This study set out to determine the duration of the immune response induced by BCG vaccination in teenagers in the UK, up to 3 years after teenage vaccination or 14 years after infant vaccination. The immune response was measured by IFN-γ response in diluted whole blood culture following 6 days stimulation with M.tb PPD, M.tb Antigen 85, or an appropriate control.
Our results show the effect of infant vaccination in the UK can persist for at least 14 years as measured by the in vitro IFN-γ response to M.tb PPD. The magnitude of the measured response, while much higher than that of unvaccinated children, was likely to be lower than the initial response to vaccination, as our study of teenagers showed evidence of highest responses at 3 months after vaccination, with a reduction to 0.6 of the 3 month response by 12 months, and a further reduction to 0.6 of the 12 month response at three years.
In addition to M.tb PPD, we measured responses to Ag85 isolated from M.tuberculosis. This antigen has homologues in all known species of mycobacteria and has been identified as a prominent target of the immune response against mycobacteria. It is currently a component of several new tuberculosis vaccines. We have evidence from a previous study that IFN-γ responses to M. bovis BCG Ag85 are highly prevalent among UK teenagers before they receive BCG vaccination (manuscript in preparation). A lack of response to M.tb Ag85 in UK cord blood samples, with increasing prevalence of response in blood from BCG-unvaccinated infants over the first year of life (manuscript in preparation), also point to a common environmental exposure to this antigen in the UK. In spite of this, a vaccine-induced increase in response to M.tb Ag85 was observed in our vaccinees, though with a different kinetic of response to that seen to M.tb PPD, the response was maintained at 12 months at its 3 month level. This is likely to follow an initial peak of response in the first month following vaccination, as studies of vaccination of adults from the UK with MVA-85A (and BCG) using ex vivo ELISPOT to Ag85A peptides (and PPD) as the measure of response found a marked fall in response after this initial period . In animal model systems it has been demonstrated that prior exposure to different environmental mycobacteria can have an effect on the response to subsequent BCG vaccination either blocking the proliferation of BCG  or masking the effect of the BCG by providing as much protection as the vaccine . In the case of Ag85, it may be that repeated environmental exposure to a source of Ag85 maintains response to this antigen and contributes to the persistence of the response to M.tb PPD, of which Ag85 is a major component, so contributing neither to masking nor blocking but to persistence of the response.
Teenagers who had received BCG vaccination in infancy showed no difference in response to Ag85 compared to unvaccinated teenagers at a low threshold of response (< 62 pg/ml), but there was a suggestion that infant BCG vaccination leads to more prevalent high (> 500 pg/ml) responses to Ag85 14 years after vaccination (50% of vaccinated versus 22% unvaccinated), though it should be emphasised this is based upon small numbers and not statistically significant. Larger group sizes would be needed to see if a BCG-induced response to Ag85 persists over many years. Exposure to an environmental species which naturally sensitises individuals to Ag85 makes a persistent vaccine effect more difficult to ascertain with increasing age.
An additional finding from this study was that IFN-γ responses to Ag85 were increased among control subjects at 3 months but not at 12 months. As the same batch of the Ag85 preparation was used at each time point this could reflect some boosting effect by the tuberculin test. Boosting effects have been documented when testing on the day of TST reading or within one month post TST though none using the same assay or at the same time point as in our study . Further human studies of this effect would provide valuable insights.
A recent study has demonstrated that vaccine-induced T-cell response to smallpox antigens can persist for as long as 67 years, in the absence of environmental boosting effects . The authors concluded that this long term memory was attributable mainly to the central memory T-cell population; the effector memory population waned much more rapidly although it could be restored by revaccination. The same study examined the IFN-γ response to PPD, by ELISpot assay, among ten individuals aged 25–63 years, who had received BCG vaccination in childhood, and found it to be maintained both in central and effector memory compartments. The authors conjectured that environmental mycobacterial exposure helped to maintain this response, but no unvaccinated controls were included in the study. Our observed reduction in the magnitude of the peripheral blood IFN-γ response could represent a contraction of the effector memory population with a residual persistent central memory population. We know that at least some individuals in this UK population are exposed to environmental mycobacteria [10, 13], which may provide a low-level boosting effect and maintain a small peripheral memory T-cell population to mount the observed IFN-γ response to mycobacterial antigens. Although the unvaccinated control group of teenagers in our study showed some prior sensitivity to M.tb PPD, as expected , the children vaccinated in infancy showed a greatly increased IFN-γ response to M.tb PPD as compared to unvaccinated controls. It may be that environmental exposure boosts pre-existing vaccine-induced responses, as well as inducing a response per se in unvaccinated individuals.
BCG is the only currently available licensed vaccine against TB, and is typically administered in infancy, including now in the UK. With the worldwide resurgence of TB, it is important to assess longevity of protection afforded by BCG. Information about the induction and maintenance of anti-mycobacterial immune mechanisms by BCG is also relevant for new TB vaccines currently in development  as several of these vaccines are designed to act as a booster vaccination following BCG, and such information may inform decisions about the best timing for such boosting.