Vaccination with Plasmodium vivax Duffy-binding protein inhibits parasite growth during controlled human malaria infection

Abstract

There are no licensed vaccines against Plasmodium vivax. We conducted two phase 1/2a clinical trials to assess two vaccines targeting P. vivax Duffy-binding protein region II (PvDBPII). Recombinant viral vaccines using chimpanzee adenovirus 63 (ChAd63) and modified vaccinia virus Ankara (MVA) vectors as well as a protein and adjuvant formulation (PvDBPII/Matrix-M) were tested in both a standard and a delayed dosing regimen. Volunteers underwent controlled human malaria infection (CHMI) after their last vaccination, alongside unvaccinated controls. Efficacy was assessed by comparisons of parasite multiplication rates in the blood. PvDBPII/Matrix-M, given in a delayed dosing regimen, elicited the highest antibody responses and reduced the mean parasite multiplication rate after CHMI by 51% (n = 6) compared with unvaccinated controls (n = 13), whereas no other vaccine or regimen affected parasite growth. Both viral-vectored and protein vaccines were well tolerated and elicited expected, short-lived adverse events. Together, these results support further clinical evaluation of the PvDBPII/Matrix-M P. vivax vaccine.

Figure 1. Flow charts of study design and participant recruitment.

(A) VAC071 Group 1 participants received the viral-vectored vaccines ChAd63 PvDBPII and MVA PvDBPII 8 weeks apart, followed by CHMI 2 to 4 weeks later. Group 2 received ChAd63 PvDBPII before the trial was temporarily halted. On restart of the trial, returning participants in Group 2 received a second dose of ChAd63 PvDBPII at 17 months, followed by MVA PvDBPII 8 weeks later. Group 3 participants received the 8-week viral-vectored vaccine regimen and underwent CHMI along with Group 2 volunteers at 2 to 4 weeks after the final vaccination. (B) VAC079 participants received protein PvDBPII vaccine in Matrix-M adjuvant (PvDBPII/M-M). Group 1 volunteers received three doses at 0-1-14 months (delayed third dose due to trial halt). Group 2 volunteers received three doses at 0-1-2 months, with CHMI at 2 to 4 weeks after the final vaccination. (C) VAC069 participants underwent blood-stage CHMI in three separate stages and acted as infectivity controls for vaccinees undergoing CHMI in parallel. (D) Shown is a summary of the three CHMIs. VAC071 Group 1 vaccinees underwent CHMI in parallel with control participants in September 2019. In January 2020 vaccinations commenced in VAC071 and VAC079, before the trials were halted in March 2020. After restart of the VAC079 trial in 2021, Group 1 participants underwent CHMI in parallel with control participants in May 2021. In October 2021, control participants underwent CHMI in parallel with vaccinees from VAC071 Groups 2 and 3 and VAC079 Group 2. f/u, follow-up.

Figure 2. Local and systemic solicited adverse events.

Solicited AEs were recorded by volunteers within 7 days following each vaccination in participant symptom electronic diaries. The maximal severity reported for each AE is shown as a percentage of the number of vaccinations administered. (A) ChAd63 PvDBPII AEs are shown; n=18 vaccinations (16 volunteers received one dose, 2 volunteers received a second dose). (B) MVA PvDBPII AEs are shown; n=8 vaccinations (8 volunteers received one dose). (C) PvDBPII/M-M AEs are shown. AEs reported after first (n=16), second (n=15), and third doses (n=12) are shown.

Figure 3. Immunological responses were elicited by PvDBPII vaccinations.

(A) Anti-PvDBPII Salvador I (SalI) strain total IgG serum concentrations are shown over time for each vaccination regimen showing geometric mean with standard deviation. Groups are aligned at the time of final vaccination (day 56). Arrows indicate vaccinations with timing of doses in each regimen indicated below in months. VV-PvDBPII indicates viral-vectored vaccines. Blue shading indicates trial halt of about 1 year, vaccinations occurring prior to the trial halt are shown to the left. Red shading indicates period of controlled human malaria infection (CHMI). IgG concentrations below 1 μg/mL, indicated by dotted line, are classified as negative responses but shown for clarity. (B) Shown are individual anti-PvDBPII (SalI) total IgG serum concentrations 14 days post-final vaccination with geometric means for each regimen. (C) Shown are the percentages of IFN-γ⁺ cells within CD4⁺ CD45RA^- CCR7^- effector memory T cells collected 14 days post-final vaccination following stimulation of peripheral blood mononuclear cells with a pool of PvDBPII (SalI) peptides, with group medians. The frequency of IFN-γ⁺ cells in sample-matched unstimulated wells was subtracted to control for non-specific activation. Baseline responses (Day 0) are shown for all volunteers. (D) Shown are the dilution factors of individual serum, taken pre-CHMI, required to inhibit DARC-PvDBPII (SalI) binding by 50% (IC₅₀) with geometric means. Baseline responses (Day 0) are shown for all volunteers. (E) Shown is the percentage of in vitro growth inhibition activity (GIA) of 10 mg/mL purified total IgG, taken pre-CHMI, against P. knowlesi parasites expressing PvDBP PvW1 allele, with medians. Baseline responses (Day 0) are shown for all volunteers. p values were calculated by Kruskal-Wallis test with Dunn’s multiple comparison post-test.

Figure 4. PvDBPII/M-M inhibits growth of P. vivax after CHMI.

(A) Individual parasitemia over time was measured by qPCR, with group means in bold lines. Timings of vaccinations are shown in brackets in months. On the day of CHMI, volunteers were administered an intravenous injection of P. vivax (PvW1 clone) blood-stage parasites. The dotted line indicates the minimum concentration of parasitemia to meet positive reporting criteria (20 genome copies [gc]/mL). (B) Shown is a comparison of parasite multiplication rate (PMR) per 48 hours between vaccinees and controls. Individual PMRs are modelled from the qPCR data over time and are shown with group median. (C) Shown is a comparison of log₁₀ cumulative parasitemia (LCP) during CHMI between vaccinees and controls with group median. LCP calculated from area under the curve (AUC) of log₁₀-transformed qPCR over time for each individual, up until day 14 after challenge when the first volunteer reached malaria diagnostic criteria across all CHMIs. p values were calculated by Kruskal-Wallis test with Dunn’s multiple comparison post-test.

Figure 5. Antibody activity correlates with in vivo parasite growth inhibition.

(A to D) The percent of in vivo parasite growth inhibition (IVGI), calculated as % reduction in PMR in vaccinees relative to the mean PMR in infectivity controls were correlated with pre-CHMI measurements of percentage of IFN-γ⁺ cells within CD4⁺ CD45RA^- CCR7^- effector memory T cells (A); anti-PvDBPII (PvW1) total IgG serum titers in arbitrary units (AU) (B); dilution factor of individual serum required to inhibit DARC-PvDBPII (PvW1) binding by 50% (IC₅₀) (C); and (D) % in vitro GIA of 10 mg/mL purified total IgG against P. knowlesi parasites expressing the PvDBP PvW1 allele. Spearman’s rank correlation coefficients and p values are shown, n=18.