Immunogenicity of the Tuberculosis Vaccine MVA85A Is Reduced by Coadministration with EPI Vaccines in a Randomized Controlled Trial in Gambian Infants
Martin O. C. Ota1,*, Aderonke A. Odutola1, Patrick K. Owiafe1, Simon Donkor1, Olumuyiwa A. Owolabi1, Nathaniel J. Brittain2, Nicola Williams3, Sarah Rowland-Jones1, Adrian V. S. Hill2, Richard A. Adegbola4 and Helen McShane2
+ Author Affiliations
1Bacterial Diseases Programme, Medical Research Council Unit, P. O. Box 273, Banjul, The Gambia.
2Jenner Institute, University of Oxford, Old Road Campus Research Building, Roosevelt Drive, Oxford OX3 7DQ, UK.
3Centre for Statistics in Medicine, Wolfson College, Linton Road, Oxford OX2, UK.
4Bill & Melinda Gates Foundation, Seattle, WA 98102, USA.
↵*To whom correspondence should be addressed. E-mail: firstname.lastname@example.org
New tuberculosis vaccines are urgently needed to curtail the current epidemic. MVA85A is a subunit vaccine that could enhance immunity from BCG vaccination. To determine MVA85A safety and immunogenicity as well as interactions with other routine vaccines administered in infancy, we randomized healthy 4-month-old infants who had received Bacille Calmette-Guérin at birth to receive Expanded Program on Immunization (EPI) vaccines alone, EPI and MVA85A simultaneously, or MVA85A alone. Adverse events were monitored throughout. Blood samples obtained before vaccination and at 1, 4, and 20 weeks after vaccination were used to assess safety and immunogenicity. The safety profile of both low and standard doses was comparable, but the standard dose was more immunogenic and therefore was selected for the second stage of the study. In total, 72 (first stage) and 142 (second stage) infants were enrolled. MVA85A was safe and well tolerated and induced a potent cellular immune response. Coadministration of MVA85A with EPI vaccines was associated with a significant reduction in MVA85A immunogenicity, but did not affect humoral responses to the EPI vaccines. These results provide important information regarding timing of immunizations, which is required for the design of infant efficacy trials with MVA85A, and suggest that modifications to the standard EPI schedule may be required to incorporate a new generation of T cell–inducing vaccines.
Globally, tuberculosis (TB) continues to be a disease of major public health importance (1) despite widespread coverage by Bacille Calmette-Guérin (BCG) vaccination; the persistent TB disease burden underlines the inadequacy of the BCG vaccine to protect against TB in adults. As such, there is an urgent need for a more effective vaccine (2). However, BCG protects against severe forms of TB in childhood and has nontargeted beneficial effects, including improving child survival, increasing responses to hepatitis B and oral polio vaccines administered in infancy, preventing allergic disorders, and protecting against leprosy (3–5). Consequently, it is logical to deploy a vaccination strategy against TB that includes BCG.
MVA85A is a TB vaccine designed to enhance BCG (6, 7). Modified vaccinia virus Ankara (MVA) is a strain of vaccinia virus that is replication-incompetent in human cells. MVA85A is a recombinant MVA strain that encodes the immunodominant mycobacterial antigen 85A (8). When administered after BCG, MVA85A can provide enhanced protection against TB challenge in animal models (9). MVA85A has been shown to be safe and highly immunogenic in clinical trials in adults in the UK, Gambia, and South Africa (6, 7, 10, 11).
The prevalence of TB in infants, children, and adolescents in most developing countries means that any effective TB vaccine should be given early in life, which coincides with the period that other childhood vaccines are given according to the Expanded Program on Immunization (EPI) schedule. The EPI program has been highly successful at increasing the coverage of existing licensed vaccines across the developing world. Integrating a new TB vaccine within the EPI schedule would facilitate deployment and uptake. However, before any vaccines can be coadministered, it is important to evaluate potential interference between the vaccines that could affect the induction of a protective immune response (12) because coadministration with EPI vaccines may diminish the potency of novel vaccine candidates (12). Here, we assessed the safety and immunogenicity of MVA85A in 4-month-old infants and, in parallel, evaluated potential interference with EPI schedule vaccines routinely administered at this age.
We screened 350 infants in total (Fig. 1). Of these, 136 (39%) failed screening. Reasons for exclusion are listed in table S1. The first 36 infants were enrolled in the low-dose group (stage 1a), and the next 36 infants were enrolled in the standard-dose group (stage 1b). After safety and immunogenicity analyses of the stage 1 data, the standard dose was selected for the remainder of the study (stage 2), in which 142 infants were enrolled. All enrolled infants in each stage were randomized to one of the three study groups (Fig. 1). Demographic details, age at BCG vaccination, body weight, age at enrollment, and sex ratio were comparable across the three study groups (table S2). Stage 3 was initially planned in the protocol as an expansion of stage 2, if necessary. However, because the interference results from stage 2 were clear, the protocol was amended and stage 3 was not performed. Study duration was from October 2006 to October 2009.
Consort flowchart. A consort diagram of the study from initial discussion: progressing from low dose to standard dose listing the number of participants for each stage and group of the study. During stage 2 of the study, 47 infants were randomized to group 2, but 48 received treatment because 1 infant inadvertently received EPI and MVA85A vaccines simultaneously. Therefore, in group 3, 48 were randomized but 47 received treatment.
Stage I safety results
To select a dose to be used for the rest of the study, we compared the safety profile between the low-dose (n = 24, excluding EPI-only controls) and the standard-dose (n = 24) MVA85A recipients in stage 1a and stage 1b, respectively (tables S3 to S5). Two days after vaccination, a significantly higher proportion of the standard-dose recipients had redness and pain, but not desquamation or swelling, at the site of vaccination in group 3 (MVA85A alone) but not group 2 (EPI + MVA85A) (table S3). None of the MVA85A recipients had pain that was greater than grade 1. Seven days after vaccination, a significantly higher proportion of subjects had redness in group 3 and desquamation in groups 2 and 3. Pain and swelling at this time point were comparable. There were no differences after vaccination in systemic adverse events (AEs) or hematological or biochemical values between stage 1a and 1b (tables S4 and S5), except for alanine transaminase, which was significantly higher in stage 1b than 1a in group 2 but not group 3, even though these values were still within normal limits (P = 0.018; table S5). There was no difference between all three groups when stages 1b and 2 were combined (table S5).
MVA85A dose response
In stage 1, the area under the curve (AUC) [interquartile range (IQR)] for the Ag85A immune response induced by low-dose MVA85A (stage 1a) was 223 (54 to 421) in group 2 and 1191 (129 to 1807) in group 3 (Table 1). The standard-dose group (stage 1b) induced significantly higher responses than low dose in group 2, as depicted by an AUC (IQR) of 1302 (742 to 3401) (P = 0.0005), and group 3, AUC of 4515 (3118 to 5543) (P = 0.0012 by Mann-Whitney U test). Responses in group 1 (EPI alone) did not differ across the three time points measured (Table
Because higher immunogenicity and a comparable safety profile were seen when results from stage 1a and 1b were compared, a decision to continue to use the standard dose in stage 2 was made and was approved by the ethical committees. Safety and immunogenicity data from stages 1b and 2 were therefore combined for the further data analyses.
MVA85A safety data
Safety data were collected from all infants (tables S3 to S5). MVA85A was safe and well tolerated. Consistent with previous data, most subjects developed mild local and systemic AEs (6, 10, 13) (tables S3 and S4). Two days after vaccination, there was a significant increase in the number of infants in group 2 with excessive crying and vomiting (P = 0.035 for both, table S4 by χ2 test). Seven days after vaccination, all systemic AEs had resolved to nonsignificant differences among all three groups. There were no significant differences in the biochemical and hematological indices among the three groups, either at baseline or 1 week after vaccination (table S5). Infants’ mean weights at 9 and 12 months after vaccination were comparable across the three groups (table S2).
Seven serious adverse events (SAEs) occurred during the study. The frequency of SAEs was comparable across the three groups. The three SAEs in group 1 were malaria, urinary tract infection, and bronchopneumonia; the two SAEs in group 2 were both bronchopneumonia; and the two SAEs in group 3 were diarrhea and bronchopneumonia. None of these SAEs was judged to be related to study vaccine.
Effect of EPI vaccines on MVA85A immunogenicity
To evaluate potential interference between EPI vaccines and MVA85A, we compared the ex vivo interferon-γ (IFN-γ) enzyme-linked immunospot (ELISPOT) response to the Ag85A peptide pool between groups 2 and 3 (Fig. 2). Baseline immune responses were comparable. There were significantly higher responses at 1 (P = 0.0138), 4 (P = 0.0127), and 20 (P = 0.0093) weeks after vaccination in group 3 than in group 2 with the Mann-Whitney U test. Comparing the median (range) of the AUC, group 3 was significantly higher than group 2 [median difference, 1374; 95% confidence interval (CI), 499 to 2354; P = 0.0010], indicating that the average response was higher over the entire period of study. Despite the effect of the EPI vaccines, the cellular immune response to MVA85A remained significantly higher 20 weeks after vaccination compared to baseline in both group 2 (P < 0.0001) and group 3 (P < 0.0001). Fig. 2 Immunogenicity of MVA85A. Immunogenicity was measured by ex vivo IFN-γ ELISPOT assay to a single pool of antigen 85A peptides. The figure compares responses in group 2 (EPI + MVA85A, filled circles) and group 3 (MVA85A alone, open circles) at baseline and 1, 4, and 20 weeks after vaccination. The horizontal lines indicate median values. The baseline median (IQR) of purified protein derivative (PPD) responses was 75 (45 to 193), 90 (25 to 181.8), and 90 (42.5 to 212.5) for groups 1, 2, and 3, respectively, and was not significantly different from each other. The median (IQR) difference AUC for the PPD responses over the 5 months after MVA85A vaccination (groups 2 and 3) or not (group 1) was 467 (−50 to 1039; P = 0.073) between groups 1 and 2, 737 (158 to 1643; P = 0.010) between groups 1 and 3, and 345 (−275 to 1088; P = 0.265) between groups 2 and 3 by Mann-Whitney U test. Interference of MVA85A vaccines with antibody responses to EPI vaccines The concentrations of antibodies specific to each vaccine component in the DTwP-Hib (diphtheria, tetanus, whole-cell pertussis, and Haemophilus influenzae type b) and hepatitis B regimen were compared between groups 2 and 1. The geometric mean (GM) or median antibody titer to each of the antigens was not significantly different between the groups (Table 2). Further analysis with antibody titers categorized according to threshold values also showed no significant difference in the proportions achieving protective levels between the two groups. View this table: In this window In a new window Table 2 Antibody level to EPI vaccines measured at the age of 5 months after three doses of DTwP-Hib and hepatitis B at ages 2, 3, and 4 months according to study groups. The table shows geometric means (GM) with 95% confidence interval (CI). Hepatitis B surface antigen and pertussis toxoid data were not normally distributed after log transformation and are therefore presented as medians and ranges. Subanalysis categorizes antibody titers according to known protective threshold levels. There are no antibody data from the infant randomized to group 3 but treated as in group 2. Note that the assay for hepatitis B had a maximum cutoff of 1000 mIU/ml. Effect is the back-transformed coefficient unless where medians were used, and it represents the ratio of the effect on antibody titer of having EPI with MVA85A concomitantly compared to EPI vaccines only. DISCUSSION This clinical trial evaluates the administration of an MVA-based TB vaccine in 4-month-old infants, as well as interactions between MVA85A and routine EPI schedule vaccines. The higher proportion of infants with local AEs in the standard-dose group compared to the low-dose group was consistent with previous studies using MVA85A and other recombinant MVAs, where there is a dose-response relationship between vector dose and local AE profile (13, 14). Most AEs were comparable between the two doses 7 days after vaccination. A significantly higher proportion of infants that received the EPI + MVA85A vaccines simultaneously (group 2) showed an increase in vomiting and crying when compared to the EPI-alone (group 1) or MVA85A-alone (group 3) groups. Because these events were not more frequent in the infants receiving MVA85A alone, compared to the group receiving EPI vaccines alone, this finding may be a result of inflammatory processes at two injection sites rather than one. It may also be due to the EPI + MVA85A group receiving a relatively higher antigenic load, because they received six vaccines compared to either five or one in the EPI and MVA85A groups, respectively (15, 16). These events were all mild and resolved by day 7 after vaccination. Both doses of MVA85A induced a significant antigen-specific immune response compared to unvaccinated controls. The peak response occurred 7 days after vaccination as previously described (6, 7, 10, 11). The cellular immune response to MVA85A was significantly lower when given simultaneously with the EPI vaccines compared to when administered alone. This difference may relate to interference between MVA85A and the EPI vaccines themselves or, more likely, the alum adjuvant they contain. Alum is the only licensed adjuvant for human use and is associated with the induction of a T helper type 2 (TH2)–dominated immune response (17, 18), which reciprocally inhibits TH1-type responses (19). The increased response to PPD in group 3 (MVA85A without EPI vaccines) is likely to be related to the enhanced response to antigen 85A seen in this group, because antigen 85A is a component of PPD. Because IFN-γ is a key effector molecule in protective immunity against intracellular organisms, including Mycobacterium tuberculosis (20, 21), our finding has implications for the use of MVA for vaccines against other pathogens, for viral vectors in general, and for TH1 type–inducing vaccines. We do not fully understand the mechanisms responsible for the interaction of these vaccines or the biological implications; however, this finding indicates that the internal milieu at the time of processing antigens might affect the magnitude of the response. The coadministration of EPI vaccines with the malaria vaccine candidate RTS,S/AS02, which contains a TH1-biased adjuvant, led to lower levels of antibody responses to the malaria antigen than in regimes that avoided coadministration (12). The benefits of introducing a new vaccine within the EPI schedule include reduced clinic visits, increased coverage, and possible enhanced immunogenicity of the vaccines (4). However, these favorable outcomes must be balanced against potential negative immunological interactions between simultaneously administered vaccines. At present, our data do not support concomitant administration of MVA85A with licensed EPI vaccines, and this finding has supported the assessment of MVA85A administered outside of the EPI schedule in an ongoing Phase IIb efficacy trial in infants (clinicaltrials.gov trial identifier NCT00953927). However, although we have demonstrated a statistically significant reduction in immunogenicity by IFN-γ response when MVA85A is co-administered with the EPI vaccines, the clinical significance of this finding is uncertain in the absence of a validated immunological correlate of protection. Further work is needed to evaluate whether such interference can be overcome by increasing vaccine dose, thus retaining the advantages of concomitant administration. This study has demonstrated that MVA85A is safe and well tolerated and induces an immune response considered protective against M. tuberculosis. It illustrates the importance of doing interference studies with new vaccines and evaluating bidirectional interference. Trials such as this are essential with new TB vaccines and T cell–inducing vaccines in general. They provide essential information needed for the design of efficacy trials. More generally, our findings highlight that modifications to the standard EPI regime may be needed to incorporate a new generation of T cell–inducing vaccines. MATERIALS AND METHODS Study site and recruitment of study participants This study was conducted at Sukuta Health Centre, Gambia. The study was approved by the Gambian government/Medical Research Council Joint Ethics Committee and Oxford Tropical Research Ethics Committee. Preliminary discussions about the trial were conducted with parents shortly after an infant’s birth, and information sheets were distributed. When these infants attended for their EPI schedule vaccines at age 3 months (visit 1), they were enrolled provided they fulfilled the following criteria: They were healthy, had received BCG within the first 4 weeks of life and had a typical BCG scar, had received their EPI schedule immunizations according to the national program (DTwP-Hib at 2 and 3 months, oral polio vaccine at birth and 2 and 3 months, hepatitis B at birth and 2 months), and their parents had given written informed consent. A blood sample was taken for baseline safety and immunogenicity. An independent statistician randomly assigned participants in a 1:1:1 ratio in blocks of six to one of the three study groups. A unique number for each participant detailing the treatment group was placed in a sealed opaque envelope and arranged serially. At 4 months of age (visit 2), when a child fulfilled the eligibility criteria, one of the study team selected the next envelope to determine study number and treatment group. The EPI vaccines given at this time point were the third doses of DTwP-Hib and hepatitis B. This trial was conducted according to the principles of the International Conference on Harmonisation–Good Clinical Practice guidelines, was externally monitored by an independent contract research organization, and is registered on clinicaltrials.gov (NCT00480454). Vaccines MVA85A was administered intradermally on the side ipsilateral to the BCG vaccination. All EPI vaccines were given intramuscularly apart from the oral polio vaccine. After MVA85A injection, the vaccination site was covered to minimize virus dissemination into the environment. Because this was the first trial with MVA85A in infants, we commenced this trial with low-dose MVA85A [2.5 × 107 plaque-forming units (PFUs)]. Once we had established safety with this dose in 24 infants, we then administered 5 × 107 PFU (the standard dose of MVA85A used previously) to the next 24 infants. A dose of 5 × 107 PFU was selected for the remainder of the study after analysis of the safety data. We obtained ethical approval before proceeding to each stage of the study. MVA85A was manufactured under Good Manufacturing Practice conditions by Impfstoffwerk Dessau-Tornau, Germany. Concomitant licensed EPI vaccines were obtained from the same source as the Ministry of Health of Gambia through the World Health Organization. Reactogenicity Each infant’s vital signs were monitored before vaccination and 30 and 60 min after vaccination. Thereafter, the dressing over the injection site was removed. Trained field workers visited the infants at home daily for the following 2 days to administer a reactogenicity questionnaire to the parents/guardians that included history of fever, vomiting, diarrhea, reduced feeding, and excessive crying. The field worker also examined the infant for expected local AEs (swelling, tenderness, redness, and desquamation at the site of injection for MVA85A) and fever. Pain at the injection site was graded on a scale of 0 to 3 (where 0 = no pain, 1 = painful to touch, 2 = partial restriction of activities, and 3 = unable to use the limb). Systemic and laboratory AEs were evaluated in all three study groups. Follow-up visits Subjects were reviewed 1 week after MVA85A administration (visit 3), corresponding to the time of peak MVA85A immune response reported previously (6, 7, 10, 11). Blood samples were obtained from all study participants before vaccination with the EPI vaccines deferred from visit 2 for group 3. Local and systemic reactogenicity in the first week after vaccination was recorded. Subsequent visits for immune responses were at the ages of 5 (visit 4) and 9 (visit 5) months. The protocol was amended to include additional visits at 12 (visit 6) and 18 (visit 7) months to monitor longevity of responses to MVA85A. Data from these visits will be reported in a separate manuscript. In all scheduled visits, solicited and unsolicited AEs were recorded and physical examinations were conducted. Blood samples were taken at each visit apart from visit 6. Parents and guardians were encouraged to take their infants to the trial clinic or Medical Research Council clinic with their study ID cards if their infant was unwell during the study. Any such visit was reported immediately to the trial team. LABORATORY METHODS MVA85A immunogenicity Immunogenicity of MVA85A was determined by measuring the frequency of IFN-γ spot-forming units (SFUs) by ex vivo ELISPOT assay (10). Briefly, peripheral blood mononuclear cells (PBMCs) suspended in 5% human AB serum in RPMI were seeded at 2 × 105 cells per 100 μl per well in duplicates and incubated in the presence of the antigen or controls. A pool of 15-mer peptides spanning the Ag85A protein (66 peptides that overlapped by 10 amino acids; 2 μg/ml each) and M. tuberculosis PPD (10 μg/ml) were used as recall antigens. Medium alone served as a negative control, whereas phytohemagglutinin (10 μg/ml) was the positive control. The number of IFN-γ SFU counted in each well was automatically entered into a database minus background (medium wells). When responses were too numerous to quantify, a maximum cutoff value of 200 spots per well or 1000 SFU per million PBMCs was assigned. EPI vaccine antibody measurement Antibody (immunoglobulin G) concentrations against diphtheria toxoid, tetanus toxoid, pertussis (toxoid and filamentous hemagglutinin), Haemophilus influenzae type b capsular polysaccharide, and hepatitis B surface antigen were measured by standard enzyme-linked immunosorbent assays (ELISAs) at the Health Protection Agency Immunoassay Laboratory, UK. For each antigen, sera were titrated against known concentrations of specific antibodies in standard sera samples. All laboratory assays were done blinded to the child’s study group. Data management and statistical analyses Data were double-entered into an Access database and manually validated. For categorical variables, data were summarized using numbers and percentages, and groups were compared with the χ2 test or, for small numbers, Fisher’s exact test. For continuous, normally distributed data, the mean and SD were used to summarize data, and groups were compared with an analysis of variance (for comparison of three groups) or a t test (for comparison of two groups). In the case of non-normally distributed data, a Kruskal-Wallis (for comparison of three groups) or Mann-Whitney U test (for comparison of two groups) was used. Comparisons between time points within a group were made with Wilcoxon matched-pairs signed-rank test. Such data were summarized with medians and 25th and 75th percentiles. In the case where a log transformation resulted in normally distributed data, the GM was presented and groups were compared with a t test. An AUC analysis was used to assess for overall differences in response over the follow-up period. Subjects were analyzed according to the treatment they received, rather than the group to which they were randomized. Analyses were done with Stata Release 10 statistical software (Stata). SUPPLEMENTARY MATERIAL www.sciencetranslationalmedicine.org/cgi/content/full/3/88/88ra56/DC1 Table S1. Reasons for exclusion of infants screened for participation. Table S2. Demographic and baseline characteristics of subjects by study group. Table S3. Comparison of number (percent) of local adverse events between study groups in stages 1a, 1b, and 2. Table S4. Comparison of number (percent) of systemic adverse events between study groups in stages 1a, 1b, and 2. Table S5. Comparison of biochemistry and hematology parameters [mean (SD)] between study groups in stages 1a, 1b, and 2. Citation: M. O. C. Ota, A. A. Odutola, P. K. Owiafe, S. Donkor, O. A. Owolabi, N. J. Brittain, N. Williams, S. Rowland-Jones, A. V. S. Hill, R. A. Adegbola, H. McShane, Immunogenicity of the Tuberculosis Vaccine MVA85A Is Reduced by Coadministration with EPI Vaccines in a Randomized Controlled Trial in Gambian Infants. Sci. Transl. Med. 3, 88ra56 (2011).
REFERENCES AND NOTES
↵ C. Dye, C. J. Watt, D. M. Bleed, S. M. Hosseini, M. C. Raviglione, Evolution of tuberculosis control and prospects for reducing tuberculosis incidence, prevalence, and deaths globally. JAMA 293, 2767–2775 (2005). Abstract/FREE Full Text
↵ G. A. Colditz, T. F. Brewer, C. S. Berkey, M. E. Wilson, E. Burdick, H. V. Fineberg, F. Mosteller, Efficacy of BCG vaccine in the prevention of tuberculosis. Meta-analysis of the published literature. JAMA 271, 698–702 (1994). Abstract/FREE Full Text
↵ M. L. Garly, C. L. Martins, C. Balé, M. A. Baldé, K. L. Hedegaard, P. Gustafson, I. M. Lisse, H. C. Whittle, P. Aaby, BCG scar and positive tuberculin reaction associated with reduced child mortality in West Africa. A non-specific beneficial effect of BCG? Vaccine 21, 2782–2790 (2003). CrossRefMedlineWeb of Science
↵ M. O. Ota, J. Vekemans, S. E. Schlegel-Haueter, K. Fielding, M. Sanneh, M. Kidd, M. J. Newport, P. Aaby, H. Whittle, P. H. Lambert, K. P. McAdam, C. A. Siegrist, A. Marchant, Influence of Mycobacterium bovis bacillus Calmette-Guérin on antibody and cytokine responses to human neonatal vaccination. J. Immunol. 168, 919–925 (2002). Abstract/FREE Full Text
↵ S. P. Zodpey, S. N. Shrikhande, S. W. Kulkarni, B. R. Maldhure, Scar size and effectiveness of Bacillus Calmette Guerin (BCG) vaccination in the prevention of tuberculosis and leprosy: A case-control study. Indian J. Public Health 51, 184–189 (2007). Medline
↵ T. Hawkridge, T. J. Scriba, S. Gelderbloem, E. Smit, M. Tameris, S. Moyo, T. Lang, A. Veldsman, M. Hatherill, L. Merwe, H. A. Fletcher, H. Mahomed, A. V. Hill, W. A. Hanekom, G. D. Hussey, H. McShane, Safety and immunogenicity of a new tuberculosis vaccine, MVA85A, in healthy adults in South Africa. J. Infect. Dis. 198, 544–552 (2008). Abstract/FREE Full Text
↵ H. McShane, A. A. Pathan, C. R. Sander, S. M. Keating, S. C. Gilbert, K. Huygen, H. A. Fletcher, A. V. Hill, Recombinant modified vaccinia virus Ankara expressing antigen 85A boosts BCG-primed and naturally acquired antimycobacterial immunity in humans. Nat. Med. 10, 1240–1244 (2004). CrossRefMedlineWeb of Science
↵ S. D’Souza, V. Rosseels, M. Romano, A. Tanghe, O. Denis, F. Jurion, N. Castiglione, A. Vanonckelen, K. Palfliet, K. Huygen, Mapping of murine Th1 helper T-cell epitopes of mycolyl transferases Ag85A, Ag85B, and Ag85C from Mycobacterium tuberculosis. Infect. Immun. 71, 483–493 (2003). Abstract/FREE Full Text
↵ H. M. Vordermeier, B. Villarreal-Ramos, P. J. Cockle, M. McAulay, S. G. Rhodes, T. Thacker, S. C. Gilbert, H. McShane, A. V. Hill, Z. Xing, R. G. Hewinson, Viral booster vaccines improve Mycobacterium bovis BCG-induced protection against bovine tuberculosis. Infect. Immun. 77, 3364–3373 (2009). Abstract/FREE Full Text
↵ R. H. Brookes, P. C. Hill, P. K. Owiafe, H. B. Ibanga, D. J. Jeffries, S. A. Donkor, H. A. Fletcher, A. S. Hammond, C. Lienhardt, R. A. Adegbola, H. McShane, A. V. Hill, Safety and immunogenicity of the candidate tuberculosis vaccine MVA85A in West Africa. PLoS One 3, e2921 (2008). CrossRefMedline
↵ T. J. Scriba, M. Tameris, N. Mansoor, E. Smit, L. van der Merwe, F. Isaacs, A. Keyser, S. Moyo, N. Brittain, A. Lawrie, S. Gelderbloem, A. Veldsman, M. Hatherill, A. Hawkridge, A. V. Hill, G. D. Hussey, H. Mahomed, H. McShane, W. A. Hanekom, Modified vaccinia Ankara-expressing Ag85A, a novel tuberculosis vaccine, is safe in adolescents and children, and induces polyfunctional CD4+ T cells. Eur. J. Immunol. 40, 279–290 (2010). MedlineWeb of Science
↵ S. Abdulla, R. Oberholzer, O. Juma, S. Kubhoja, F. Machera, C. Membi, S. Omari, A. Urassa, H. Mshinda, A. Jumanne, N. Salim, M. Shomari, T. Aebi, D. M. Schellenberg, T. Carter, T. Villafana, M. A. Demoitié, M. C. Dubois, A. Leach, M. Lievens, J. Vekemans, J. Cohen, W. R. Ballou, M. Tanner, Safety and immunogenicity of RTS,S/AS02D malaria vaccine in infants. N. Engl. J. Med. 359, 2533–2544 (2008). CrossRefMedlineWeb of Science
↵ M. Walther, F. M. Thompson, S. Dunachie, S. Keating, S. Todryk, T. Berthoud, L. Andrews, R. F. Andersen, A. Moore, S. C. Gilbert, I. Poulton, F. Dubovsky, E. Tierney, S. Correa, A. Huntcooke, G. Butcher, J. Williams, R. E. Sinden, A. V. Hill, Safety, immunogenicity, and efficacy of prime-boost immunization with recombinant poxvirus FP9 and modified vaccinia virus Ankara encoding the full-length Plasmodium falciparum circumsporozoite protein. Infect. Immun. 74, 2706–2716 (2006). Abstract/FREE Full Text
↵ V. S. Moorthy, E. B. Imoukhuede, S. Keating, M. Pinder, D. Webster, M. A. Skinner, S. C. Gilbert, G. Walraven, A. V. Hill, Phase 1 evaluation of 3 highly immunogenic prime-boost regimens, including a 12-month reboosting vaccination, for malaria vaccination in Gambian men. J. Infect. Dis. 189, 2213–2219 (2004). Abstract/FREE Full Text
↵ B. Watson, A. Cawein, B. L. McKee, J. G. Hackell, Safety and immunogenicity of acellular pertussis vaccine, combined with diphtheria and tetanus as the Japanese commercial Takeda vaccine, compared with the Takeda acellular pertussis component combined with Lederle’s diphtheria and tetanus toxoids in two-, four- and six-month-old infants. Pediatr. Infect. Dis. J. 11, 930–935 (1992). MedlineWeb of Science
↵ F. Zepp, A. Schuind, C. Meyer, R. Sänger, A. Kaufhold, P. Willems, Safety and reactogenicity of a novel DTPa-HBV-IPV combined vaccine given along with commercial Hib vaccines in comparison with separate concomitant administration of DTPa, Hib, and OPV vaccines in infants. Pediatrics 109, e58 (2002). Abstract/FREE Full Text
↵ L. Bungener, F. Geeraedts, W. Ter Veer, J. Medema, J. Wilschut, A. Huckriede, Alum boosts TH2-type antibody responses to whole-inactivated virus influenza vaccine in mice but does not confer superior protection. Vaccine 26, 2350–2359 (2008). CrossRefMedlineWeb of Science
↵ D. Spazierer, H. Skvara, M. Dawid, N. Fallahi, K. Gruber, K. Rose, P. Lloyd, S. Heuerding, G. Stingl, T. Jung, T helper 2 biased de novo immune response to Keyhole Limpet Hemocyanin in humans. Clin. Exp. Allergy 39, 999–1008 (2009). CrossRefMedlineWeb of Science
↵ A. K. Abbas, K. M. Murphy, A. Sher, Functional diversity of helper T lymphocytes. Nature 383, 787–793 (1996). CrossRefMedline
↵ J. L. Flynn, J. Chan, K. J. Triebold, D. K. Dalton, T. A. Stewart, B. R. Bloom, An essential role for interferon γ in resistance to Mycobacterium tuberculosis infection. J. Exp. Med. 178, 2249–2254 (1993). Abstract/FREE Full Text
↵ M. J. Newport, C. M. Huxley, S. Huston, C. M. Hawrylowicz, B. A. Oostra, R. Williamson, M. Levin, A mutation in the interferon-γ–receptor gene and susceptibility to mycobacterial infection. N. Engl. J. Med. 335, 1941–1949 (1996). CrossRefMedlineWeb of Science
Acknowledgments: We thank the Gambian government, the EPI program of Gambia, and staff of Sukuta Health Centre for their collaboration. We thank K. Flanagan, J. Adetifa, and the Sukuta field team for assisting us in this trial. We appreciate M. Sonko, K. Manneh, and S. Sanneh for their field work and A. Bojang and J. Sutherland for laboratory assistance. We also thank K. Bojang for safety monitoring; V. Thomas and J. Mueller for internal monitoring and support; A. Lawrie for coordination; S. Vermaak for help with data analysis; and C. McKenna for study monitoring. We are grateful to the parents and guardians who allowed their infants to participate in this trial. Funding: The study was funded by the Medical Research Council (UK) and the European Commission (EU 6th Framework; TBVAC). Neither funder had any role in the design of the study or the preparation of the manuscript. H.M. is a Wellcome Trust Senior Clinical Research Fellow, and A.V.S.H. is a Wellcome Trust Principal Research Fellow. Author contributions: M.O.C.O., H.M., A.V.S.H., S.R.-J., and R.A.A. designed the study. A.A.O., O.A.O., M.O.C.O., R.A.A., and S.R.-J. provided clinical help. M.O.C.O., A.A.O., N.J.B., and H.M. managed the project. P.K.O., M.O.C.O., H.M., A.V.S.H., and R.A.A. performed laboratory assays and analysis. N.W., S.D., N.J.B., H.M., and M.O.C.O. performed the statistical analysis. M.O.C.O., H.M., N.J.B., A.A.O., O.A.O., A.V.S.H., S.D., N.W., S.R.-J., and R.A.A. wrote the manuscript. Competing interests: A.V.S.H. and H.M. are named inventors on a patent filing related to MVA85A (International patent number WO 2006/072787 A1: Compositions for immunising against mycobacteria), and are shareholders in a joint venture, Oxford-Emergent Tuberculosis Consortium, formed for the future development of this vaccine. The other authors declare that they have no competing interests.