Browse the corpus

The worldwide burden of P. falciparum malaria remains a major public health concern [1], with approximately 207 million cases and 627 000 deaths worldwide in 2012 [2]. The preerythrocytic P. falciparum vaccine RTS,S, formed from fusion of the circumsporozoite protein (CS) to the surface-antigen of hepatitis B virus, is the most advanced malaria vaccine in development. However, it confers only limited, relatively short-lived protection in African infants [3–5]. Analysis of the immunological correlates of immunity induced by RTS,S suggests that high levels of antibodies against CS on the sporozoite correlate with protection, with a possible minor contribution from low levels of induced CD4+ T cells [6–8]. While these clinical results are the most effective to date in a field setting, there remains a need to improve on this limited clinical efficacy [9, 10], either through modifications to RTS,S or by developing vaccine strategies that combine numerous antigens or vaccine platforms.

levels of induced CD4+ T cells [6–8]. While these clinical results are the most effective to date in a field setting, there remains a need to improve on this limited clinical efficacy [9, 10], either through modifications to RTS,S or by developing vaccine strategies that combine numerous antigens or vaccine platforms. Increasingly, data from animal models and vectored immunizations demonstrate a correlation between CD8+ T cells and immunity to liver-stage parasites, even in the absence of antibodies [11–17]. Clinical vaccine development had been hampered by the limited ability of traditional subunit vaccine strategies, namely adjuvanted protein constructs, to induce high enough numbers of antigen-specific CD8+ T cells that may confer protection [18]. However, more recently, adenoviral-vectored malaria vaccines administered in heterologous prime-boost regimens with a modified vaccinia virus Ankara (MVA) boost have been capable of inducing good humoral and T-cell responses that include high levels of CD8+ T cells [17–21]. These CD8+ T-cell responses have been associated with clinical efficacy [17]. Given concerns regarding the effect of preexisting immunity on the immunological potency of human adenoviruses, simian adenoviruses (ChAd) are being developed as alternative, potent vectors [22]. Indeed, prime-boost vaccination with ChAd63 and MVA expressing the leading preerythrocytic antigen, ME-TRAP, is clinically the most potent inducer of CD8+ T cells in humans and the most effective malaria vaccine besides RTS,S, demonstrating efficacy, defined as sterile protection or delay, in 8 of 14 malaria-naive volunteers (57%) following sporozoite challenge [17].

VA expressing the leading preerythrocytic antigen, ME-TRAP, is clinically the most potent inducer of CD8+ T cells in humans and the most effective malaria vaccine besides RTS,S, demonstrating efficacy, defined as sterile protection or delay, in 8 of 14 malaria-naive volunteers (57%) following sporozoite challenge [17]. Given that CS is expressed during both the sporozoite and liver stages of P. falciparum infection and therefore is possibly susceptible to both humoral and cell-mediated immunity at both stages, we assess here the efficacy of ChAd63-MVA expressing CS. If effective, this vaccine could then be combined with ChAd63-MVA expressing ME-TRAP or RTS,S, to improve clinical efficacy. Following a phase 1a study of ChAd63-MVA CS in malaria-naive volunteers, in which the regimen was shown to be safe and immunogenic (de Barra et al, submitted), we performed a study of controlled human infection with Plasmodium sporozoites (also known as “controlled human malaria infection” [CHMI]) [23], using the standard challenge model involving infectious bites from 5 mosquitoes, to compare the efficacy of ChAd63-MVA CS with that of ChAd63-MVA ME-TRAP.

a et al, submitted), we performed a study of controlled human infection with Plasmodium sporozoites (also known as “controlled human malaria infection” [CHMI]) [23], using the standard challenge model involving infectious bites from 5 mosquitoes, to compare the efficacy of ChAd63-MVA CS with that of ChAd63-MVA ME-TRAP. METHODS Participants The study was conducted at the Centre for Clinical Vaccinology and Tropical Medicine, University of Oxford (Oxford, United Kingdom), and at the National Institute for Health Research (NIHR) Wellcome Trust Clinical Research Facility, part of the University of Southampton and University Hospital Southampton National Health Service (NHS) Foundation Trust (Southampton, United Kingdom). The challenge procedure was performed as previously described [24], using 5 infectious bites from P. falciparum strain 3D7–infected Anopheles stephensi mosquitoes. This took place at the Alexander Fleming Building, Imperial College (London, United Kingdom), and mosquitoes were supplied by the Department of Entomology, Walter Reed Army Institute of Research (WRAIR; Washington, DC). Healthy, malaria-naive men and non-pregnant women aged 18–45 years were invited to participate in the study. All volunteers gave written informed consent prior to participation, and the study was conducted according to the principles of the Declaration of Helsinki and in accordance with good clinical practice. There was no selection of volunteers on the basis of preexisting neutralizing antibodies to the ChAd63 vector before enrollment. The full list of inclusion and exclusion criteria is given in the Supplementary Materials.

ding to the principles of the Declaration of Helsinki and in accordance with good clinical practice. There was no selection of volunteers on the basis of preexisting neutralizing antibodies to the ChAd63 vector before enrollment. The full list of inclusion and exclusion criteria is given in the Supplementary Materials. Ethical and Regulatory Approval All necessary approvals for the study were granted by the United Kingdom National Research Ethics Service, Committee South Central–Oxford A (reference 12/SC/0037), and the United Kingdom Medicines and Healthcare Products Regulatory Agency (reference 21584/0293/001-0001). The study was additionally reviewed by the Western Institution Review Board (Seattle, WA; reference 20120266) at the request of the PATH Malaria Vaccine Initiative and was approved. The Genetically Modified Organisms Safety Committee of the Oxford University Hospitals NHS Trust (reference GM462.11.65) authorized recombinant vaccine use. The trial was registered with ClinicalTrials.gov (reference NCT01623557). The local safety committee provided safety oversight, and good clinical practice compliance was independently monitored by an external organization (Appledown Clinical Research, Great Missenden, United Kingdom).

uthorized recombinant vaccine use. The trial was registered with ClinicalTrials.gov (reference NCT01623557). The local safety committee provided safety oversight, and good clinical practice compliance was independently monitored by an external organization (Appledown Clinical Research, Great Missenden, United Kingdom). ChAd63 and MVA Vaccines Generation, manufacture, and quality control monitoring of the recombinant ChAd63 and MVA vectors encoding ME-TRAP and CS have been previously described [de Barra et al, submitted; 25]. The antigen ME-TRAP contains a fusion protein of a multi-epitope string (ME), followed by preerythrocytic thrombospondin-related adhesion protein (TRAP) from P. falciparum strain T9/96 [17]. The poor immunogenicity of the standard full-length CS insert (CSO) previously used in clinical trials by our group [26–29] suggested that there may be an important difference in the intrinsic immunogenicity of CSO, compared with that of the ME-TRAP insert. For this study, we used information from multiple sources [30–32] to design a novel CS antigen that omits the extreme C-terminus of the protein that encodes the glycophosphatidylinositol anchor sequence and may down-modulate CS immunogenicity [de Barra et al, submitted; 33].

O, compared with that of the ME-TRAP insert. For this study, we used information from multiple sources [30–32] to design a novel CS antigen that omits the extreme C-terminus of the protein that encodes the glycophosphatidylinositol anchor sequence and may down-modulate CS immunogenicity [de Barra et al, submitted; 33]. Study Design This was a Phase I/IIa open-label, vaccine and CHMI trial (Figure 1). Volunteers chose whether to participate as vaccinees (groups 1 and 2) or unvaccinated controls undergoing CHMI alone (group 3). Vaccinees were randomly allocated to groups 1 or 2. All vaccinations were administered intramuscularly into the deltoid, with the ChAd63 and MVA-vectored vaccines administered in alternating arms. ChAd63-vectored vaccines were administered on day 0, and MVA boost was administered on day 56. Details of dosing, clinical follow-up and safety monitoring are given in Supplementary Information. An interval of 1–14 days was allowed between vaccination and follow-up visits after vaccination. CHMI was performed on day 77. Throughout this article, “study day” refers to the nominal time point for a group and not the actual day of sampling. Figure 1. Flow of study design and volunteer recruitment. Twenty volunteers were excluded following screening for the following reasons: psychiatric history (n = 3), no medical screening letter returned (n = 3), multiple medical problems (n = 2), excessive alcohol use (n = 2), syncope (n = 1), connective tissue disease (n = 1), iron deficiency (n = 1), raised alanine aminotransferase level (n = 1), poor venous access (n = 1), gastrointestinal problems under investigation (n = 1), family history of heart disease (n = 1), lost to follow-up (n = 1), unavailable during challenge (n = 1), and history of recreational drug use (n = 1). Furthermore, 7 volunteers withdrew consent after screening but before enrollment. All immunizations were administered intramuscularly with sequential vaccines administered into the deltoid of alternating arms. No enrolled volunteers withdrew from the study and all volunteers completed study visits as scheduled. Abbreviations: ChAd63, simian adenovirus 63; CS, circumsporozoite protein; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara; pfu, plaque-forming units; vp, viral particles.

ed volunteers withdrew from the study and all volunteers completed study visits as scheduled. Abbreviations: ChAd63, simian adenovirus 63; CS, circumsporozoite protein; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara; pfu, plaque-forming units; vp, viral particles. Ex Vivo Interferon γ (IFN-γ) Enzyme-Linked Immunosorbent Spot (ELISPOT) Analysis Ex vivo (18-hour stimulation) ELISPOT assays for ME-TRAP and CS were performed on fresh (ie, not previously frozen) peripheral blood mononuclear cells (PBMCs) from blood samples obtained on days 0, 14, 28, 56, and 63 after vaccination and on 1 day before and 7, 35, and 90 days after CHMI. Antigens were tested in duplicate with 250 000 freshly isolated PBMCs added to each well. Details about the ELISPOT methods are available in the Supplementary Materials. Total Immunoglobulin G (IgG) Enzyme-Linked Immunosorbent Assay (ELISA) Antibody responses were assessed using serum samples collected on days 0, 28, 56, and 63 after vaccination and 1 day before and 35 and 90 days after CHMI. Antibody responses to TRAP were measured by an IgG ELISA performed at the Jenner Institute (Oxford; Supplementary Materials). Antibody responses to CS were measured by an IgG ELISA performed at the WRAIR International Reference Center for Malaria Serology (Supplementary Materials) [34]. Parasite Quantitative Polymerase Chain Reaction (qPCR) qPCR for P. falciparum was conducted as described previously [35] (see Supplementary Materials).

Total Immunoglobulin G (IgG) Enzyme-Linked Immunosorbent Assay (ELISA) Antibody responses were assessed using serum samples collected on days 0, 28, 56, and 63 after vaccination and 1 day before and 35 and 90 days after CHMI. Antibody responses to TRAP were measured by an IgG ELISA performed at the Jenner Institute (Oxford; Supplementary Materials). Antibody responses to CS were measured by an IgG ELISA performed at the WRAIR International Reference Center for Malaria Serology (Supplementary Materials) [34]. Parasite Quantitative Polymerase Chain Reaction (qPCR) qPCR for P. falciparum was conducted as described previously [35] (see Supplementary Materials). Criteria for Malaria Diagnosis Diagnosis of malaria following CHMI was defined as positive findings of thick film microscopy, with at least 1 morphologically normal malaria trophozoite seen by ≥1 experienced microscopist. qPCR was simultaneously performed, although investigators directly involved in clinical management were blinded to these results. For volunteers with positive findings of thick film microscopy but no symptoms consistent with P. falciparum infection, investigators were unblinded to the qPCR results, with the volunteer treated only if any preceding samples had >500 parasites/mL. For volunteers with symptoms or signs that, in the opinion of the clinical investigators, likely represented malaria (eg, fever, rigors, or severe symptomatology), despite negative findings of thick film microscopy and no alternative cause, investigators were unblinded to the qPCR results. If any volunteer's preceding samples had >500 parasites/mL, the volunteer was treated for malaria. A vaccinee was classified as a participant who demonstrated a delay to patency/treatment if treatment was started >2 times the standard deviation in days after the mean time to treatment of unvaccinated control volunteers. This corresponds to clearance of an estimated >95% of preerythrocytic-stage parasites [36].

aria. A vaccinee was classified as a participant who demonstrated a delay to patency/treatment if treatment was started >2 times the standard deviation in days after the mean time to treatment of unvaccinated control volunteers. This corresponds to clearance of an estimated >95% of preerythrocytic-stage parasites [36]. Statistical Analysis Data were analyzed using GraphPad Prism, version 5.03 for Windows (GraphPad Software, La Jolla, California). Individual, geometric mean (GM), or median responses for measurements within each group are described. Parasite densities were log transformed to remove skewness, with 1 added to each value to allow transformation of zero values. Significance testing of differences between groups used either a 2-tailed t test or the 2-tailed Mann–Whitney test (or the Kruskal–Wallis test, for comparisons of >2 groups) for nonparametrically distributed data. Correlations were assessed using the Spearman rank correlation coefficient. Time to treatment was analyzed using Kaplan–Meier survival curves, and between-group comparisons were made using the log-rank test.

–Whitney test (or the Kruskal–Wallis test, for comparisons of >2 groups) for nonparametrically distributed data. Correlations were assessed using the Spearman rank correlation coefficient. Time to treatment was analyzed using Kaplan–Meier survival curves, and between-group comparisons were made using the log-rank test. RESULTS Recruitment and Vaccinations Recruitment took place between March and June 2012. Thirty healthy malaria-naive adult volunteers (10 women and 20 men) were enrolled as vaccinees across 2 sites in the United Kingdom. Six further volunteers (5 women and 1 man) were enrolled to undergo CHMI as unvaccinated infectivity controls (Figure 1). The mean age of volunteers was 26.4 years (range, 19–40 years). Vaccinations began in April 2012, CHMI occurred in July 2012, and all follow-up visits were completed by November 2012. All vaccinees received their immunizations as scheduled. All doses of vaccines were the same as those used in the comparable phase 1a studies [de Barra et al, submitted; 25]. All volunteers underwent CHMI 15–21 days after MVA immunization (ie, on days 71–77). Vaccine Safety and Reactogenicity No unexpected or serious adverse events (AEs) related to vaccination occurred. The local and systemic (Supplementary Figure 1) reactogenicity profile of each vaccine was similar to phase 1a data [de Barra et al, submitted; 25].

RESULTS Recruitment and Vaccinations Recruitment took place between March and June 2012. Thirty healthy malaria-naive adult volunteers (10 women and 20 men) were enrolled as vaccinees across 2 sites in the United Kingdom. Six further volunteers (5 women and 1 man) were enrolled to undergo CHMI as unvaccinated infectivity controls (Figure 1). The mean age of volunteers was 26.4 years (range, 19–40 years). Vaccinations began in April 2012, CHMI occurred in July 2012, and all follow-up visits were completed by November 2012. All vaccinees received their immunizations as scheduled. All doses of vaccines were the same as those used in the comparable phase 1a studies [de Barra et al, submitted; 25]. All volunteers underwent CHMI 15–21 days after MVA immunization (ie, on days 71–77). Vaccine Safety and Reactogenicity No unexpected or serious adverse events (AEs) related to vaccination occurred. The local and systemic (Supplementary Figure 1) reactogenicity profile of each vaccine was similar to phase 1a data [de Barra et al, submitted; 25]. T-Cell Immunogenicity to ChAd63-MVA CS and ME-TRAP T-cell responses followed the expected kinetics after ChAd63 receipt [de Barra et al, submitted; 17, 25, 35, 37, 38], with peak responses seen 28 days after ChAd63 receipt (group 1 [CS]: GM, 343 spot-forming cells (SFCs)/million PBMCs [95% CI, 191–617]; group 2 [ME-TRAP]: GM, 553 SFCs/million PBMCs [95% CI, 330–925]). The peak T-cell response after boost was seen at day 63 after receipt of MVA CS for group 1 (GM, 1017 SFCs/million PBMCs [95% CI, 630–1641]) and at 1 day before CHMI after MVA ME-TRAP receipt for group 2 (GM, 2027 SFCs/million PBMCs [95% CI, 1472–2792]; Figure 2A and 2B). There was no significant difference in T-cell responses between day 63 after vaccination and 1 day before CHMI for either group. Figure 2. Antigen-specific T-cell responses to vaccination measured by interferon γ enzyme-linked immunosorbent spot assay. Kinetics of T-cell responses after vaccination with ChAd63-MVA encoding either circumsporozoite protein (CS; group 1; A) or ME-TRAP (group 2; B). Each line represents an individual volunteer. **P < .01 and ***P < .001, by the Kruskal–Wallis test with the Dunn multiple comparison test. C, Median T-cell frequencies for both antigens by group. Mean T-cell frequencies at day 28 after vaccination were 304 and 673 spot-forming cells (SFCs) after ChAd63-MVA CS or ME-TRAP receipt, respectively, and at day 63 peaked at 1378 and 2068 SFCs after ChAd63-MVA CS or ME-TRAP receipt, respectively. Abbreviations: ChAd63, simian adenovirus 63; CHMI, controlled human malaria infection; controls, unvaccinated volunteers undergoing CHMI; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara; PBMC, peripheral blood mononuclear cell.

CS or ME-TRAP receipt, respectively. Abbreviations: ChAd63, simian adenovirus 63; CHMI, controlled human malaria infection; controls, unvaccinated volunteers undergoing CHMI; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara; PBMC, peripheral blood mononuclear cell. Responses to both antigens were well maintained, with GMs of 285 SFCs/million PBMCs (95% CI, 156–520) to CS and 659 SFCs/million PBMCs (95% CI, 418–1036) to ME-TRAP 16 weeks after MVA receipt in groups 1 and 2, respectively (Figure 2C). T-cell responses among infectivity controls showed a GM of 110 SFCs/million PBMCs (95% CI, 40–304) to CS and a GM of 85 SFCs/million PBMCs (95% CI, 31–231) to ME-TRAP 1 day before CHMI. These responses did not change significantly during follow-up (Figure 2C). Detailed mapping of T-cell responses to the ME-TRAP antigen are outlined in the Supplementary Materials. Detailed mapping of T-cell responses to CS peptides was not performed because this was described recently in detail with several HLA class I–restricted epitopes [39].

Responses to both antigens were well maintained, with GMs of 285 SFCs/million PBMCs (95% CI, 156–520) to CS and 659 SFCs/million PBMCs (95% CI, 418–1036) to ME-TRAP 16 weeks after MVA receipt in groups 1 and 2, respectively (Figure 2C). T-cell responses among infectivity controls showed a GM of 110 SFCs/million PBMCs (95% CI, 40–304) to CS and a GM of 85 SFCs/million PBMCs (95% CI, 31–231) to ME-TRAP 1 day before CHMI. These responses did not change significantly during follow-up (Figure 2C). Detailed mapping of T-cell responses to the ME-TRAP antigen are outlined in the Supplementary Materials. Detailed mapping of T-cell responses to CS peptides was not performed because this was described recently in detail with several HLA class I–restricted epitopes [39]. Antibody Immunogenicity of ChAd63-MVA CS and ME-TRAP Anti-CS IgG antibody responses were measured in all vaccinees (Figure 3A). Anti-CS IgG antibodies were detected in ME-TRAP vaccinees (group 2) because of the inclusion of 4 copies of the N-acetylneuraminic acid phosphatase (NANP) repeat from the CS antigen in the ME string. In group 1, anti-CS IgG responses peaked 21 days after MVA receipt, with a median level of 2.1 µg/mL. In group 2, anti-CS IgG responses also peaked 21 days after MVA, but 8 of 14 volunteers in this group did not have a measurable response, giving a median level of 0 µg/mL. Anti-TRAP IgG antibody responses were assessed in group 2 only (Figure 3B) and also peaked 21 days after MVA ME-TRAP receipt (median, 1475 ELISA units). A weak relationship between anti-CS IgG antibody responses and CS-specific T-cell responses 1 day before CHMI was observed in group 1 (r = 0.5; P = .08, by 2-tailed Spearman correlation; Figure 3C). Exposure to CHMI did not induce significant levels of anti-CS or TRAP antibodies among infectivity controls (Figure 3A and 3B). Figure 3. Antibody responses to vaccination measured by enzyme-linked immunosorbent assay (ELISA). A, Anti- circumsporozoite protein (CS) immunoglobulin G (IgG) antibody responses after vaccination with ChAd63-MVA CS (group 1; red) or ME-TRAP (group 2; blue). Lines represent group medians. ***P = <.001 and *P = <.05, by the Friedman test comparing responses before and after vaccination with the Dunn post hoc test. B, Anti-TRAP IgG antibody responses after vaccination with ChAd63 ME-TRAP (group 2). ***P = .0002, by the 2-tailed Wilcoxon matched pairs test. C, Correlation between anti-CS IgG antibodies and CS-specific T-cell immunogenicity the day before challenge in group 1. Spearman r = 0.5; P = .08. Abbreviations: ChAd63, simian adenovirus 63; CHMI, controlled human malaria infection; controls, unvaccinated volunteers undergoing CHMI; ELISPOT, enzyme-linked immunosorbent spot assay; EU, ELISA units; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara; PBMC, peripheral blood mononuclear cell; SFC, spot-forming cell.

ontrolled human malaria infection; controls, unvaccinated volunteers undergoing CHMI; ELISPOT, enzyme-linked immunosorbent spot assay; EU, ELISA units; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara; PBMC, peripheral blood mononuclear cell; SFC, spot-forming cell. ChAd63-MVA Efficacy Among All Regimens Following Sporozoite Challenge The infectivity controls (group 3) and 27 of 30 vaccinees were diagnosed with malaria. One volunteer (7%) in group 1 (who received ChAd63-MVA CS) and 2 volunteers (13%) in group 2 (who received ChAd63-MVA ME-TRAP) were sterilely protected (Figure 4A). The control volunteers (group 3) were diagnosed after a median time of 10.3 days, mean time of 10.5 days (range 8.0--14.0, SD 2.2). Three vaccinees (20%) in group 1 and 5 vaccinees (33%) in group 2 demonstrated a delay in time to treatment, relative to controls. There was no significant difference between unvaccinated controls and vaccinees in the protocol-specified end point of time to treatment for malaria (Figure 4A). However, when comparing the time to collection of the first sample after CHMI with either >500 parasites/mL (Figure 4B) or >20 parasites/mL (Figure 4C), a significant difference was seen between unvaccinated controls and vaccinees receiving ChAd63-MVA ME-TRAP (P = .01 and P = .005, respectively). Figure 4. Efficacy of ChAd63-MVA circumsporozoite protein (CS) and ME-TRAP immunization following Plasmodium falciparum 3D7 sporozoite challenge. Kaplan–Meier survival analyses. Log-rank test for significance. A, Kaplan–Meier survival analysis of time to treatment. Median time, 12.0 days for group 1 (CS), 12.5 days for group 2 (ME-TRAP), and 10.3 days for unvaccinated controls. B, Kaplan–Meier survival analysis of time to first sample with >500 parasites/mL detected by quantitative polymerase chain reaction (qPCR). Median time, 10.5 days for group 1 (CS), 12.0 days for group 2 (ME-TRAP), and 7.5 days for unvaccinated controls. C, Kaplan–Meier survival analysis of time to first sample with >20 parasites/mL detected by qPCR. Median time, 7.5 days for group 1 (CS), 9.0 days for group 2 (ME-TRAP), and 7.0 days for unvaccinated controls. Abbreviations: CHMI, controlled human malaria infection; controls, unvaccinated volunteers undergoing CHMI; ME-TRAP, multiple epitope–thrombospondin related adhesion protein.

20 parasites/mL detected by qPCR. Median time, 7.5 days for group 1 (CS), 9.0 days for group 2 (ME-TRAP), and 7.0 days for unvaccinated controls. Abbreviations: CHMI, controlled human malaria infection; controls, unvaccinated volunteers undergoing CHMI; ME-TRAP, multiple epitope–thrombospondin related adhesion protein. qPCR Data Primary analysis comparing the mean parasite density 7.5 days after CHMI (a measure of the liver to blood inoculum) showed a significant reduction when vaccinees receiving ChAd63-MVA ME-TRAP but not ChAd63-MVA CS were compared with unvaccinated control volunteers (P = .01 and P = .08, respectively, by the Mann–Whitney U test; Figure 5). The same comparison performed using negative binomial regression gave P values of .03 and .05, and a similar result was seen when the liver to blood inoculum was estimated 7.5 days after CHMI by using simple linear regression (P = .01 and P = .05, by the Mann–Whitney U test). Mean total number of parasites 7.5 days after CHMI was a strong predictor of the time to treatment (hazard ratio [HR], 1.003974 [95% CI, 1.002272–1.00568], by Cox proportional hazards regression analysis; P ≤ .0001). Figure 5. Comparison of mean parasite density, measured by quantitative polymerase chain reaction, 7.5 days after controlled human malaria infection (CHMI) between vaccinees and control volunteers. P values were determined by the Mann–Whitney U test. Abbreviations: ChAd63, simian adenovirus 63; Control, unvaccinated volunteers undergoing CHMI; CS, circumsporozoite protein; group 1, ChAd63-MVA CS recipients; group 2, ChAd63 ME-TRAP recipients; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara.

Mann–Whitney U test. Abbreviations: ChAd63, simian adenovirus 63; Control, unvaccinated volunteers undergoing CHMI; CS, circumsporozoite protein; group 1, ChAd63-MVA CS recipients; group 2, ChAd63 ME-TRAP recipients; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara. Exploratory analysis of parasite densities by using area under the curve (AUC) analysis showed that parasite density over the first 3 replication cycles in infected volunteers was a significant predictor of the time to treatment (HR, 1.000015 [95% CI, 1.000008–1.000022], by Cox proportional hazards regression analysis; P < .000; Figure 6). Over the first, second, and third blood-stage replication cycles, there was a significant reduction in parasite densities among ChAd63-MVA ME-TRAP vaccinees, as measured by AUC analysis (ie, log [parasite density + 1]), compared with unvaccinated controls, when vaccinees who achieved sterile protection were included in the analysis (cycle 1, P = .01; cycle 2, P = .03; and cycle 3 P = .05; by the 2-tailed t test, for all comparisons). Parasite densities in vaccinees receiving ChAd63 CS were significantly less than those in controls over the first blood-stage replication cycle only (P = .05 log [parasite density + 1], by the 2-tailed t test). AUC analysis showed that, compared with controls, ChAd63-MVA ME-TRAP resulted in a 79% reduction in parasitemia during cycle 1, whereas ChAd63-MVA CS caused a 69% reduction. Figure 6. Comparison of areas under the curve (AUCs) of parasite densities, measured by quantitative polymerase chain reaction (PCR), between vaccinees and control volunteers. A, Group mean log-transformed PCR data. The AUC of parasite density over the first 3 replication cycles in infected volunteers was a significant predictor of the time to diagnosis (hazard ratio, 1.000015 [95% confidence interval, 1.000008–1.000022], by Cox proportional hazards regression analysis; P < .000). B, AUC analysis of parasite densities, comparing controls to vaccinees at days 6.5–8 (the first cycle after hepatocyte release), days 8.5–10 (the second cycle), and days 10.5–12 (the third cycle) after controlled human malaria infection (CHMI). Means of log [parasite density + 1] were compared for each vaccine group to those of controls, using a 2-tailed t test.

ring controls to vaccinees at days 6.5–8 (the first cycle after hepatocyte release), days 8.5–10 (the second cycle), and days 10.5–12 (the third cycle) after controlled human malaria infection (CHMI). Means of log [parasite density + 1] were compared for each vaccine group to those of controls, using a 2-tailed t test. Abbreviations: ChAd63, simian adenovirus 63; controls, unvaccinated volunteers undergoing CHMI; CS, circumsporozoite protein; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara; SP, sterile protection.

ring controls to vaccinees at days 6.5–8 (the first cycle after hepatocyte release), days 8.5–10 (the second cycle), and days 10.5–12 (the third cycle) after controlled human malaria infection (CHMI). Means of log [parasite density + 1] were compared for each vaccine group to those of controls, using a 2-tailed t test. Abbreviations: ChAd63, simian adenovirus 63; controls, unvaccinated volunteers undergoing CHMI; CS, circumsporozoite protein; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara; SP, sterile protection. ChAd63-MVA Safety Among All Regimens Following Sporozoite Challenge No unexpected clinical or laboratory AEs were observed in vaccinees after CHMI, and there was no significant difference in the number of AEs related to CHMI between groups (P = .72; Supplementary Figure 3A). The total duration of symptoms in volunteers with symptomatic malaria ranged from 1 to 19 days (median, 6 days), with no significant difference between groups (P = .33; Supplementary Figure 3B). There was no difference between groups in the time that individuals were symptomatic before treatment (P = .43; Supplementary Figure 3C) or the number of symptoms present at time of treatment (P = .65) in volunteers with a diagnosis of malaria (Supplementary Figure 3D). Two of the 33 volunteers (6%) in whom malaria was diagnosed after CHMI had no symptoms of malaria at diagnosis. Of the volunteers with a malaria diagnosis, 28 (85%) experienced at least 1 AE after challenge that was severe in intensity (Supplementary Figure 3E). One volunteer in group 1 was admitted for inpatient management of vomiting secondary to antimalarial therapy (atovaquone/proguanil) 1 day after malaria diagnosis and was discharged the next day with no sequelae. Blood samples obtained 9, 35, and 90 days after CHMI and within 24 hours of diagnosis demonstrated transient hematological and biochemical abnormalities at frequencies and severities expected following P. falciparum infection (Supplementary Figure 3F) [40].

ia diagnosis and was discharged the next day with no sequelae. Blood samples obtained 9, 35, and 90 days after CHMI and within 24 hours of diagnosis demonstrated transient hematological and biochemical abnormalities at frequencies and severities expected following P. falciparum infection (Supplementary Figure 3F) [40]. Associations Between Immunological Outcomes and Vaccine Efficacy In group 1 but not group 2, IgG antibody responses to CS correlated significantly and negatively with qPCR-determined densities 7.5 days after CHMI (group 1: Spearman r = −0.6 [P = .03]; group 2: Spearman r = −0.3 [P = .34]; Figure 7A and 7B). A marginal negative correlation was seen in group 2 between IgG antibody responses to ME-TRAP and qPCR findings 7.5 days after CHMI (Spearman r = −0.5; P = .05; Figure 7C). No significant correlation was seen between IFN-γ ELISPOT findings for CS or ME-TRAP and qPCR findings 7.5 days after CHMI for group 1 or 2 (Figure 7D and 7E), in concordance with previous data in which ELISPOT-determined responses did not correlate with vaccine efficacy [17]. Phenotyping of the T-cell responses by flow cytometry was performed, and results will be reported in a subsequent article. Figure 7. Associations between immunological outcomes and vaccine efficacy. Correlation between parasite density at day 7.5, measured by quantitative polymerase chain reaction (qPCR), and levels of anti–circumsporozoite protein (CS) immunoglobulin G (IgG) antibody in group 1 (CS; Spearman r = −0.6; P = .03; A) and group 2 (ME-TRAP; Spearman r = −.3; P = .34; B). C, Correlation between parasite density at day 7.5, measured by qPCR, and anti-TRAP IgG antibody responses in group 2 (ME-TRAP; Spearman r = −0.5; P = .05). D, Correlation between interferon γ (IFN-γ)–secreting T-cell frequency to CS measured by enzyme-linked immunosorbent spot (ELISPOT) parasite density at day 7.5 (parasite/mL measured by qPCR) in group 1 (CS; Spearman r = −0.2; P = .50. E, Correlation between IFN-γ–secreting T-cell frequency to ME-TRAP measured by ELISPOT and parasite density at day 7.5 (parasite/mL measured by qPCR) in group 2 (ME-TRAP; Spearman r = 0.1; P = .6).

bent spot (ELISPOT) parasite density at day 7.5 (parasite/mL measured by qPCR) in group 1 (CS; Spearman r = −0.2; P = .50. E, Correlation between IFN-γ–secreting T-cell frequency to ME-TRAP measured by ELISPOT and parasite density at day 7.5 (parasite/mL measured by qPCR) in group 2 (ME-TRAP; Spearman r = 0.1; P = .6). Abbreviations: Black filled points, sterilely protected vaccinees; EU, enzyme-linked immunosorbent assay units; group 1, ChAd63-MVA CS; group 2, ChAd63-MVA ME-TRAP; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; PBMC, peripheral blood mononuclear cell; SFC, spot-forming cell; unfilled points, vaccinees demonstrating delay to start of antimalarial therapy in comparison to unvaccinated control volunteers.

nits; group 1, ChAd63-MVA CS; group 2, ChAd63-MVA ME-TRAP; ME-TRAP, multiple epitope–thrombospondin-related adhesion protein; PBMC, peripheral blood mononuclear cell; SFC, spot-forming cell; unfilled points, vaccinees demonstrating delay to start of antimalarial therapy in comparison to unvaccinated control volunteers. DISCUSSION In this first head-to-head comparison of the 2 leading preerythrocytic antigens, ME-TRAP and CS, delivered in the same vaccine platform, ME-TRAP had greater clinical efficacy, with sterile protection achieved in 13% of vaccinees (2 of 15) and a delayed time to diagnosis in 33% (5 of 15). This efficacy is slightly less than that recently reported in another CHMI study of ChAd63-MVA ME-TRAP [17], despite the induction of similar, very high frequency of antigen-specific T cells (peak median IFN-γ–secreting T cell count, 2027 in this study vs 2436 SFCs/million PBMCs in the previous study). Because the median time to diagnosis for unvaccinated control volunteers in this study was 1.5 days shorter than that of the previously reported CHMI study [17], it is possible that a larger challenge inoculum in this CHMI study could explain the small, suggested difference in efficacy results (there were no other differences in study methods). By the same reasoning, this could mean that the efficacy attained with ChAd63-MVA CS (sterile protection was achieved in 7% [1 of 15], and a delayed time to diagnosis was achieved by 20% [3 of 15]) underestimates that which may have been seen under less stringent CHMI conditions. Indeed, given that the infectious dose experienced by individuals in malaria-endemic countries is generally considerably less than that administered in CHMI studies [23], efficacy may prove to be greater in field studies.

20% [3 of 15]) underestimates that which may have been seen under less stringent CHMI conditions. Indeed, given that the infectious dose experienced by individuals in malaria-endemic countries is generally considerably less than that administered in CHMI studies [23], efficacy may prove to be greater in field studies. ChAd63-MVA CS induced moderate to high IFN-γ–expressing T-cell responses, but anti-CS IgG levels were markedly lower than that seen with in a sporozoite CHMI trial assessing RTS,S, in which 50% of vaccinees (18 of 36) receiving RTS,S/AS01B and 32% (14 of 44) receiving RTS,S/AS02A achieved sterile protection (2.1 µg/mL with ChAd63 MVA CS vs 144 mg/mL with RTS,S/AS01B and 83 mg/mL with RTS,S/AS02A) [41]. The correlation between anti-CS antibodies and time to treatment suggests this may, surprisingly, be contributing to the mechanism of efficacy even at very low levels. This study provides the first evidence that sterile immunity can be generated with viral vectors encoding CS alone [41], although it is notable that some sterile efficacy has been reported using combinations of DNA and adenoviral vectors encoding CS and AMA1 [18].

g to the mechanism of efficacy even at very low levels. This study provides the first evidence that sterile immunity can be generated with viral vectors encoding CS alone [41], although it is notable that some sterile efficacy has been reported using combinations of DNA and adenoviral vectors encoding CS and AMA1 [18]. Kaplan–Meier analysis of time to diagnosis between vaccinees and unvaccinated controls and numerous analyses of the qPCR data demonstrated significant efficacy for ChAd63-MVA ME-TRAP alone. There was no such statistically significant difference for the ChAd63-MVA CS vaccines using the same analysis. However, the AUC analysis, comparison of parasitemia at 7.5 days after CHMI, the evidence of sterile protection, and a delay to diagnosis in certain vaccinees all support the view that ChAd63-MVA CS led to a reduction (by approximately 69%–79%, depending on the analysis) in the number of parasites released from the liver. Because ChAd63-MVA ME-TRAP was, by use of the same measures, estimated to reduce the liver parasite burden by 79%–84%, it appears that relatively large reductions in liver-stage infection are required to significantly influence clinical outcomes after mosquito bite CHMI, as suggested previously [34, 35]. As this study shows, it can be difficult to quantify the efficacy of preerythrocytic vaccines that do not provide sterile immunity. We would argue that, given the necessarily small numbers of participants in CHMI studies and the importance of CHMI studies to deselect novel vaccine strategies and antigens [23], detailed analysis of qPCR data should be routinely performed to ensure that promising signals suggestive of clinically important efficacy are correctly identified.

ven the necessarily small numbers of participants in CHMI studies and the importance of CHMI studies to deselect novel vaccine strategies and antigens [23], detailed analysis of qPCR data should be routinely performed to ensure that promising signals suggestive of clinically important efficacy are correctly identified. Our data, importantly, compare the efficacy of ChAd63-MVA containing CS or ME-TRAP and, together with previous data comparing these antigens in DNA-MVA [25] and fowlpox-MVA regimes [26, 27, 42], support ME-TRAP as currently the most promising liver-stage antigen for inclusion in a future multistage vaccine. However, given the efficacy we have demonstrated here and the possibility that immunization with ME-TRAP and CS could prove to be more efficacious than either antigen alone, our next priority is to clinically assess the combination of ChAd63-MVA ME-TRAP and ChAd63-MVA CS in a CHMI trial. We suggest that detailed analyses of parasite kinetics should be routinely performed in future CHMI vaccine studies to allow detection of smaller but biologically important differences in vaccine efficacy that could influence future vaccine development.

Our data, importantly, compare the efficacy of ChAd63-MVA containing CS or ME-TRAP and, together with previous data comparing these antigens in DNA-MVA [25] and fowlpox-MVA regimes [26, 27, 42], support ME-TRAP as currently the most promising liver-stage antigen for inclusion in a future multistage vaccine. However, given the efficacy we have demonstrated here and the possibility that immunization with ME-TRAP and CS could prove to be more efficacious than either antigen alone, our next priority is to clinically assess the combination of ChAd63-MVA ME-TRAP and ChAd63-MVA CS in a CHMI trial. We suggest that detailed analyses of parasite kinetics should be routinely performed in future CHMI vaccine studies to allow detection of smaller but biologically important differences in vaccine efficacy that could influence future vaccine development. Supplementary Data Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.

ordjournals.org). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author. Supplementary Data Notes Acknowledgments. We thank M. Smith and R. Lopez-Ramon, for clinical assistance; S. French, for logistical support; J. Furze, M Cottingham, and L. Coughlan, for laboratory assistance; the staff at the Southampton NIHR Wellcome Trust Clinical Research Facility; the study volunteers; and staff at the WRAIR Malaria Serology ELISA Reference Laboratory, for performing the NANP ELISA. Disclaimer. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Financial support. This work was supported by the PATH Malaria Vaccine Initiative, the United Kingdom National Institute of Health Research, through the Oxford Biomedical Research Centre (grant A91301, Adult Vaccine), and the Wellcome Trust (grants 084113/Z/07/Z and 45488/Z/05 to A. V. S. H. and grant 097940/Z/11/Z to S. H. H.). Potential conflicts of interest. A. V. S. H. and S. C. G. are named inventors on patent applications covering malaria vectored vaccines and immunization regimens. S. C. and A. N. are employees of and/or shareholders in Okairos, which is developing vectored vaccines for malaria and other diseases. All other authors report no potential conflicts.

erest. A. V. S. H. and S. C. G. are named inventors on patent applications covering malaria vectored vaccines and immunization regimens. S. C. and A. N. are employees of and/or shareholders in Okairos, which is developing vectored vaccines for malaria and other diseases. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

Malaria remains one of the leading causes of mortality globally [1], and there is urgent need for a vaccine. The majority of deaths are in children <5 years old, with this age group accounting for approximately 306 000 deaths in 2015. The enormous economic and social consequences of malaria have been well documented [2]. Efforts to develop effective vaccines are complicated by the complex immunology of malaria parasite infection, and no reliable natural model of complete immunity exists. Despite this, a small number of candidate vaccines have shown partial efficacy against experimental and natural human infection, with the current leading vaccine being the recombinant protein in adjuvant, RTS,S/AS01. RTS,S targets circumsporozoite protein (CS), which is expressed by the Plasmodium falciparum sporozoite at the preerythrocytic stage and was the first subunit vaccine to show high rates of sterile efficacy, typically 50%, in controlled human malaria infection (CHMI) studies [3]. In a large African phase 3 trial, this vaccine had an efficacy against clinical malaria of 55.8% (97.5% confidence interval [CI], 50.6%–60.4%) in children aged 5–17 months and 31.3% (23.6%–38.3%) in infants aged 6–12 weeks over the first year after vaccination [4, 5]. Vaccine efficacy wanes over time but can be enhanced by a fourth dose [6]. Analysis of the immunological correlates of efficacy of this vaccine suggest that vaccine-induced antibodies targeting CS are the most important mediators of RTS,S-induced protection against malaria [3], although no antibody level threshold has been shown to be predictive of efficacy. The rate at which anti-CS antibodies wane is similar to the rate at which efficacy declines [7, 8], suggesting that anti-CS antibodies may also be associated with the duration of protection. A number of factors, including age at vaccination, human immunodeficiency virus status, and high baseline anti-CS antibody titers influence anti-CS antibody titers after vaccination with RTS,S [9].

which efficacy declines [7, 8], suggesting that anti-CS antibodies may also be associated with the duration of protection. A number of factors, including age at vaccination, human immunodeficiency virus status, and high baseline anti-CS antibody titers influence anti-CS antibody titers after vaccination with RTS,S [9]. The preerythrocytic stage of P. falciparum infection presents an attractive target for an efficacious human vaccine because sufficient reduction in the number of viable merozoites reaching the blood from the liver will prevent parasitization of red blood cells and initiation of the symptomatic blood stage of infection. Anti-CS antibodies can target sporozoites for destruction prior to hepatocyte invasion. Because sporozoites travel from the skin to liver within minutes, it may be difficult for a vaccine to achieve complete protection against P. falciparum based solely on antibodies to sporozoites. The liver stage of infection provides a longer window of opportunity for cell-mediated immunity to recognize and destroy infected hepatocytes. Chimpanzee adenovirus 63 (ChAd63) with a multiepitope string fused to thrombospondin-related adhesion protein (ME-TRAP) insert and modified vaccinia virus Ankara (MVA) with the ME-TRAP insert are viral-vectored vaccines, and when they are administered in a prime-boost sequence at an 8-week interval, they are a leading candidate vaccine strategy targeting the liver stage of infection [10]. The ChAd63 and MVA viral vectors deliver the recombinant ME-TRAP insert, which generates a potent cellular immune response against the liver-stage P. falciparum antigen, TRAP, of greater magnitude than the cellular response induced by RTS,S/AS01. This strategy showed durable partial efficacy in 2 phase 2a sporozoite challenge trials in the United Kingdom [11, 12], using the 3D7 parasite as a challenge strain. The viral vector–encoded P. falciparum TRAP allele is from the heterologous T9/96 strain, and induced T-cell responses correlate with efficacy [11]. Therefore, these are effectively heterologous strain CHMI studies. Interestingly, a higher level of efficacy of 67% (95% CI, 33%–83%) against P. falciparum infections detected by polymerase chain reaction (PCR) was observed in a phase 2b trial in Kenyan adults [13]. Again, T cells to TRAP peptides correlated with vaccine efficacy, but the short duration of malaria transmission and follow-up at this trial site precluded analysis of the durability of vaccine-induced protection [13].

s detected by polymerase chain reaction (PCR) was observed in a phase 2b trial in Kenyan adults [13]. Again, T cells to TRAP peptides correlated with vaccine efficacy, but the short duration of malaria transmission and follow-up at this trial site precluded analysis of the durability of vaccine-induced protection [13]. This heterologous prime-boost strategy showed potent cellular immunogenicity in adults in the United Kingdom [11], as well as adults and infants in malaria-endemic areas [13–15] (Ewer et al, unpublished data) and has an excellent track record of safety and tolerability in these populations. Analysis of the potential utility of combining antisporozoite and anti–liver-stage vaccines have suggested a likely additive or synergistic effect [16], in keeping with findings from preclinical studies [17, 18]. In this phase 1/2a, open-labeled, CHMI study, we assessed the safety, immunogenicity, and efficacy of a vaccine schedule combining these 2 distinct candidate vaccine types in a staggered immunization regimen: one that induces very high titer antibodies to CS, using RTS,S/AS01B, and another that induces potent T-cell responses to TRAP, using viral vectors.

I study, we assessed the safety, immunogenicity, and efficacy of a vaccine schedule combining these 2 distinct candidate vaccine types in a staggered immunization regimen: one that induces very high titer antibodies to CS, using RTS,S/AS01B, and another that induces potent T-cell responses to TRAP, using viral vectors. METHODS Participants Recruitment and vaccination was conducted at 3 United Kingdom sites, in Oxford, Southampton and London. The CHMI procedure was performed as previously described [19] at Imperial College, London, using 5 infectious bites from Anopheles stephensi mosquitoes infected with P. falciparum strain 3D7. All subjects were infected with a single batch of mosquitoes at the initial CHMI and with a second single batch at the repeat CHMI. Infected mosquitoes were supplied by the Department of Entomology, Walter Reed Army Institute of Research (Washington D.C.). Healthy, malaria-naive males and nonpregnant females aged 18–45 years were invited to participate in the study. All volunteers gave written informed consent prior to participation, and the study was conducted according to the principles of the Declaration of Helsinki and in accordance with good clinical practice (GCP).

Healthy, malaria-naive males and nonpregnant females aged 18–45 years were invited to participate in the study. All volunteers gave written informed consent prior to participation, and the study was conducted according to the principles of the Declaration of Helsinki and in accordance with good clinical practice (GCP). Ethical and Regulatory Approval Necessary approvals for the study were granted by the United Kingdom National Research Ethics Service, Committee South Central–Oxford A (reference 13/SC/0208), the Western Institution Review Board (reference 20130698), and the United Kingdom Medicines and Healthcare Products Regulatory Agency (reference 21584/0317/001-0001). The trial was registered with ClinicalTrials.gov (reference NCT01883609). The Local Safety Committee provided safety oversight, and GCP compliance was independently monitored externally by the Clinical Trials and Research Governance Team of the University of Oxford.

ts Regulatory Agency (reference 21584/0317/001-0001). The trial was registered with ClinicalTrials.gov (reference NCT01883609). The Local Safety Committee provided safety oversight, and GCP compliance was independently monitored externally by the Clinical Trials and Research Governance Team of the University of Oxford. Study Design This phase 2a, open-labeled, partially randomized challenge trial consisted of 4 cohorts. Allocation to study group occurred at screening and was based on subject preference. Any subjects without a preference were randomly assigned to vaccine group 1 or vaccine group 2. Group 1 (n = 20) received 5 vaccinations (RTS,S/AS01B 50 µg at 0, 4, and 8 weeks, ChAd63 ME-TRAP 5 × 1010 virus particles at 2 weeks, and MVA ME-TRAP 2 × 108 plaque-forming units at 10 weeks); group 2 (n = 20) received 3 vaccinations (RTS,S/AS01B 50 µg at 0, 4, and 8 weeks); and group 3 (n = 6) received no vaccinations. All vaccinations were administered intramuscularly into the deltoid region of the arm. In each volunteer, all RTS,S/AS01B injections were given in one arm, and all viral vector injections were given in the contralateral arm. All subjects underwent initial CHMI by mosquito bite at the same time (week 12 after first vaccination for vaccinated subjects). Following CHMI, a diagnosis of blood-stage malaria parasite infection was made in subjects with symptoms suggestive of malaria and positive findings of thick film microscopy or, if either thick film was negative or symptoms were absent, in subjects with a qPCR result of >500 parasites/mL [12]. Vaccinated subjects who had not developed blood-stage malaria by day 21 after CHMI were deemed to exhibit sterile protection and were invited to undergo repeat CHMI 6 months later, for which an additional control group (group 4) was recruited.

ptoms were absent, in subjects with a qPCR result of >500 parasites/mL [12]. Vaccinated subjects who had not developed blood-stage malaria by day 21 after CHMI were deemed to exhibit sterile protection and were invited to undergo repeat CHMI 6 months later, for which an additional control group (group 4) was recruited. Further details of the study sites, inclusion/exclusion criteria, vaccines, clinical follow-up, safety monitoring, malaria treatment and diagnosis, immunological and molecular methods, and statistical analysis can be found in the Supplementary Materials.

ptoms were absent, in subjects with a qPCR result of >500 parasites/mL [12]. Vaccinated subjects who had not developed blood-stage malaria by day 21 after CHMI were deemed to exhibit sterile protection and were invited to undergo repeat CHMI 6 months later, for which an additional control group (group 4) was recruited. Further details of the study sites, inclusion/exclusion criteria, vaccines, clinical follow-up, safety monitoring, malaria treatment and diagnosis, immunological and molecular methods, and statistical analysis can be found in the Supplementary Materials. RESULTS Participants Eighty subjects were screened for eligibility, and 48 subjects were identified as eligible. Twenty subjects were allocated to group 1 to receive RTS,S/AS01B and viral vectors encoding ME-TRAP. Seventeen subjects were allocated to group 2 to receive RTS,S/AS01B only. Six unvaccinated controls were recruited to group 3 for the initial CHMI, and 5 subjects were allocated to group 4 for the repeat CHMI. Vaccinations took place between 2 September 2013 and 13 November 2013. Prior to CHMI, 3 subjects withdrew from group 1, and 1 subject withdrew from group 2. There were no withdrawals due to safety concerns, and no predefined study stopping or holding rules were activated. CHMI was performed on 25 and 26 November 2013, and repeat CHMI was performed on 13 May 2014 (Figure 1). Figure 1. Flow of study design and volunteer recruitment. Twenty-seven subjects were excluded because of inclusion/exclusion criteria. Three subjects withdrew consent after screening but before enrollment. Two subjects were deemed eligible as control subjects but only after group 3 enrollment was complete. They were kept as backup subjects in case of last-minute withdrawals from group 3 but never underwent controlled human malaria infection (CHMI). Seventeen subjects expressed a preference as to which vaccine group to be allocated to and were assigned accordingly. Twenty subjects expressed no preference for vaccine group allocation, and were therefore randomized to group by the study statistician. Abbreviations: ChAd63, chimpanzee adenovirus serotype 63; ME-TRAP, multiple-epitope thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara.

assigned accordingly. Twenty subjects expressed no preference for vaccine group allocation, and were therefore randomized to group by the study statistician. Abbreviations: ChAd63, chimpanzee adenovirus serotype 63; ME-TRAP, multiple-epitope thrombospondin-related adhesion protein; MVA, modified vaccinia virus Ankara. Protective Efficacy Against CHMI A total of 39 subjects participated in the initial CHMI (17 subjects from group 1, 16 subjects from group 2, and 6 subjects from group 3), which was conducted over 2 days. Three subjects in group 1 and 4 subjects in group 2 received a diagnosis of malaria before day 21 after challenge, resulting in a sterile efficacy of 82.4% (95% CI, 64%–100%) and 75% (95% CI, 54–96), respectively (Figure 2). The median time to diagnosis was 14.5 days in group 1 and 13.25 days in group 2. All 6 control subjects received a diagnosis of malaria, with a median time to diagnosis of 12.25 days (range, 11–13 days; SD, 0.7 days). Both vaccine regimens demonstrated a significantly reduced risk of malaria parasite infection over controls in the per protocol analysis (group 1 hazard ratio [HR], 0.065; P < .0001; group 2 HR, 0.12; P < .0001 for group 2), but there was no significant difference in efficacy between vaccine regimen (HR, 0.65; P .57). Eight protected subjects from group 1 and 6 protected subjects from group 2 agreed to undergo repeat CHMI. A single subject each from group 1 and group 2 received a diagnosis of malaria, on day 17 and day 14.5, respectively, and all 5 control subjects developed malaria, with a mean time to diagnosis of 12.4 days (median, 12.5 days; range, 11.5–13.5 days; SD, 0.8 days). Figure 2. Efficacy of RTS,S/AS01B plus chimpanzee adenovirus 63 and modified vaccinia virus Ankara–vectored vaccine (ChAd63-MVA) expressing a multiepitope string fused to thrombospondin-related adhesion protein (ME-TRAP) and RTS,S/AS01B alone following Plasmodium falciparum 3D7 sporozoite challenge. A, Kaplan–Meier survival analysis of the time to treatment following initial controlled human malaria infection (CHMI). Mean time to diagnosis (± standard deviation [SD]) was 12.2±0.7 days for unvaccinated controls. Seventeen of 17 subjects (100%) in group 1 and 14 of 16 subjects (87.5%) in group 2 had no diagnosis by day 21 or received a diagnosis after the control mean time + 2 SD. B, Kaplan–Meier survival analysis of the time to the first sample with >20 parasites/mL detected by quantitative polymerase chain reaction (qPCR).

s. Seventeen of 17 subjects (100%) in group 1 and 14 of 16 subjects (87.5%) in group 2 had no diagnosis by day 21 or received a diagnosis after the control mean time + 2 SD. B, Kaplan–Meier survival analysis of the time to the first sample with >20 parasites/mL detected by quantitative polymerase chain reaction (qPCR). Mean time to end point (±SD) was 7.4±0.7 days for unvaccinated controls. Sixteen of 17 subjects (94.1%) in group 1 and 15 of 16 subjects (93.8%) in group 2 did not reach this end point or did so after the control mean time + 2 SD. C, Kaplan–Meier survival analysis of the time to the first sample with >500 parasites/mL detected by qPCR. Mean time to end point (±SD) was 9.8±0.8 days for unvaccinated controls. Seventeen of 17 subjects (100%) in group 1 and 15 of 16 subjects (93.8%) in group 2 did not reach this end point or did so after the control mean time + 2 SD. D, Kaplan–Meier survival analysis of the time to treatment following repeat CHMI in protected subjects. Significance testing was performed by the log-rank test.

Seventeen of 17 subjects (100%) in group 1 and 15 of 16 subjects (93.8%) in group 2 did not reach this end point or did so after the control mean time + 2 SD. D, Kaplan–Meier survival analysis of the time to treatment following repeat CHMI in protected subjects. Significance testing was performed by the log-rank test. Safety The safety profile of a 3-dose regimen of RTS,S/AS01B and of ChAd63-MVA ME-TRAP when given separately to malaria-naive adults has been described previously [3, 10–12, 20], and a similar reactogenicity profile was observed after vaccination in this study. The majority of adverse events (AEs) following vaccinations in both group 1 and group 2 were mild in severity and self-limiting. There were no serious AEs related to vaccination, and no suspected, unexpected serious adverse reactions (SUSARs). Solicited and unsolicited AEs following vaccination are detailed in Supplementary Tables 1–12.

verse events (AEs) following vaccinations in both group 1 and group 2 were mild in severity and self-limiting. There were no serious AEs related to vaccination, and no suspected, unexpected serious adverse reactions (SUSARs). Solicited and unsolicited AEs following vaccination are detailed in Supplementary Tables 1–12. Humoral Response to Vaccination Anti-TRAP IgG antibodies were measured in group 1 subjects only (Figure 3), and geometric mean titers (GMTs) peaked on the day before challenge, at 947 ELISA units (EU; 95% CI, 617–1455). No association was detected between anti-TRAP IgG levels and efficacy (Spearman r = −0.25; P = .3). Anti-CS antibodies were measured at key time points in all vaccinated subjects. Serum anti-CS antibody levels peaked on the day before challenge in both vaccinated groups, with peak GMTs of 1733 EU (95% CI, 1240–2422) and 1824 EU (95% CI, 1330–2502) in groups 1 and 2, respectively. There was no significant difference in anti-CS antibody GMTs between group 1 and group 2 on the day before challenge (P > .99, by the Mann–Whitney test). Anti-CS antibody GMTs on the day before challenge were significantly higher in protected subjects (1985 EU [95% CI, 1584–2487]), compared with those in nonprotected subjects (1177 [95% CI, 627–2209]; P = .035, by the Mann–Whitney test; Figure 3). There was a correlation between anti-CS antibody titer and parasite density on day 7.5 (Spearman r = −0.4; P = .018). There was no significant difference in the avidity of anti-CS antibodies between protected and nonprotected volunteers at any time point, but avidity significantly increased between day 28 and the day before challenge in protected but not nonprotected volunteers (P = .001 and P > .99, respectively, by the Wilcoxon matched pairs test). Avidity also increased between day 56 and the day before challenge in protected but not nonprotected volunteers (P < .0001 and P = .375, respectively, by the Wilcoxon matched pairs test). In the protected vaccinated subjects who underwent repeat CHMI, avidity on the day before rechallenge remained significantly higher than at day 28 (P = .002, by the Mann–Whitney test). Figure 3. Antibody responses to vaccination, measured by enzyme-linked immunosorbent assay (ELISA).

y the Wilcoxon matched pairs test). In the protected vaccinated subjects who underwent repeat CHMI, avidity on the day before rechallenge remained significantly higher than at day 28 (P = .002, by the Mann–Whitney test). Figure 3. Antibody responses to vaccination, measured by enzyme-linked immunosorbent assay (ELISA). A, Anti-thrombospondin adhesion protein (TRAP) immunoglobulin G (IgG) antibody responses after vaccination with RTS,S/AS01B plus chimpanzee adenovirus 63 and modified vaccinia virus Ankara–vectored vaccine (ChAd63-MVA) expressing a multiepitope string fused to TRAP (ME-TRAP; group 1 subjects only). Lines represent group medians. B, Anti–circumsporozoite protein (CS) IgG antibody responses after vaccination with RTS,S/AS01B plus ChAd63-MVA ME-TRAP (group 1; blue) or RTS,S alone (group 2; black). Line represents group median. C, Comparison of anti-CS IgG antibody responses between group 1 (blue) and group 2 (black) as measured on the day before controlled human malaria infection (CHMI). P > .999, by the Mann–Whitney test. Comparison of anti-CS IgG antibody responses in volunteers who were or were not sterilely protected. Lines represent geometric means. D, Correlation between anti-CS IgG titers on the day before challenge and parasite density on day 7 after challenge. Spearman r = −0.4; P = .018. E, Avidity of total IgG against the NANP repeat region of circumsporozoite protein. Significant increase in avidity between day 28 and the day before challenge in protected but not nonprotected volunteers. P = .001, by the Wilcoxon matched pairs test. Avidity of total IgG remained significantly higher at time of the second CHMI (RC-1) than at day 28. P = .002, Mann–Whitney test. Lines represent geometric mean. Abbreviations: C+, elapsed time after CHMI, in days; C-1, day before CHMI; PCR, polymerase chain reaction; RC-1, day before second CHMI.

n matched pairs test. Avidity of total IgG remained significantly higher at time of the second CHMI (RC-1) than at day 28. P = .002, Mann–Whitney test. Lines represent geometric mean. Abbreviations: C+, elapsed time after CHMI, in days; C-1, day before CHMI; PCR, polymerase chain reaction; RC-1, day before second CHMI. Cellular Response to Vaccination T-cell responses against ME-TRAP were measured in all group 1 subjects by an ex vivo interferon γ (IFN-γ) enzyme-linked immunosorbent spot (ELISPOT) assay (Figure 4). Peak responses after ChAd63 ME-TRAP vaccination were detected 21 days later (Geometric Mean, 539 spot-forming cells [SFCs] per million peripheral blood mononuclear cells [PBMCs]; 95% CI, 300–968 SFCs per million PBMCs). Peak responses after MVA ME-TRAP vaccination were detected 7 days later (median, 1520 SFCs per million PBMCs; interquartile range [IQR], 699–3305 SFCs per million PBMCs). T-cell responses against ME-TRAP were well maintained over time, with a median of 464 SFCs per million PBMCs (IQR, 231–933 SFCs per million PBMCs) 90 days after initial challenge and 342 SFCs per million PBMCs (IQR, 143–815 SFCs per million PBMCs) in participating subjects the day before repeat CHMI. Figure 4. Antigen-specific T-cell responses to vaccination, measured by interferon γ (IFN-γ) enzyme-linked immunospot assay (ELISPOT). A, Median T-cell responses to multiepitope string fused to thrombospondin-related adhesion protein (ME-TRAP). B, Median T-cell responses to circumsporozoite protein (CS) peptide pools are shown for group 1 (RTS,S/AS01 and ME-TRAP; blue line) and group 2 (RTS,S/AS01; black line). Abbreviations: ChAd63, chimpanzee adenovirus 63; MVA, modified vaccinia virus Ankara; PBMC, peripheral blood mononuclear cell; SFC, spot-forming cell.

B, Median T-cell responses to circumsporozoite protein (CS) peptide pools are shown for group 1 (RTS,S/AS01 and ME-TRAP; blue line) and group 2 (RTS,S/AS01; black line). Abbreviations: ChAd63, chimpanzee adenovirus 63; MVA, modified vaccinia virus Ankara; PBMC, peripheral blood mononuclear cell; SFC, spot-forming cell. T-cell responses against CS were measured in all vaccinated subjects by an IFN-γ ELISPOT assay (Figure 4). Responses peaked in group 1 on the day before challenge (4 weeks after final dose of RTS,S/AS01B; median, 36 SFCs per million PBMCs; IQR, 12–176 SFCs per million PBMCs), with a median response of 12 SFCs per million PBMCs (IQR, 12–70 SFCs per million PBMCs) in group 2 at the same time point. No association between IFN-γ ELISPOT responses to TRAP or CS and vaccine efficacy was detected (Spearman r = −0.01 [P = .98] and r = −0.001 [P = .0996] for TRAP and CS, respectively).

llion PBMCs), with a median response of 12 SFCs per million PBMCs (IQR, 12–70 SFCs per million PBMCs) in group 2 at the same time point. No association between IFN-γ ELISPOT responses to TRAP or CS and vaccine efficacy was detected (Spearman r = −0.01 [P = .98] and r = −0.001 [P = .0996] for TRAP and CS, respectively). Flow cytometry using intracellular cytokine staining (ICS) was performed for CS and hepatitis B virus surface antigen (HBsAg) on day 42 after the first vaccination and on the day before challenge, using cryopreserved PBMCs. In this assay, responses were measured as the number of cells per million CD4+ or CD8+ T cells expressing at least 2 markers from among CD154 (CD40 ligand), IFN-γ, interleukin 2, and tumor necrosis factor α (Figure 5A and 5B). CS-specific CD4+ T-cell responses peaked on day 42 (2 weeks after the second dose of RTS,S) in both groups, and no significant differences were detected between groups 1 and 2 either on day 42 or the day before challenge (Figure 5C). CS-specific CD8+ T-cell responses were not detected at any significant frequency. A positive association was detected between the number of polyfunctional CD4+ T cells at day 42 and the level of anti-CS IgG in serum on the day before challenge (Spearman r = 0.4; P = .03; Figure 5D). Vaccination with RTS,S increased the frequency of HBsAg-specific CD4+ polyfunctional T cells in both groups at all time points after vaccination (Figure 5E and 5F). Figure 5. T-cell responses determined by flow cytometry on cryopreserved peripheral blood mononuclear cells before and after vaccination for circumsporozoite protein (CS) and hepatitis B virus surface antigen (HBsAg). Polypositivity indicates number of cells per million expressing ≥2 of the following markers: CD154 (CD40 ligand), interferon γ (IFN-γ), interleukin 2 (IL-2), and tumor necrosis factor α (TNF-α). A and B, Number of CS-specific polypositive CD4+ or CD8+ T cells per million in groups 1 and 2, respectively. C, Comparison of CD4+ polypositive T cells at peak time point after vaccination (day 42) and the day before controlled human malaria infection (CHMI) for groups 1 (G1) and 2 (G2). D, Correlation between peak CS-specific CD4+ polypositive frequency and anti-CS IgG level on the day before challenge (r = 0.4; P = .03, by the Spearman test). E and F, Number of HBsAg-specific polypositive CD4+ or CD8+ T cells per million in groups 1 and 2, respectively.

infection (CHMI) for groups 1 (G1) and 2 (G2). D, Correlation between peak CS-specific CD4+ polypositive frequency and anti-CS IgG level on the day before challenge (r = 0.4; P = .03, by the Spearman test). E and F, Number of HBsAg-specific polypositive CD4+ or CD8+ T cells per million in groups 1 and 2, respectively. Abbreviations: C+, elapsed time after CHMI, in days; C-1, day before CHMI; IgG, immunoglobulin G.

infection (CHMI) for groups 1 (G1) and 2 (G2). D, Correlation between peak CS-specific CD4+ polypositive frequency and anti-CS IgG level on the day before challenge (r = 0.4; P = .03, by the Spearman test). E and F, Number of HBsAg-specific polypositive CD4+ or CD8+ T cells per million in groups 1 and 2, respectively. Abbreviations: C+, elapsed time after CHMI, in days; C-1, day before CHMI; IgG, immunoglobulin G. Flow cytometry was also performed on freshly isolated PBMCs, using CS peptides (for groups 1 and 2) and ME-TRAP peptides (for group 1 only) on the day before challenge. Group 1 responses to TRAP T9/96 and 3D7 were comparable across all cytokines and CD107a (P < .0001, by the Kruskal–Wallis test with the Dunn correction), with all volunteers exhibiting at least 1 positive cytokine response to both TRAP strains (Figure 6A and 6B). A positive response to CS was observed in 15 of 17 group 1 volunteers (83%), compared with just 9 of 16 volunteers in group 2 (56%), with a significantly higher frequency of IFN-γ–producing CD4+ T cells in group 1 (Figure 6C). Figure 6. Intracellular cytokine staining of peripheral blood mononuclear cells (PBMCs) 1 day before controlled human malaria infection (CHMI; 27 days after the final RTS,S vaccination and 13 days after vaccination with modified vaccinia virus Ankara [MVA] expressing a multiepitope string fused to thrombospondin-related adhesion protein [ME-TRAP]), showing the CD107a expression frequency and the frequencies of cytokine-secreting cells as a percentage of the frequency of parent CD4+ and CD8+ T cells. Geometric mean of each response is shown in response to stimulation with TRAP T9/96 peptides (homologous to vaccine insert) by group 1 (A), TRAP 3D7 peptides (homologous to CHMI challenge strain) by group 1 (B), and circumsporozoite (CS) peptides by groups 1 and 2 (C). D, Ex vivo interferon γ enzyme-linked immunospot (ELISPOT)–determined responses of group 1 and 2 volunteers to CS peptides split into 3 peptide pools and a combined pool, with background subtracted. Dotted line shows the median background ELISPOT response, setting the positive response threshold. Data are for 17 individuals in group 1 and 16 in group 2. Data points represent individual volunteers. Abbreviations: IFN-γ, interferon γ; IL-2, interleukin 2; SFC, spot-forming cell; TNF-α, tumor necrosis factor α.

cted. Dotted line shows the median background ELISPOT response, setting the positive response threshold. Data are for 17 individuals in group 1 and 16 in group 2. Data points represent individual volunteers. Abbreviations: IFN-γ, interferon γ; IL-2, interleukin 2; SFC, spot-forming cell; TNF-α, tumor necrosis factor α. Ex vivo IFN-γ ELISPOT analysis revealed a trend toward higher responses to CS peptides in group 1 as compared to group 2 (P = .0517, by the Mann–Whitney test on combined groups); when assessed by peptide pool, a significant trend toward higher responses was observed in pool 1 (P = .0380, by the Mann–Whitney test). This is likely due to CS epitope(s) present in the ME string of ChAd63 ME-TRAP and MVA ME-TRAP. Ex vivo IFN-γ data suggest that this epitope lies toward the N-terminus of CS, as identified by a significantly higher group 1 response to peptide pool 1. The ME string contains 2 epitopes present in pool 1: CD8 epitope cp26 KPKDELDY and CD4 epitope DPNANPN, as part of a longer ME sequence, DPNANPNNVDPNANPNV (Table 1). As the main differences in ICS-determined IFN-γ production were in the CD4+ T-cell compartment, epitope DPNANPN could be responsible for the enhanced CS responses in group 1. This epitope is not present in RTS,S, so it was solely induced by ChAd63.MVA ME-TRAP prime boost vaccination. Table 1. Comparison of Peptide Sequences Present in the Multiepitope (ME) String Fused to Thrombospondin-Related Adhesion Protein and the T-Cell Region of RTS,S

onsible for the enhanced CS responses in group 1. This epitope is not present in RTS,S, so it was solely induced by ChAd63.MVA ME-TRAP prime boost vaccination. Table 1. Comparison of Peptide Sequences Present in the Multiepitope (ME) String Fused to Thrombospondin-Related Adhesion Protein and the T-Cell Region of RTS,S Epitope Sequence CS Amino Acid Position (Length) Epitope Type Present in ME String Present in RTS,S Present in ELISPOT CS Peptides ELISPOT Pool Number DPNANPNVDP NANPNV 111–126 (16) CD4 Yes No Yes, DNANPN only 1 NMPNDPN RNV 286–293 (8) CD8 Yes Yes Yes, PNDPN RNV only 1 YL NKIQNSL 319–327 (9) CD8 Yes Yes Yes, full length 2 KPKDELDY 353–360 (8) CD8 Yes Yes Yes, full length 3 Data are from Lalvani et al [21]. Abbreviations: CS, circumsporozoite protein; ELISPOT, enzyme-linked immunospot. DISCUSSION Both RTS,S/AS01B and ChAd63-MVA encoding ME-TRAP have previously demonstrated partial efficacy in CHMI trials [3, 11, 12, 20], but this is the first study in which RTS,S/AS01B and ChAd63-MVA ME-TRAP have been given to subjects in the same vaccine regimen. In this study, we have shown that administering these vaccines sequentially is safe, with no SUSARs and no vaccine-related serious AEs. The reactogenicity profile observed in the subjects who received the combined vaccine regimen (group 1) was similar to that observed when RTS,S/AS01B or ChAd63-MVA ME-TRAP were given alone in a malaria-naive adult population [3, 11, 12, 20].

cines sequentially is safe, with no SUSARs and no vaccine-related serious AEs. The reactogenicity profile observed in the subjects who received the combined vaccine regimen (group 1) was similar to that observed when RTS,S/AS01B or ChAd63-MVA ME-TRAP were given alone in a malaria-naive adult population [3, 11, 12, 20]. Furthermore, we have demonstrated that these vaccine candidates remain immunogenic when the regimens are combined. Anti-CS antibodies were not significantly different between group 1 and group 2 on the day before challenge, and peak numbers of TRAP-specific T cells in group 1 were similar to those observed with ChAd63-MVA ME-TRAP administered alone in a previous study [11]. GMTs of anti-CS antibodies were significantly higher in protected subjects on the day before challenge, but there was no correlation between any TRAP- or CS-specific T cell counts or TRAP-specific IgG and protection.

p 1 were similar to those observed with ChAd63-MVA ME-TRAP administered alone in a previous study [11]. GMTs of anti-CS antibodies were significantly higher in protected subjects on the day before challenge, but there was no correlation between any TRAP- or CS-specific T cell counts or TRAP-specific IgG and protection. In this study, we observed a high level of protective efficacy in both vaccine arms. A higher proportion of subjects in group 1 remained protected following CHMI than in group 2 (82.4% vs 75%), although this difference was not statistically significant (P = .57). This high level of vaccine efficacy was also seen to be durable at 6 months, with 87.5% and 83.3% of initially protected subjects who underwent repeat CHMI remaining protected in groups 1 and 2, respectively. In addition, a higher proportion of subjects in group 1 reached the secondary efficacy end points of delayed time to malaria diagnosis and delayed time to PCR-confirmed parasitemia, compared with group 2. The trends observed in this study for initial challenge, rechallenge, and effects on the prepatent period are encouraging for further evaluation of the group 1 regimen, but the numbers in this study are small, and the differences observed not statistically significant. In 2013, a CHMI study of the cryopreserved whole sporozoite (PfSPZ) vaccine reported sterile efficacy of 100% in the high-dose regimen, consisting of 5 doses of 1.35 × 105 parasites [22]. However, the vaccinee numbers in the high-dose group were small (n = 6), and only 5 of 6 unvaccinated controls (83.3%) developed blood-stage infection, raising concerns over the infectivity of the parasites used in that CHMI. The results observed in the trial we present in this article therefore are amongst the highest published sterile vaccine efficacy in any CHMI study in which all control subjects were infected.

ted controls (83.3%) developed blood-stage infection, raising concerns over the infectivity of the parasites used in that CHMI. The results observed in the trial we present in this article therefore are amongst the highest published sterile vaccine efficacy in any CHMI study in which all control subjects were infected. The level of protective efficacy observed in the RTS,S/AS01B alone group (75%) is higher than has been reported in most prior CHMI studies of this vaccine regimen [3, 20]. The mean time to patency in the control group of 12.2 days indicates that this was not an unusually weak challenge, and the vaccination and CHMI methods used in this trial are largely comparable to those in other CHMI studies of this dosing schedule of RTS,S/AS01 [3]. Practical limitations on study size are a factor for both this study and prior CHMI studies of RTS,S, resulting in a relatively small historical data set. In light of this, it is possible that the higher efficacy seen in the RTS,S alone group in this trial is a chance finding due in part to small numbers, or that further CHMI studies of RTS,S/AS01 in malaria-naive subjects, including further evaluation of differing dosing regimens and schedules, would further clarify the efficacy of the vaccine in this setting. The study was designed to have 84% power to detect a significant (P < .05) increase in sterile efficacy in group 1 to 90% and 69% power to detect a significant increase to 85%, compared with group 2. This power calculation assumed an expected 50% sterile efficacy in group 2 [3]. The increase in efficacy to 82.4% in group 1 from 75% in group 2 observed in this trial was not statistically significant (P = .69), but the power to detect a statistically significant improvement was very limited. Practical limitations of CHMI trials makes conducting large studies difficult, and designing future studies with sufficient power would be complicated, assuming an efficacy of 75% in an RTS,S/AS01B alone group. One alternative approach is to wait longer after immunization, to allow vaccine efficacy to wane and thereby provide greater power to detect additive or synergist effects of combination vaccines. Further consideration of this issue and of the practical limitations of CHMI studies with challenge 3–4 weeks after the last vaccination in future trial designs is warranted.

mmunization, to allow vaccine efficacy to wane and thereby provide greater power to detect additive or synergist effects of combination vaccines. Further consideration of this issue and of the practical limitations of CHMI studies with challenge 3–4 weeks after the last vaccination in future trial designs is warranted. We undertook a rechallenge of protected subjects 6 months after the initial CHMI, and 7 of 8 group 1 subjects (87.5%) and 5 of 6 group 2 subjects (83.3%) remained protected. By simply calculating the product of the percentage efficacies in the 2 CHMIs, one can estimate vaccine efficacy at 6–7 months after the immunizations as a measure of durable sterile protection. For group 1, this is 72% [(14/17) ×( 7/8) × 100], and for group 2, it is 62.5% [(12/16) × (5/6) × 100)]. Again, the durable protection rate at this time point in group 2 appears higher than in previous rechallenge trials with RTS,S/AS01B administered 3 times [3], and the group 1 protection rate is even higher. This durability at 6 months is also encouraging for continued investigation of combination vaccine approaches and supports the consideration of delayed CHMI as an approach to evaluating improvements in efficacy provided by vaccines that confer substantial short-term efficacy.

the group 1 protection rate is even higher. This durability at 6 months is also encouraging for continued investigation of combination vaccine approaches and supports the consideration of delayed CHMI as an approach to evaluating improvements in efficacy provided by vaccines that confer substantial short-term efficacy. In this trial, we present data from a combined vaccine regimen in which subjects received 5 vaccinations over a 10-week period. A priority for future studies should be to evaluate the effect of simplifying the vaccination schedule. A study evaluating the concomitant administration of RTS,S with viral vectors expressing ME-TRAP, thereby reducing the total number of vaccinations in a more practical schedule for potential deployment, is currently underway. These results are encouraging for further evaluation of malaria vaccine regimens that combine viral vectors with protein subunits and also vaccine regimens that target multiple stages of the malaria parasite life cycle. Supplementary Data Supplementary materials are available at http://jid.oxfordjournals.org. Consisting of data provided by the author to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the author, so questions or comments should be addressed to the author.

In this trial, we present data from a combined vaccine regimen in which subjects received 5 vaccinations over a 10-week period. A priority for future studies should be to evaluate the effect of simplifying the vaccination schedule. A study evaluating the concomitant administration of RTS,S with viral vectors expressing ME-TRAP, thereby reducing the total number of vaccinations in a more practical schedule for potential deployment, is currently underway. These results are encouraging for further evaluation of malaria vaccine regimens that combine viral vectors with protein subunits and also vaccine regimens that target multiple stages of the malaria parasite life cycle. Supplementary Data Supplementary materials are available at http://jid.oxfordjournals.org. Consisting of data provided by the author to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the author, so questions or comments should be addressed to the author. Supplementary Data Notes Acknowledgments. We thank M. Smith, R. Lopez-Ramon, and A. Cook, for clinical assistance; N. Lella and S. Pawluk, for logistical support; J. Furze, for laboratory assistance; the staff at the National Institute of Health Research (NIHR) Wellcome Trust Clinical Research Facilities at the Southampton and Hammersmith Hospital sites; S. Davidson and colleagues at the WRAIR Entomology laboratories, for supply of infected mosquitoes for controlled human malaria infection; the WRAIR Malaria Serology Laboratory, for additional serologic analysis; B. Angus, for local safety oversight; the study volunteers; and Philippe Moris at GlaxoSmithKline (GSK), for CMI testing.

d colleagues at the WRAIR Entomology laboratories, for supply of infected mosquitoes for controlled human malaria infection; the WRAIR Malaria Serology Laboratory, for additional serologic analysis; B. Angus, for local safety oversight; the study volunteers; and Philippe Moris at GlaxoSmithKline (GSK), for CMI testing. Disclaimer. The views expressed are those of the author(s) and not necessarily those of the PATH Malaria Vaccine Initiative, the United Kingdom National Health Service, the United Kingdom NIHR, or the Department of Health. Financial support. This work was supported by the PATH Malaria Vaccine Initiative and by the United Kingdom NIHR, through the NIHR Oxford Biomedical Research Centre, the Southampton NIHR Wellcome Trust Clinical Research Facility, and the Imperial College NIHR Wellcome Trust Clinical Research Facility.

Disclaimer. The views expressed are those of the author(s) and not necessarily those of the PATH Malaria Vaccine Initiative, the United Kingdom National Health Service, the United Kingdom NIHR, or the Department of Health. Financial support. This work was supported by the PATH Malaria Vaccine Initiative and by the United Kingdom NIHR, through the NIHR Oxford Biomedical Research Centre, the Southampton NIHR Wellcome Trust Clinical Research Facility, and the Imperial College NIHR Wellcome Trust Clinical Research Facility. Potential conflicts of interest. A. V. S. H. and S. C. G. are named inventors on patent applications covering malaria vectored vaccines and immunization regimens. D. M., M. L., and R. W. B. are employees of GSK, which is developing vectored vaccines for malaria and other diseases. S. N. F. acts on behalf of the University of Southampton/University Hospital Southampton National Health Service Foundation trust as chief and principal investigator for clinical trials sponsored by vaccine manufacturers, including GSK, but receives no personal payments for the work. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.