rate susceptibility testing with DNA microarrays.
Cape Province of South Africa. J. Clin. of reinfection tuberculosis after successful J. Appl. Microbiol. 96(1), 59–68 (2004).
Microbiol. 42(2), 891–894 (2004). treatment is higher than rate of new 37 Watterson SA, Wilson SM, Yates MD,
16 Espinal MA. Time to abandon the standard tuberculosis. Am. J. Respir. Crit. Care Med. Drobniewski FA. Comparison of three
retreatment regimen with first-line drugs for 171(12), 1430–1435 (2005). molecular assays for rapid detection of
failures of standard treatment. Int. J. Tuberc. • Notes that infection after cure is more rifampin resistance in Mycobacterium
Lung Dis. 7(7), 607–608 (2003). likely to progress to disease than after tuberculosis. J. Clin. Microbiol. 36(7),
initial infection. 1969–1973 (1998).
17 Deriemer K, Garcia-Garcia L,
Bobadilla-del-Valle M et al. Does DOTS 27 Wayne LG, Sohaskey CD. Nonreplicating 38 van Rie A, Warren R, Mshanga I et al.
work in populations with drug-resistant persistence of Mycobacterium tuberculosis. Analysis for a limited number of gene
tuberculosis? Lancet 365(9466), 1239–1245 Annu. Rev. Microbiol. 55, 139–163 (2001). codons can predict drug resistance of
(2005). 28 Mitchison DA, Coates AR. Predictive Mycobacterium tuberculosis in a high-
•• Discusses overall value of directly observed in vitro models of the sterilizing activity of incidence community. J. Clin. Microbiol.
treatment short-course. anti-tuberculosis drugs. Curr. Pharm. Des. 39(2), 636–641 (2001).
18 Suarez PG, Floyd K, Portocarrero J et al. 10(26), 3285–3295 (2004). 39 Maher D, Raviglione M. Global
Feasibility and cost-effectiveness of 29 Moore DA, Mendoza D, Gilman RH et al. epidemiology of tuberculosis. Clin. Chest
standardised second-line drug treatment for Microscopic observation drug Med. 26(2), 167–182 (2005).
chronic tuberculosis patients: a national susceptibility assay, a rapid, reliable 40 Cohen T, Murray M. Modeling epidemics of
cohort study in Peru. Lancet 359(9322), diagnostic test for multidrug-resistant multidrug-resistant M. tuberculosis of
1980–1989 (2002). tuberculosis suitable for use in resource- heterogeneous fitness. Nat. Med. 10(10),
19 Van Deun A, Salim MA, Das AP, Bastian I, poor settings. J. Clin. Microbiol. 42(10), 1117–1121 (2004).
Portaels F. Results of a standardised regimen 4432–4437 (2004). • Excellent example of the use of modeling.
for multidrug-resistant tuberculosis in 30 Victor TC, van Helden PD, Warren R. 41 Blower SM, Chou T. Modeling the
Bangladesh. Int. J. Tuberc. Lung Dis. 8(5), Prediction of drug resistance in emergence of the ‘hot zones’: tuberculosis
560–567 (2004). M. tuberculosis: molecular mechanisms, and the amplification dynamics of drug
20 Tinne JE, Henderson JL. Primary tools, and applications. IUBMB Life resistance. Nat. Med. 10(10), 1111–1116
streptomycin-resistant tuberculosis in a 53(4–5), 231–237 (2002). (2004).
newborn child. Simple method of assessing 31 Johnson R, Jordaan AM, Pretorius L et al.
streptomycin resistance. Lancet 2(27), Ethambutol resistance testing by mutation Website
901–904 (1950). detection. Int. J. Tuberc. Lung Dis. 10(1),
101 WHO website provides extensive
21 Zitrin CM, Lincoln EM. Initial 68–73 (2006).
epidemiological information and
tuberculous infection due to drug- • Notes that ethambutol resistance
treatment guidelines
resistant organisms. With a review of the is common.
www.who.int/tb
world literature on initial infection due 32 Louw GE, Warren RM, Donald PR et al.
to isoniazid-resistant tubercle bacilli. Frequency and implications of
Affiliations
J. Pediatr. 58, 219–225 (1961). pyrazinamide resistance in managing
previously treated tuberculosis patients. • Paul D van Helden, PhD
22 Steiner P, Rao M, Mitchell M, Steiner M. Molecular Biology and Human Genetics,
Primary drug-resistant tuberculosis in Int. J. Tuberc. Lung Dis.10(7), 802–807
(2006). MRC Centre for Molecular and Cellular
children. Emergence of primary drug-resistant Biology, DST/NRF Centre of Excellence for
strains of M. tuberculosis to rifampin. Am. Rev. 33 van Rie A, Warren R, Richardson M et al. Biomedical TB Research, Faculty of Health
Respir. Dis. 134(3), 446–448 (1986). Classification of drug-resistant tuberculosis Sciences, Stellenbosch University,
23 Schaaf HS, Vermeulen HA, Gie RP, Beyers N, in an epidemic area. Lancet 356(9223), PO Box 19063, Tygerberg 7505, South Africa
Donald PR. Evaluation of young children in 22–25 (2000). Tel.: +27 219 389 401
household contact with adult multidrug- •• Shows that most drug-resistant cases in a Fax: +27 219 389 476
resistant pulmonary tuberculosis cases. high-incidence area are from transmission. pvh@sun.ac.za
Pediatr. Infect. Dis. J. 18(6), 494–500 (1999). 34 Lloyd-Smith JO, Schreiber SJ, Kopp PE, • Peter R Donald, DCH, MD
Getz WM. Superspreading and the effect Paediatrics and Child Health,
24 Schaaf HS, Gie RP, Kennedy M, Beyers N,
of individual variation on disease Faculty of Health Sciences, Stellenbosch
Hesseling PB, Donald PR. Evaluation of
emergence. Nature 438(7066), 355–359 University, PO Box 19063,
young children in contact with adult Tygerberg7505, South Africa
multidrug-resistant pulmonary tuberculosis: (2005).
•• Suggests that a few source cases Tel.: +27 219 389 506
a 30-month follow-up. Pediatrics 109(5), Fax: +27 219 389 138
765–771 (2002). propagate epidemics.
prd@sun.ac.za

www.future-drugs.com 765
van Helden, Donald, Victor et al.

• Thomas C Victor, PhD • H Simon Schaaf, M Med, MD • Gerhard Walzl, M Med, PhD
Molecular Biology and Human Genetics, Paediatrics and Child Health, Molecular Biology and Human Genetics,
MRC Centre for Molecular and Cellular Faculty of Health Sciences, Stellenbosch MRC Centre for Molecular and Cellular
Biology, DST/NRF Centre of Excellence for University, PO Box 19063, Biology, DST/NRF Centre of Excellence for
Biomedical TB Research, Faculty of Health Tygerberg7505, South Africa Biomedical TB Research, Faculty of Health
Sciences, Stellenbosch University, PO Box Tel.: +27 219 389 506 Sciences, Stellenbosch University,
19063, Tygerberg 7505, South Africa Fax: +27 219 389 138 PO Box 19063, Tygerberg7505, South Africa
Tel.: +27 219 389 251 prd@sun.ac.za Tel.: +27 219 389 158
Fax: +27 219 389 476 • Eileen G Hoal, PhD Fax: +27 219 389 476
tv@sun.ac.za Molecular Biology and Human Genetics, gwalzl@sun.ac.za
MRC Centre for Molecular and Cellular • Robin M Warren, PhD
Biology, DST/NRF Centre of Excellence for Molecular Biology and Human Genetics,
Biomedical TB Research, Faculty of Health MRC Centre for Molecular and Cellular
Sciences, Stellenbosch University, Biology, DST/NRF Centre of Excellence for
PO Box 19063, Tygerberg7505, South Africa Biomedical TB Research, Faculty of Health
Tel.: +27 219 389 412 Sciences, Stellenbosch University,
Fax: +27 219 389 476 PO Box 19063, Tygerberg7505, South Africa
egvh@sun.ac.za Tel.: +27 219 389 073
Fax: +27 219 389 476
rw1@sun.ac.za

766 Expert Rev. Anti Infect. Ther. 4(5), (2006)