Efficacy of RTS,S/AS01_E Only Seen in Baseline Parasitemic and Not Baseline Aparasitemic Plasmodium falciparum-Exposed, Drug-Treated Kenyan Adults

Abstract

Background

RTS,S/AS01 vaccine efficacy (VE) was previously shown as lower in African adults than in malaria-naive US adults, potentially due to concurrent Plasmodium falciparum infections. We investigated whether treatment of infection prior to vaccination would lead to improved VE and immunogenicity.

Methods

A phase 2b study in Kenyan adults evaluated the efficacy of RTS,S/AS01_E in conjunction with antimalarial chemopreventive drugs. Participants, grouped by baseline presence or absence of P. falciparum infections, were randomized to receive RTS,S/AS01_E or rabies vaccine. Four groups received antimalarial drugs prior to immunization and were followed for 6 months to assess P. falciparum infection. We included an additional group not treated with antimalarial drugs for immunological assessment.

Results

VE (RTS,S/AS01_E vs rabies vaccine) was 34.8% (95% confidence interval [CI], 8.9% to 53.4%) and −24.0% (95% CI, −97% to 22.4%) in baseline P. falciparum-positive and P. falciparum-negative participants, respectively. In RTS,S/AS01_E recipients, there were no statistical differences in anticircumsporozoite (anti-CS) antibody titers in baseline P. falciparum-positive or P. falciparum-negative participants, or in susceptibility to infection during the postvaccination follow-up period. Drug treatment did not improve anti-CS antibody titers.

Conclusions

Treating P. falciparum infections during vaccination does not result in increased VE. Anti-CS antibody responses to vaccination do not differ with baseline P. falciparum infection status, drug treatment, or susceptibility to P. falciparum infections.

Clinical Trials Registration

NCT04661579; PACTR202006896481432.

A regimen of malaria vaccine which showed 87% efficacy in malaria naïve US adults induced only an efficacy of 35% in baseline parasitemic, and no efficacy in baseline aparasitemic Kenyan adults living in high malaria transmission settings.

To inform whether circumsporozoite (CS)-based Plasmodium falciparum vaccines being used to protect against clinical malaria disease in young children living in sub-Saharan Africa [1, 2] could have a role in parasite elimination, it is critical to evaluate whether they protect against infection in all age groups that serve as reservoirs of parasites contributing to ongoing transmission.

Proof-of-concept evidence for RTS,S was initially derived in malaria-naive adults in the United States using controlled human malaria infection (CHMI) challenge models where vaccine efficacy (VE) of 52%–63% was achieved with the RTS,S/AS01_B formulation in moderate-sized trials [3, 4]. However, when RTS,S was tested in Kenyan adults living in high malaria transmission regions [5], VE after 9 weeks of active and passive case detection against smear-positive P. falciparum was only 29.5% (95% confidence interval [CI], −15.4% to 56.9%; P = .164).

It is recognized that immunologic hyporesponsiveness, T-cell exhaustion, and B-memory-cell dysfunction, as measured by decreased immune response to pathogens, is influenced by concurrent asymptomatic submicroscopic P. falciparum parasitemia [6–12]. This immunologic hyporesponsiveness may impede the development of protective immune responses following immunization. Higher VE achieved with adjustments in vaccination schedule and dosage (87% in fractional-dose regimen compared to 63% for the standard-dose regimen) in more recent CHMI studies [4, 13, 14] provided an added incentive to test the delayed fractional-dose regimen in African adults. This phase 2b trial tested the efficacy and immunogenicity of the fractional-dose regimen of RTS,S in adults with or without baseline malaria infections and treated with antimalarial chemotherapy.

METHODS

The protocol for this phase 2b, randomized, open-label, controlled, single-center study was approved by the Kenya Medical Research Institute (KEMRI) Scientific and Ethics Review Unit and the Walter Reed Army Institute of Research (WRAIR) Institutional Review Board. The trial was conducted in accordance with International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines and Good Clinical Practice. Written informed consent was obtained from each subject before study initiation.

Study Design

Participants with or without polymerase chain reaction (PCR)-positive P. falciparum parasitemia at baseline were randomized into 5 groups to receive specified or no antimalarial treatment/prophylaxis and RTS,S/AS01_E or a comparator rabies vaccine (Figure 1, Figure 2, and Supplementary Tables 1, 2, and 3). The objective of the study was to assess VE in participants who were either positive (groups 1 and 4) or negative (groups 2 and 5) for P. falciparum by PCR at baseline. Antimalarials were administered prior to each vaccination, and VE was assessed by time to first P. falciparum infection as measured by PCR. Participants in group 3, included to serve as immunologic controls, were positive for P. falciparum by PCR at baseline and received RTS,S/AS01_E but not antimalarial drugs.

Figure 1.

Study design. Abbreviations: ADI, active detection of infection; DHA, dihydroartemisinin; Pf, Plasmodium falciparum.

Figure 2.

Consort diagram. Abbreviation: ATP, according to protocol.

Participants

The trial was conducted at the KEMRI-Walter Reed Project's Kombewa Clinical Research Center in healthy adults recruited from the village of Seme and Kisumu West subcounties of Kisumu County, Kenya. Eligible participants included men and nonpregnant women aged 18–55 years with no serious acute or chronic illness as determined by medical history, physical examination, record review, or laboratory screening tests. Participants with below stage 3 human immunodeficiency virus (HIV), who were otherwise healthy and on antiretroviral therapy, were included. Participants were stratified by HIV and P. falciparum infection status and randomized to 1 of 5 study groups (Figure 2 and Supplementary Table 1).

Study Treatments

Participants in groups 1, 2, 4, and 5 received antimalarial treatments prior to each of the 3 vaccinations (Figure 1 and Supplementary Table 2). Three daily doses of dihydroartemisinin (40 mg) and piperaquine tetraphosphate (320 mg; half-life 23–30 days) were administered 26 days prior to the first 2 vaccinations. Three daily doses of artemether (20 mg) and lumefantrine (120 mg; half-life 3–4 days) were administered 5 days prior to the third vaccination. A single low-dose primaquine phosphate (15 mg) was also given with the first dose of either dihydroartemisinin/piperaquine tetraphosphate or artemether/lumefantrine to clear circulating mature, sexual-stage gametocytes that may interfere with the PCR assay.

Study Vaccines and Vaccination

RTS,S/AS01_E is manufactured by GlaxoSmithKline (Rixensart, Belgium). The pediatric formulation for vaccine doses 1 and 2 contained 25 μg of RTS,S and adjuvant AS01_E (25 μg of MPL, 25 μg of QS21, and liposomes) in 0.5 mL. At dose 3, participants were administered a fractionated dose of RTS,S/AS01_E (0.1 mL; one-fifth of the antigen and adjuvant dose). Participants randomized to the control comparator group received 3 equal doses of rabies vaccine (Abhayrab, Human Biologicals Institute, Andra Pradesh, India). All vaccines were administered intramuscularly in the deltoid muscle of the arm on a 0-, 1-, and 7-month schedule.

Safety Assessment

Unsolicited adverse events (AEs) occurring within 28 days following administration of each dose of vaccine and, where applicable, courses of per protocol scheduled antimalarial treatment were recorded from the total vaccination cohort (TVC) comprising any enrolled subject receiving at least 1 dose of vaccine. Serious adverse events (SAEs) and pregnancies in participants were recorded until study end. Additionally, solicited local and systemic AEs were collected from a reactogenicity cohort comprising of the first 50 participants in groups 1 and 2, and for all participants in group 3.

Efficacy Assessment

One week prior to dose 3, all participants in groups 1, 2, 4, and 5 were presumptively treated with artemether/lumefantrine prior to the start of active case detection. Participants in group 3 did not receive antimalarial treatment, were not included in the active detection of infection (ADI) phase of the study, and were followed for immunogenicity only. Efficacy evaluation included both ADI with blood draws every 3 weeks and passive case detection in all participants presenting with symptoms consistent with malaria. The primary analysis, defined as the time to first P. falciparum infection, detected by a P. falciparum/pan-Plasmodium 18S rRNA PCR [15], was conducted on the according-to-protocol (ATP) cohort for efficacy, participants receiving all 3 vaccinations with no major protocol deviations that were determined to potentially interfere with the efficacy assessment of the study vaccine. Based on previous experience with assessing gametocyte carriage [16], a cycle threshold of 31, equivalent to <1.28 parasites/µL was used as the cutoff for a positive result. A cross-sectional prevalence assessment of P. falciparum parasitemia by PCR at study end (approximately 500 days after study start) was also conducted.

Immunogenicity Assessments

The ATP cohort for immunogenicity included all participants included in the TVC with no major protocol deviations that were determined to potentially interfere with the immunogenicity assessment of the study vaccine. Antibody levels and avidity against the central repeat region (NANP) and the carboxy-terminal domain of the CS protein region were measured using validated standard enzyme-linked immunosorbent assays (ELISAs) performed at the ImmunCore/Malaria Serology Laboratory at WRAIR [13]. Antibody concentrations against the hepatitis B surface antigen (HBs) were performed at the Human Immunology Laboratory at the International AIDS Vaccine Initiative using an in-house validated assay according to ICH guideline Q2(R1) [17], based on the Monolisa Anti-HBs PLUS kit (Bio-Rad). Antirabies immunoglobulin G (IgG) titers were performed at Kansas State Veterinary Diagnostic Laboratory using a validated ELISA based on the Platelia Rabies Kit II (Bio-Rad).

Statistical Analyses

The statistical analyses were conducted following the principles specified in ICH Topic E9 [18] in accordance with the predefined analysis plan. The analyses were performed after database lock using SAS 9.4 (SAS Institute, Inc).

The analysis of efficacy end points for the primary and secondary objectives and end points were defined by time to a positive PCR for the presence of P. falciparum parasites meeting the definition of an event, from a blood sample collected at any time during the ADI phase of the trial in the ATP cohort for efficacy. The impact of the event-driven study design was to continue to follow up until 92 first-infection events were observed in aggregate between groups 1 and 4, and 72 first-infection events were observed in the aggregate of groups 2 and 5, up to a maximum of 12 months of ADI. The primary and secondary end points were assessed using Cox proportional hazards regression (time to first PCR-positive infection) with a covariate for group assignment (groups 1 and 4 for primary end point, and groups 2 and 5 for secondary end point).

The primary immunogenicity analysis was based on the ATP cohort for immunogenicity. A secondary analysis, based on the TVC, was performed to complement the ATP analysis.

For the safety analysis, data from all participants in the safety analysis population were included. All analyses were descriptive and performed on the intent-to-treat set that included all participants who received ≥ 1 dose of a study vaccine.

RESULTS

Demographic Characteristics

From November 2020 to May 2021, 1191 volunteers (> 98% Luo ethnicity) were assessed for eligibility to participate, and 620 participants were randomized into the study groups in accordance with their baseline P. falciparum status (Figure 2 and Supplementary Tables 1 and 2). Overall, 605 participants were administered study vaccine. Reasons for participants not receiving study vaccine included pregnancy, noncompliance with study procedures, withdrawal of consent, alcohol abuse, and personal reasons. Of the 605 participants who received study vaccine, 582 (93.9%) participants received all scheduled vaccinations, and 539 (89.1%) participants completed the study. The mean age was 32.91 (SD 8.19) years; 323 (53.4%) were female and 282 (46.6%) were male; 103 (17%) participants were positive for HIV infection; 507 (83.8%) participants owned a bed net and slept under it daily; and 546 (93.5%) participants were fully compliant with antimalarial treatment (Supplementary Table 3).

Vaccine Efficacy

The ADI phase commenced 14 days after the third dose of vaccine was administered on day 197 in participants from groups 1, 2, 4, and 5. All individuals in the 4 groups had negative P. falciparum PCR test results on the days they were vaccinated.

Participants in groups 2 and 5 (negative baseline parasitemia) were recruited more quickly than those in groups 1 and 4 (positive baseline parasitemia), so the period at risk for new P. falciparum infection during the ADI phase of the study largely overlapped but was not identical (Supplementary Figure 1). Also, in the rabies vaccine groups (representative of the baseline malaria attack rate in the region), the attack rate during the ADI phase of the study was 1.48 events/person-years at risk (PYAR) in group 4 and 0.62 events/PYAR in group 5 (Table 1).

Table 1.

Vaccine Efficacy, First Infection Events of Plasmodium falciparum During the Active Detection of Infection Period by HIV Status and Overall for Groups 1 and 4 and for Groups 2 and 5 (Total Cohort for Efficacy)

	First or Only Infection								Vaccine Efficacy
	n	Number of events	PYAR	Rate	n	Number of events	PYAR	Rate	Efficacy, %	LL, %	UL, %	P Value
	Group 1 (RTS,S)				Group 4 (rabies)
Overall unadjusted	152	60	59.76	1	152	80	54.13	1.48	34.5	8.3	53.1	.014
Adjusted by HIV									35.5	9.7	53.9	.011
	Group 2 (RTS,S)				Group 5 (rabies)
Overall unadjusted	118	38	49.63	0.77	123	34	54.88	0.62	−24	−97	24.4	.373
Adjusted by HIV									−24	−97	22.2	.369

Events is the number of participants having at least 1 confirmed P. falciparum infection detected by PCR in each group. PYAR is the sum of follow-up period (censored at the first occurrence of a confirmed P. falciparum infection) expressed in years. Rate expressed as events per PYAR at risk. Efficacy is the vaccine efficacy adjusted by sex (Cox regression model, 1 − adjusted hazard ratio). P values for vaccine efficacy adjusted by HIV.

Abbreviations: LL, 95% lower confidence limit; n, total number of participants at risk in each group; PYAR, person years at risk; UL, 95% upper confidence limit.

VE Against Incident P. falciparum Infection (RTS,S/AS01_E vs Rabies Vaccine) in Baseline P. falciparum PCR Positives

The primary objective was to determine VE in those who were P. falciparum PCR positive at baseline. Overall, in the ATP cohort during the 6-month period following the last dose of the vaccine, 60 PCR-detectable infection events were recorded for group 1, and 80 events were recorded for group 4 during ADI. The PYAR was 59.53 for group 1 and 53.67 for group 4, and the incidence rate was lower for group 1 (1.01 events/PYAR) compared to group 4 (1.49 events/PYAR). When compared to the rabies vaccine, the overall estimated VE of RTS,S/AS01_E was 34.8% (95% CI, 8.9%–53.4%; P = .012; Table 1). Similar results were observed for the total cohort for efficacy (results not shown). When adjusted by HIV status, the estimated VE was 35.9% (95% CI, 10.3%–54.2%; P = .009). HIV status had no statistically significant impact on VE (P = .247; Supplementary Figure 2). No other potential confounding covariates at baseline that reached the cutoff of P < .1 were included in the adjusted model. The time to PCR-positive parasitemia is shown in a Kaplan-Meier survival plot (Figure 3A) with evidence for a difference in event-free survival time between groups 1 and 4.

Figure 3.

Vaccine efficacy shown as Kaplan-Meier survival plots comparing the event-free survival between (A) groups 1 and 4 (primary end point), and (B) groups 2 and 5 (secondary end point). Lines represent proportion of subjects experiencing an event over days and the shaded areas represent the 95% confidence intervals. Abbreviation: Pf, Plasmodium falciparum.

VE Against Incident P. falciparum Infection (RTS,S/AS01_E vs Rabies Vaccine) in Baseline P. falciparum PCR Negatives

The secondary objective was to determine VE in those who were P. falciparum PCR negative at baseline. Overall, in the ATP cohort, 38 events were recorded for group 2, and 34 events were recorded for group 5. The PYAR was 49.63 years for group 2 and 54.88 years for group 5. The incidence rates for group 2 and 5 were similar (0.77 and 0.62 events/PYAR, respectively). The unadjusted estimated VE was not significant, −24% (95% CI, −97% to 22.4%; P = .373; Table 1) and was similar when adjusted by HIV status. Similar results were observed for the total cohort for efficacy (results not shown). When adjusted by HIV status, the estimated VE was −24% (95% CI, −97% to 22.2%; P = .369; Supplementary Figure 2). The time to PCR-positive parasitemia is shown in a Kaplan-Meier survival plot (Figure 3B) with no evidence of a difference in survival times between groups 2 and 5. P. falciparum PCR performed on a cross-sectional sampling taken at study close-out visits showed 26.5% overall P. falciparum prevalence with no statistically significant differences between groups.

Immunogenicity

Anti-CS NANP repeat and C-terminus geometric mean IgG titers, and antibody avidity were measured in all participants immunized with RTS,/AS01_E (groups 1, 2, and 3). Prior to RTS,S/AS01_E vaccination, groups 1, 2, and 3 had approximately 78%, 67%, and 74% of participants, respectively, with anti-CS NANP repeat antibody responses and 89%, 78%, and 94% of participants, respectively, with anti-CS C-terminus antibody responses above the lower limit of quantification.

RTS,S/AS01_E was immunogenic in all groups based on geometric mean titer (GMT) to both the NANP repeat and the C-terminus domain measured at baseline and after each immunization. There was no significant effect on anti-CS antibody titers to either the NANP repeat or the C-terminus domain postimmunization in individuals who were baseline positive (groups 1 and 3) or negative (group 2) for P. falciparum infection (Figure 3) or between male and female participants. The anti-NANP antibody titers decreased over time but increased after the fractional dose at month 7, but not to the levels comparable to those seen after the second dose. Despite parasite clearance (group 1) or chemoprevention (group 2) with antimalarial medications, the antibody kinetics were comparable among all groups (Figure 4). There was no difference in anti-NANP (Supplementary Table 5) or anti–C-terminus (Supplementary Table 6) avidity indices across groups 1, 2, and 3.

Figure 4.

Anticircumsporozoite NANP and C-terminus antibody kinetics. A, Anti-NANP repeat titers from groups 1 and 2 who received RTS,S and antimalarial medications compared to (B) titers from group 3 who received RTS,S but no antimalarial medications. C, Anti–C-terminus titers from groups 1 and 2 who received RTS,S and antimalarial medications compared to (D) titers from group 3 who received RTS,S but no antimalarial medications. The dots represent the geometric mean titers at each sampling point and the error bars represent the 95% confidence intervals around the titers. Abbreviations: ELISA unit, enzyme-linked immunosorbent assay unit (equivalent to geometric mean titer); neg, negative; Pf, Plasmodium falciparum; pos, positive; Rx, medication.

After dose 3, anti-NANP and anti–C-terminus antibody titers in participants immunized with RTS,S/AS01_E showed no differences between adults with or without P. falciparum parasites in the blood at baseline (Figure 5). Nor were there differences in anti-CS antibody responses postimmunization in HIV-positive or HIV-negative participants in either group 1 or 2 (Supplementary Figure 3).

Figure 5.

Anti-CS NANP and C-terminus titers in groups 1, 2, and 3 after dose 3 (day 225). Anti-NANP titers are shown as Box plots (A) and reverse cumulative distribution curve (B). Anti–C-terminus titers are shown as Box and whiskers plots (C) and reverse cumulative distribution curve (D). Box and whiskers plots show individual titers as dots. The box brackets the 1st and 3rd quartiles (25th to 75th percentile) with the median (50th percentile) in the middle. The vertical lines are known as whiskers, they are 1.5x the interquartile range (IQR), which is the difference between the 3rd and 1st quartile. The whiskers will end either at a point 1.5x IQR below or above a quartile or at the last observation, if it is less than 1.5 IQR away. The endpoints are denoted by horizontal lines. Abbreviations: CS, circumsporozoite; EU, enzyme-linked immunosorbent assay unit (equivalent to geometric mean titer); neg, negative; Pf; Plasmodium falciparum; pos, positive; Rx, medication.

Ancillary analyses were performed to evaluate the association of anti-CS antibodies with VE. Anti-CS GMTs measured 1 month after dose 3 in groups 1 and 2 were analyzed with respect to whether they experienced a P. falciparum PCR-positive event during the 6-month ADI period. Both anti-CS NANP and C-terminus antibody GMTs were equivalent in noninfected compared to infected participants (Figure 6).

Figure 6.

Box and whiskers plots showing anti-CS NANP and C-terminus titers after dose 3 (day 225), stratified by Plasmodium falciparum infection outcome during the active detection of infection phase: anti-NANP titers within group 1, P. falciparum PCR positive at baseline (A) and group 2, P. falciparum PCR negative at baseline (B); anti–C-terminus titers within group 1 (C) and group 2 (D). Box and whiskers plots show individual titers as dots. The box brackets the 1st and 3rd quartiles (25th to 75th percentile) with the median (50th percentile) in the middle. The vertical lines are known as whiskers, they are 1.5x the interquartile range (IQR), which is the difference between the 3rd and 1st quartile. The whiskers will end either at a point 1.5x IQR below or above a quartile or at the last observation, if it is less than 1.5 IQR away. The endpoints are denoted by horizontal lines. Abbreviations: CS, circumsporozoite; EU, enzyme-linked immunosorbent assay unit (equivalent to geometric mean titer); neg, negative; Pf, P. falciparum.

Antibody titers against HBs, a component of the RTS,S antigen, were similar in baseline parasitemic versus nonparasitemic individuals (group 1 vs 2) and in individuals who received antimalarial drug treatment or not (group 1 vs 3) (Supplementary Figure 4). Consistent with prior studies, a marked seroresponse increase from baseline to protective thresholds of anti-HBs titer was achieved in all RTS,S-vaccinated groups. There was no significant difference in antirabies IgG titers in participants based on baseline parasitemia status (group 4 vs group 5) (Supplementary Figure 4).

Safety

The RTS,S/AS01_E regimen was well tolerated. The number of participants per group and the number of local and systemic elicited AEs, the number of unsolicited AEs, and the number of SAEs are shown in Supplementary Table 4.

The proportion of participants experiencing local and systemic solicited AEs was similar across groups, and most participants had AEs of mild severity. The most frequently reported local solicited AE was pain at the site of injection. The most frequently reported systemic solicited AE was headache. Unsolicited AEs were mild in severity in most participants, and all AEs were unrelated to study treatment.

Overall, 13 SAEs were reported in 10 participants (3 participants in group 1; 2 participants in group 2; 1 participant in group 3; 1 participant in group 4; 3 participants in group 5). One SAE resulted in death (alcoholic coma), and 1 led to discontinuation. No SAE was considered related to study treatment.

Seven participants became pregnant during the immunization phase of the study (5 received at least 1 dose of RTS,S/AS01_E and 2 received at least 1 dose of rabies vaccine) and were withdrawn from any further immunizations but were followed for safety until delivery or termination of the pregnancy. There were no pregnancies in women who completed all vaccinations. There was 1 participant in group 3 whose pregnancy ended in spontaneous abortion, determined as not related to vaccination.

DISCUSSION

The introduction of 2 malaria vaccines, RTS,S/AS01_E and R21/Matrix-M, to prevent clinical malaria disease in young children in sub-Saharan Africa has been a welcomed supplemental tool. Studies in African adults are necessary to assess how these and future vaccines may be harnessed to accelerate P. falciparum elimination in populations with different age structures, different histories of exposure and acquired immunity, and under settings with different malaria transmission intensities.

This study used the pediatric formulation of RTS,S/AS01_E, containing half the active ingredients of the adult formulation RTS,S/AS01_B, as a prototype for CS-based vaccines, which was administered using a delayed fractional third-dose regimen. We chose to use the pediatric formulation as a result of similar efficacy demonstrated in CHMI studies [13] and the desire to streamline vaccine delivery in the event efficacy is achieved across age groups [5].

This study was designed to support eventual use-case scenarios which combine a vaccine that prevents infection with antimalarial medications that clear or prevent parasitemia. There are several lines of evidence suggesting the benefit of combining drug treatment with vaccination. Studies combining RTS,S/AS01_E with seasonal malaria chemoprevention in young African children resulted in a synergistic effect in averting clinical malaria episodes [19, 20]. There is evidence that acute clinical malaria and/or asymptomatic malarial parasitemia may suppress the induction of protective antibody-mediated immune responses to vaccines [6–12]. Data from a recent study with an irradiated sporozoite vaccine suggested that VE was only achieved when drug treatment was implemented at study start [21]. However, our hypothesis that antimalarial chemotherapy could enhance the overall effectiveness of RTS,S/AS01-vaccinated Kenyan adults was not supported by the trial data. Antimalarial drug treatment did not substantively improve VE to the levels achieved in CHMI studies in P. falciparum-naive adults [3, 4, 13], nor in comparison to the earlier study performed in the same Kenyan site in which no drug treatment was given during the vaccination phase [5]. The most surprising result in our study was the lack of VE in the baseline P. falciparum-negative participants, although the individuals mounted the same level of anti-CS antibody titers postvaccination as compared to those in the P. falciparum-positive participants, indicating that this is not due to vaccine nonresponsiveness. This finding aligns with a pediatric study in which VE against molecularly detected infections was higher in children who were P. falciparum PCR-positive at first vaccination versus those who were negative [22]. In our study, any comparison of differences between the baseline P. falciparum-positive participants with the baseline P. falciparum-negative participants must take into consideration that (1) baseline P. falciparum status represents a single point in time, (2) baseline P. falciparum status is not randomly assigned, and (3) there was a temporal delay in the recruitment of baseline P. falciparum-positive compared to P. falciparum-negative enrollees (Supplementary Figure 1). One possible explanation is that baseline P. falciparum-negative participants already have a substantive natural immunity against P. falciparum infection and VE was masked by the stronger natural immunity. We are currently using trial samples to interrogate disparate baseline acquired and innate immunity between baseline positive and negative individuals. Another explanation is disparate exposure to P. falciparum-infected mosquitoes; we were unable to show marked differences in geospatial analysis of domicile between these groups 1 and 4 compared to groups 2 and 5 (Supplementary Figure 5).

Antibody responses to CS protein and HBs were similar with or without drug treatment, with or without baseline P. falciparum infection, and, most surprisingly, whether a P. falciparum PCR-positive event occurred during the ADI period. This is contrary to results from CHMI studies in malaria-naive adults, which showed that anti-CS antibodies are associated with protection [3, 4, 13, 23]. This suggests that, in the context of natural exposure with preexisting immunity, the correlates of protection are more complicated than in CHMI studies.

This study has several limitations in identifying approaches that could be applied to overcome immune hyporesponsiveness and increase vaccine-induced protection against infection. Due to budget constraints, the end point measurements were assessed only in groups that included the provision of antimalarial medications during vaccine doses. Additionally, the dosage of RTS,S/AS01 used in this study was derived from studies in malaria-naive adults, and it may be necessary to use a higher dose in malaria-experienced adults, akin to the recommendation of high-dose influenza vaccines recommended for the elderly.

In conclusion, a vaccine comprising only the CS protein is unlikely to be sufficient to accelerate P. falciparum malaria elimination due to modest efficacy in preventing infection, including in adults. This is likely due to the need for antibodies to neutralize the parasite in the short period of time prior to it entering the liver-stage development phase. Therefore, multiantigen/multistage vaccine approaches that include a CS component, as well as 1 or more blood-stage and/or sexual-stage (transmission-blocking) antigens, have risen to the top of future vaccine strategies.

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.

Notes

Acknowledgements. We are grateful to the study volunteers for their participation in this clinical trial. We thank GlaxoSmithKline (GSK) Biologicals SA for RTS,S/AS01_E. We also acknowledge the following members of the study team, without whose hard work and dedication during a pandemic this study would not have been possible: KEMRI (Walter Otieno, Eric Rono, Janet Oyieko, Valentine Sing'oei, Eliasa Bett, Dorothy Okello, Rachel Aguttu, Jacob Jagongo, Kennedy Otieno, Raymond Miyere, Roseline Apamo, Roseline Ohore, Alex Arika, Linah Ooro, Ludia Mbathi, Beatrice Orando, and the field team), PATH (Krystal Graham, Margaret Toher, Lionel Martellet, Linda Hoang, Trevor Lutzenhiser, Heather Richards, Richard Okwanyo, Shannon Shanahan, Rebecca Sanders, Kelsey Mertes, Allison Clifford, Kit Dubel, Lalaine Anova, Kerry Laurino), FHI-Clinical (Daniel Joffe, Michelle Pentikainen, Neo Choabi, Victorine Owira, Chantel Friend, Nicola van Zyl, Greg Tippett, Thando Madonsela, Monique Goldblatt, Loga Ganesh), DF/Net (Naydene Slabbert, Megan Baer, Gavin Robertson, Brian Postle, Joe Jiang) and GSK team. We also thank Arianna Marini, Marija Zaric, Heejin Yun, Rachel Bromell, Morolayo Ayorinde, and Claire Streatfield at the Human Immunology Laboratory at the International AIDS Vaccine Initiative for the validation of the HBs antibody test and for generating the anti-HBs data for this study, and Kansas State Veterinary Diagnostic Laboratory for performance of the rabies antibody assay.

Disclaimer. Material has been reviewed by the Walter Reed Army Institute of Research. There is no objection to its presentation and/or publication. The opinions or assertions contained herein are the private views of the author, and are not to be construed as official, or as reflecting true views of the Department of the Army or the Department of Defense. The investigators have adhered to the policies for protection of human subjects as prescribed in AR 70–25.

Financial support. This work was supported by 3 grants to PATH from the Bill and Melinda Gates Foundation (grant numbers INV-007217 for clinical trial and supportive work, and INV-008924 and INV-042281 for time spent working on this study by PATH staff [authors, S. C., K. I., S. G., M. R., C. G., L. D. M., J. J. A., C. F. O., and C. K. L.; others in acknowledgments, K. G., M. T., L. H., T. L., H. R., R. O., S. S., K. B., K. L., and L. A.]).