Serological Responses to an Avian Influenza A/H7N9 Vaccine Mixed at the Point-of-Use With MF59 Adjuvant: A Randomized Clinical Trial

Importance  Human infections with avian influenza A/H7N9 have resulted in high morbidity and mortality in China.

Objective  To compare safety and immunogenicity of different doses of influenza A/Shanghai/2/13 (H7N9) vaccine mixed with or without the MF59 adjuvant.

Design, Setting, and Participants  Multicenter, randomized, double-blind, phase 2 trial at 4 US sites enrolled 700 adults aged 19 to 64 years beginning in September 2013; 6-month follow-up was completed in May 2014.

Interventions  The H7N9 inactivated virus vaccine was administered intramuscularly on days 0 and 21 at nominal doses of 3.75, 7.5, 15, or 45 µg of hemagglutinin (actual doses approximately 50% higher) with or without the MF59 adjuvant. A total 99, 100, or 101 participants were randomized to each group (7 groups; N = 700).

Main Outcomes and Measures  Proportions achieving day 42 antibody titer of 40 or greater or seroconversion (a minimum 4-fold increase to titer ≥40) with the hemagglutination inhibition assay; vaccine-related serious adverse events through month 13; and solicited postvaccination symptoms through day 7.

Results  Hemagglutination inhibition antibodies were minimal after participants received an unadjuvanted vaccine. After receiving 2 doses of H7N9 vaccine at a dosage of 3.75 µg plus the MF59 adjuvant, day 42 seroconversion occurred in 58 participants (59%; 95% CI, 48%-68%). The peak seroconversion occurred at day 29 in 62 participants (62%; 95% CI, 52%-72%). The day 42 geometric mean titer was 33.0 (95% CI, 24.7-44.1). Higher antigen doses were not associated with increased response. For the neutralizing antibody assays, after receiving 3.75 µg of H7N9 vaccine plus the MF59 adjuvant, day 42 seroconversion occurred in 81 participants (82%; 95% CI, 73%-89%). The day 42 geometric mean titer was 81.4 (95% CI, 66.6-99.5). There was no statistically significant difference in day 42 hemagglutination inhibition seroconversion after mixing adjuvant with either the first or both 15 µg doses (n = 34 [35%; 95% CI, 25%-45%] vs n = 47 [47%; 95% CI, 37%-58%], respectively; P = .10). Recent receipt of seasonal influenza vaccination and older age were associated with attenuated response. No vaccine-related serious adverse events occurred. Solicited postvaccination symptoms were generally mild with more local symptoms seen in participants who received the adjuvant.

Conclusions and Relevance  Point-of-use mixing and administration of 2 doses of H7N9 vaccine at the lowest tested antigen dose with MF59 adjuvant produced seroconversion in 59% of participants. Although these findings indicate potential value in this approach, the study is limited by the absence of antibody data beyond 42 days and the absence of clinical outcomes.

Trial Registration  clinicaltrials.gov Identifier: NCT01938742

In 2013, human infections with a novel influenza subtype, A/H7N9, were identified in China.¹ Infected patients had live poultry exposures. The virus produced no disease in poultry; however, it caused severe pneumonia in humans, with hospitalization and mortality rates of 67% and 33%, respectively.^1,2 Limited human-to-human spread of A/H7N9 in households was documented³ without sustained transmission. A second wave of infections occurred in winter 2013-2014; through June 27, 2014, a total of 450 laboratory-confirmed cases with 165 deaths (36.6%) were reported to the World Health Organization.⁴

Influenza A/H7N9 has generated particular concern because its hemagglutinin and polymerase basic 1 genes contain Q226L and E627K mutations, respectively, that are associated with adaptation to human airway epithelia and respiratory droplet spread between ferrets.⁵ In ferret experiments, A/Shanghai/2/13 H7N9 virus was able to infect and transmit by direct contact or aerosol.⁶ Based on gene sequencing, H7N9 viruses are expected to be sensitive to neuraminidase inhibitors (oseltamivir and zanamivir) and resistant to adamantines.⁴

The H7N9 virus had not been previously documented in humans, although other H7-containing viruses have produced human disease.^7,8 Efforts to develop vaccines against H7-containing subtypes were hampered by the poor immunogenicity of this hemagglutinin.⁹ An oil-in-water adjuvant, MF59, improves immunogenicity^10,11 and enhances efficacy¹² of inactivated influenza vaccines. Seasonal vaccines containing MF59 were licensed for the elderly in 1997 by the European Union,¹⁰ and subsequently were licensed in more than 30 countries with an excellent safety profile and more than 65 million doses delivered.¹³

The US Department of Health and Human Services Biomedical Advanced Research and Development Authority has procured for the National Pre-pandemic Influenza Vaccine Stockpile both H7N9 antigen and MF59 adjuvant. In this study, we evaluated the safety, tolerability, and immunogenicity of a rapid mix-and-match approach that paired the stockpiled antigen from one manufacturer with stockpiled MF59 from another manufacturer.

The protocol and informed consent forms were approved by the National Institute of Allergy and Infectious Diseases (NIAID) Division of Microbiology and Infectious Diseases, the US Food and Drug Administration, and institutional review boards at the participating NIAID Vaccine and Treatment Evaluation Units (VTEUs). The participants provided written informed consent.

The trial was a randomized, double-blind, active-controlled, multicenter, phase 2 clinical trial conducted in healthy participants aged 19 to 64 years. Full inclusion and exclusion criteria are available in the protocol document (Supplement 1). The study objectives were to assess the safety, reactogenicity, and immunogenicity of point-of-use mixing and immunization with H7N9 antigen and MF59 adjuvant. The main outcome measures were (1) percentages of participants achieving antibody seroconversion (a minimum 4-fold increase to titer ≥40) or titer of 40 or greater in the hemagglutination inhibition assay at day 42; (2) vaccine-related serious adverse events (SAEs) through month 13; and (3) solicited local and systemic postvaccination symptoms through day 7. A hemagglutination inhibition titer of 40 or greater is generally believed to be associated with protection against influenza disease.^14,15

Given the seriousness of the H7N9 threat, the NIAID-funded VTEUs conducted the trial in a time-sensitive manner. Rapid screening and enrollment began September 18, 2013, and was completed October 21, 2013. Participants were recruited at 4 VTEUs in the United States (Emory University, Atlanta, Georgia; Cincinnati Children’s Hospital and Medical Center, Cincinnati, Ohio; University of Iowa, Iowa City; University of Texas Medical Branch, Galveston) and randomized with equal allocation to 1 of 7 groups (Figure 1).

The race and ethnicities of participants were classified according to their self-reported responses using options provided by study investigators. Ethnicity options were Hispanic and non-Hispanic. Race options were American Indian/Alaskan Native, Asian, Hawaiian/Pacific Islander, black/African American, multiracial, unknown, and white. Race and ethnicity were assessed to track diversity of participants and permit ad hoc group–specific assessments of outcomes if needed.

The vaccine was administered on days 0 and 21 and blood for immune response measurements was collected on days 0, 8, 21, 29, and 42. Telephone calls for assessment of SAEs and AEs were scheduled on days 2, 23, 81, 141, 201, and 386. Safety and tolerability were evaluated by collection of SAEs or new chronic medical conditions throughout the 13-month study; unsolicited AEs through day 42; and, using a memory aid, solicited local and systemic symptoms for 7 days after vaccination.

The avian influenza A/Shanghai/2/13 (H7N9) split virus inactivated monovalent vaccine was prepared from influenza viruses propagated in embryonated chicken eggs and was manufactured by Sanofi Pasteur. Licensed influenza vaccine lot release testing specifications were followed with a modification to the potency (hemagglutinin content) test. The traditional single radial immunodiffusion assay could not initially be used to assess the hemagglutinin content because calibrated reagents for it were not available. The hemagglutinin content of bulk vaccine was therefore initially determined by reversed-phase, high-performance liquid chromatography yielding the nominal hemagglutinin doses of 3.75, 7.5, 15, or 45 µg. When the hemagglutinin contents derived from the single radial immunodiffusion concentrations became available later, they were approximately 50% higher (5.75, 11.25, 24.75, and 74.25 µg of hemagglutinin). In this report, we have maintained identification of study groups based on the nominal hemagglutinin doses.

The MF59 adjuvant was manufactured by Novartis Vaccines Inc. As administered, it included 9.75 mg of squalene. The vaccine and the MF59 adjuvant were stored at 4°C. Pharmacists from each site performed mixing under a laminar flow hood using aseptic technique according to USP 797 guidelines. Once mixed, the adjuvanted preparations were stored at room temperature and used within 8 hours. After mixing, the 45-µg dose of H7N9 vaccine without adjuvant preparation was stored at 4°C. For all doses except the 45-µg dose, a final volume of 0.5 mL was administered intramuscularly to the deltoid within 30 minutes of drawing into the syringe. The final volume was 0.75 mL for the 45-µg dose.

The randomization sequence was generated by the trial statistician using SAS version 9.3 (SAS Institute Inc), which was the software used for all analyses, with randomly chosen block sizes of 7 or 14. Upon enrollment, each participant was assigned a randomization number from the electronic data entry system that corresponded to a treatment on a randomization list available to the unblinded vaccine administrator.

Hemagglutination inhibition and neutralizing antibody responses were measured as previously described at a single central laboratory (Southern Research, Birmingham, Alabama).^16,17 Serum samples were tested against the homologous influenza A/Shanghai/2/13 (H7N9) reassortant virus obtained from the US Centers for Disease Control and Prevention. Serum samples were tested in duplicate and the geometric mean titer (GMT) of replicate results was used for analysis. The initial dilution was defined as 1:10 per US Food and Drug Administration recommendations; serum samples without activity were scored as 5.

The sample size of 700 participants (100 per group) was selected to obtain preliminary estimates in a timely manner. Although the study was not designed to test a specific hypothesis, the planned sample size provided 90% power to detect an AE occurring at a rate of at least 3 in 1000, and 80% power to detect an absolute difference of 20%, assuming the reference group had a response rate of 50%. After each follow-up visit was completed by all participants, hemagglutination inhibition antibody was examined in a subset of participants due to the need to inform public health decision makers.

Results of the interim analyses were not made available to the principal investigator, clinical sites, or statisticians. Analyses of primary safety end points were primarily descriptive. Group-specific or pairwise comparisons were considered for the primary immunogenicity end points at day 42 for the intention-to-treat (ITT) analysis subset. For each comparison of interest, a χ² test was used to compare the percentage of participants with seroconversion, whereas a nonparametric test (Kruskal-Wallis or Mann-Whitney) was used for comparisons of titer magnitude. Statistical significance was considered at an α level of .05, without adjustment for multiple comparisons; all tests were 2-sided. Because missing data were minimal, imputation was not performed.

In this study, all participants who achieved a hemagglutination inhibition titer of 40 or greater also met the seroconversion definition; therefore, the results are reported as hemagglutination inhibition seroconversion. For neutralizing antibody, some participants achieved titers of 40 or greater without meeting the seroconversion definition; therefore, neutralizing antibody seroconversion and achievement of titer of 40 or greater are reported separately.

As a prespecified exploratory analysis, a multivariable logistic regression model was fit to examine the relationship between day 42 seroconversion and study group, as well as to assess the effects of the following covariates: age (19-34, 35-49, or 50-64 years), sex, body mass index (calculated as weight in kilograms divided by height in meters squared; <30 vs ≥30), waist circumference, prior receipt of seasonal influenza vaccine for prior, current, neither, or both seasons (eg, none for 2012-2013 or 2013-2014; 2012-2013 only; or 2013-2014 plus 2012-2013 or 2013-2014 but not 2012-2013), and NIAID VTEU unit location. A forward stepwise selection algorithm was used to determine which of these covariates to include in the final model. At each step, independent variables were considered for addition or removal from the model at a significance level of α = .10. Model fit was assessed using residual χ² and Hosmer-Lemeshow tests.

The modified ITT analysis included all data for participants who received at least 1 vaccine dose and had valid hemagglutination inhibition results prior to vaccination and for at least 1 postvaccination visit. The per-protocol analysis included data from participants in the ITT analysis with the following exclusions: (1) all data for participants found to be ineligible at baseline; (2) data from days 29 and 42 if the second vaccination was not received during the recommended vaccination period; (3) data following receipt of nonstudy vaccine or corticosteroids; or (4) data from any study visit substantially outside of the recommended vaccination period. Because minimal differences were observed between the per-protocol and modified ITT analysis subsets, results are presented for the modified ITT analysis only.

There were 767 participants screened and 700 enrolled. A total of 692 participants (98.9%) were included in the day 42 ITT analysis. Reasons for screening failures (n = 67) or exclusion from the ITT population appear in Figure 1.

Demographic and baseline characteristics were balanced across groups with the exception of race (P = .02; Table 1). Nineteen percent of participants were 50 to 64 years of age. Overall, 49% of participants had received influenza vaccination in the prior season (2012-2013) and 24% in both the current (2013-2014) and prior (2012-2013) seasons. Body mass index was classified as overweight in 31% of participants and obese in 29% (Table 1). Forty-four percent of participants had waist circumferences associated with increased health risks.¹⁸

Nearly all participants (99%) had undetectable baseline hemagglutination inhibition antibody. After the priming dose, only 1% to 3% of participants in any group (from groups 1-7) achieved hemagglutination inhibition seroconversion at day 21 (Table 2). However, participants from groups 1 through 4 who received an adjuvanted priming dose had a strong effect of the second dose regardless of antigen dose (Figure 2A). Day 42 hemagglutination inhibition seroconversion was 59%, 58%, and 47% for the 3.75-µg, 7.5-µg, and 15-µg groups that received the M59 adjuvant, respectively (P = .21; Table 2).

Using MF59 with only the priming dose of 15 µg of the vaccine (group 4) was not different for hemagglutination inhibition seroconversion (P = .07) or GMT (P = .13) compared with giving MF59 with both 15-µg doses (group 3) (Table 2 and Figure 2A). However, mixing MF59 with only the 15-µg second dose (group 5) was inferior to group 3 (P < .001 for both seroconversion and GMT) and group 4 (P = .02 for seroconversion and P < .001 for GMT). Figure 2A illustrates a clear rank order: the groups that received adjuvant with both doses (lighter blue curves) or the M59 adjuvant with the first vaccine dose (solid orange curve) followed by the group that received M59 adjuvant with the second vaccine dose (dotted orange curve), and then the groups that did not receive M59 (darker blue curves).

Overall, 18% of participants had detectable baseline neutralizing antibody titers. Postvaccination trends for neutralizing antibody were similar to those observed for hemagglutination inhibition but titers were 2 to 3 times higher (Table 2 and Figure 2B). Neutralizing antibody developed sooner after vaccination than hemagglutination inhibition antibody, but the 2 correlated well once hemagglutination inhibition developed; the Spearman correlation coefficient for day 29 was 0.83 (eFigure 1 in Supplement 2).

Using the stepwise selection procedure, the covariates of age, region, and prior receipt of seasonal influenza vaccine were included, along with study group, in a multivariable logistic regression for day 42 hemagglutination inhibition seroconversion. This analysis found prior receipt of seasonal influenza vaccine in the current (2013-2014), prior (2012-2013), and both seasons was associated with diminished day 42 hemagglutination inhibition seroconversion (prior season only: odds ratio [OR], 0.35 [95% CI, 0.21-0.56], P < .001; current and prior seasons: OR, 0.29 [95% CI, 0.17-0.51], P < .001) (Figure 3 and eTable 1A in Supplement 2). This effect was not simply due to higher vaccination rates in older (immunosenescent) participants because age was adjusted for in the model. Furthermore, the same effect was seen for neutralizing antibody (eTable 1B in Supplement 2) and in a stratified analysis of the subset of 19- to 34-year-old participants.

We also observed lower day 42 hemagglutination inhibition seroconversion rates in older participants (eTable 2 in Supplement 2). For hemagglutination inhibition seroconversion across all groups and compared with participants aged 19 to 34 years, there was an OR of 0.48 (95% CI, 0.31-0.75) for participants aged 35 to 49 years (P = .001) and 0.27 (95% CI, 0.17-0.51) for those aged 50 to 64 years (P < .001). Consistent with the multivariable logistic regression results, hemagglutination inhibition GMT decreased across the 3 age groups (Figure 4).

The multivariable logistic regression model for day 42 hemagglutination inhibition seroconversion suggested that participants in the Midwest had lower seroconversion compared with participants in the South (OR, 0.58 [95% CI, 0.39-0.86], P = .006), but this was not corroborated for day 42 neutralizing antibody seroconversion (OR, 0.81 [95% CI, 0.55-1.19], P = .28). For the analyses reported in this subsection, all crude data and ORs are provided in eTable 1 in Supplement 2.

As of May 2014, 4 SAEs occurred (unstable angina, postoperative urinary retention after cholecystectomy, groin abscess, and ankle fracture) and were determined to be unrelated to the vaccine. One participant developed Hashimoto disease that was determined to be unrelated to the vaccine because a high level of prevaccination antibodies to thyroid peroxidase was observed. Overall, 35% of participants reported 1 or more AEs; most were mild in severity and not related to the vaccine. There were 5 severe AEs; none were related to the vaccine. Month 13 telephone assessments for AEs and SAEs are ongoing and will be reported separately.

Multivariable logistic regression modeling indicated participants who received a vaccine dose with MF59 were more likely to report local reactogenicity (arm reactions) compared with those who received the vaccine without the M59 adjuvant (eg, for first vaccination, 315/400 [78.8%] vs 161/300 [53.7%]; P < .001 for both the first and second vaccination). There was no difference in local reactions based on antigen dosage or systemic symptoms based on adjuvant receipt or antigen dosage. Participants in all study groups were more likely to report symptoms after the first vaccination compared with the second vaccination (eFigure 2 in Supplement 2).

Without the MF59 adjuvant, the highest antigen dosage of 45 µg induced minimal antibody responses against H7N9. However, after mixing with the MF59 adjuvant, the lowest antigen dosage of 3.75 µg achieved a hemagglutination inhibition titer of 40 or greater in 62% of participants at peak (86% for neutralizing antibody). The groups receiving 3.75 µg or 7.5 µg plus MF59 met 1 of the 2 criteria suggested by the US Food and Drug Administration for accelerated approval of pandemic vaccines.¹⁹

The significant antigen dose–sparing effect of MF59 is an important finding, potentially allowing for protection of many more people with limited vaccine. The study did not determine the optimal antigen dose to combine with MF59 because the lowest dose (3.75 µg; actual delivered dose, 5.75 µg of hemagglutinin) produced the maximum antibody seroconversion. This is an area for future research.

The very poor immunogenicity of the H7 antigens evaluated herein and previously,^9,20-23 the requirement for 2 doses, and the improved antibody response with an oil-in-water adjuvant are similar to the experience with H5N1 vaccines.^24-27

Perhaps the most novel finding of this study occurred when the M59 adjuvant was mixed with only the first of 2 doses (group 4), in which the day 42 antibody response was not significantly lower than after 2 adjuvanted doses (group 3). This finding suggests the possibility of adjuvant-sparing regimens and a critical role of MF59 in priming naive lymphocytes to differentiate into a higher quantity and quality of memory B and T cells. Further studies of adjuvant-sparing regimens are needed and, if confirmed, could translate into important clinical implications for the overall vaccine field.

Multivariable logistic regression modeling indicated that recent receipt of seasonal vaccine was associated with diminished antibody responses, suggesting interference from preexisting immunity. Reduced response against new influenza exposures following prior vaccination or infection has been noted previously. Original antigenic sin was described in the 1950s as an exposure to a new influenza strain with resulting preferential induction of antibodies to a previously encountered related strain and diminished response to the new strain.²⁸ The Hoskins paradox,²⁹ described in the 1970s as an impaired protective effect with repeated seasonal vaccinations, was later disputed. Significant negative interactions of prior season vaccination on vaccine efficacy during the 2010-2011 and 2011-2012 influenza seasons have been reported.^30-33 The receipt of prior seasonal vaccine was associated with reduced hemagglutination inhibition response in clinical trials of seasonal influenza vaccines,^34,35 the 2009 pandemic monovalent vaccine,^36,37 and the H5N1 pandemic vaccine.³⁸ Our results and other work³⁹ suggest that adjuvants can be helpful to partially overcome this dampening effect. The mechanisms and clinical relevance of seasonal vaccination interference with H7N9 vaccine response are unclear and deserve further investigation. Cross-reactive antibody binding to the conserved HA2 (stem), cellular responses, or both might play a role. The detection of low levels of neutralizing antibody at baseline, and their more rapid increase relative to hemagglutination inhibition antibodies after vaccination, are likely due to preexisting long-lived plasma cells and memory B cells, respectively, that recognize HA2.

Increasing age was associated with substantially reduced antibody response, a known barrier to protection of elderly persons with seasonal vaccines that has been partially addressed with higher dosage^40,41 and use of adjuvants.^10,42 Excess mortality was associated with obesity during the 2009 pandemic.⁴³ Sixty percent of participants in this study were overweight or obese according to body mass index categories, but we did not observe an effect of increased body mass index or waist circumference on antibody response.

Another primary objective of the study was to assess safety and tolerability of vaccination after field mixing of A/Shanghai/2/13 (H7N9) antigen with MF59. No safety or tolerability concerns were identified, although this study had limited power to detect uncommon events.

This study has several important limitations. The study did not assess antibody longevity beyond day 42. Because the at-risk period during a pandemic could last for several months, this information is needed to determine requirements for additional doses (boosters). The antigenic relatedness of the A/Shanghai/2/13 vaccine to the strain causing a future H7N9 pandemic may be suboptimal for protection. Testing of serum samples from vaccinees against drifted H7N9 strains will be required to assess needs for heterologous boosting. The study excluded adults aged 65 years or older, an age group highly affected by H7N9 influenza in China. Another limitation was the use of research pharmacies to formulate MF59 with the H7N9 antigen; for a future pandemic, it may or may not be feasible for community or national chain pharmacies to perform point-of-use mixing.

While our trial was ongoing, a brief article reported a phase 1 trial of an H7N9 vaccine given with adjuvant that enhanced antibody response.⁴⁴ Just prior to submission of this report, a publication of a phase 1 trial indicated that another adjuvanted H7N9 vaccine was immunogenic.⁴⁵ Both vaccines had poor immunogenicity without an adjuvant. The former study assessed virus-like particles produced with recombinant baculovirus technology, and the adjuvant was saponin-based ISCOMATRIX. The latter study used a novel synthetic seed or cell culture technology that may expedite production of vaccine antigen, and the adjuvant used was MF59.

Point-of-use mixing and administration of 2 doses of H7N9 vaccine at the lowest tested antigen dose with MF59 adjuvant produced seroconversion in 59% of participants. Even though these findings indicate potential value in this approach, the study is limited by absence of antibody data beyond 42 days and absence of clinical outcomes.

Corresponding Author: Mark J. Mulligan, MD, Hope Clinic of the Emory Vaccine Center, Emory University, 500 Irvin Court, Decatur, GA 30030 (mark.mulligan@emory.edu).

Author Contributions: Dr Mulligan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Mulligan, Bernstein, Winokur, Rupp, Hill, Bellamy.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Mulligan, Bernstein, Winokur, Rupp, Anderson, Bellamy.

Critical revision of the manuscript for important intellectual content: Mulligan, Bernstein, Anderson, Rouphael, Dickey, Stapleton, Edupuganti, Spearman, Ince, Noah, Hill.

Statistical analysis: Mulligan, Hill, Bellamy.

Obtained funding: Mulligan, Bernstein, Winokur, Rupp, Noah, Hill.

Administrative, technical, or material support: Mulligan, Bernstein, Winokur, Anderson, Rouphael, Dickey, Stapleton, Ince, Noah.

Study supervision: Mulligan, Bernstein, Winokur, Anderson, Rupp, Noah, Hill.

Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Mulligan reported serving on a data and safety monitoring committee for VaxInnate Inc and receiving a personal fee for this service. No other disclosures were reported.

Funding/Support: This project has been funded in whole or in part with federal funds from the National Institute of Allergy and Infectious Diseases, the National Institutes of Health, and the US Department of Health and Human Services under contracts HHSN272200800005C (Emory University), HHSN272200800006C (Cincinnati Children’s Hospital Medical Center), HHSN272200800008C (University of Iowa, Iowa City), HHSN272200800002C (Baylor University College of Medicine and subcontractor University of Texas Medical Branch), HHSN272201200003I and HHSN27200003 (Battelle and subcontractor Southern Research Inc), and HHSN272200800013C (EMMES Corporation). Additional support was provided by the Georgia Research Alliance, Children’s Healthcare of Atlanta, and the National Center for Advancing Translational Sciences of the National Institutes of Health under awards UL1TR000454 and UL1TR000442. The authors and participating faculty and staff were compensated for their work on this project through the US government contracts to their institutions listed above. The vaccine and adjuvant provided by the US Department of Health and Human Services Biomedical Advanced Research and Development Authority from the National Pre-pandemic Influenza Vaccine Stockpile and were manufactured by Sanofi Pasteur (H7N9 vaccine) and Novartis Vaccines (M59 adjuvant).

Role of the Funders/Sponsors: The funders/sponsors participated in the design and monitoring of the study; and the review and approval of the manuscript. The funders/sponsors did not participate in the collection, management, analysis, and interpretation of the data; the preparation of the manuscript; or the decision to submit the manuscript for publication.

Group Information: The DMID 13-0032 H7N9 Vaccine Study Group includes the authors listed in the byline and the following participating investigators and staff from our institutions who contributed to the conduct of the study and data collection: Karen Mask, MPH, RN, Allison Beck, PA-C, Lilin Lai, MD, Nayoka Rimann, BS, Colleen Kelley, MD, Melinda Ogilvie, BS, Eileen Osinski, BS, Dawn Battle (Hope Clinic of the Emory Vaccine Center); Andres Camacho-Gonzalez, MD, Anita McElroy, PhD, MD, Andi Shane, MD, MPH, MSc, Larry Anderson, MD, Kathy Stephens, RN, MSN, Brooke Hartwell, RN, BS, Teresa Ball, RN, BA, Laila Hussani, BS, Theda Gibson, MS, Melanie Johnson, BS, Bethany Sederdahl, BS, Natasha Mann, BA (Emory Children’s Center); Robert Frenck, MD, Rebecca Brady, MD, Tara Foltz, BA, Amy Cline, BSN, Sarah McCartney, BSN, Margery Huron, BSN (Cincinnati Children’s Hospital Medical Center); Jeffrey Meier, MD, Margo Schilling, MD, Nancy Wagner, RN, Geraldine Dull, BA, Kathy Flanders, ARNP, Dan Zhao, RN, Mary Reidy, RN, Gretchen Cress, RN, Nikki Gerot, BA (University of Iowa); Diane Barrett, MS, Carrie Harrington, RN, Amy McMahan, LVN, Marianne Shafer, BA, Lori Simon, BSN (University of Texas Medical Branch); Barbara Taggart, BS, Valerie Johnson, BS, Donna Bowen, AS, Shixiong Li, MS, Candi Looney, BS, MBA, Megan May, BS, Rachel May, BS, Lawanda Parker, BS, Bridgette Myers, BS, Nertaissa Cochran, BS, Michelle Bell, Logan Haller, PMP (Southern Research Institute); Claire Stablein, BS, Sara Marshall, PhD, Megan McDonough, MPH, Fenhua He, MS, Kuo Guo, MS (EMMES Corporation); and Linda Lambert, PhD, Wendy Buchanan, MS, Valerie Riddle, MD, Suzanne Murray, RN, BSN, Richard Gorman, MD (National Institute of Allergy and Infectious Diseases DMID Respiratory Diseases Branch).

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

Previous Presentation: This work was presented in part at the Second World Health Organization Integrated Meeting on Development and Clinical Trials of Influenza Vaccines that Induce Broadly Protective and Long-Lasting Immune Responses; May 5-7, 2014; Geneva, Switzerland.

Additional Contributions: The investigators at the 4 National Institute of Allergy and Infectious Diseases Vaccine and Treatment Evaluation Units thank the study participants who made this study possible. We are grateful for the manuscript review expertise provided by following colleagues at the US Department of Health and Human Services: Robin Robinson, PhD, Rick Bright, PhD, Michael O’Hara, PhD, Corrina Pavetto, MS, Bai Yeh, MBA, Vittoria Cioce, PhD, James King, MD, and Karen Biscardi, MS. These individuals were not specifically compensated for these contributions.