Vaccines to prevent tuberculosis infection rather than disease: Physiological and immunological aspects

There is increasing enthusiasm and optimism that a vaccine could be developed that prevents infection rather than disease. In this article I discuss the fact that despite this optimism nothing has been produced so far that seems to have this capability, and moreover even the borderline between when infection ends and disease begins has not even been defined. To be effective such a vaccine, or at least the immunity it would generate, would have to work within the confines of the pulmonary physiological systems, which are complex. To date much of the emphasis here has turned away from T cell mediated immunity and towards establishing specific antibodies in the lungs. Here, I argue that with the exception of a possible exclusionary function, most claims of a protective role of antibody are completely over-blown. Finally, even if we had a potential "anti-infection" vaccine, how would we test and validate it?