Acute Respiratory Infections (Update September 2009)
Tuberculosis
Introduction
An estimated one third of the world population (two billion people) is exposed to the risk of tuberculosis (TB), whose main causative agent, Mycobacterium tuberculosis (Mtb), infects approximately 9 million new individuals and causes 1.7 million deaths every year [334] [335] . This problem is compounded by the global emergence of multi drug-resistant (MDR) Mtb strains [336] , the evidence that diabetes predisposes people to TB infection, and the increased susceptibility of HIV-infected individuals to TB, which all make the need for improved TB control even more urgent.
The currently available TB vaccine is bacillus Calmette-Guérin (BCG), a M bovis derivative which is routinely used in children because it is effective at protecting them against severe extrapulmonary forms of the disease, such as TB meningitis. BCG, which also elicits protection against leprosy [337] , does not however appear able to provide protection against pulmonary TB in adults [338] [339] . BCG vaccination may provide protection only against primary infection but has a limited effect on reactivation TB in already infected individuals or on TB re-infection in adults. It also has been shown that the "take" of BCG may be impaired due to previous exposure to environmental mycobacteria [340] such as M avium [341] , which are frequent in tropical countries. There is, therefore, an urgent need to develop better or improved TB vaccines, but it would be unethical and impractical to test new vaccine strategies that would not include BCG in early infancy [342] [343] . Therefore, most new TB vaccine development is being done in a BCG prime-vaccine boost strategy.
Disease Burden
The global death toll of TB was 1.7 million deaths in 2006, 200 000 of which were from HIV-associated TB [344].
M tb currently infects about 2 billion people worldwide and causes an estimated 8.8 million new cases every year, especially in the sub-Saharan African continent, in Southeast Asia and in Eastern Europe [345] . In view of underreporting and lack of systematic surveillance, the real incidence of TB is suspected to actually be perhaps as high as 14 million new TB cases each year. About one half of the new cases occur in China, India, Pakistan, Bangladesh, Indonesia and The Philippines. In Africa, the single most important factor determining the increased incidence of TB in the last 10 years is HIV infection: in some regions, up to 75% of new TB cases are in HIV-infected people [346] . Paradoxically, TB seems to be severely aggravated in these dually infected patients when active antiretroviral therapy is initiated.
In addition, about 500 000 new cases of multiple-drug resistant TB (MDR-TB) are reported each year worldwide, with high incidence in parts of Russia, Latvia and Estonia (up to 10% of new TB cases) and Azerbaidjan (22% of new cases). The newly discovered extensively drug-resistant Mtb strains (XDR-TB) emerged in 2005 in KwaZulu Natal (South Africa) among HIV-TB patients, probably as a consequence of lack of observance of therapy [347] [348] [349] . XDR-TB strains are now been found in >45 different countries, mostly in HIV-infected patients [350] . XDR-TB is resistant to almost all drugs used to treat TB, including isoniazid, rifampicin, fluoroquinolones and amikacin, kanamycin or capreomycin [351].
TB is highly contagious. Left untreated, each patient with active TB will infect on average between 10 and 15 people every year. TB spreads readily from person to person, due to the production of small particle droplets when a patient coughs and to the low dose of bacilli needed to initiate infection. Transmission is common in households, schools, hospitals, prisons, crowded work places, refugee camps and shelters. In fact, the incidence of TB in industrialized countries, which generally increased from the early 1980s to the early 1990s, has been decreasing ever since. As an example, in the USA, 13 767 cases were reported in 2006, a 3.2% decrease from 2005 and a 53% decrease from 1992 [352] . However, TB is a poverty-related disease: it has long been recognized that war, malnutrition, population displacement and crowded living and working conditions favor the spread of TB among humans, whereas periods of improvement in societal conditions and hygiene favor its rapid decline.
The global spread of the disease is also facilitated by migrations and movements of populations: thus, in industrialized countries, about one-half of TB cases occur in foreign-born or migrant persons. The incidence of TB in 2003 in France was for example 5.6 cases per 100,000 population in the native population, 31.7 per 100,000 in persons born in North African countries and 187.7 per 100,000 in persons born in TB-endemic sub-Saharan African countries [353] . But the major bottleneck for higher success rates in controlling TB is the fact that currently only about 40% of all sputum-positive TB are detected. Thus, the majority of TB cases remains untreated or is treated only at a very late and highly infectious stage, causing enormous individual hardship as well as creating a public health time bomb. For all these reasons, added to the relative ineffectiveness of the current BCG vaccine, the development of improved TB vaccines has become a necessity for adequate control and elimination of the disease.
Bovine tuberculosis, caused by M bovis, is a zoonotic disease and was the cause of many human deaths in the 1930s and 1940s through consumption of contaminated milk and dairy or meat products. Compulsory eradication programs were introduced in many countries based on the slaughter of infected cattle detected by the intradermal tuberculin skin test. However, probably due to a wildlife reservoir, the incidence of TB in cattle has exponentially increased over the last two decades in certain countries, especially Great Britain, constituting a potential public health problem and calling for the development of more specific diagnostic reagents [354].
Bacteriology
Mycobacterium tuberculosis (M tb), the agent of human TB, was discovered in 1882 by Robert Koch and for a long time called after his name (the 'Koch bacillus'). All members of the Mycobacterium genus share the property of acid-fastness (Ziehl-Neelsen staining), due to their mycolic acid-rich cell wall structure. They include M. tuberculosis, M. africanum, and M. ulcerans, which are primary human pathogens, M. bovis, the agent of TB in cattle and other animals, which also can cause disease in humans, and a great many nontuberculous or environmental species, some of which can be pathogenic in humans such as those belonging to the M. avium-intracellulare complex.
TB bacilli usually multiply first in the lung alveoli and alveolar ducts and in draining lymph nodes, where they are engulfed by dendritic cells (DCs). They also multiply in the macrophages that were attracted from the bloodstream and are armed to kill the bacteria upon uptake in their phagosomes. However, Mycobacteria can evade the phagosome-lysosome fusion pathway, multiply in the host macrophages and kill them, progressively creating a primary tubercle. The outcome of infection is controlled by CD4+ and CD8+ T cells, both of which are necessary for the maintenance of a latent state of infection [355] [356] . Delayed cutaneous hypersensitivity develops and together with other cellular immune reactions, leads to the caseous necrosis of the primary complex. Bacilli eventually spread to many parts of the body such as liver, spleen, meninges, bones, kidneys and lymph nodes, where they can either be a source of overtly disseminated TB or, more commonly, remain dormant. Occasional decline in cell-mediated immunity leads to reactivation TB, most frequently seen in adults as a pulmonary disease with infiltration or cavity in the apex of the lung. This is the most common and most infectious form of TB.
CD4+ T-cells play a major role in containment of TB infection; progressive TB is usually associated with a Th2 T-cell response, whereas a pure Th1 response, including production of IL-2 and IFN-?, mediates protection [357] . CD8+ T cells and macrophages are involved in the control of mycobacterial infection [358] [359] . The production of Th1 associated cytokines such as IFN?, TNF? and IL-12 appears to be an essential component of resistance to Mtb infection. The most effective vaccination strategies in animal models have been those that stimulate T cell responses, both CD4+ and CD8+, to produce these cytokines.
The tuberculin skin test has long been used as evidence of TB infection or as a sign of adequate response to BCG vaccination, although no clear relationship between delayed-type hypersensitivity and protective immunity could be established [346] . A number of antigens found in M. tuberculosis, including Ag85, MPT64, ESAT-6 and CFP10, have been identified which may play a major role in cellular immunity and the induction of a protective IFN-? response.
Vaccine
BCG
By culturing a M. bovis isolate from a cow for a period of 13 years and a total of 231 passages, Calmette, a physician, and Guérin, a veterinarian, created an attenuated variant of M. bovis, Bacille Calmette-Guérin (BCG). BCG was first tested in infants as an oral vaccine in 1921. New methods of administration were later introduced, such as intradermal, multiple puncture, and scarification [360] . BCG vaccination was included in the WHO Expanded Programme on Immunization (EPI) as of 1974: since then, approximately 100 million children received a BCG vaccine each year, and, to date, well over 4 billion people have been vaccinated with BCG since the historical immunization in 1921.
However, WHO stopped recommending BCG vaccination at birth as of 2007, at least for infants at risk for HIV infection, as data showed that about 417 per 100 000 infants developed disseminated BCG disease due to antenatal infection with HIV [361] . Disseminated BCG disease typically presents with the same symptoms as severe TB and shows a CFR of 75%-86% [362] [363].
As shown by sequencing, the original BCG strain lost the RD1 region of the M. bovis genome in the course of the selection process. This region encodes major TB antigens including culture filtrate protein 10 (CFP-10) and early secretory antigen target 6 (ESAT-6) [364] [365] [366] [367] . Major BCG vaccine strains in use today differ even further from the original BCG strain and from each other, with "stronger" strains (Pasteur 1173 P2, Danish 1331) being more reactogenic and, presumably, more immunogenic, than "weaker" strains (Glaxo 1077, Tokyo 172) [346] [368].
No other widely used vaccine is as controversial as BCG. Its effects in large randomized, controlled, and case-control studies, have been widely disparate, from excellent protection against TB to no protection. Most studies have demonstrated that BCG vaccines afford a higher degree of protection against severe forms of TB, such as meningitis and disseminated TB, than against moderate forms of the disease. The efficacy of neonatal BCG vaccination also wanes with age [369] . Studies that evaluated meningitis or miliary TB demonstrated that BCG can provide good protection against these serious forms of TB in young children, with reported efficacy ranges from 46-100%. In contrast, efficacy against pulmonary TB, which is more prevalent in adolescents and adults, has ranged from 0-80%. Overall, BCG is only at best credited with a 50% efficacy [370] [371].
Efficacy of BCG vaccination also appears to vary with geographic latitude - the farther from the equator, the more efficacious the vaccine. Presumably, exposure to nonpathogenic mycobacteria, which is more intense in warm climates, induces a degree of protective immunity in exposed populations, masking potential protection from BCG.
Vaccination with BCG nevertheless remains the standard for TB prevention because of its efficacy in preventing life-threatening forms of TB in infants and young children, and also because it is the only vaccine available, is inexpensive, and requires only one encounter with the baby. The fact that BCG does not protect against pulmonary TB in adults has however prompted the search for new, improved TB vaccines. Dozens of TB vaccine candidates have been tested in recent years in animal models, including subunit protein and peptide vaccines, DNA vaccines, rationally attenuated Mtb strains, recombinant BCG, and live vectors expressing immunodominant Mtb antigens (for reviews, see [372] [373] [374] [375]).
Among the many innovative new approaches that have been tested, one has been to reengineer BCG strains to endow them with the capacity of expressing immunodominant Mtb antigens. This approach has yielded a series of vaccine candidates, such as:
- A recombinant BCG vaccine (BCG30) that was engineered at University of California at Los Angeles (USA) to express the 30 kD major secretory protein Ag85B [376] [377] . The vaccine was tested in Phase I trial in the USA and elicited markedly enhanced central and effector memory Ag85B-specific CD4+ Th1 and CD8+ T cell immunity compared with the parental BCG Tice strain and induced a significant number of Ag85B-specific T cells capable of inhibiting intracellular mycobacteria [378]. The development of this vaccine is currently on hold for regulatory concerns.
- A BCG::RD1 recombinant, in which the RD1 segment of the M. tuberculosis genome has been reintroduced, resulting in the expression of ESAT-6 and Ag85A proteins. This new BCG strain, developed at the Pasteur Institute, Paris, showed increased persistence and improved protection against challenge with virulent M tb in animal models [379] but meets with safety concerns for its use in humans.
- Another recombinant BCG, VPM 1002 (previously known as rBCG: [delta] ureC-Hly), was engineered at the Max Planck Institute for Infection Biology in Berlin (Germany) to express listeriolysin O, which increases MHC class I presentation [380] , and its urease gene was deleted in order to prevent neutralization of the acidic pH in phagosomes. This recombinant BCG was found to be devoid of pathogenicity for SCID mice and provided greatly improved protection against aerosol TB in the mouse model. This vaccine is currently phase I clinical trials.
- AERAS-422, developed by the Aeras Global TB Vaccine Foundation, which expresses Mtb antigens 85A, 85B, and Rv3407, together with perfringolysin, a pore-forming protein similar to listeriolysin, is currently in late preclinical development and planned to enter into Phase I trials by the end of 2010 [381].
The VPM 1002 and AERAS-422 vaccines have been found to be more efficacious than classical BCG against Mtb challenge in animal models, including against challenge with the Beijing strain of Mtb which is relatively insusceptible to BCG [382].
Live attenuated Mtb mutants
Another approach has relied on the engineering of live attenuated mycobacterial strains, either auxotrophic mutants of Mtb [383] [384] , or regulatory mutants such as the PhoP knock-out mutant [385] . PhoP is a gene which controls the expression of several virulence genes of Mtb. The PhoP/PhoR Mtb mutant has shown very good protection in mice and was one of the rare new TB candidate vaccines to show significant improvement over BCG in the guinea pig aerosol model [386] . Concern that auxotrophic mutants could revert to full virulence in vivo has led to the engineering of mutants bearing two independent unlinked deletions [387].
Subunit and DNA vaccines
Due to safety concerns, in particular in immunocompromised persons, as well as to technical challenges regarding manufacture and reproducibility, live mycobacteria vaccines are not the product of choice of most vaccine manufacturers. Many new TB vaccines approaches are therefore focused on recombinant subunit vaccines, DNA vaccines [388] , or live-vector recombinant vaccines that express a variety of M tb antigens.
An analysis of the Mtb proteome led to the identification of about 45 potential vaccine antigens. Among those, purified mycobacterial antigens such as Ag85A, Ag85B, HSP65, the R8307 protein, a 36kD proline-rich mycobacterial antigen, or the 19kD and 45kD proteins have been found to induce protection levels in mice similar to that obtained with BCG when presented in combination with Th1-inducing adjuvants [389] [390] . Subunit vaccines based on the same proteins in adjuvants used in combination with BCG have resulted in better protection against experimental challenge of cattle with M bovis than BCG vaccine on its own [391].
Most promising among these were Ag85 (A and B), and the Mtb9.8, Mtb9.9, Mtb11, Mtb32, Mtb39, Mtb41 and ESAT-6 proteins [392] [393] [394] [395] , which were prioritized for vaccine development based on their ability to stimulate PBMC responses from more than 50% of healthy PPD+ donors who presumably have contained their infection due to protective CD4+ T cell responses [342] [396].
Some of these proteins have been fused together, such as Mtb32 and Mtb39, yielding M72, which was found to be highly immunogenic and protective in monkeys when formulated in the AS02A adjuvant from GSK, or antigens 85B and ESAT-6, whose fusion product also induced strong protective immune responses in monkeys [397] [398] . ESAT-6 could advantageously be replaced by TB10.4 [399] . M72 is being developed by Corixa Corp and GSK and was found to be safe and immunogenic in Phase I clinical trial [400].
Antigen 85A, a mycolyl-transferase that is involved in cell wall biosynthesis, is also considered a leading candidate for inclusion in a TB vaccine. In addition, a multi-epitope polypeptide, as well as nonproteinic antigens such as mycolic acids and carbohydrate moieties, are being developed as candidate antigens.
Live recombinant TB vaccines.
Modified vaccinia virus Ankara (MVA) was engineered to express M tb Ag85A (MVA85A) and tested successfully in a BCG prime-recombinant MVA boost strategy in the mouse challenge model, the guinea pig aerosol challenge model and the cynomolgous macaque challenge model [386] , before undergoing Phase I clinical studies first in the UK, then in a high TB endemicity setting in The Gambia [343] [401] . The MVA recombinant was found to boost BCG-primed and naturally acquired anti-mycobacterial immunity. Interestingly, the MVA85A vaccine induced up to 30-fold greater cellular immune responses in BCG-primed than in BCG-naive individuals, even if BCG had been administered as long as 38 years before the MVA boost. Another Phase I clinical trial recently took place in South Africa [402] . The BCG/MVA85A prime-boost strategy was also tested in cattle, resulting in significantly wider Ag85A-specific T cell responses and higher frequencies of Ag85-specific IFN-? secreting cells [403] . Currently, the prime-boost strategy is being tested in a Phase IIa trial in South Africa, and a Phase IIb trial is planned to begin in early 2009 on 4 months-old infants who had a BCG vaccination at birth. The MVA85A vaccine is currently the most advanced candidate TB vaccine [404].
Live, nonreplicative adenoviruses expressing M tb antigens have been developed by the Aeras Global TB Vaccine Foundation (USA) and Crucell (Netherlands). Advanced TB vaccine candidates in this category include AERAS 407, a replication-defective adenovirus 35 construct that expresses Mtb Ag85A, Ag85B and TB10.4, and went through Phase I trial in South Africa and South America; and AERAS X05, a Shigella-delivered recombinant double-stranded RNA encoding Ag85A, Ag85B and Rv3407, which is expected to enter clinical trials shortly. The immunogenicity of a BCG-AdAg85A prime-boost regimen in cattle was found to be significantly greater than that of either vaccine separately [405].
It is clear that, from now on, human clinical studies will act as the principal driving force for the development of new TB vaccines. Testing of such a wide variety of vaccine types using different immunization strategies directed against a sole pathogen is unique in the history of vaccine development. It will make the valid comparison of clinical data most challenging [406] . Therefore, it will be all the more important that this effort be tightly coordinated to provide maximal comparability and transparency. WHO is working with all stake holders in the field to standardize key parameters such as trial entry criteria, endpoints, immunoassays, etc. The main players in this area include the US National Institute for Allergy and Infectious Diseases (NIAID), the Aeras Global TB Vaccine Foundation, supported by the Bill and Melinda Gates Foundation, the Welcome Trust, a network of European researchers supported by the European Commission, the European and Developing Countries Clinical Trial Partnership, and pharmaceutical manufacturers including GlaxoSmithKline (GSK), IDRI-Corixa, Crucell, SanofiPasteur and Emergent Biosolutions. Discussions are ongoing between these major players to develop a plan for the design and funding of Phase III efficacy trials in appropriate field trial sites in Africa and Asia [375].