Recombinant Full-length Plasmodium falciparum Circumsporozoite Protein–Based Vaccine Adjuvanted With Glucopyranosyl Lipid A–Liposome Quillaja saponaria 21: Results of Phase 1 Testing With Malaria Challenge

Abstract

Background

Malaria is preventable yet causes >600 000 deaths annually. RTS,S, the first marketed malaria vaccine, has modest efficacy, but improvements are needed for eradication.

Methods

We conducted an open-label, dose escalation phase 1 study of a full-length recombinant circumsporozoite protein vaccine (rCSP) administered with adjuvant glucopyranosyl lipid A–liposome Quillaja saponaria 21 formulation (GLA-LSQ) on days 1, 29, and 85 or 1 and 490 to healthy, malaria-naive adults. The primary end points were safety and reactogenicity. The secondary end points were antibody responses and Plasmodium falciparum parasitemia after homologous controlled human malaria infection.

Results

Participants were enrolled into 4 groups receiving rCSP/GLA-LSQ: 10 µg × 3 (n = 20), 30 µg × 3 (n = 10), 60 µg × 3 (n = 10), or 60 µg × 2 (n = 9); 10 participants received 30 µg rCSP alone × 3, and there were 6 infectivity controls. Participants experienced no serious adverse events. Rates of solicited and unsolicited adverse events were similar among groups. All 26 participants who underwent controlled human malaria infection 28 days after final vaccinations developed malaria. Increasing vaccine doses induced higher immunoglobulin G titers but did not achieve previously established RTS,S benchmarks.

Conclusions

rCSP/GLA-LSQ had favorable safety results. However, tested regimens did not induce protective immunity. Further investigation could assess whether adjuvant or schedule adjustments improve efficacy.

Clinical Trials Registration

NCT03589794

Malaria remains a major cause of preventable disease and death worldwide, causing 247 million cases and 619 000 deaths in 2021 alone [1]. While malaria deaths decreased by half between 2000 and 2015, progress stagnated in 2015 and reversed in 2019 [1]. Insecticide-resistant mosquitoes, antimalarial drug resistance, and parasite evasion of diagnosis by rapid diagnostic testing all threaten to further reverse these gains and reinforce the need for a highly effective malaria vaccine as an additional tool for malaria prevention and eradication.

The circumsporozoite protein (CSP) coats the sporozoite surface and is a major focus of malaria vaccine efforts. The first World Health Organization–recommended malaria vaccine, RTS,S targets the last 18 NANP repeats in the central repeat region and the carboxyl terminus, which includes multiple T-cell epitopes in the Th2R and Th3R regions [2]. Over 4 years, RTS,S demonstrated 36% efficacy against clinical malaria in infants 5–17 months of age in sub-Saharan Africa [3]. Another vaccine, R21, targets the same CSP regions as RTS,S and is given with a different adjuvant [4]. In early testing, seasonal administration of R21 with a booster 1 year after the primary series had an efficacy of 63%–77%, depending on the adjuvant dose [5]. While these results are promising, the long-term performance of and booster frequency needed for R21 remain unclear. With 3.3 billion people at risk of malaria worldwide [1], yearly boosters pose significant logistical and supply chain challenges, so the search for additional and longer-lasting vaccine options continues.

Recently isolated and optimized monoclonal antibodies targeting the CSP junctional region, located between the amino terminus and central repeat region and not included in the RTS,S vaccine [6, 7], provided 82%–100% efficacy against controlled human malaria infection (CHMI) and 75%–88% efficacy against Plasmodium falciparum infection over a 6-month malaria season in Mali [8–11]. Given the cost and resource challenges of monoclonal antibody administration, active immunization with a highly effective vaccine that protects for ≥2 years would be more efficient and cost-effective. Therefore, targeting the junctional region with a full-length recombinant CSP (rCSP) vaccine could improve on the protection provided by RTS,S.

We initiated a phase 1 first-in-human clinical trial of a full-length rCSP expressed in a Pseudomonas fluorescens platform [12] in malaria-naive adults in Baltimore, Maryland. The vaccine was coadministered with glucopyranosyl lipid A–liposome Quillaja saponaria 21 formulation (GLA-LSQ), which promotes antigen-specific Type 1 T helper cell (T_H1) responses and cytotoxic T-cell production [13, 14]. GLA is a synthetic lipid A derivative less heterogeneous than the naturally-derived monophosphoryl lipid (MPL) A contained in AS01_E, the RTS,S vaccine adjuvant [13]. Both AS01_E and GLA-LSQ contain liposomal QS-21 [15]. The first trial segment tested safety and immunogenicity in 3 groups: 10-µg rCSP/GLA-LSQ, 30-µg rCSP/GLA-LSQ, and 30-µg rCSP alone [16]. The vaccine had a favorable safety and tolerability profile. All vaccine doses stimulated substantial increases in anti-rCSP immunoglobulin (Ig) G titers. Here, we report final study results of safety, immunogenicity, and efficacy against CHMI.

METHODS

Trial Design and Participants

We conducted an open-label, dose escalation phase 1 study to determine safety, immunogenicity, and efficacy against CHMI of rCSP administered with GLA-LSQ at the University of Maryland School of Medicine Center for Vaccine Development and Global Health (CVD) in Baltimore. Safety and immunogenicity testing for the first 3 groups of 10 volunteers each have been published [16]. For remaining groups, additional inclusion and exclusion criteria applied, given the CHMI procedures. The study enrolled healthy adults aged 18–45 years who agreed to use effective contraception, avoid travel to a malaria-endemic region, and comply with study procedures. Additional exclusion criteria included history of malaria, previous residence in or recent travel to a malaria-endemic region, hemoglobinopathy, history of anaphylaxis, splenectomy, recent immunosuppression or antimalarial drug use, abnormal screening electrocardiogram, or increased cardiovascular risk, as determined using the method described by Gaziano et al [17], among others. The clinical protocol can be accessed in the Supplementary Materials.

Ethical Considerations

All volunteers provided written informed consent. The University of Maryland, Baltimore Institutional Review Board approved the study protocol. The Division of Microbiology and Infectious Diseases, National Institute of Allergy and Infectious Diseases (NIAID), appointed a safety monitoring committee and independent safety monitor for safety oversight. The study was conducted in accordance with International Conference on Harmonization of Good Clinical Practices and the Declaration of Helsinki. CVD investigators conducted the clinical trial and safety assessments. CVD laboratory staff performed immunogenicity assays, and laboratory staff at Walter Reed Army Institute of Research conducted follow-up comparator assays. The Emmes Company provided database design and performed all statistical analyses. The NIAID registered the trial at ClinicalTrials.gov under NCT03589794.

Blinding

Although the study was open label, laboratory personnel conducting immunogenicity and ultrasensitive real-time polymerase chain reaction (usPCR) testing were blinded to study group assignment. Technicians analyzing immunogenicity samples were also blinded to subject identification and visit number.

Procedures

The rCSP vaccine antigen (Ajinomoto Althea) is a full-length P. falciparum rCSP based on the 3D7 clone sequence expressed in a soluble form in the P. fluorescens cell line. GLA-LSQ (Infectious Disease Research Institute), also known as AP 10–602, is a nanoliposomal formulation containing the immunological adjuvants GLA, a synthetic monophosphoryl lipid A–like molecule, and LSQ, a liposomal formulation of QS-21, a saponin. Additional details of vaccine formulation and reconstitution are in the Supplementary Materials.

Six groups were vaccinated: a low-dose, standard-interval group (safety only) (LLL[S]/A; n = 10) and a low-dose, standard-interval group (safety + efficacy) (LLL[SE]/A; n = 10) both received 10 µg rCSP/GLA-LSQ; a medium-dose, standard-interval group received 30 µg rCSP/GLA-LSQ (MMM/A; n = 10); a medium-dose, standard-interval, no-adjuvant group received 30 µg rCSP alone (MMM; n = 10); and a high-dose, standard-interval group (HHH/A; n = 10) and a high-2-dose, delayed-interval (HHd/A) group (n = 9) both received 60 µg of rCSP/GLA-LSQ. GLA-LSQ contained 5 µg of GLA and 2 µg of LSQ. Another group of 6 participants did not receive any vaccine product and served as infectivity controls to verify that challenge mosquitoes could transmit malaria and to calculate an efficacy estimate. Supplementary Table 1 details the study product received and study procedures for each group.

Vaccinated participants received their assigned study product as a single 0.5-mL intramuscular upper deltoid injection. The HHd/A group received vaccine on days 1 and 490; all other groups, on days 1, 29, and 85. Participants were monitored for ≥30 minutes after each vaccination, then recorded vaccine reactogenicity for the next 7 days. Study staff evaluated participants in clinic for adverse events (AEs) 7 and 28 days after each vaccination. In the HHd/A group, clinic follow-up occurred on day 8 and subsequently by phone, owing to coronavirus disease 2019 (COVID-19) pandemic–related research restrictions. When in-person research resumed, the protocol was amended, and willing participants in the HHd/A group were rescreened around day 460 and then received second vaccinations on day 490. Further details regarding follow-up are in the Supplementary Materials.

At 28 days (±3 days) after final vaccinations, which all occurred in the same week, participants in the HHd/A, HHH/A, and LLL(SE)/A groups underwent homologous CHMI using 5 Anopheles stephensi mosquitoes (Sanaria) infected with the NF54 P. falciparum strain over 2 days (10 vaccinees and 3 controls each day). NF54 is the parent strain of the 3D7 clone and is considered a homologous challenge strain for 3D7-based vaccines. Further CHMI details are in the protocol in the Supplementary Materials. Beginning 5 days after CHMI, participants were evaluated in clinic daily for malaria symptoms and for blood specimen usPCR testing. If a participant had 2 positive usPCR results, they were treated for 3 days with artemether-lumefantrine or atovaquone-proguanil in case of artemether-lumefantrine intolerance. After completing treatment, participants discontinued daily follow-up and returned for P. falciparum usPCR testing 28 days after CHMI to document clearance.

Participants provided blood samples for safety laboratories and humoral immune response measurements on the day of each vaccine and 7 days later, 28 days after the second and third vaccines (if applicable), and 84 days after the last vaccine. Blood samples for humoral immune response measurements were collected 168 days after third vaccines, if applicable (see the protocol in the Supplementary Materials for a list).

High-throughput enzyme-linked immunosorbent essays (ELISAs) were performed to measure anti-rCSP–specific immunoglobulins IgG, IgM, and IgA in serum (see the Supplementary Materials for details). P. falciparum parasites were detected with usPCR using PCR primers based on NF54 P. falciparum 18S ribosomal RNA [18, 19], with a lower limit of detection of 16 parasites per milliliter [20, 21]. Samples were tested in duplicate with appropriate controls. A standard curve was used to calculate parasite density from the cycle threshold. Data were analyzed using LightCycler 96 Software. Baseline and day of CHMI serum samples were analyzed at WRAIR for comparative ELISA IgG testing with RTS,S benchmarks, using full-length CSP, CSP repeat (NANP) (Biomatik), and CSP C-terminal region (Pf16; Atlantic Peptides) as capture antigens [22, 23]. The avidity index (AI) was calculated as the ratio of antigen-specific antibody titers obtained with ELISA in the presence (sample optical density [OD] [urea]) and absence (sample OD [phosphate-buffered saline]) of a chaotropic reagent (4 mol/L urea).

The primary study end points were solicited local and systemic reactions and related severe laboratory AEs within 7 days after vaccination, related severe unsolicited AEs and SAEs within 28 days after vaccination, and SAEs and AEs of special interest during follow-up. Events were assessed for severity and relatedness and graded per protocol (see the Supplementary Materials). Secondary study end points included anti-rCSP antibody titers and P. falciparum asexual parasitemia after CHMI.

Statistical Analysis

This phase 1 trial primarily evaluated safety and was not powered for statistical comparisons. CHMI was intended to provide an efficacy estimate. Assuming that 6 infectivity controls had a 0.99 chance of infection, if 10 vaccinees had a 0.22 chance of infection (vaccine efficacy [VE] of 78%), the study would have had ≥80% power to detect a difference in proportions infected. All continuous variables were summarized using descriptive statistics, including mean and/or geometric mean (GM), standard deviation, median, maximum, and minimum. In addition, the GM fold rise (GMFR) and the proportion of individuals who seroconverted (≥4-fold increase in antibody titer from baseline) with 95% confidence intervals were calculated for each treatment group and time point.

The intention-to-treat population was used for analysis of VE, estimated as 1 − risk ratio, where the risk ratio was the ratio of proportion of participants in the vaccination group to controls in whom parasitemia developed within 28 days after CHMI. Confidence intervals for VE were calculated from exact unconditional confidence limits for relative risk, based on the Farrington-Manning score statistic. Kaplan-Meier survival curves depicted time to malaria infection for each of 3 vaccine groups compared with the infectivity control group, and comparisons were made using log-rank tests with P values based on permutation methods. All analyses were performed using SAS software (version 9.4). Additional details for statistical analyses are provided in the Supplementary Materials.

RESULTS

Study Population

Between 8 January 2019 and 5 August 2021, 114 participants were screened, and 65 were enrolled into 7 groups: LLL(S)/A (n = 10), MMM/A (n = 10), MMM (n = 10), HHd/A (n = 9), HHH/A (n = 10), LLL(SE)/A (n = 10), and infectivity controls (n = 6) (Figure 1 and Table 1). All enrolled vaccine group participants received ≥1 injection and were included in safety analyses. LLL(S)/A, MMM/A, and MMM groups were vaccinated first to confirm safety and inform dosing for remaining groups. The next group to enroll was intended to be a high-dose group similar to HHH/A; however, owing to the COVID-19 pandemic, all in-person study activities were paused on 23 March 2020 after first vaccinations in 9 participants. During the pause, safety follow-up occurred via telephone and no blood sampling occurred until activities resumed on 15 March 2021. Five participants discontinued during the pause, and the remaining 4 participants were rescreened and continued a revised, high 2-dose delayed interval schedule (HHd/A). We then reenrolled a new HHH/A group and an LLL(SE)/A group. Four participants who had completed all vaccinations did not undergo CHMI because of scheduling (HHd/A group and LLL(SE)/A group), relocation (HHH/A), or illness (HHH/A).

Safety

Solicited adverse reactions were previously reported for the LLL(S)/A, MMM/A, and MMM groups [16]. No serious AEs (SAEs) occurred among the 35 additional participants enrolled. Within 7 days of vaccination, 14 of 29 participants (48%) in the HHd/A, HHH/A, and LLL(SE)/A groups experienced ≥1 systemic solicited AE (Supplementary Table 2), slightly higher than the 6 of 30 (20%) in groups previously reported. No AEs were classified as severe, compared with 2 unrelated severe systemic solicited AEs in previously reported groups (Figure 2 and Supplementary Table 3) [16]. No participants reported fever within 7 days of vaccination. Nineteen of 29 participants (66%) in the HHd/A, HHH/A, and LLL(SE)/A groups experienced ≥1 local solicited AE, versus 13 of 30 (43%) in the groups previously reported. Local injection site pain and tenderness were the most commonly reported local solicited AEs, and all were mild (Figure 2 and Supplementary Table 3). No participant reported quantitative graded injection site induration or swelling. Rates of solicited AEs were similar among all vaccinated groups.

Twenty-five of 35 participants (71%) in the HHd/A, HHH/A, LLL(SE)/A, and infectivity control groups experienced ≥1 unsolicited AE, compared with 18 of 30 (60%) in the groups already reported (Supplementary Table 2). Two of 35 (6%) reported ≥1 severe unsolicited AE, similar to the 2 of 30 (7%) reported in previously described groups, none considered related to study product (Supplementary Table 4). One participant of 29 (3%) in the HHd/A, HHH/A, and LLL(SE)/A groups reported a related unsolicited AE (mild systolic blood pressure increase) within 28 days of vaccination. No other related unsolicited AEs were reported. No participants were discontinued owing to moderate or severe unsolicited AE. Rates of unsolicited AEs were similar for all dosing groups.

Twelve of 29 participants (41%) in the HHd/A, HHH/A, and LLL(SE)/A groups experienced ≥1 postbaseline clinical safety laboratory AE, compared with 6 of 30 (20%) in previously reported groups. None were related or severe. Increased alanine aminotransferase (ALT) was the most common, recorded in 6 of 29 participants (21%) in the HHd/A, HHH/A, and LLL(SE)/A groups. All laboratory abnormalities were considered unrelated to study treatment and were mild, except in 1 participant in the LLL(SE)/A group with moderate ALT increase (Figure 3 and Supplementary Table 5).

Efficacy

Twenty-six participants underwent CHMI: 3 HHd/A, 8 HHH/A, 9 LLL(SE)/A, and 6 infectivity controls. All developed malaria infection, resulting in a VE of zero (Supplementary Table 6). The median time to malaria infection was 8.0 days across all treatment groups (Supplementary Table 7), as depicted in the Kaplan-Meier survival curves (Figure 4), and peak parasite densities did not differ between groups (P = .37 by Kruskal-Wallis test).

Immunogenicity

Anti-rCSP IgG titers were previously reported for the LLL(S)/A, MMM/A, and MMM groups [16], presented again for comparison. Overall, participants had increasing anti-rCSP IgG titers after each vaccination. The GMFR in anti-rCSP IgG titers was highest 28 days after the last vaccination (Figure 5 and Supplementary Table 8), similar to findings from previously reported groups [16]. Postvaccination responses were primarily of IgG class (Supplementary Tables 8–10). Interestingly, CHMI did not boost IgG responses. The HHH/A group had the highest anti-rCSP IgG GMFR at all time points. The highest anti-rCSP IgM GMFR were also seen 28 days after last vaccination in all groups, with the MMM/A group having the highest GMFR for all time points, except for 7 days after dose 1 (Supplementary Table 9). For anti-rCSP IgA, the highest GMFR was measured 7 days after last vaccination for all groups except the LLL(S)/A and the MMM groups, where it was highest 7 days after dose 2 (Supplementary Table 10). At all time points except 7 days after dose 2, the HHH/A group had the highest anti-rCSP IgA GMFR. Time trends of individual anti-rCSP IgG titers are shown in Supplementary Figure 1. Detailed anti-rCSP IgG, IgA, and IgM results are in the Supplementary Materials.

Comparison of Anti-CSP IgG Titers to RTS,S Benchmarks

Serum samples from participants who underwent CHMI (baseline and day of CHMI samples) were tested at WRAIR for comparative analysis against RTS,S benchmark values. For all antigens evaluated (full-length CSP, CSP repeat [NANP], and CSP C-terminal region), GM IgG OD1 titers of rCSP vaccinated groups were notably lower than RTS,S benchmarks (Table 2 and Supplementary Figure 2). In contrast, AI values for the rCSP groups were comparable to RTS,S benchmarks for full-length CSP plate and CSP C-terminal region antigens, although not all participants in each rCSP group had a calculable AI (Table 2 and Supplementary Figure 3). An AI would not be calculable if either the sample OD (phosphate-buffered saline) or the sample OD (urea) fell below the detection limit (titer <50). Detailed results are in the Supplementary Materials.

DISCUSSION

In this phase 1 study of rCSP/GLA-LSQ, all vaccine formulations demonstrated favorable safety and tolerability. All AEs related to vaccination were mild to moderate in intensity. The most common solicited AEs were pain, tenderness, and fatigue. AE rates were similar between vaccinated groups and reflect a safety profile similar to those of commonly used, licensed vaccines. Unfortunately, the vaccine did not demonstrate efficacy, as all participants developed malaria after CHMI.

No reliable immunological correlates of protection exist for malaria vaccines. Results of RTS,S immunogenicity suggest an association between anti-CSP IgG antibody titers and protection, but this relationship is not linear, nor does it suggest an absolute protective threshold [24–28]. We saw a vaccine dose-response relationship to anti-rCSP IgG levels, with the highest GMFR in the HHH/A group. However, mean anti-rCSP IgG at challenge was lower than RTS,S benchmarks.

Interestingly, avidity of anti-CSP full-length IgG and anti-CSP C-terminal region IgG antibodies were comparable to RTS,S benchmarks, but anti-CSP repeat region IgG antibody affinity was lower than the RTS,S benchmark. Findings of prior studies suggest a positive relationship between avidity and RTS,S-induced protection, although these studies differ as to whether anti–C-terminal antibody avidity or anti-CSP repeat region antibody avidity was associated with protection [25, 29]. The lack of clinical protection despite the high avidity of full-length and C-terminal region antibodies elicited by the rCSP/GLA-LSQ vaccine, and the inferior antibody concentration versus benchmark values, suggest that the magnitude of antibodies elicited by rCSP/GLA-LSQ may have been insufficient (eg, below a protective threshold).

Preclinical studies did show promise for this vaccine. Mice vaccinated with 20–25 µg of rCSP given with 5 µg of GLA-LSQ had >80% inhibition of liver-stage parasite development and 40% sterile protection in a mouse challenge model using transgenic Plasmodium berghei sporozoites containing P. falciparum CSP [12]. However, these mice received 5 µg of GLA with 10 µg of LSQ versus the 5-µg GLA with 2-µg LSQ dose in this clinical trial. In addition, studies of monoclonal antibodies to the junctional region have shown that they cannot be generated in mice owing to a lack of a counterpart to the human VH3-30 gene family that encodes these antibodies [30]. Further study into the immunological mechanisms of protection of CSP-based vaccines, such as systems serology studies, are needed to determine binding and functional features of human protective immunity.

Our results are limited by small sample size, though this was intentional to mitigate potential safety concerns in this first-in-human study. Despite this limitation, with 20 vaccinated participants, some evidence of efficacy against CHMI should have been seen if the vaccine had any protective effect. The lack of VE suggests that the vaccine construct, doses, and/or schedules given did not induce an immune response that provided sterile protection. Multiple potential explanations exist for this lack of efficacy. Suboptimal adjuvant dosing may have played a role. While GLA-LSQ has compositional similarities to AS01_E, GLA is less heterogeneous than MPL, making it difficult to compare 5 µg of GLA used in this study with the 25 µg of MPL contained in AS01_E. However, GLA-LSQ contains 2 µg of QS-21, compared with 25 µg in AS01_E. Furthermore, CSP antigen presentation as a virus-like particle might increase immunogenicity, although previous studies show equipoise on this topic [31, 32].

Regimen optimization, including a possible delayed third dose, may be necessary. For instance, another full-length CSP vaccine FMP013/ALFQ induced better protection (56% vs 11%) with a 4-month delay in the third dose (given at 6 vs 2 months) [33]. While preclinical studies confirmed the sequence, length, and function of rCSP [12], crystal structure visualization confirming the full conformation of CSP is challenging owing to size and central repeat region instability [34]. Other potential explanations include rCSP modifications when combined with GLA-LSQ, incorrect epitope folding/presentation, and triggering of antibodies targeting nonprotective epitopes or other ineffectual immune system pathways. Because a correlate of protection has not been identified for CSP-based vaccines, full-length CSP ELISA was chosen for initial immunogenicity assays for this first-in-human study. A follow-up study is underway examining responses to peptides across the length of CSP, including the junctional region, and will be reported separately. Cellular immune responses to rCSP/GLA-LSQ were not evaluated, as none have been associated with protection by CSP-based vaccines [35–37]. Additional studies exploring alternative antigen delivery, adjuvant dose, or vaccine schedules may improve efficacy.

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

Acknowledgments. We thank the participants from the Baltimore, Maryland, area for their time and commitment to this trial. We also thank Karen Kotloff, MD, principal investigator (PI) of the Vaccine and Treatment Evaluation Unit contract, for her input and guidance during the trial, and the Division of Microbiology and Infectious Diseases at the NIAID, for their invaluable support and advice. Our special thanks to Faith Pa’ahana-Brown, Jennifer Winkler, and Lisa Chrisley for study coordination; the Applied Immunology Section for clinical specimen handling and analysis; Sophie Harper, Aly Kwon, and Brenda Dorsey for regulatory oversight and quality assurance; and Panagiota Komninou for data entry.

Author contributions. D. J. F. K., A. X. M., E. Y. H. N., G. A. D., and M. B. L. participated in the conceptualization of the vaccine trial. D. J. F. K., A. A. B., M. A. T., E. Y. H. N., G. A. D., and M. B. L. executed the vaccine trial, participant follow-up and trial oversight. J. A. R. coordinated immunogenicity testing at WRAIR. M. S., E. S. B. L., and M. F. P. performed immunogenicity laboratory assays. K. A. S., S. J., and B. S. performed parasite diagnostic assays. C. C. provided database design and management, and J. B. and J. S. L. completed the statistical analyses. D. J. F. K. and M. B. L. wrote the original draft manuscript and verified the underlying data. All authors contributed to review and editing of the manuscript. All authors read and approved the final version of the manuscript.

Disclaimer. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases (Vaccine and Treatment Evaluation Unit contract HHSN272201300022I [PI, K. Kotloff; protocol PI, M. Laurens]; contracts HHSN272201200005I and HHSN272201800009I [support for stability testing]; and grant K23AI155838 to D. J. F. K.); the National Institutes of Health (T32 Fellowship Training Program in Vaccinology to Drs Myron Levine, Kathleen Neuzil, and Marcelo Sztein project T32 AI007524 [D. J. F. K.]); Burroughs Wellcome Fund/American Society of Tropical Medicine and Hygiene (postdoctoral fellowship in tropical infectious diseases to D. J. F. K.); the Passano Foundation (clinician-investigator award to D. J. F. K.); and the NIAID Malaria Vaccine Production Support Services (contract AI-N01-054210 [support for rCSP production and testing]).

References

Author notes