Specific central memory CD4+ T cells, defined by CCR7 expression, were virtually undetectable 2 months after vaccination. A change to central

memory phenotype may occur at a later post-vaccination time and this will be explored in future studies. In MVA85A-vaccinated subjects from the UK, Ag85A-specific T-cell proliferation peaked 6 months post-vaccination 32. Interestingly, in mice MVA-induced CD8+ T cells mostly convert to a central memory phenotype within weeks of immunization 44, suggesting that the rate of conversion to central memory cells may differ between species. In other human studies, we have also consistently observed predominant effector phenotypes of human mycobacteria-specific CD4+ T cells in infants 33, 45 and adults 20. Mycobacteria-specific CD4+ T cells from children with latent M.tb infection or active TB 16, and chronically HIV-infected adults with latent M.tb infection 46, FK506 also display this phenotype. Long-lived central memory cells prevail when Ag is cleared after vaccination, e.g. after tetanus toxoid vaccination 42, whereas chronic CMV, EBV or HIV infection is associated with predominance of effector memory cells 47. One might hypothesize

that chronic exposure to mycobacterial Ag is responsible for our observed phenotype. Adolescents with latent M.tb infection, one potential source of such chronic exposure, were not enrolled to into our study. Additional studies are required to dissect this further. No serious adverse events were recorded, and mild local reactions at the vaccination site were Syk inhibitor predominant. These reactions were commonly reported in the first week after vaccination, did not interfere with daily activities and did not persist. Systemic reactions were uncommon and included mild flu-like symptoms. Clinically, there were no major differences between the adolescents’ and the children’s experience in the trial with a slightly increased incidence of non-vaccine-related systemic events reflecting

this younger age group’s increased risk of transient viral illnesses. This complements the good safety profile of MVA85A found in healthy adults from the same region 25, the United Kingdom 36 and The Gambia 24, as well as other recombinant MVA being tested in clinical trials 40, 48. Together these small phase I/II trials demonstrate a very promising safety profile of MVA85A, which is now being assessed in larger groups of participants, in an infant, phase IIb safety and efficacy study. In conclusion, MVA85A was found to be safe and highly immunogenic in TB-naïve, HIV-uninfected adolescents and children. The vaccine induced durable, polyfunctional CD4+ T-cell responses with a CCR7− effector memory phenotype. These data support future studies to evaluate the efficacy of this vaccine to prevent TB.