Clinical trial in healthy malaria-naïve adults to evaluate the safety, tolerability, immunogenicity and efficacy of MuStDO5, a five-gene, sporozoite/hepatic stage Plasmodium falciparum DNA vaccine combined with escalating dose human GM-CSF DNA

Abstract

When introduced in the 1990s, immunization with DNA plasmids was considered potentially revolutionary for vaccine development, particularly for vaccines intended to induce protective CD8 T cell responses against multiple antigens. We conducted, in 1997-1998, the first clinical trial in healthy humans of a DNA vaccine, a single plasmid encoding Plasmodium falciparum circumsporozoite protein (PfCSP), as an initial step toward developing a multi-antigen malaria vaccine targeting the liver stages of the parasite. As the next step, we conducted in 2000-2001 a clinical trial of a five-plasmid mixture called MuStDO5 encoding pre-erythrocytic antigens PfCSP, PfSSP2/TRAP, PfEXP1, PfLSA1 and PfLSA3. Thirty-two, malaria-naïve, adult volunteers were enrolled sequentially into four cohorts receiving a mixture of 500 μg of each plasmid plus escalating doses (0, 20, 100 or 500 μg) of a sixth plasmid encoding human granulocyte macrophage-colony stimulating factor (hGM-CSF). Three doses of each formulation were administered intramuscularly by needle-less jet injection at 0, 4 and 8 weeks, and each cohort had controlled human malaria infection administered by five mosquito bites 18 d later. The vaccine was safe and well-tolerated, inducing moderate antigen-specific, MHC-restricted T cell interferon-γ responses but no antibodies. Although no volunteers were protected, T cell responses were boosted post malaria challenge. This trial demonstrated the MuStDO5 DNA and hGM-CSF plasmids to be safe and modestly immunogenic for T cell responses. It also laid the foundation for priming with DNA plasmids and boosting with recombinant viruses, an approach known for nearly 15 y to enhance the immunogenicity and protective efficacy of DNA vaccines.

Keywords: DNA vaccine; GM-CSF; Plasmodium falciparum; clinical trials; controlled human malaria infection; malaria; malaria challenge; malaria vaccine; vaccine safety.

Figure 1. Flow diagram of volunteers. 102 volunteers provided informed consent and 57 passed screening. Thirty-three were excluded because: HLA type not needed (11), abnormal laboratory results (9), finding on medical history (3), finding on physical exam (3), HIV positive (1), conflicting concomitant medication (1), leaving area (1), military supervisor denied approval (1), non-compliance with appointment schedule (1), problem with method of compensation (1), enrolled in another protocol (1) and 12 withdrew of their own accord without specifying a reason. During the interval between screening and the first immunization, ten additional volunteers were dropped because: military supervisor denied approval (2), military transfer (3), pregnancy (1) or withdrew of their own accord without specifying a reason (4). Forty-six of the remaining 47 enrolled volunteers were assigned sequentially to cohorts 1–4. One volunteer in cohort 3 withdrew after the first immunization (personal reasons not related to adverse events) and was not replaced, one volunteer withdrew from the infectivity controls in cohort 1 (personal reasons not related to adverse events) and was replaced from the enrolled volunteers (denoted by an asterisk*), and one volunteer withdrew from the infectivity controls of cohort 4 prior to challenge (personal reasons not related to adverse events).

Figure 2. Frequency of Mild and Moderate Adverse Events in Cohorts 1, 2, 3 and 4 after each immunization. Local adverse events are shown after each DNA immunization, and are grouped according to mild (grade 1) or moderate (grade 2) severity. Increasing doses of hGM-CSF-encoding plasmid did not affect tolerability.

Figure 3. Anti-dsDNA and anti-hGM-CSF antibody responses following DNA immunization. Anti-dsDNA levels (IU/mL) (A) and anti-hGM-CSF levels (optical density) (B) were measured at pre-immunization, at 4, 8, 12, 24 and 52 weeks. Cohorts 1, 2, 3 and 4 are color-coded. The upper black line in each panel represents the division between normal and borderline positive values for these tests. There are no data for the four-week time point for anti-dsDNA for cohorts 1 and 2 (data points used in the graph are averages of the values at 0 and 8 weeks). Anti-dsDNA levels were measured by the National Naval Medical Center Clinical Laboratory using fresh sera, and the anti-hGM-CSF antibodies were measured by Vical Inc., using frozen sera.

Figure 4. ELISA antibody responses by antigen and by cohort. (A) The four graphs provide geometric mean antibody titers to CSP, TRAP/SSP2, EXP1 and LSA1, respectively, as measured by ELISA pre-immunization (Pre-Imm), two weeks after the third DNA immunization/pre-challenge (Pre-Ch), and 25 d and 38 d after challenge (Ch+25, Ch+38). Cohorts are color-coded, with solid lines representing immunized volunteers, dotted lines representing infectivity control volunteers. (B) The four graphs provide geometric means of antibody titers to CSP, TRAP/SSP2, EXP1 and LSA1 for each cohort, respectively. Antigens are color-coded, with solid lines representing immunized volunteers, dotted lines representing infectivity control volunteers.

Figure 5. Immunofluorescent antibody responses to sporozoites, liver and blood stages. The figures show IFA titers against sporozoites, liver stages and blood stages, measured at pre-immunization (Pre), week 10 (pre-challenge) and week 14 (25 d after challenge). Time points are color-coded in different shades of black and gray. Cohorts 1, 2, 3 and 4 were tested against sporozoites and blood stages and cohort 1 against liver stages. Infectivity controls were measured at the same time points. *Volunteer not tested.

Figure 6. Vaccine efficacy by Kaplan-Meier Plot. The figure present parasitemia-free survival curves (Kaplan-Meier) for immunized (solid lines) and infectivity controls (dotted lines) for the four cohorts, respectively, based on microscopic examination of peripheral blood smears. Log rank test was used to test the statistical significance between the two groups. There was no significant difference in time to parasitemia between immunized volunteers and control volunteers.

Figure 7. Days to parasitemia of combined HLA A02 and A03 volunteers. Days to parasitemia are shown in HLA A02 and A03 volunteers, comparing vaccinees (gray bars) and infectivity controls (white bars). Since the mean day of onset of parasitemia was different for each of the four challenges, days prior to or following the group-specific mean day of onset were calculated for each volunteer in order to standardize. Data were then combined on a common timeline with mean day of onset for each cohort set to zero. There were no statistically significant differences in time to parasitemia between cohorts (p = 0.52, Wilcoxon) or between immunized volunteers and infectivity controls (p = 0.32, Mann-Whitney).

Figure 8. Schematic diagram of the DNA gene products. Each panel presents the native protein (above) and the protein expressed by the DNA (below) for the five vaccine antigens and for hGM-CSF. The amino acid length is provided in parentheses to the right of the diagram. Identical colors indicate identical sequences. SS = native signal sequence. TPA = human tissue plasminogen activator leader sequence (increases expression in mammalian cells) (yellow boxes). (A) The PfCSP DNA vaccine includes the full length native sequence CSP gene. The locations of the endogenous PfCSP signal sequence, the PfCSP repeat region, the PfCSP glycosylphosphatidylinositol (gpi) anchor and a 23 amino acid C-terminus insertion derived from the transcriptional terminator of the bovine growth hormone gene (dark blue box) are shown. (B) The PfSSP2 DNA vaccine includes the full length native sequence SSP2 gene. The locations of the endogenous PfSSP2 signal sequence, the PfSSP2 repeat region and the PfSSP2 transmembrane region are shown. (C) The PfEXP1 DNA vaccine includes the full length native sequence EXP1 gene. The location of the endogenous PfEXP1 signal sequence is shown. (D) The PfLSA1 DNA vaccine includes the full N- and C- termini of the native sequence LSA1 gene minus the repeat region. The locations of the endogenous PfLSA1 signal sequence and the PfLSA1 repeat region are shown. The 727 amino acid deletion encompassing the PfLSA1 repeat region is represented by dashes. (E) The PfLSA3 DNA vaccine includes the full length native sequence LSA3 gene. The locations of the PfLSA3 repeat regions are shown. (F) The hGM-CSF DNA plasmid includes the full length native sequence human GM-CSF gene. A 3 amino acid insertion generated during the cloning of the GM-CSF plasmid is located at the N-terminal end and is represented by a light blue box (the first few amino acids of the native GM-CSF protein, MWLQSLLLL, were replaced by MALWILQSLLLL in the DNA construct).