The Bacille Calmette-Guerin (BCG) vaccine has been for over 80 years the only vaccine used to protect against tuberculosis. Following the World Health Organization recommendations, infants under 1 year of age are vaccinated in many countries. Every year 100 million new born children are vaccinated and the global coverage is about 80%. Nevertheless protective efficacy of BCG varies between 0 and 80%! A 60-year follow-up study of American Indian and Alaska natives reported last year a long-term efficacy of 52%.

Why does BCG efficacy vary?
Epidemiological data and animal experiments point to the following causes:
- strain variation in BCG preparations
- environmental factors
- genetic and nutritional differences between populations

Strain variation in BCG preparations

Deletion analyses of the genomes of different strains of BCG have shown that BCG vaccines have lost important genes during the laboratory attenuation process from the parental strain. This leads to absence (ESAT-6, CF)-10, MPT64) or no expression (MPB70, MPB83) of important proteins present in the parental strain.

Environmental factors
In the tropical regions low efficacy of BCG could be due to population sensitization by environmental mycobacteria, prior to BCG vaccination.

New tuberculosis vaccines
Low BCG efficacy fired the research for a more effective vaccine. About 200 such vaccines have been developed by various research groups, and some are already in clinical trials. These include:
- live attenuated Mtb vaccines
- recombinant BCG
- vaccinia-vectored vaccines
- DNA vaccines
- subunit and fusion proteins in novel adjuvants
- killed BCG and M.bovis

Live attenuated Mtb vaccines
Recently a Mtb strain with defects in mycobacterial lipids (Mtb drrC) has been shown more protective than BCG in mice, but more investigations are needed to ensure that it will not reverse to a virulent state.

Recombinant BCG
An effort is made to introduce additional copies of existing genes or reintroducing some of the genes lost in the attenuation process. rBCG30 is such a vaccine, overexpressing the 30KDa major secreting protein of Mtb, and is already in phase I clinical trial since 2004. Other candidates are on way.

Vaccinia-vectored vaccines
Live vectors, such as the vaccinia virus or recombinant Salmonella, can be modified to express Mtb proteins. Specifically MVA85A (vaccinia virus expressing Ag85A) is already in phase I clinical trial with promising early results.

DNA vaccines DNA vaccines containing plasmids coding for mycobacterial antigens have also been tested. Such candidate antigens are the mycolyl-transferase family (Ag85 complex) and heat shock proteins 60, 65,70. Potency may be increased by coimmunization with cytokines or embedding DNA in various adjuvants. However these vaccines appear to need protein boosting to provide satisfactory protection.

Subunit and fusion proteins in novel adjuvants
Various candidates exist in this category. One of them Mtb72F (fusion protein of Mtb32 and Mtb39 antigens) entered clinical trials in 2004.

Killed BCG and M.bovis
BCG has to survive in the host and propagate in order to confer immunity. So killed BCG is not protective. However, heat-killed BCG proved to be as protective as live attenuated BCG, when combined with the Eurocine L3 (TM) adjuvant. Similar results were obtained with formalin killed M.bovis.

Despite BCG and widespread vaccination, the number of new TB cases is still increasing. It is clear that BCG is not effective in preventing adult pulmonary TB. Since BCG given to neonates is effective in reducing disseminated TB in infants, a reasonable first step would be to introduce new vaccines as booster vaccines for BCG, as a transition strategy, especially in endemic countries.

Images from Greece...