Abstract
Background. The tocopherol-based oil-in-water emulsion adjuvant system family (AS03) improves antigen sparing with split-virion H5N1 influenza vaccines, representing an important development for pandemic preparedness. In this phase 1/2 randomized, controlled, observer-blinded study in 680 adults, we assessed the immunogenicity and safety of A/Indonesia/5/05 H5N1 (IBCDC-RG2, clade 2.1) prepandemic candidate vaccines produced at 2 separate manufacturing sites.
Methods. Two doses, each of which contained 3.75 μg of hemagglutinin antigen, were given 21 days apart either without adjuvant or with an adjuvant system containing 11.86 mg or 5.93 mg of tocopherol (AS03).
Results. The AS03-adjuvanted A/Indonesia/05/2005 (NIBRG-14) vaccines were significantly more immunogenic than nonadjuvanted vaccine in homologous assays. Neutralizing cross-clade immunogenicity against clades 1, 2.2, and 2.3 was demonstrated at day 42 with all vaccines; at 6 months, seroconversion rates were highest for clade 2.2 (60.7%) and for clade 1 (38.3%). Adjuvantation was associated with increased short-term injection-site reactions (pain) in 80% of participants, with such reactions assessed as being of grade 3 severity for 4.0% of doses. No other safety or reactogenicity concerns were identified over 6 months of follow-up.
Conclusions. Humoral responses against the adjuvanted 3.75-μg hemagglutinin antigen vaccines from both manufacturing sites fulfilled European and US licensure criteria for immunogenicity for influenza vaccines.
Clinical trial registration. ClinicalTrials.gov identifier NCT00510874.
The emergence of swine-origin A/H1N1/09 infection in humans in the spring of 2009 highlighted the rapidity with which new influenza viruses with pandemic potential can spread [1]. Moreover, infection with avian A/H5N1 influenza viruses continues to be reported in humans, and these highly pathogenic viruses remain a serious and complex public health challenge [2, 3].
The ability to meet demand for vaccines to address an impending influenza pandemic will depend heavily sure criteria for immunogenicity [6]. Formulations that reduce the amount of antigen required per dose (ie, dose sparing) are a priority [6–13]. Given the ongoing mutation of H5N1 strains and the expected drift of any novel influenza virus over time, a further challenge is to develop cross-protective vaccines. The use of stockpiled prepandemic vaccines, while vaccines matched to the “ true” pandemic strain are manufactured, has been suggested as a strategy for pandemic preparedness [4, 14].
The use of an adjuvant is a promising antigen-sparing strategy [12, 13, 15]. Aluminum-salt adjuvanted split-virion H5N1 vaccines have not demonstrated convincing improvements in immunogenicity [10]. Oil-in-water emulsions have contributed to improved immune responses to H5N1 antigens at low doses of influenza antigen. A tocopherol-based oil-in-water emulsion adjuvant system (AS03) formulated with inactivated split-virion A/Vietnam/1194/04 H5N1 (NIBRG-14) antigen enables immune responses in adults that fulfill licensure criteria with as little as 2 doses of 3.75 μg of hemagglutinin antigen (HA) given 21 days apart [16–18]. AS03-adjuvanted H5N1 A/Vietnam/1194/04 (NIBRG-14) vaccine (Prepandrix; the GlaxoSmith-Kline group of companies) is currently the only product approved in Europe for prepandemic use against H5N1 influenza in adults ≥18 years of age [19, 20]. Finally, an important component of any pandemic preparedness involving adjuvant is the ability to use the adjuvant with vaccine antigens prepared at multiple manufacturing sites, to ensure supply in case of vaccine shortages.
In the present study, we compared the immunogenicity and safety of 2 doses (3.75 μg of HA) of recombinant A/Indonesia/ 5/05 (IBCDC-RG2) (clade 2.1) split-virion vaccines produced at 2 separate manufacturing sites and combined with 2 different formulations of an oil-in-water emulsion containing tocopherol (AS03) or no adjuvant.
Methods
This randomized, phase 1/2, controlled, observer-blinded study was designed to assess the immunogenicity and safety of 2 doses of A/Indonesia/5/05 H5N1 (IBCDC-RG2) prepandemic vaccine given without adjuvant or adjuvanted with 1 of 2 formulations of AS03. The study was conducted in the United States and Canada, and eligible participants were healthy volunteers or individuals with controlled chronic illness, all of whom were 18–64 years of age inclusive. Women of childbearing age were required to use reliable contraception.
The protocol and consent forms were approved by an independent ethics committee or by local or central institutional review boards. The study was conducted in accordance with Good Clinical Practice, the Declaration of Helsinki, the US Code of Federal Regulations for the Protection of Human Subjects, the Canadian TriCouncil Policy Statement on Ethical Conduct for Research Involving Humans, and all relevant local regulations. All participants provided informed, written consent.
Vaccines and schedule. The vaccine was a split-virion, inactivated A/Indonesia/05/2005 (IBCDC-RG2) clade 2.1 strain (Centers for Disease Control and Prevention, Atlanta) containing 3.75 μg of HA, and it was manufactured by Glaxo-SmithKline (GSK) Biologicals at facilities in Quebec, Canada (H5N1-Quebec), or Dresden, Germany (H5N1-Dresden). Vaccine was given without adjuvant or with AS03 formulations containing 11.86 mg of tocopherol (known as AS03A)or 5.93 mg of tocopherol (known as AS03B). Participants were randomized 1:2:2:2:2 to receive H5N1-Quebec nonadjuvanted, H5N1q adjuvanted with AS03A (H5N1-Quebec AS03A), H5N1-Quebec adjuvanted with AS03B (H5N1-Quebec AS03B), H5N1-Dresden adjuvanted with AS03A (H5N1-Dresden-AS03A), or H5N1-Dresden adjuvanted with AS03B (H5N1-Dresden-AS03B).
A blocked randomization list was generated by GSK Biologicals. A minimization algorithm was used to randomize study participants by site and age (groups 18–40 and 41–64 years of age), with each age stratum limited to ≤60% of participants in any treatment group.
Participants received 1 dose of vaccine intramuscularly on day 0 (in the deltoid muscle of the nondominant arm) and on day 21 (in the opposite deltoid muscle). Vaccines were prepared and administered by unblinded staff who took no further part in the study. Subsequent assessments were performed by blinded observers. Participants attended study sites for screening (days -21 to 0) and on days 0, 7, 21, 28, 42, and 182. Blood samples were collected on days 0, 21, 42, and 182, for serological testing, and on days 0, 7, and 42, for determination of clinical safety parameters. A safety visit was conducted via telephone on day 84.
Objectives and outcome measurements. The primary immunogenicity objective was to assess the superiority of adjuvanted H5N1-Quebec vaccine compared with nonadjuvanted H5N1-Quebec, as demonstrated by vaccine-homologous hemagglutination inhibition (HAI) seroconversion rates (SCRs) and geometric mean titers (GMTs) at 42 days after administration of the primary dose. The primary safety objective was to describe solicited and unsolicited adverse events for up to 84 days after administration of the primary dose.
Secondary objectives were to assess the equivalence of vaccine manufactured at 2 sites, to compare immunogenicity of vaccine constituted with 2 different doses of adjuvant, neutralizing antibody (NA) responses (subset of participants), cross-reactivity against recombinant A/Vietnam/1194/04 (clade 1), the persistence of immune responses 182 days after administration of the primary dose, and the occurrence of serious or medically attended adverse events (AEs) throughout the study.
Immunogenicity. HAI antibody titers against A/Indonesia/ 5/05 (IBCDC-RG2) and recombinant A/Vietnam/1194/2004 were determined using an established HAI assay method, which was modified to use horse rather than avian erythrocytes [21- 24]. All serum samples were tested in duplicate.
HAI end points were based on the criteria of the European Union Committee for Medicinal Products for Human Use (CHMP) and the US Food and Drug Administration Center for Biologics Evaluation and Research (CBER) for the accelerated approval of pandemic influenza vaccines [9, 11, 25]. The end points were SCR, which was defined as the percentage of participants who had pre-and postvaccination titers of <1:10 and ≥1:40, respectively, or who showed a ≥4-fold increase in the postvaccination titer if the titer at baseline was ≥1:10; GMTs and geometric mean fold increase; and the seroprotection rate (SPR), which was defined as the percentage of participants with postvaccination titers of ≥1:40. To fulfill criteria for licensure, the lower limits of the 95% confidence intervals (CIs) for the SCR needed to be ≥40%, and for SPR, they needed to be ≥70% (according to CBER criteria, although point estimates exceeding these values fulfill CHMP criteria); the point estimate of the geometric mean fold increase needed to be >2.5 (a CHMP criterion only).
Neutralization assays specific for A/Indonesia/5/05 (IBCDC-RG2, clade 2.1), recombinant A/Vietnam/1194/2004 (clade 1), recombinant A/Anhui/1/2005 (clade 2.3), and recombinant A/ turkey/Turkey/1/2005 (clade 2.2) were performed according to methods described elsewhere [21, 23]. Serum samples were tested in triplicate. The 50% neutralization titers were calculated according to Reed and Muench [26]. All serological testing was performed at GSK Biologicals. Participants were considered to be seronegative if no neutralizing activity was present at a base dilution of 1:28; samples found to have negative results at this dilution were assigned a titer of 14 for calculations. Because no protective neutralizing correlate is established, we defined vaccine response by the conventional ≥4-fold increase in titer relative to the value noted at baseline, and the SCR was defined as the percentage of participants achieving this increase.
Reactogenicity and safety. Participants were observed at the study site for 30 min after vaccination. Solicited local and general symptoms were recorded by participants using diary cards during the day of vaccination and for 6 subsequent days. Solicited local events (redness, swelling, and pain at the injection site) and solicited general events (fever, headache, fatigue, joint pain, muscle aches, shivering, and sweating) were graded for severity by use of a standard scale, and oral temperatures were self-recorded. All other AEs were graded as grade 1 (ie, easily tolerated) to grade 3 (ie, preventing normal activity). All solicited local events were considered to be vaccine related. Investigators provided causality assessments for solicited general events.
Physical examination of the axillary and supraclavicular lymph nodes was performed at baseline and on days 7, 21, 28, and 42 (and on the latter day only if more than minimal findings were present at day 28 or at the discretion of the investigator). Unsolicited AEs were assessed from day 0 to day 84. Serious AEs (SAEs) and medically attended events were assessed prospectively from day 0 to day 182. All AEs were coded by preferred term and primary system organ class, by use of theMedical Dictionary for Regulatory Activities (MedDRA).
Statistical analyses. The immunogenicity analysis was performed on the according-to-protocol (ATP) cohort, which comprised participants who had complete data for the primary immunogenicity end points and no major protocol deviations. Exclusions from the ATP cohort were determined, and the database was locked before unblinding. The primary safety analyses were performed on the total vaccinated cohort (TVC), which comprised participants who received ≥1 dose of vaccine and for whom any postvaccination data were available. Participants who did not receive the vaccine that they were randomized to receive were excluded from the ATP analyses, and they were analyzed on the basis of the vaccine that they actually received, for TVC analysis.
Sample size calculations were based on previous experience with H5N1 antigens [20]. The target sample size of 675 accounted for the primary evaluation to demonstrate the superiority of adjuvanted versus nonadjuvanted vaccine for the HAI SCR and the GMT ratio; both parameters needed a statistically significant result atP≤.05 (2-sided) for success. The sample size provided 95% power for each test, yielding an overall power of ∼90% for the simultaneous tests. Adjuvant activity of the 2 formulations of AS03 (ASO3A or ASO3B) was analyzed sequentially, and adjuvanted vaccine was considered to be superior to nonadjuvanted vaccine if there was a ≥15% difference in SCRs and a ≥0.3-log10 difference in HAI GMT (GMT ratio, ≥2.0) [25]. The Quebec-and Dresden-manufactured adjuvanted vaccines were considered to be equivalent if the 95% confidence interval of the GMT ratio was between 0.67 and 1.5. Point estimates and 95% confidence intervals were calculated for all immunogenicity end points.
Solicited events were tabulated per participant, including severity scores and the duration of symptoms. Descriptive summaries with 95% confidence intervals included participants with any solicited event, with grade 3 events, and with unsolicited AEs. Fisher's exact tests were used to assess the rates of local and general solicited events.
Results
Enrollment began on 28 July 2007, and the study was completed on 4 July 2008. All 680 enrolled participants received study vaccine (Figure 1). The groups were balanced for demographic and clinical characteristics at baseline (Table 1). A total of 371 participants (54.6%) were 18–40 years of age, and 309 (45.4%) were 41–65 years of age.
Participant flow. Tocopherol-based oil-in-water adjuvant system: AS03A, 11.86 mg of tocopherol; AS03B, 5.93 mg of tocopherol.
Participant flow. Tocopherol-based oil-in-water adjuvant system: AS03A, 11.86 mg of tocopherol; AS03B, 5.93 mg of tocopherol.
Characteristics of Study Participants at Baseline, by Antigen and Adjuvant System
Characteristics of Study Participants at Baseline, by Antigen and Adjuvant System
Immunogenicity
HAI assay. Antibody responses to all vaccines are shown in Table 2 and Figure 2. All participants in the primary analysis were seronegative for A/Indonesia/5/2005 (IBCDC-RG2) antibodies on day 0 (prevaccination). At day 42, both of the adjuvanted groups, but not the nonadjuvanted group, fulfilled the licensure criteria for the SCR, the SPR, and the geometric fold increase. The difference between the H5N1-Quebec nonadjuvanted group and the H5N1-Quebec AS03A and H5N1-Quebec AS03B groups for the vaccine-homologous SCR and GMT ratio demonstrated the superiority of each adjuvanted vaccine. Compared with values at baseline, increased antibody titers in both adjuvanted groups persisted at 182 days, and in the group with the higher adjuvant level (the AS03A group), titers continued to fulfill the licensure criterion for SCR.
Hemagglutinin Inhibition Assay Responses in the According-to-Protocol Immunogenicity Cohort
Hemagglutinin Inhibition Assay Responses in the According-to-Protocol Immunogenicity Cohort
Homologous and heterologous neutralizing antibody responses with A/Indonesia/5/05 H5N1 (IBCDC-RG2) vaccine in randomly selected subsets of the according-to-protocol immunogenicity cohort. Tocopherol-based oil-in-water adjuvant system: AS03A, 11.86 mg of tocopherol; AS03B, 5.93 mg of tocopherol. *A 4-fold increase in titer relative to the prevaccination titer; **, 21 days after the second vaccination. GMT, geometric mean titer.
Homologous and heterologous neutralizing antibody responses with A/Indonesia/5/05 H5N1 (IBCDC-RG2) vaccine in randomly selected subsets of the according-to-protocol immunogenicity cohort. Tocopherol-based oil-in-water adjuvant system: AS03A, 11.86 mg of tocopherol; AS03B, 5.93 mg of tocopherol. *A 4-fold increase in titer relative to the prevaccination titer; **, 21 days after the second vaccination. GMT, geometric mean titer.
Cross-reactive antibody responses against recombinant A/ Vietnam/1194/04 were detected in all groups at day 42 (Table 2). The SCRs were 61.8% (95% CI, 53.3%-69.8%) for H5N1-Quebec AS03A and 58.9% (95% CI, 50.5%-67.0%) for H5N1-Quebec AS03B, which fulfilled the licensure criteria intended for the homologous strain. The difference between the nonadjuvanted group and the AS03A and AS03B groups for the SCR and GMT ratio demonstrated the superiority of each adjuvanted group.
Dresden-and Quebec-manufactured vaccines (pooled data from each manufacturing site, each with 2 adjuvant formulations) were found to have equivalent HAI responses against A/ Indonesia/5/2005 (IBCDC-RG2) (Table 2). Although not a powered protocol-specified analysis, equivalence could be demonstrated for the vaccines incorporating Dresden and Quebec antigens for each adjuvant formulation separately.
NAs. Approximately two-thirds of participants had undetectable NAs against A/Indonesia/5/2005 (IBCDC-RG2) at baseline. The vaccine-homologous SCRs in the AS03A adjuvanted groups at days 42 and 182 were 91.3%-97.9% and 78.3%-91.5%, respectively (Figure 2). Although neutralizing GMTs in the adjuvanted groups decreased between days 42 and 182, they remained approximately 13-to 19-fold greater than values at baseline and 4-fold higher than GMTs in the H5N1-Quebec nonadjuvanted group; results with Dresden-manufactured antigen were similar.
The majority of participants were seropositive for NAs against the clade 1 A/Vietnam/1194/2004 at baseline. At days 42 and 182, the SCR relative to the value noted at baseline in the H5N1-Quebec AS03A group was 53.2% and 38.3%, respectively, and in the H5N1-Quebec AS03B group, it was 45.5% and 13.3%, respectively. Exploratory analyses in the H5N1-Quebec AS03A group demonstrated NA responses against recombinant A/Anhui/1/2005 and recombinant A/turkey/Turkey/ 1/2005, with SCRs of 78.9% (95% CI, 71.2%-85.3%) and 88.8% (95% CI, 82.5%-93.5%) at day 42, respectively. At day 182, NA seroconversion persisted against recombinant A/turkey/ Turkey/1/2005 in 60.7% (95% CI, 52.1%-68.9%) of participants and against recombinant A/Anhui/1/2005 in 2.1% (95% CI, 0.4%-6.1%) of participants.
Safety
Solicited local and general events. Pain at the injection site was the most common solicited event in all groups in the 7 days after vaccination. Pain was less frequent in the nonadjuvanted vaccine group (18 [23.1%] of 78 individuals) than in the adjuvanted groups for vaccine from both manufacturing sites (for AS03A, 133 [87.5%] of 152 individuals; for AS03B, 130 [86.1%] of 151 individuals) (Figure 3). Redness and swelling were much less common than pain in the adjuvanted vaccine groups, and they were absent in the nonadjuvanted group. Reducing adjuvant content did not have a marked effect on local reactogenicity overall, although the incidence of grade 3 local pain decreased. No participant sought medical attention for any solicited local event. The incidence of solicited events associated with Dresden-and Quebec-manufactured vaccine antigens was similar. Solicited general events were more common in adjuvanted versus nonadjuvanted vaccine groups (Figure 3), with muscle aches, headache, fatigue, and joint pain reported. All solicited general events resolved without medical attention. There was a single report of a body temperature ≥39°C in a participant who received H5N1-D/AS03B. Overall, the incidence of solicited events was higher in the adjuvanted vaccine groups than in the nonadjuvanted vaccine group, which was significant for pain at the injection site (P<.001), fatigue (P<.001 s), increased body temperature (P=.009), joint pain (P=.016), and muscle aches (P=.001) (by Fisher's exact tests, with 5×2 group reaction comparison and with Quebec and Dresden vaccine data pooled).
Solicited local and general adverse events in the postvaccination period from day 0 to day 7 with Quebec-and Dresden-manufactured A/Indonesia/5/05 H5N1 (IBCDC-RG2) vaccine in the total vaccinated cohort. CI, confidence interval; nonadj, nonadjuvanted; temp, temperature.
Solicited local and general adverse events in the postvaccination period from day 0 to day 7 with Quebec-and Dresden-manufactured A/Indonesia/5/05 H5N1 (IBCDC-RG2) vaccine in the total vaccinated cohort. CI, confidence interval; nonadj, nonadjuvanted; temp, temperature.
Unsolicited AEs and clinical laboratory assessments. The overall incidence of ≥1 unsolicited AE over 84 days after the first vaccine exposure was 44.9% (35 events in 78 individuals), for nonadjuvanted vaccine, and 50.7% (77 events in 152 individuals) and 47.0% (71 events in 152 individuals), for H5N1-Quebec AS03A and AS03B adjuvanted vaccines, respectively (Table 3). There were 15 SAEs reported in 6 participants in the adjuvanted groups; these were common medical events and were not considered by the investigators to be vaccine related. The majority of SAEs were multiple surgical complications in a single subject. Four participants reported unsolicited symptoms that were grade 3 and needed medical attention (ie, heat exhaustion, fatigue, muscle strain, and nasal congestion). There was no clear association between unsolicited AEs or SAEs and the receipt of adjuvanted vaccines. No participant discontinued the study because of an AE. No association of abnormal clinical laboratory values with any test article or with adjuvanted vaccines overall was noted.
Unsolicited Adverse Events from Day 0 to Day 84 with A/Indonesia/5/05 H5N1 (IBCDC-RG2) Vaccine (Dose 1 and 2 Pooled) in the Total Vaccinated Cohort
Unsolicited Adverse Events from Day 0 to Day 84 with A/Indonesia/5/05 H5N1 (IBCDC-RG2) Vaccine (Dose 1 and 2 Pooled) in the Total Vaccinated Cohort
Unsolicited AEs suggestive of enlargement or tenderness of the lymph nodes occurred somewhat more frequently in recipients of adjuvanted vaccine formulations (Table 4). None of these events were reported to be severe or serious, and all resolved spontaneously. Most adenopathy observed in protocolmandated physical examinations (Table 4) involved axillary nodes ipsilateral to the injection site and was grade 1 (either firm, mobile nodes that were<2.5 cm and nontender or nodes that were tender only with firm pressure). Node enlargement did not increase in frequency or escalate in severity after the second dose, and it was also noted in recipients of nonadjuvanted vaccine.
Evaluation of Lymphadenopathy after Vaccination
Evaluation of Lymphadenopathy after Vaccination
Analgesic use,summarized from participant reports of concomitant medication use, was not signifacantly different between groups. The frequency of any analgesic use in the days after each dose of vaccine for each group was 20.6% (95% CI, 14.6%-27.9%) in the H5N1-Quebec group, 28.2% (95% CI, 23.2%-33.7%) in the H5N1-Quebec AS03A group, 22.7% (95% CI, 18.1%-27.9%) in the H5N1-Quebec AS03B group, 27.9% (95% CI, 22.8%-33.3%) in the H5N1-Dresden-AS03A group, and 26.4% (95% CI, 21.4%-31.8%) in the H5N1-Dresden AS03B group.
Discussion
This study demonstrates that two 3.75-μg HA doses of A/Indonesia/5/2005 (IBCDC-RG2) clade 2.1 inactivated split-virion vaccines from 2 manufacturing facilities, adjuvanted with a tocopherol-based oil-in-water emulsion containing either 11.86 mg of tocopherol or 5.93 mg of tocopherol, induce vaccinehomologous HAI antibody titers that fulfill US and European pandemic influenza vaccine licensure criteria for the SCR, the SPR, and the geometric mean fold increase at 42 days after primary vaccination [9, 11, 25]. At 182 days after primary vaccination, responses continued to fulfill licensure criteria for SCR [9, 11, 25]. Cross-clade antibody responses were demonstrated against recombinant A/Vietnam/1194/2004 (clade 1). The adjuvanted vaccines were superior to nonadjuvanted vaccine, on the basis of the SCR and GMT at 42 days for vaccine homologous and heterologous clade HAI antibody titers.
The rate of homologous neutralizing seroconversion was high at 42 days (range, 91.3%-97.9%), and persisted at 182 days in 91.5% of the H5N1-Quebec AS03A group and 78.3% of the H5N1-Quebec AS03B group. The cross-clade neutralizing response against recombinant A/Vietnam/1194/2004 (clade 1) was demonstrated, with seroconversion rates of 53.2% and 38.3% at 42 and 182 days, respectively, noted with H5N1-Quebec AS03A. Exploratory evaluation of neutralizing immunogenicity against recombinant A/turkey/Turkey/1/2005 (clade 2.2) and recombinant A/Anhui/1/2005 (clade 2.3) demonstrated seroconversion rates of 71%-82% at 42 days. These results support findings of a previous trial that showed strong and persistent immune responses with as low as 3.75-μgHA doses of AS03-adjuvanted A/Vietnam/1194/2004 (NIBRG-14) split-virion vaccine [17, 18].
The highly antigen-sparing property of AS03-adjuvanted H5N1 split-virion vaccines increases the number of doses available from the existing influenza vaccine manufacturing infrastructure, thereby facilitating vaccine availability for mass vaccination programs. In addition, strong cross-clade immunogenicity indicates that the adjuvanted vaccines could be used in the event of a World Health Organization prepandemic alert while waiting for the pandemic strain vaccine to be produced. In such strategies, an initial dose could be used to prime a population, maximizing immune responses to a subsequent dose that is homologous for the emergent pandemic strain [27, 28].
Adjuvantation of H5N1 vaccines does increase local reactions at the site of the injection [7, 10, 29]. Paint at the injection site was reported by ***80% of participants across adjuvanted vaccine groups, compared with 19.2% of participants in the nonadjuvanted group. Redness and swelling were much less common than pain; neither symptom was reported with nonadjuvanted vaccine. Reducing the adjuvant content slightly decreased the incidence of some solicited AEs, but a significant dose-dependent effect was found only for the incidence of grade 3 pain. Fortunately, >95% of local reactions were graded by participants as grade 1 or 2, and reactions did not increase with a second dose. A high participant adherence rate of 95% suggests that reactions were not viewed as excessively burdensome. In a previous study, AS03-adjuvanted H5N1 split-virion vaccine was administered to >5000 healthy adults without raising any safety concerns [30]. Transient local injection-site pain may be viewed as an acceptable burden for a dose-sparing vaccine, which facilitates delivery of vaccine to broad global populations at risk and potentially with improved immunogenicity and efficacy than nonadjuvanted vaccines. The acceptability of this transient adverse event to public health planners and the public may vary with the mortality and morbidity associated with the pandemic threat and the security of vaccine supply.
Although reducing the content of adjuvant had modest benefit in reducing injection-site reactions, this advantage was potentially offset by some reduced durability of immunogenicity observed with the vaccine with lower adjuvant dose. Ad hoc analyses (data not shown) showed that although the effect of reduced adjuvant content on immunogenicity was marginal in the group 18–40 years of age, the antibody responses in persons 41–64 years of age were significantly lower than those noted with the higher dose, both in terms of GMT and SCR. This age-dependent phenomenon requires further investigation; however, based on the available data, formulation of H5N1 split-virion vaccines with 11.86 mg of tocopherol AS03A may provide optimal protection across the age spectrum of an entire adult population.
The limitations of this and other candidate pandemic influenza vaccine trials include the impossibility of assessing efficacy in advance of an outbreak, the fact that current licensure criteria are based on seasonal vaccines, and the lack of a general standardization for safety and reactogenicity reporting among sponsors [9, 11, 25]. In addition, although the CHMP recommends that neutralizing antibodies be measured for candidate pandemic influenza vaccines, no neutralizing titer correlate of protection has been established.
In summary, we showed that a 2-dose series of 3.75 μg of HA H5N1 A/Indonesia/5/2005 (IBCDC-RG2) split-virion vaccine formulated with a tocopherol-based oil-in-water emulsion adjuvant system provided robust, durable immunogenicity against vaccine virus, with antigen made by 2 different processes, as well as clade and subclade cross-reactive responses. Antigen requirement was reduced relative to internal controls and reported nonadjuvanted vaccines [7, 10, 13, 31]. Local transient reactogenicity was noted with the adjuvanted vaccines, but no significant safety concerns were identified.
Acknowledgments
We thank the National Institute for Biological Standards and Control (Potters Bar, United Kingdom), for providing the vaccine virus strain for the assays and reference standards, and also the Centers for Disease Control and Prevention (Atlanta, Georgia), for supplying the recombinant A/Indonesia/5/2005 strain. In addition, the guidance and support of the US Department of Health and Human Services were greatly appreciated.
We are indebted to the participating study volunteers, clinicians, nurses, and laboratory technicians at the study site and the sponsor's project staff, for their support and contributions throughout the study. We are grateful to the clinical and serological laboratory teams of GlaxoSmithKline Biologicals for their contribution to these studies. Finally, we thank Annick Moon (Independent, Oxford, United Kingdom), who provided medical writing services for the development of the manuscript, and Isabelle Camby (GlaxoSmithKline Biologicals), who provided editorial assistance and manuscript coordination.
This manuscript was developed according to the publication practices recommended by the International Committee of Medical Journal Editors. Annick Moon developed the manuscript according to the recommendations, documentation, and outline provided by the lead authors (ie, J.M.L. and L.F.). All authors critically reviewed the proposed drafts of the manuscript, and their comments were taken into account and incorporated by the corresponding author. All authors approved the content of the final version of the manuscript before it was submitted by the corresponding author. Isabelle Camby provided support to coordinate the circulation of the manuscript to all co-authors, to collect comments received from the co-authors, and to make sure that recommendations of the International Committee of Medical Journal Editors were fulfilled. All the steps of this process occurred and were tracked and documented via the Datavision platform.
