Conceived and designed the experiments: NL ST EM. Performed the experiments: BJW MD GL. Analyzed the data: NL BJW ST EM LPV. Contributed reagents/materials/analysis tools: ST EM MD GL. Wrote the paper: NL BJW LPV ST EM.
NL, ST, MD and L-PV are employees of Medicago Inc., which sponsored the preclinical and clinical trials. This does not alter the authors' adherence to all the PLOS one policies on sharing data and materials.
The recent swine H1N1 influenza outbreak demonstrated that egg-based vaccine manufacturing has an Achille's heel: its inability to provide a large number of doses quickly. Using a novel manufacturing platform based on transient expression of influenza surface glycoproteins in
ClinicalTrials.gov NCT00984945
The recent swine H1N1 influenza pandemic (pH1N1) revealed the limitations of the current influenza vaccine manufacturing technologies. On May 26, 2009, the WHO recommended rapid development of vaccines and the first reassortant virus was available on May 27 2009
It is ironic that the world's attention had been focused on H5N1 viruses for more than a decade when pH1N1 emerged. It is important that the global sense of relief at the relatively benign nature of the pH1N1 pandemic should not lull us into complacency regarding H5N1 influenza and other potential pandemic strains. Although human-to-human spread has eluded H5N1 viruses to date, a total of 498 human cases have been reported to the WHO with a crude fatality rate of 59% (as of May 6, 2010). Continued hyper-mutation of H5N1 strains is occurring worldwide in avian populations and reassortment with an influenza strain actively circulating in humans is always a serious threat. Jackson S and colleagues have recently shown that such reassortment can readily occur between human H3N2 and avian H5N1 strains in ferrets
We have recently described a plant-based manufacturing technology that can produce vaccine doses within one month of the sequencing of a pandemic strain
The protocol for this trial and supporting CONSORT checklist are available as supporting information; see
The H5 VLP vaccine was produced as described in D'Aoust
This clarified extract was then passed through a Poros HQ column equilibrated at pH 7.5 with 50 mM Tris-HCl -0.01% Tween 80. The flow-through was captured on a Poros HS column equilibrated in 50 mM NaPO4, 0.01% Tween 80 (pH 6.0)(Applied Biosystems, USA). After washing with 50 mM NaPO4, 65 mM NaCl, 0.01% Tween 80 (pH 6.0), the VLPs were eluted with 50 mM NaPO4, 500 mM NaCl, 0.01% Tween 80 (pH 6.0) and then captured on a Poros EP 250 coupled to bovine fetuin (30 mg fetuin/mL Poros EP 250 matrix)(Desert Biologicals, Australia) as recommended by the manufacturer and equilibrated in 50 mM NaPO4, 150 mM NaCl (pH 6.0). The column was washed with 50 mM NaPO4, 400 mM NaCl, (pH 6.0) and the VLPs were eluted first with 1,5 M NaCl, and then water containing 0.0005% Tween 80. The purified VLPs were concentrated by TFF on a 300,000 Da MWCO membrane, diafiltered against formulation buffer (100 mM PO4, 150 mM NaCl, 0.01% Tween 80 at pH 7.4) and passed through a 0.22 µm filter for sterilisation.
Electron microscopy was performed as previously described by D'Aoust
SDS–PAGE analysis of VLP preparations was performed on pre-cast gels, Criterion™ XT 4–12% Bis-Tris (Bio-Rad Laboratories Hercules, CA). Samples were mixed with 4X SDS sample buffer with DTT (EMD Chemicals Inc., Gibbstown) and 2.5 µg of protein was loaded per lane. Gels were processed according to the manufacturer's instructions and stained with BioSafe™ Coomassie G-250 (Bio-Rad Laboratories Hercules, CA).
Endotoxin levels were determined by the
Detection of residual DNA was performed with the PicoGreen® fluorescent dye assay (Invitrogen Canada, Burlington, ON) and measured by fluorometry using Lambda DNA for the standard curve (Invitrogen Canada, Burlington, ON).
The ferret study was approved by the Institutional Animal Care and Use Committee (IACUC) of the Southern Research Institute (SRI Birmingham, AL). Male Fitch ferrets were castrated, descented and demonstrated to be seronegative negative for representative circulating human influenza A strains prior to shipment to Southern research Institue(, 6–8 months old, 0.8–1.6 kg, Triple F Farms, Sayre, PA). The ferrets were vaccinated twice intramuscularly on days 0 and 21 with H5 VLP vaccine (0.7, 1.8, 3.7 or 11.0 µg HA formulated with alum (Alhydrogel®: 0.5 mg aluminium per 0.5 mL dose) or with placebo (PBS + alum). Eight animals per group in the 1.8 or 3.7 µg vaccine and placebo groups were challenged intranasally 45 days after the 21 day boost with a lethal dose of A/Vietnam/1203/04 H5N1 clade 1 virus (10 X Ferret Lethal Dose50). Animals were monitored for weight loss, temperature and loss of activity weekly during vaccination and daily during challenge. All surviving animals were euthanized 14 days post-challenge. Three challenged animals in each group were sacrificed 3 days after challenge and their lungs and nasal turbinate tissues were collected, weighted, snap frozen in liquid nitrogen until used for virus titres. Homogenized samples were serially diluted 10-fold and inoculated into viable 10 to 11 day old embryonated hen's eggs and viral titers were measured by the Egg Infectious Dose50 (EID50) assay. Data are expressed as log10EID50/mL using the Reed-Muench method
The Hemagglutination inhibition (HI) assay was performed according to WHO recommendation
Microneutralization (MN) assays were performed as previously described
Single radial hemolysis (SRH) was performed at the University of Siena, Italy, against whole inactivated A/Indonesia/5/2005 (H5N1) virus (subclade 2.1 strain: CBER). The SRH was modified slightly from the method described by Schild et al
Ninety-six-well plates were coated overnight at 4°C with 50 µL of either avidin (1 mg/mL) derived from egg white: (Sigma-Aldrich, St-Louis, MO) or recombinant avidin expressed in corn as a source of plant glycans (1 mg/L: Sigma-Aldrich) in 50 mM sodium carbonate buffer (pH 9.6). The presence of plant-specific xylose and fucose on the recombinant corn avidin was established by Western blot. Plates were washed with PBS containing 0.1% Tween-20 (PBS-T) and blocked with 1% casein in PBS-T (blocking solution) for 1 h at 37°C. The plates were then incubated with serial two-fold dilutions of sera in blocking solution for 1 h at 37°C. The plates were washed (PBS-T) and incubated for 1 h at 37°C with HRP-conjugated donkey anti-human IgG (H+L) (Jackson ImmunoResearch Inc.) or HRP-conjugated goat anti-human IgE (Sigma-Aldrich). Both conjugates were diluted at 1∶10,000 in blocking solution. HRP activity was detected by addition of 100 µl SureBlue TMB Microwell Peroxidase Substrate (KPL, Kirkegaard & Perry Laboratories, Inc.). After 20 minutes at room temperature, the enzymatic reaction was stopped with 100 µl 1N HCl and OD was determined at 450 nm. Differences pre- and post-immunization were calculated as the highest titer giving a difference >0.1 OD. Rabbit antibodies raised against xylose or fucose residues (AgriSera) were used as controls. ‘Positive’ reactions were defined as a two-fold difference between the titer for corn avidin versus egg avidin.
This randomized, double-blind, placebo-controlled Phase I clinical trial was performed at the McGill University Health Centre (MUHC) Vaccine Study Center, (Pierrefonds, QC) to assess the safety and immunogenicity of the plant-produced H5 VLP vaccine (registered at ClinicalTrials.gov: NCT00984945). The trial was approved by both the Canadian Biologic and Genetic Therapies Directorate and the Research Ethics Committee of the MUHC (Final approval, document Dated of September 2nd 2009.
Forty eight healthy adults aged 18–60 years were recruited. The principal exclusion criteria included significant medical or neuropsychiatric illness, immunosuppression or immunodeficiency; ongoing febrile illness; history of autoimmune disease; history of H5N1 vaccination; any vaccination within a 30 day period prior to enrolment, or planned vaccination between the first vaccination up to blood sampling at Day 42; use of any investigational or non-registered product within 90 days prior to study enrolment or planned use during the study; systemic glucocorticoid therapy; coagulation disorders or treatment with anticoagulants; history of allergy to constituents of H5 VLP (H5N1) vaccine or tobacco; history of severe allergic reactions or anaphylaxis; receipt of a blood transfusion or immunoglobulins within 90 days of enrolment; pregnancy; lactation; or cancer or treatment for cancer within 3 years of vaccine administration. Written informed consent was obtained from all subjects.
H5 VLP vaccine was produced as described above and formulated with 1% Alhydrogel® prior to vaccination (0.5 mg aluminium per 0.5 mL dose). Groups of 12 subjects received vaccines containing 5, 10 or 20 µg of HA (as assessed by SRID assay). The placebo group received PBS buffer (100 mM phosphate, 150 mM NaCl, 0.01% Tween 80) + alum (as above).
Starting at the lowest dose of H5 vaccine (5 µg/dose), groups of 16 subjects were randomized using block permutation to receive either active vaccine (n = 12) or placebo (n = 4) and the study was completed in 3 waves. At each dose level, safety data for the first 7 days after vaccination was reviewed by an independent panel before the next group of 12+4 subjects was dosed. The same staggered approach was used for booster immunizations. All subjects were immunized by a nurse masked to group assignment. All immunizations were administered in the deltoid muscle and subjects were observed for at least 2 hours after each immunization for any signs or symptoms of local or systemic reaction. Vital signs were assessed hourly during this period. Serum was collected before and 21 days after each immunization and aliquots were held at 2 to 8°C for biochemistry and haematologic analyses. Additional aliquots were stored at −20°C until analyzed for immunologic assays. At the time of writing, a 6-month follow-up period is ongoing.
A memory aid, rulers and thermometers were given to all subjects to record adverse events occurring up to Day 21 after each vaccination. The occurrence of solicited local and systemic reactions was recorded daily for 7 days following each dose (See
HI titers in the ferret study were assessed using the Student's
The H5 VLP vaccine was produced in accordance with Good Manufacturing Procedures (GMP) for Phase 1 clinical grade material. The vaccine contained the HA protein of the A/Indonesia/5/05 H5N1 anchored in plasma membrane from
A. Cross-section showing internal differences. B. Transmission electron microscopy images of influenza viruses and plant-made VLPs.
The VLP vaccine induced detectable HI titers at all doses tested (0.7 to 11((g) (
Vaccine Dose |
Vaccination | HI titers |
|||
Indo/5/05 (clade 2.1) | Turkey/Turkey/1/05 (clade 2.2) | Anhui/1/05 (clade 2.3) | VN/1203/04 (clade 1) | ||
0.7 µg VLP | 1st dose |
|
|||
2nd dose |
|||||
1.8 µg VLP | 1st dose |
||||
2nd dose |
|||||
3.7 µg VLP | 1st dose |
||||
2nd dose |
|||||
11.0 µg VLP | 1st dose |
|
|||
2nd dose |
All vaccines were formulated with Alhydrogel 1% (0.5 mg per dose).
HI titers measured 21 days after vaccination.
HI titers measured 14 days after boost vaccination.
Geometric Mean Titer measured on all animals. HI negative animals were given an arbitrary value of 4 or 5 depending on starting dilution in the HI assay.
The H5 VLP vaccine also induced detectable cross-reactive HI antibodies in ferrets. Not surprisingly, cross-reactive HI titers were highest after the second dose (
HI titers were overall somewhat lower for the clade 1 challenge strain (A/Vietnam/1203/04) than for clade 2 strains. Twenty-one days after the second dose, HI titers to A/Vietnam/1203/04 were detectable in only 75–87.5% of the challenged ferrets (See
The challenge with A/Vietnam/1203/04 in the 1.8 and 3.7 µg and placebo groups revealed complete protection from clinical illness in the vaccinated animals only. In the first days after challenge, ferrets in the placebo group had a marked increase in fever followed by a decrease in body temperature while vaccinated animals had no important body temperature fluctuations (
Ferrets were immunized twice with the H5 VLP vaccine (A/Indonesia/5/05) or placebo (alum) and were challenged with 10 FLD50 of the A/Vietnam/1203/04 H5N1 strain 45 days after the booster injection. A. Mean temperature (5 ferrets per group) B. Percent weight loss (at day 6, 3 ferrets were found dead and the 2 remaining were euthanized due to ≥20% body weight loss C. Survival D. Activity score.
Experiment | Pre-challenge HI GMT |
Mean viral load |
Mean % body weight change | Survival | ||
Indo/5/05 | VN/1203/04 | URT | lungs | |||
After 2nd dose | ||||||
1.8 µg VLP | 429 (8/8) | 24 (6/8) | 3.5 (1/3) |
2.0 (2/3) | +2.1 | 5/5 |
3.7 µg VLP | 382 (8/8) | 28 (7/8) | 2.5 (2/3) |
<1.5 (0/3) | −0.3 | 5/5 |
PBS | <8 (0/8) | <8 (0/8) | 3.2 (3/3) | 2.25 (1/3) | −17.7 | 0/5 |
HI titers measured on sera taken 14 days after last vaccination.
Values are expressed as log10EID50/ml, mean calculated on positive animals only. Individual values by dose and location: 1.8 µg URT (3.5, < LOD,< LOD) lungs (1.98, 2.0, <LOD); 3.7 µg URT (2.5, 2.5, < LOD) lung (< LOD, < LOD,< LOD), placebo URT (3.75, 3.33, 2.5) lungs (2.25, 2 x< LOD).
Virus titration performed on animals sacrificed 3 days post-challenge.
The H5 VLP vaccine was well tolerated. Although the sample size was limited for this first-in-human study, 12 subjects per treatment group should permit the capture of AEs having 10% or 5% incidences with probabilities of ∼72% and ∼46% respectively. There were no Serious AEs reported up to Day 42 (21 days after second dose). The 6-month follow-up period is still ongoing at time of writing. Pain at injection site, redness and headache were the most commonly reported local and systemic reactions (
Symptoms are graded on the following scale: mild = subject is aware of the AE but it causes no limitation of usual activities, moderate = subject is aware of the AE and the event causes some limitation of usual activities and severe = AE is of such severity that the subject is unable to carry out usual activities.
First dose | Second dose | |||||||
Adverse event | 5 µg | 10 µg | 20 µg | Placebo | 5 µg | 10 µg | 20 µg | Placebo |
|
||||||||
Redness | 2 (16.7; 0.16–1.00) | 4 (33.3; 0.40–1.00) | 9 (75.0; 0.66–1.00) | 3 (25.0; 0.29–1.00) | 8 (66.7; 0.63–1.00) | 2 (16.7; 0.16–1.00) | 5 (41.7; 0.48–1.00) | 5 (41.7; 0.48–1.00) |
Swelling | 1 (8.3; 0.03–1.00) | 0 | 4 (33.3; 0.40–1.00) | 2 (16.7; 0.16–1.00) | 2 (16.7; 0.16–1.00) | 5 (41.7; 0.48–1.00) | 1 (8.3; 0.03–1.00) | 3 (25.0; 0.29–1.00) |
Pain | 11 (91.7; 0.72–1.00) | 8 (66.7; 0.63–1.00) | 8 (66.7; 0.63–1.00) | 7 (58.3; 0.59–1.00) | 9 (75.0; 0.66–1.00) | 9 (75.0; 0.66–1.00) | 8 (66.7; 0.63–1.00) | 6 (50.0; 0.54–1.00) |
|
||||||||
Fever | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
Headache | 4 (33.3; 0.40–1.00) | 6 (50.0; 0.54–1.00) | 2 (16.7; 0.16–1.00) | 4 (33.3; 0.40–1.00) | 6 (50.0; 0.54–1.00) | 5 (41.7; 0.48–1.00) | 1 (8.3; 0.03–1.00) | 2 (16.7; 0.16–1.00) |
Joint aches | 0 | 1 (8.3; 0.03–1.00) | 1 (8.3; 0.03–1.00) | 0 | 0 | 2 (16.7; 0.16–1.00) | 0 | 1 (8.3; 0.03–1.00) |
Fatigue | 0 | 4 (33.3; 0.40–1.00) | 1 (8.3; 0.03–1.00) | 1 (8.3; 0.03–1.00) | 4 (33.3; 0.40–1.00) | 3 (25.0; 0.29–1.00) | 2 (16.7; 0.16–1.00) | 2 (16.7; 0.16–1.00) |
Muscle aches | 2 (16.7; 0.16–1.00) | 5 (41.7; 0.48–1.00) | 2 (16.7; 0.16–1.00) | 3 (25.0; 0.29–1.00) | 1 (8.3; 0.03–1.00) | 3 (25.0; 0.29–1.00) | 0 | 0 |
Feeling of general discomfort | 2 (16.7; 0.16–1.00) | 2 (16.7; 0.16–1.00) | 0 | 1 (8.3; 0.03–1.00) | 2 (16.7; 0.16–1.00) | 2 (16.7; 0.16–1.00) | 0 | 0 |
Chills | 0 | 2 (16.7; 0.16–1.00) | 0 | 1 (8.3; 0.03–1.00) | 1 (8.3; 0.03–1.00 | 1 (8.3; 0.03–1.00 | 0 | 0 |
Data are number(%; 95% CI). Adverse events up to 7 days after vaccination are reported.
All 48 subjects were included in the safety and immunogenicity analysis (
A total of 48 subjects were enrolled in the study and randomized in a 1∶1∶1∶1 ratio to receive two vaccinations of either 5, 10 or 20 µg of H5 VLP or placebo mixed with Alhydrogel. Vaccinations were administered 21 days apart.
H5 VLP Vaccine | Placebo | |||
5 µg | 10 µg | 20 µg | ||
|
||||
N | 12 | 12 | 12 | 12 |
Mean (std. dev.) | 45 (8) | 42 (11) | 33 (13) | 40 (12) |
(Min.;Max.) | (29; 56) | (21; 57) | (21; 53) | (21; 59) |
|
||||
Male | 5 (41.7%) | 3 (25.0%) | 7 (58.3%) | 6 (50.0%) |
Female | 7 (58.3%) | 9 (75.0%) | 5 (41.7%) | 6 (50.0%) |
Parameter | H5 VLP vaccine | Placebo | ||
5 µg HA, n = 12 | 10 µg HA, n = 12 | 20 µg HA, n = 12 | n = 12 | |
|
||||
HI | ||||
GMT | 4.0 | 4.0 | 4.3 (3.7–5.1) | 4.0 |
Number of subject with positive response (%) | 0 | 0 | 8.3 (5.5–57.2) | 0 |
Seroprotection (%) | 0 | 0 | 0 | 0 |
|
||||
GMA | 6.2 (4.4–8.8) | 7.8 (5.0–12.1) | 7.3 (4.6–11.4) | 5.4 (4.4–6.6) |
Number of subject with positive response (%) | 8.3 (5.5–57.2) | 16.7 (2.1–48.4) | 16.7 (2.1–48.4) | 0 |
Seroprotection (%) | 0 | 16.7 (2.1–48.4) | 16.7 (2.1–48.4) | 0 |
|
||||
GMT | 5.0 | 5.0 | 5.0 | 5.0 |
Number of subject with positive response (%) | 0 | 0 | 0 | 0 |
|
||||
|
||||
GMT | 4.7 (3.3–6.7) | 5.5 (4.0–7.6) | 7.5 (4.4–12.9) | 4.5 (4.0–10.0) |
Number of subject with positive response (%) | 8.3 (0.2–38.5) | 25 (5.5–57.2) | 50 (21.1–78.9) | 8.3 (0.2–38.5) |
Seroprotection (%) | 0 | 0 | 8.3 (0.2–38.5) | 0 |
Seroconversion (%) | 0 | 0 | 8.3 (0.2–38.5) | 0 |
GMI | 1.2 (0.8–1.7) | 1.4 (1.0–1.9) | 1.7 (1.0–3.2) | 1.1 (0.9–1.3) |
|
||||
GMA | 9.7 (6.6–14.4) | 10.0 (6.3–15.6) | 11.2 (6.4–19.7) | 6.3 (4.4–8.9) |
Number of subject with positive response (%) | 41.7 (15.2–72.3) | 33.3 (9.9–65.1) | 41.7 (15.2–72.3) | 16.7 (2.1–48.4) |
Seroprotection (%) | 8.3 (0.2–38.5) | 16.7 (2.1–48.4) | 25.0 (5.5–57.2) | 0 |
Seroconversion (%) | 16.7 (2.1–48.4) | 8.3 (0.2–38.5) | 25.0 (5.5–57.2) | 16.7 (2.1–48.4) |
GMI | 1.6 (1.1–2.5) | 1.3 (1.1–1.6) | 1.5 (0.7–3.5) | 1.2 (0.9–1.6) |
|
||||
GMT | 6.9 (4.5–10.6) | 5.0 | 7.6 (5.1–11.2) | 5.0 |
Number of subject with positive response (%) | 25 (5.5–57.2) | 0 | 33.3 (9.9–65.1) | 0 |
Seroconversion (%) | 8.3 (0.2–38.5) | 0 | 16.7 (2.1–48.4) | 0 |
|
||||
|
||||
GMT | 11.9 (5.6–25.5) | 18.2 (9.8–33.8) | 29.5 (10.7–80.9) | 4.0 |
Number of subject with positive response (%) | 66.7 (34.9–90.1) | 100 (73.5–100) | 75 (42.8–94.5) | 0 |
Seroprotection (%) | 16.7 (2.1–48.4) | 25.0 (5.5–57.2) | 50.0 (21.1–78.9) | 0 |
Seroconversion (%) | 16.7 (2.1–48.4) | 25.0 (5.5–57.2) | 58.3 (27.7–84.8) | 0 |
GMI | 3.0 (1.4–6.4) | 4.5 (2.4–8.4) | 6.8 (2.3–20.3) | 1.0 |
|
||||
GMA | 15.5 (9.3–26.0) | 16.4 (9.4–28.7) | 26.2 (13.7–50.1) | 7.1 (4.4–11.4) |
Number of subject with positive response (%) | 50.0 (21.1–78.9) | 50.0 (21.1–78.9) | 75.0 (42.8–94.5) | 25.0 (5.5–57.2) |
Seroprotection (%) | 41.7 (15.2–72.3) | 41.7 (15.2–72.3) | 75.0 (42.8–94.5) | 8.3 (0.2–38.5) |
Seroconversion (%) | 41.7 (15.2–72.3) | 50.0 (21.1–78.9) | 58.3 (27.7–84.8) | 25.0 (5.5–57.2) |
GMI | 2.5 (1.5–5.0) | 2.1 (1.4–3.7) | 3.6 (2.4–8.7) | 1.3 (0.8–2.4) |
|
||||
GMT | 16.8 (8.9–31.7) | 28.3 (14.9–53.8) | 48.1 (20.1–114.9) | 5.7 (4.7–7.1) |
Number of subject with positive response (%) | 83.3 (51.6–97.9) | 100 (73.5–100) | 91.7 (61.5–99.8) | 16.7 (2.1–48.4) |
Seroconversion (%) | 41.7 (15.2–72.3) | 50.0 (21.1–78.9) | 66.7 (34.9–90.1) | 0 |
Note: Data in parenthesis are 95% CI. HI denotes Hemagglutination Inhibition assay, SRH Single Radial Hemolysis assay, MN MicroNeutralisation assay, GMT Geometric Mean Titer. GMI, Geometric Mean of the Increase.
HI titers rose to ≥40 in 16.7, 25 and 50% of the subjects after the second dose in the 5, 10 and 20 µg groups respectively (
GMI and seroconversion rates were calculated for the 3 serological assays used: HI, MN and SRH. The HI assay provides an estimate of IgG that can prevent agglutination of erythrocytes driven by HA of both wild and attenuated viruses. This assay has limited sensitivity for avian strains
Using HI results, the seroconversion rates after the second dose were 16.7, 25 and 58.3% for the 5, 10 and 20 µg groups respectively (
Using HI results, seroprotection rates after the second dose were 16.7, 25 and 58.3% in the 5, 10 and 20 µg groups respectively. Seroprotection rates in these same groups at day 42 were higher by both MN (41.7, 50 and 66.7% respectively) and SRH (41.7, 41.7 and 75.0% respectively) (
H5N1 strains | ||||||||||||
A/Indonesia/5/05 (clade 2.1) | A/turkey/Turkey/1/05 (clade 2.2) | A/Anhui/1/05 (clade 2.3) | A/Vietnam/1203/04 (clade 1) | |||||||||
Outcome | 5 µg | 10 µg | 20 µg | 5 µg | 10 µg | 20 µg | 5 µg | 10 µg | 20 µg | 5 µg | 10 µg | 20 µg |
|
||||||||||||
Seroconversion (%) | 16.7 | 25.0 | 58.3 | 0 | 8.3 | 0 | 8.3 | 8.3 | 25.0 | 0 | 0 | 0 |
≥1∶40 (%) | 16.7 | 25.0 | 50.0 | 0 | 8.3 | 0 | 8.3 | 0 | 16.7 | 0 | 0 | 0 |
GMT | 11.9 | 18.2 | 29.5 | 4.9 | 6.0 | 7.0 | 6.0 | 7.1 | 8.7 | 4.3 | 4.5 | 4.8 |
|
||||||||||||
Seroconversion (%) | 25.0 | 25.0 | 58.3 | 0 | 0 | 8.3 | ||||||
≥1∶40 (%) | 25.0 | 25.0 | 58.3 | 0 | 0 | 8.3 | ||||||
GMT | 14.2 | 28.3 | 48.7 | 6.55 | 6.74 | 12.1 |
A good correlation was observed between the immune response measured by the HI and SRH assays (Spearman rs = 0.616,
Cross-reactive antibodies towards H5N1 strains both within (subclade) and across clades were detected with the HI and MN assays (
The presence of glycans bearing the plant-specific α-1,3-fucose and β-1,2-xylose residues and terminal N-acetylglucosamine in the H5 VLP vaccine was confirmed by Western blot and mass spectrometry (data not shown). During the Phase I clinical trial, we sought to determine if IM administration of the vaccine induced either IgG or IgE responses to these plant-specific glycans. Although subjects with histories of severe allergies were excluded from the trial, subjects reporting mild to moderate allergies to plant derivatives (eg: seasonal allergies, hay fever, allergy to ragweed, grape or eggplant) were included.
Group | At screening, number of subjects reporting ongoing allergy (number;%) | |
Any known allergy | Allergy to any plant derivative |
|
5 µg H5 VLP | 5 (41.7%) | 5 (41.7%) |
10 µg H5 VLP | 5 (41.7%) | 3 (25.0%) |
20 µg H5 VLP | 4 (33.3%) | 4 (33.3%) |
Placebo | 8 (66.6%) | 7 (58.3%) |
Total in vaccine groups | 14 (38.8%) | 12 (33.3%) |
Total in all groups | 22 (45.8%) | 19 (39.5%) |
as expressed by seasonal allergies, hay fever, allergy to ragweed or allergy to grapes or eggplant.
Group | Before immunisation (D0) | After two immunisations (D42) | ||||
Number of subjects with detectable IgGs to plant-specific sugar moieties | Number of subject with an increase in IgGs to plant-specific sugar moieties | Number of subject with an increase in IgEs to plant-specific sugar moieties | ||||
Total subjects | In subjects who reported known allergy to plant component | Total subjects | In subjects who reported known allergy to plant component | Total subjects | In subjects who reported known allergy to plant component | |
5 µg H5 VLP | 1/12 | 1/5 | 2/12 | 1/5 |
0/12 | 0/5 |
10 µg H5 VLP | 3/12 | 1/3 | 1/12 | 0/3 | 0/12 | 0/3 |
20 µg H5 VLP | 2/12 | 0/4 | 2/12 | 0/4 | 0/12 | 0/4 |
Placebo | 1/12 | 0/7 | 1/12 | 0/7 | 0/12 | 0/7 |
Total | 7/48 (14,6%) | 2/19 (10,5%) | 6/48 (12,5%) | 1/19 (5,2%) | 0/48 (0%) | 0/19 (0%) |
Not the same subject who had detectable Abs to plant-specific sugar moieties before immunisation.
The 2009-10 H1N1 pandemic demonstrated that traditional egg-based manufacturing technologies are not currently in a position to provide the volume of vaccines required to respond to a global pandemic strain in a timely fashion. In fact, it is unlikely that egg-based production will ever be able to respond rapidly enough to influence the first wave of a rapidly spreading pandemic virus. Our recently described candidate H5 VLP vaccine produced in plants
Safety is a critical issue for vaccines since public confidence can make or break immunization programs
None of the two subjects who had both known allergies to plant derivatives and detectable levels of IgGs to plant-specific N-glycans before vaccination had increases in this type of IgGs, but one with known allergies in the VLP group showed an increase, as did 4 other subjects which had no known allergies to plant derivatives. Thus 13,9% of the VLP subjects showed a detectable increase while 8,3% of the placebo subjects showed a similar increase, and this difference has no statistical significance (
Prior to the Phase I trial, ferret studies of the H5 VLP vaccine demonstrated the induction of good levels of neutralizing antibodies both within- and across clades as well as complete protection from cross-clade lethal challenge. It is interesting that both clinical and virologic data in the ferret study revealed excellent protection from the A/Vietnam challenge despite relatively low HI antibody titres (
As a secondary outcome, the Phase I trial provided a ‘first look’ at the immunogenicity of the H5 VLP vaccine at doses of 5, 10 and 20 µg adjuvanted with alum. Our decision to use an alum-adjuvanted formulation in the Phase I study was based, in part, on lower antibody responses in ferrets exposed to the candidate VLP vaccine without alum (data not shown). Furthermore, several human trials of egg-based split and whole inactivated virion H5N1 vaccines had suggested modest benefits of aluminium adjuvants
Immunogenicity criteria for licensing candidate influenza vaccines vary slightly between jurisdictions but are largely based on serologic responses measured by two assays, Hemagglutination Inhibition (HI) and Single Radial Hemolysis (SRH). Although the HI assay has long been the standard serologic test for influenza, it may be less sensitive for avian strains such as H5N1
Using this approach, the 12 subjects in the 20 µg group met 2/3 of the CHMP criteria based on HI testing (seroconversion >40%, GMI >2.5) and 3/3 based on the SRH assay (seroprotection >70%, seroconversion >40%, GMI >2.5). As noted above however, the SRH assay also detected higher rates of apparent seroconversion and seroprotection in the placebo group than the HI test (
Overall, these results are reassuring for the continued clinical development of this candidate H5 VLP vaccine. Although evidence is accumulating that newer adjuvants can significantly boost H5N1 responses when combined with split-virus antigens
(DOCX)
(DOCX)
The authors are grateful Dr. Noah and her colleagues from the Southern Research Institute for the lethal challenge study in ferrets, to Dierdre McCormick and all the clinical staff at the McGill Vaccine Evaluation Center in Pierrefonds for their support and contribution during the clinical trial, to Gail Clement who supported and monitored the clinical trial and who provided helpful advice for this first-in-human trial, and Medicago's employees who developed and produced this first plant-made VLP to be tested in humans. The authors are also indebted to participating study volunteers. We would also like to address special thanks to Dr Scott Halperin for helpful comments and advice, the review of the clinical protocol and his participation to the independent safety monitoring board. Our thanks also go to Dr Gaston De Serres who was also a member of the independent safety monitoring board. The authors are also thankful to employees from Anapharm (Quebec, Canada) who performed database management and statistical analysis for the phase 1 trial.