New developments and Perspectives

Chapter 20: New developments and Perspectives

Viviana Ritacco, Sylvia Cardoso Leão and Juan Carlos Palomino

20.1. The scenario

“The history of tuberculosis (TB) has been one of scientific, medical and political failure.”

With this disturbing statement, The Lancet’s editors introduced an issue dedicated to TB that was released on the occasion of the World TB Day 2006 (Zumla 2006). According to the Global TB Control Report released one year later by the World Health Organization (WHO), the good news is that “the worldwide TB epidemic has leveled off for the first time since the disease was declared a public health emergency in 1993.” The bad news is that “at the current rate of progress, the 1990 prevalence and mortality rates will not be halved worldwide by 2015.” The Global Plan to Stop TB needs to triple investment in order to achieve such a goal (World Health Organization 2007).

TB is the only disease ever declared a global emergency by the WHO. Paradoxically, although we count on effective – and proven cost-effective – interventions for its control, TB continues to cause great mortality and suffering, especially in poor and less-developed countries. Its association with the HIV/AIDS pandemic forms a lethal combination. In addition, multidrug resistant (MDR) TB and the recently-described extensively drug resistant (XDR) TB – with further resistance to key second-line drugs and virtually incurable – severely complicate the management and control of the disease worldwide (Dorman 2007, Shah 2007). As repeatedly stated, one third of the world’s population is latently infected with Mycobacterium tuberculosis and 10 % of these people will develop active disease at some point in their life. Almost 8.8 million new cases of TB were reported in 2005, and 1.6 million deaths were attributed to the disease. Asia and Sub-Saharan Africa accounted for 7.4 million new cases of TB worldwide (World Health Organization 2007).

Yet, it was not long ago that we envisaged – and proudly announced – the elimination of TB by the end of the last millennium. Indeed, in the late ’70s and early ’80s, it was thought that TB could be eradicated from most developed and industrialized countries. TB was already regarded as a disease from the past and started to be neglected by medical doctors, scientists and agencies in charge of its control. However, this never became a reality, mainly due to the appearance of antibiotic resistance, and therefore, TB continues to be the big killer it was in the pre-antibiotic era. This unexpected re-emergence of TB in the ’90s served not only to strengthen control measures but also to fuel research on TB.

Substantial scientific advances were made in knowledge about the agent and the disease in that decade. First, the complete genome sequence of the tubercle bacillus was deciphered. Second, molecular epidemiology came into being, shedding a new light on mechanisms of TB transmission. Third, cellular mechanisms involved in M. tuberculosis resistance to several drugs were discovered. Fourth, new tools for speeding TB diagnosis and assessing drug susceptibility were sought while old methods were re-discovered and/or re-formulated. Lastly, research on drug and vaccine development exploded.

As for more recent advances, an analysis published in Nature Medicine highlights the most interesting scientific findings made in TB over the last three years, as identified by TB researchers (Anonymous 2007). The three articles that were considered to have the highest impact in the field were:

  • the description – for the first time in 40 years – of a new drug, a diarylquinoline that acts on an entirely new mycobacterial target, the proton pump of adenosine triphosphate (ATP) synthase (Andries 2005)
  • the report from a rural South African area of a deadly outbreak of what thereafter became known as XDR-TB (Gandhi 2006)
  • the description of a novel recombinant bacille Calmette-Guérin (BCG) vaccine candidate that works by over-expressing not an antigen, but a membrane-perforating enzyme (Grode 2005).

The main topics of these three papers underline the most pressing gaps in the translation of TB research. Other articles in the top list were related to vaccine candidates, virulence factors, genomics, new drugs, bacterial survival, and metabolism.

20.2. Bacillus and disease under the light of molecular epidemiology

The accelerated expansion that TB research underwent during the ’90s is reflected by the way in which molecular epidemiology is regarded nowadays. In fact, molecular strain typing of M. tuberculosis is no longer a novelty. What is more, it is taken for granted as if it always had been there. In less than two decades, it has changed our view of TB transmission dynamics, challenging traditional dogmas and answering unsolved epidemiological questions (Mathema 2006). Nowadays, molecular epidemiology is embedded within almost every aspect of TB research, laboratory diagnosis, clinical management, and control interventions. Indeed, molecular epidemiology could be regarded as a paradigm of TB translational research.

Soon after standardization was agreed on DNA restriction fragment length polymorphism (RFLP) with IS6110, its use was generalized worldwide (van Embden 1993). Other DNA strain typing methods were developed, IS6110 RFLP became the gold standard of M. tuberculosis strain genotyping, and a new discipline – molecular epidemiology – came into being. The considerable amount of information gathered in national and international M. tuberculosis genotype databases throughout the world enabled the analysis of global TB dissemination and promoted phylogeographical studies (Brudey 2006) that, in turn, incorporated more sophisticated and accurate markers of M. tuberculosis evolution (Arnold 2007).

A common conviction of previous times was that the genome of the tubercle bacillus was extremely stable and homogeneous. Phenotypical discrimination between strains – and even between species within the M. tuberculosis complex – was limited to phage typing and comparison of drug resistance patterns, which were the only tools available for differentiation. Molecular epidemiology refuted such belief by revealing the existence of wide M. tuberculosis polymorphisms in population-based studies (van Soolingen 2001).

Molecular epidemiology tools also enabled the identification, description and differentiation of rare species within the M. tuberculosis complex, such as “M. canettii”, M. microti, M. caprae, and M. pinnipedii. These species had previously been overlooked, mainly because they were difficult to distinguish by conventional biochemical tests (see chapter 8). The differentiation between these M. tuberculosis complex taxons contributed to the triggering of evolutionary studies (Gutierrez 2005). In turn, differentiation to the species level by spoligotyping (Kamerbeek 1997) – a user-friendly genotyping tool applied worldwide – turned out to have practical implications on medical management and epidemiology. More recently, basic studies on genomics have been applied for the design of a clinical test – which is already available in the market – for the rapid identification and differentiation of M. tuberculosis complex species, including BCG strains, in clinical isolates (see chapter 14).

Genotyping tools have gained a well-deserved place in national TB control programs (see Chapter 9). Indeed, molecular epidemiology studies have a direct impact on the surveillance of TB transmission and serve to adjust control strategies, not only in industrialized countries (van Doorn 2006) but also in medium- and poor-resource countries (Crampin 2006, Godreuil 2007, Palmero 2003, Prodinger 2007, Ramazanzadeh 2006, Villarino 2006).

DNA strain typing is also a powerful tool for quality control of culture in diagnostic laboratories (Martínez 2006). Its role in national TB programs of medium- and poor-resource countries has not been sufficiently stressed. Still today, the issue is annoying for certain bacteriologists, who still feel that culture is infallible and tend to be reluctant to acknowledge laboratory error. For example in a study performed in Argentina, false positive cultures due to laboratory cross contamination were confirmed by genotyping in 25 out of 26 suspected events investigated in 12 laboratories within the national TB network between 1996 and 2003. The contamination rate of positive cultures was 3 % – similar to rates reported in industrialized countries – in the only network laboratory that performed continuous surveillance of its occurrence (Alonso 2007).

Other challenging issues raised by molecular epidemiology studies are related to reinfection (Chiang 2005, Shen 2006, van Rie 2005) and multiple infection (Garcia de Viedma 2005, Shamputa 2006, van Rie 2005), loss of strain fitness associated with drug resistance (Gagneux 2006, Toungoussova 2004, van Doorn 2006), differential virulence and immunopathogenesis (Dormans 2004, Lopez 2003, Manca 2004, Manca 2005, Reed 2004, Reed 2007), tissue or organ affinity (Caws 2006), vaccine development and protection (Abebe 2006, Castanon-Arreola 2005, Grode 2005, see chapter 9). Even the results of genomics rely upon the lineage of the few strains than have so far been sequenced (see chapter 4).

As described in chapter 9, a re-formulated version of the Variable Number of Tandem Repeats (VNTR) typing based on 15 Mycobacterial Interspersed Repetitive Units (MIRU) loci is emerging as the tool of choice – and probably the next gold standard – for M. tuberculosis genotyping, at least for the near future (Supply 2006). Unfortunately, this stimulating prospect poses a practical problem for laboratories in medium- and low-resource countries that managed to perform strain typing during the ’90s. The manual VNTR-MIRUs procedure is still too labor-demanding, while high-throughput techniques – entirely or partially automated – are not affordable for most laboratories in such countries, at least in the short or medium term. In addition, these laboratories have little chance of translating the data gathered in existing databases on RFLP and spoligotype information into the new MIRUs language, at a reasonable pace. Thus, as long as science continues to advance, the scientific gap between industrialized and developing countries will widen.

In spite of the impressive advances made in the field with the existing tools, the ideal method for strain typing has not yet been achieved (see chapter 9). Microarray techniques are envisaged as the true future tools for DNA strain typing. Hopefully, VNTR-MIRUs and/or microarray genotyping methods will become standardized and their use generalized so that the increased demand for these techniques will contribute towards lowering their cost to become affordable worldwide.

20.3. New perspectives in diagnosis

It is now 125 years since the tubercle bacillus was described by Robert Koch. Disappointingly, the diagnosis of the disease still relies on the same microscopy technique based on the specific Ziehl-Neelsen staining of the bacillus, which was already available soon after that fundamental discovery. Much more progress needs to be made in obtaining better and faster diagnostic methods. Indeed, in high-burden resource-poor countries, where TB is a major public health problem, the diagnosis of active disease is mainly performed by direct microscopic examination of sputum-smear samples. This technique, although simple and inexpensive, lacks sensitivity in comparison to M. tuberculosis culture. Several modifications – mainly based on concentration and centrifugation techniques – have been proposed to improve the sensitivity of sputum-smear microscopy, with varying results. A recent systematic review and meta-analysis has shown that specificity does not vary substantially between different methods, but sensitivity can be improved. By comparison with direct smears, centrifugation and overnight sedimentation are more sensitive when preceded by any of several chemical methods, including the bleach method (Steingart 2006a). No other major improvement has been obtained in the classical staining method based on the Ziehl-Neelsen technique developed many years ago.

Fluorescent microscopy proved to be faster and more sensitive than conventional microscopy based on Ziehl-Neelsen staining, and is the standard diagnostic method in high-income countries (Steingart 2006b). It has the additional advantage of demanding less effort from the laboratorist, thus reducing fatigue and human error. As for low-income countries, the eventual introduction of fluorescent microscopy should be evaluated carefully because it requires a more expensive microscope and a more complex technique. It should be noted, however, that the main burden of fluorescent microscopy lies in the maintenance of the mercury lamp, rather than in the initial cost of the equipment. Lately, an inexpensive device has been released onto the market that can be adapted to any fluorescence microscope. It has a long-life, low power consumption ‘Royal Blue’ Luxeon light-emitting diode that is used in place of the high-cost, short-lived, and environment-unfriendly mercury vapor lamp (Anthony 2006). In fact, this lamp lasts 200 times as long, and costs 10 times less than the vapor mercury lamp. This form of illumination is suitable for the detection of auramine O-stained bacilli and may become an affordable alternative for improving diagnostic microscopy in laboratories serving poor-income settings with a high load of smear examinations (Van Hung 2007).

Fluorescein diacetate staining was recently evaluated for assessing bacilli viability in sputum smears. It has been proposed for the early and accurate detection of TB treatment failure in poor-income settings with a high TB burden (Hamid Salim 2006).

Cultivation of M. tuberculosis is the gold standard for the diagnosis of active TB in the laboratory. This has been traditionally performed in egg- or agar-based solid media. Although slow and time-consuming, it is relatively simple to perform and rather inexpensive in most settings. Newer alternative methods based on liquid culture media and giving faster results – such as the BACTEC radiometric method, BACTEC MGIT960 and BactT/Alert – have proven useful, especially in medium- and high-income countries with the necessary resources to use them routinely (Scarparo 2002). It is now standard recommendation that the combination of a solid and a liquid culture medium gives the best sensitivity in recovering mycobacteria in primary culture (Tenover 1993).

The development of many new nucleic-acid amplification techniques, such as the polymerase chain reaction (PCR), opened new possibilities for the rapid diagnosis of many infectious diseases (Yang 2004). However, when applied to TB diagnosis, only two methods, the Amplicor Mycobacterium Tuberculosis Test (Amplicor) (Roche Diagnostic Systems, Inc., New Jersey) and the Amplified Mycobacterium Tuberculosis Direct Test (MTD) (Gen-Probe, California), both commercially available, have received approval by the US Food and Drug Administration (FDA) for direct application in clinical samples (Centers for Disease Control 2000). Compared to culture and clinical status, nucleic-acid amplification tests have high sensitivity and specificity in smear-positive samples. However, lower values are obtained in smear-negative specimens, precluding their use as a screen to rule out the disease. The current recommendation is that molecular tests should always be interpreted in conjunction with the patient’s clinical data (Pfyffer 2003). More recent approaches combining PCR amplification and fluorescent probe detection in the same tube, such as real-time PCR technology, are promising to give improved sensitivities and specificities. More evaluations in target populations are needed to assess the real impact on the diagnosis of the disease (Espy 2006, Savelkoul 2006).

Until recently, the tuberculin skin test was the only available test for the diagnosis of latent TB infection. Based on the detection of delayed-type hypersensitivity to purified protein derivative (PPD) obtained from M. tuberculosis, it measures the size of the induration produced after sub-cutaneous inoculation of a standardized dose. Since PPD is actually a raw mixture of several antigens shared by M. tuberculosis, M. bovis BCG and several non-tuberculous mycobacteria (NTM), a positive tuberculin skin test result could be due to latent TB infection, previous BCG vaccination, or previous exposure to NTM. Additionally, it has various disadvantages, such as variability in the interpretation by different readers, the need of some experience to correctly interpret the result, and the requirement for the patient to return after 48-72 hours for test reading.

In order to overcome these disadvantages, immune-based blood tests have been recently introduced to detect latent TB infection (see Chapter 13). Interferon-γ (IFN-γ) assays measure the amount of IFN-γ produced or the actual number of IFN-γ-producing T lymphocytes in response to specific antigens. Both approaches are based on the fact that T cells sensitized with tuberculous antigens will produce IFN-γ when they are re-exposed ex vivo to mycobacterial antigens; a high amount of IFN-γ production is then presumed to correlate with TB infection (Pai 2004). The current IFN-γ assays use more specific antigens to M. tuberculosis than PPD: the early secretory antigen target 6 (ESAT-6), and the culture filtrate protein 10 (CFP-10). Both proteins are coded by genes located in the region of difference 1 (RD1) of the M. tuberculosis genome and are not shared with M. bovis BCG or most NTM, with the exception of M. marinum, M. sulzgai and M. kansassi (Andersen 2000). There are many studies evaluating IFN-γ assays in different populations. Based on published evidence, the T-SPOT TB assay seems to be more sensitive than the QuantiFERON-TB Gold test in detecting latent as well as active TB infection. However, the absence of a real gold standard for the diagnosis of latent TB infection prevents a more definitive conclusion. Further studies comparing these two assays are needed, especially in immunosuppressed patients (Richeldi 2006).

It should be remarked that TB diagnosis in endemic countries has so far gained little benefit from scientific progress, namely biotechnological developments (Perkins 2006a, Perkins 2006b). There is an urgent demand for a field-friendly test, ideally, a point-of-care one able to diagnose the disease on the spot in order to avoid delays in diagnosis, thus, preventing further transmission and reducing complications. This type of test is particularly useful when patients do not return for care and would greatly benefit people in settings such as prisons, homeless shelters, and clinics for migrant workers who have no ready access to, or do not seek, public health service assistance. The ideal diagnostics for TB should be available in one hour and should require no electricity, refrigeration, fresh water supply or highly trained personnel (Keeler 2006).

Serological tests – aimed at the detection of either antigens specific to, or antibodies directed against M. tuberculosis – would in principle provide a choice platform for a point-of-care diagnostic tool. However, as commented in chapter 13, there are still many unsolved hindrances to this approach. In particular, in the development of tests for antibody detection, careful attention should be paid to the selection of the target group and the control population groups for performance evaluation. Inclusion and exclusion criteria should be quite stringent regarding age range, geographical location, previous exposure to M. tuberculosis and environmental mycobacteria, tuberculin skin test status, previous BCG vaccination, and unrelated pathologies for differential diagnosis. Also, the sensitivity of the test should be evaluated in the actual target population, namely patients with (pulmonary and/or extrapulmonary) paucibacillary TB rather than in acid fast bacilli smear-positive pulmonary cases. Unfortunately, some serological tests are being marketed in developing countries without a proper on-site assessment. Another issue to consider is the genetic diversity, not only between hosts but also between M. tuberculosis lineages, which might render a promising test only suitable for a restricted geographical region (Lopez 2003). In order to improve performance, a comprehensive set of purified, well-characterized antigens should be investigated, searching for differences in patterns of response rather than comparing responses to individual candidate antigens.

A quite different approach that utterly fulfills the requirements of the point-of-care diagnosis is based on the electronic nose technology, which is able to detect and identify tiny amounts of virtually every substance in a few minutes. The device can be assembled as the sensory part of a portable artificial intelligence system, able to detect several microbes simultaneously through their specific “odors.² Such a system could be used to investigate the agent either directly in the patient’s breath or in a swab of a specimen obtained from any bodily site. This highly suited technology has already been reported for other bacterial human pathogens but is still awaiting development for TB diagnosis (Dutta 2006).

Lastly, diagnostic methods currently under development and expected to be available in two to four years include a dipstick PCR, the detection of mycobacterial proteins in urine, and a blood antibody test, among others (Marris 2007).

20.4. The problem of drug resistance detection

Traditionally, drug resistance in TB was assessed by culturing M. tuberculosis on solid media in the presence of antibiotics and measuring growth by enumeration of colonies. This methodology, although simple to perform and rather inexpensive, is quite slow and laborious, requiring several weeks to give the final results (Heifets 1999). Many alternative approaches and methods have been proposed, some of which have already been presented in Chapter 19 (Palomino 2007, Piersimoni 2006). The most important consideration before they can be implemented in the routine diagnostic laboratory is that they are better and faster than the currently available methods and that they have been properly evaluated and have shown high accuracy in target populations.

Several molecular tools have also been developed and proposed as rapid methods to detect drug resistance (Garcia de Viedma 2003). They search for genetic determinants of resistance rather than for the resistance phenotype, and involve molecular nucleic acid amplification by PCR and detection of amplified products for specific mutations correlating with drug resistance. Molecular methods have several advantages over culture-based techniques: shorter turnaround time, no need for growth of the organism, the possibility for direct application in clinical samples, less biohazard risks, and feasibility for automation. However, not all molecular mechanisms of drug resistance are known. In most cases, molecular methods have been directed towards detecting resistance to rifampicin for two major reasons. First, rifampicin resistance is a good surrogate marker for treatment failure and, in settings with a high prevalence of drug resistance, for multidrug resistance. Second, the associated mutations are well defined, restricted to a short chromosomal segment, and their prevalence is sufficiently known worldwide (see Chapter 19). Current developments aim at the simultaneous identification of M. tuberculosis and detection of resistance to two or more key drugs. The desideratum would be to achieve identification and multiple drug resistance detection directly on clinical specimens, thus avoiding the delay implied in culturing the bacilli (Cavusoglu 2006, Kim 2006, Marin 2004, Park 2006, Sekiguchi 2007, Somoskovi 2006, Yang 2005). Desoxyribonucleic acid (DNA) sequencing of amplified products remains the reference standard to which new molecular tools are compared.

20.5. On drug development

Associated with the problem of drug resistance is the search for new anti-tuberculosis drugs. As mentioned in previous chapters of the book, almost no new anti-tuberculosis drug classes have been developed over the last 40 years. In fact, once the industrialized countries felt confident in accomplishing TB control, the leading pharmaceutical industries lost interest in the development of anti-tuberculosis drugs. This lack of investment became evident when the HIV/AIDS pandemic emerged, soon followed by MDR-TB and the unavoidable interactions between anti-tuberculosis and anti-retroviral drugs (see Chapter 17). In view of this, interest in the discovery of new drugs against TB was awakened towards the end of the millennium. Many candidate compounds have been considered in the last decade, but very few of them have entered into further evaluations. These potentially useful anti-tuberculosis drugs are currently in different stages of the evaluation pipeline.

The Global Alliance for TB Drug Development ( was established in 2000 to promote the development of new anti-tuberculosis compounds. Its goal is to bring a new anti-tuberculosis drug onto the market by 2010. The most advanced program on new drugs is examining moxifloxacin, and is about to enter phase III clinical trials in multiple centers. The program aims at using this fluoroquinolone instead of ethambutol or isoniazid in the first-line drug scheme of anti-tuberculosis treatment, in order to shorten the current 6-month duration of the treatment (Burman 2006). A similar program is being carried out in Africa, where gatifloxacin, another fluoroquinolone, is also substituted for ethambutol. This combination treatment is currently in Phase III clinical trials aimed at shortening the standard regimen to four months (Anonymous 2006). PA-824 is another pro-drug of the nitroimidazole class that has already passed animal model testing in combined therapy, and is currently undergoing phase I clinical trials (Nuermberger 2006). Interestingly, a protein was described in M. tuberculosis that is involved in both intracellular drug activation and resistance to this drug candidate (Manjunatha 2006). Other promising drugs are in the pre-clinical phase of evaluation. OPC-67683 is a nitro-dihydro-imidazooxazole derivative that inhibits mycolic acid biosynthesis, is free of mutagenicity and is highly active against TB in vitro and in mice (Matsumoto 2006). The diarylquinoline known as R207910, TMC207 or compound J is a new compound exhibiting a completely novel mode of action – inhibition of ATP synthase – and very high activity against M. tuberculosis, M. leprae and M. ulcerans. This promising compound is being tested in phase IIa trials on TB treatment (Andries 2005, Lounis 2006).

The availability of such a spectrum of new drug candidates offers great promise but also entails a great challenge. The role played by each drug must be explored within the frame of a multidrug treatment regimen. In the immediate future, a complete series of clinical trials will be needed to find the optimal treatment scheme of ultra-short duration, i.e. 2 months or even shorter. Innovative strategies must be designed in order to meet this challenge (Spigelman 2006). Glickman et al. have calculated the probability of developing a new anti-tuberculosis drug by 2010 (Glickmann 2005). Applying a Monte Carlo simulation model, they evaluated drug development from the perspective of a public-private partnership, taking into account several factors such as the expected number of successful compounds, the expected costs of each stage of development and the development costs for successful and unsuccessful compounds. As for the currently-available candidate drugs in all stages of development, the probability of at least one successful compound being generated was less than 5 %. Obviously, many more efforts and funding are required to reach the objectives of developing new and successful anti-tuberculosis drugs in the near future.

Research and development is also needed on innovative drug formulations and drug delivery systems aimed at increasing compliance and achieving a high local drug concentration while minimizing systemic side effects. In this respect, the growing field of inhalation therapy offers a very promising new prospect (Chow 2007, Shoyele 2006). A technology based on porous particles for pulmonary drug delivery is already in use for insulin. This technology – presented as a simple, low-cost, disposable, dry-powder inhaler – can be applied to the delivery of anti-tuberculosis drugs (Edwards 2006).

20.6. On vaccine development

Vaccine development is a problematic issue for many reasons. It is not appealing for the industry, because it demands a huge investment, takes a long time and the risk of failure is high. Besides, even in the best scenario, profit margins are meager and the risk of legal prosecution in the event of side effects is high (Rosenthal 2006).

Moreover, major obstacles also lie in the initial vaccine design itself, such as the difficulty in inducing a potent and long-lasting cellular immune response in humans, due to our poor understanding of host-parasite interactions. Vaccines that promise to be more potent than BCG were designed following a rational approach: they are recombinant BCGs over-expressing major M. tuberculosis antigens. Grode et al. used an entirely novel approach, as commented on thoroughly in Chapter 10. Instead of using an antigen, these authors selected the membrane-perforating listeriolysin to construct a recombinant BCG. This hemolytic enzyme, produced by Listeria monocytogenes, allows the agent to escape from the phagosomes of infected host cells. This improves the access of mycobacterial antigens to the major histocompatibility complex Type I pathway, thus resulting in better CD8+ T cell stimulation. At least in the mouse model, recombinant BCG secreting listeriolysin elicited an enhanced protection against a strain of the M. tuberculosis Beijing/W genotype family, while parental BCG failed to do so consistently (Grode 2005). This latter finding should be highlighted because the selection of strains used for challenging any vaccine candidate is not a minor issue. Future vaccines must prove able to protect against the most prevalent, transmissible and/or virulent lineages worldwide, not merely against laboratory-domesticated strains (Lopez 2003).

Another promising approach has gone even further in proving efficacy. The design is based on the fact that viral vectors, such as poxviruses, are powerful at boosting previously primed T-cell responses against intracellular pathogens. McShane et al. successfully applied such an approach by using BCG as the priming immunization – eventually maintaining the beneficial protective effects of BCG against disseminated disease – and a recombinant modified vaccinia virus Ankara expressing antigen 85A (MVA85A) for boosting (McShane 2004). Both CD4+ and CD8+ T cells were successfully boosted in preclinical experiments with MVA85A and also in human volunteers. The vaccine candidate is now in clinical trials in the United Kingdom and Africa. Results from one of these trials showed that the recombinant viral vector vaccine is a strong booster of BCG-primed and naturally acquired antimycobacterial immunity. In fact, this is the first clinical trial showing successful results with a novel subunit TB vaccine. Besides, the strategy is feasible and practical in low-resource high-burden countries (McShane 2005). Most importantly, this pioneer study also raises highly sensitive protocol issues and, in particular, ethical issues (Ibanga 2006).

A provocative finding was reported by researchers at the Institut Pasteur, where the BCG vaccine was first developed in 1924. They gathered a large body of