Safety and Immunogenicity of Cell-Based Quadrivalent Influenza Vaccine: A Randomized Trial Open Access

This phase 3, randomized, observer-blind, comparator-controlled, multicenter safety and immunogenicity study of QIVc versus US-licensed QIV was conducted during the Northern Hemisphere 2019–2020 influenza season at 47 centers in the United States. Eligible subjects were randomly assigned to receive either QIVc or QIV in a 2:1 ratio, stratified by months of age (6–23 and 24–47) to ensure that neither age group comprised <30% of the study population. Subjects received 1 or 2 doses of study vaccine 28 days apart, depending on vaccination history, according to CDC recommendations.²⁴  Subjects received 2 doses of study vaccine if at least 2 doses of influenza vaccine had not been previously received either in the same or different influenza seasons or if influenza vaccination history was unknown. A blood sample was collected before vaccination on day 1 and 28 days after last vaccination. The study comprised a treatment period extending from the first vaccination through 28 days after the last vaccination and a safety follow-up period that concluded 180 days after last vaccination.

This study was designed and conducted in accordance with the International Conference on Harmonization Harmonized Tripartite Guidelines for Good Clinical Practice and the ethical principles laid down in the Declaration of Helsinki. Applicable local regulations were followed, and the protocol and informed consent form were approved by the Institutional Review Boards of each study site. Parents or legal guardians of all subjects provided written informed consent. The study was sponsored by Seqirus and registered at ClinicalTrials.gov (NCT04074928).

Eligible subjects were healthy males and females aged 6 through 47 months at enrollment, with no acute severe febrile illness, history of Guillain-Barre Syndrome or other demyelinating diseases, known or suspected immunodeficiency or bleeding disorders, or influenza vaccination within 6 months before study entry (see Supplemental Information for complete list of eligibility criteria).

The study vaccines, QIVc (Flucelvax Quadrivalent, Lot 261303, Seqirus Inc., Holly Springs, NC 27540, USA) or QIV (Afluria Quadrivalent, Lot P100100543 and P100118460 for 0.5 mL dose, and P100114135 for 0.25 ML dose, Seqirus Pty Ltd., Parkville, Victoria, 3052, Australia) were administered intramuscularly in the thigh for subjects 6 through 11 months of age or in the deltoid region for subjects 12 through 47 months of age. QIVc was administered as a 0.5 mL dose. US-licensed QIV was administered as a 0.25 mL dose for subjects aged 6 through 35 months and as a 0.5 mL dose for subjects aged 36 through 47 months, following United States Prescribing Information dosing recommendations. Each 0.5 mL dose of study vaccine contained 15 µg HA from each of the 4 influenza strains selected for the 2019–2020 season (Supplemental Table 4).

The primary objective was to demonstrate that vaccination with QIVc elicits an immune response that is not inferior to that of US-licensed QIV in subjects 6 months through 47 months of age. Serum antibody titers were measured from samples collected on day 1 before vaccination and 28 days after last vaccination (day 29 or day 57, depending on subject’s vaccination history) by hemagglutination inhibition (HI) assay for A/H1N1, B/Yamagata, and B/Victoria strains and by microneutralization (MN) assay for the A/H3N2 strain, using cell-derived target viruses. Pre- and postvaccination geometric mean titers (GMT) were calculated for each vaccine group. Seroconversion (SCR) was defined as the percentage of subjects in each group with either a prevaccination HI (or MN) titer <1:10 and a postvaccination HI (or MN) titer ≥1:40, or a prevaccination HI (or MN) titer ≥1:10 and a ≥4-fold increase in postvaccination HI (or MN) titer. The endpoints of the primary objective were day 29 or 57 postvaccination GMTs and SCRs for each of the 4 vaccine strains. The GMT ratio was defined as the postvaccination HI or MN GMT for QIV divided by the postvaccination HI or MN GMT for QIVc. The SCR difference was defined as SCR QIV minus SCR QIVc.

Secondary immunogenicity objectives were to measure the immune response to each vaccine in all subjects using egg- and cell-derived target viruses in HI assays for A/H1N1, B/Yamagata, and B/Victoria and MN assay for A/H3N2. Endpoints included day 1 and day 29 or 57 GMTs, SCRs, geometric mean ratios (GMR; geometric mean of the within-subject ratio of day 29 or 57 postvaccination GMT divided by the day 1 GMT), and the percentages of subjects with titers ≥1:40 at days 1 and 29 or 57 for each strain. In addition, responses to A/H1N1, B/Yamagata, and B/Victoria were evaluated using MN assays in a subset of subjects randomly selected (nominally 20%) in the per protocol set. Exploratory objectives included evaluation of A/H3N2 antibody responses in all subjects using HI assay and cell-mediated immune (CMI) responses assessed in a separate subset of subjects.

Assessment of safety included the collection of solicited and unsolicited adverse events (AEs). The occurrence of solicited AEs within 7 days after each study vaccination was recorded on a diary card by subject’s parent or legal guardian. Solicited local AEs included injection site induration, erythema, ecchymosis, and tenderness; solicited systemic AEs included change of eating habits, sleepiness, vomiting, diarrhea, irritability, shivering, and fever (≥38°C). All unsolicited AEs occurring between Day 1 and 28 days after last vaccination were recorded, and serious AEs (SAEs), new onset of chronic disease, or AEs leading to study withdrawal were collected for 180 days after last vaccination.

Accounting for a 10% dropout rate and a 2:1 randomization schedule, a sample size of 2418 subjects was estimated to provide at least 90% power to demonstrate noninferiority for all 8 coprimary immunogenicity endpoints using a one-sided α of 0.025 for each comparison. In addition, 84 subjects were to be enrolled to study CMI responses. Thus, it was planned that approximately 2502 subjects would be enrolled into the study. No adjustment for multiple comparisons was made.

All statistical analyses for HI or MN titers were performed on logarithmically transformed (base 10) values. Individual HI or MN titers below the lower limit of quantitation (LLOQ; <10) were set to half of that limit. Consistent with US Food and Drug Administration (FDA) recommendations,²⁵  QIVc was considered noninferior to US-licensed QIV if (1) the upper bound of the two-sided 95% confidence interval (CI) of the GMT ratio for each of the 4 vaccine strains did not exceed 1.5; and (2) if the upper bound of the two-sided 95% CI for the difference in seroconversion for each of the 4 vaccine strains did not exceed 10%. To determine the GMT ratio (adjusted analysis), a general linear model was fitted on the log-transformed postvaccination HI (or MN) titers as the outcome variable and the following terms for covariates: vaccine treatment, log transformed prevaccination HI (or MN) titer, age stratum, gender, influenza vaccination history, and study site. Each of the 4 strains were analyzed separately. Binary data were summarized for each group using unadjusted estimates and are reported together with 2-sided exact 95% CIs (Clopper-Pearson). The 2-sided 95% CI difference between the seroconversion rates of treatment groups was calculated by Miettinen-Nurminen method. Safety data were described by the number and percentage of subjects experiencing AEs within predefined categories of adverse events.

The full analysis set (FAS) was used for analysis of subjects’ demographic and clinical characteristics and included all randomized subjects who received a study vaccination. The primary and secondary immunogenicity analyses were based on the per-protocol set, defined as all subjects who received study vaccination(s) per protocol, had pre- and postvaccination serology titer results, and did not have protocol deviations medically assessed as potentially affecting immunogenicity results. The primary safety analysis was based on all subjects who received at least 1 dose of study vaccine and provided any evaluable follow-up safety data. Analyses were conducted with SAS Version 9.3 (SAS Institute, Inc., Cary, NC).

A total of 2414 healthy subjects 6 months through 47 months of age were enrolled (Fig 1). Twelve subjects did not receive study vaccine and thus, the FAS included 2402 subjects. Of the 2402 subjects, 80 subjects were enrolled for exploratory evaluation of CMI (beyond the scope of this publication), and 2322 subjects were evaluated for immunogenicity. The immunogenicity per-protocol set consisted of 1667 subjects: 1092 and 575 subjects in the QIVc and QIV groups, respectively.

The baseline characteristics of subjects exposed to the study vaccine were well balanced across the vaccine groups (Table 1). Overall, the mean age was 28.1 months, and 37.2% were younger than 24 months of age at enrollment. The majority of subjects were White or Black or African American; 27.5% were of Hispanic ethnicity, and 51.6% were previously vaccinated. Of the enrolled subjects, 86.2% completed the study. The most common reason for discontinuing from the study was lost to follow-up (11.1% of subjects).

The primary objective of the study was achieved as all 8 coprimary endpoints met the noninferiority criteria (Table 2). For each of the 4 vaccine strains, the upper bound of the two-sided 95% CI for the GMT ratio did not exceed 1.5 (Fig 2A), and the upper bound of the two-sided 95% CI of the difference in the SCR between the vaccines did not exceed 10% (Fig 2B).

The secondary immunogenicity endpoints of QIVc and QIV using cell-derived or egg-derived target viruses in HI assays for A/H1N1 and both B strains and MN for A/H3N2 are presented in Supplemental Tables 5 and 6. For both sets of target virus assays, GMTs increased from day 1 to day 29 or 57 for each of the vaccine strains, as reflected by the GMR of post to prevaccination titers. In the assays using egg- derived target viruses, no notable differences in immune responses were observed for QIVc compared with QIV across all endpoints, including GMT, GMR, seropositivity rates (percentage of subjects with titer ≥1:10), percentage of subjects with titer ≥1:40, and SCR. Using cell- derived target viruses, immune responses to A/H3N2 and B/Victoria were not different for the 2 vaccines, whereas for A/H1N1 and B/Yamagata, higher postvaccination GMTs, GMRs, percentage of subjects with titer ≥1:40, and SCRs were observed for QIVc compared with QIV. Immunogenicity results for subjects who received 1 or 2 doses of vaccine are provided in Supplemental Tables 7 and 8, respectively. For a subset of subjects in each vaccine group, antibody titers for A/H1N1, B/Yamagata, and B/Victoria were measured in a MN assay using cell-derived target viruses. As was observed with the HI assays for each vaccine group, greater immune responses were observed for A/H1N1 and B/Yamagata than for B/Victoria strain (Supplemental Table 9).

Exploratory analyses of A/H3N2 antibody responses measured in HI assays were performed in all subjects. Post-hoc testing of the QIVc A/H3N2 strain (A/Indiana/08/2018) to determine HA titers with or without oseltamivir confirmed that hemagglutinin and not neuraminidase mediated agglutination. Using egg-derived target viruses, no differences were seen in the immune responses of QIVc and QIV, whereas postvaccination GMT and SCR were higher in the QIVc vaccine group compared with QIV using cell-derived target viruses in the assay (Supplemental Table 10). The noninferiority criteria defined for the primary endpoints of GMT ratios and SCR differences were also met for the A/H3N2 vaccine strain in the HI assay.

The safety assessment of subjects showed that rates of solicited and unsolicited AEs were similar between the 2 vaccine groups; any solicited AE after any vaccination was reported in 63.7% and 65.9% of subjects receiving QIVc and QIV, respectively, and any unsolicited AE was reported during the treatment period in 26.2% and 25.7%, respectively (Table 3). For both vaccine groups, the majority of solicited AEs were mild to moderate in severity. The most common solicited local AEs were tenderness and erythema at the injection site, and most common solicited systemic AEs were irritability and sleepiness. Fever (≥38°C) was reported for 6.8% and 6.9% of subjects in the QIVc and QIV groups, respectively with temperatures ≥40°C reported in <1% of subjects in either group (Fig 3).

The most frequently reported unsolicited AEs during the treatment period for both vaccine groups were upper respiratory tract infection and pyrexia. Subjects reporting unsolicited AEs assessed as at least possibly related to study vaccine were similar for QIVc (4.4%) and QIV (4.5%), with injection site bruising and irritability being the most common events (Supplemental Table 11). At least 1 medically attended AE were reported by 10.8% of QIVc and 9.3% of QIV recipients (Supplemental Table 12). Adverse events leading to new onset chronic disease were reported by 1.4% of QIVc and 1.6% of QIV recipients with more than 1 subject reporting asthma (0.2%), seasonal allergy (0.1%), ear infection (0.1%), and atopic dermatitis (0.1%) in the QIVc group and cardiac murmur (0.2%) in the QIV group. SAEs were reported in <1% of subjects in each vaccine group and none of the SAEs were assessed as related to study vaccination (Supplemental Table 13). One subject in the QIVc group had a SAE of new onset seizures 17 days after study vaccination and withdrew from the study. Two subjects in the QIVc group had SAEs with a fatal outcome; 1 subject was diagnosed with adenoviral encephalopathy 27 days after receipt of second dose of QIVc, and 1 subject died of injuries sustained in a traffic accident.

This is the first phase 3 safety and immunogenicity study to evaluate cell culture–derived influenza vaccine starting at 6 months of age, the youngest population for whom seasonal influenza vaccine is recommended.^10–12  The study’s success criteria for the primary immunogenicity endpoints of noninferiority of QIVc compared with a US-licensed QIV were achieved for all 4 vaccine strains and were met using either cell- or egg-derived target viruses in the antibody assays, confirming the robustness of the results.

The immunogenicity results of this study complement data from 2 previous clinical studies with QIVc in the pediatric population. Noninferior immunogenicity of QIVc compared with trivalent formulations was demonstrated in subjects 4 through 17 years of age, and in Nolan et al, antibody titers measured in a subset of children 2 through 8 years of age showed immune responses to all 4 vaccine strains.^23,26  The latter study demonstrated protection against influenza in children 2 through 17 years of age with vaccine efficacy consistent across all age subgroups; for subjects 2 through 3 years of age, the overall vaccine efficacy was 62.7% (95% CI 38.1% to 77.5%).²³ 

The immunogenicity results from this study are also consistent with those observed in studies conducted in infants and young children with other quadrivalent inactivated influenza vaccines. Although differences in the years of study conduct, composition of seasonal vaccines, laboratory sites, and assay reagents do not allow for direct comparison of measured antibody titers from these clinical trials, immune responses were evident in subjects 6 through 35 months of age.^27,28  As was observed in our study, postvaccination HI antibody responses to B/Victoria were lower than those to B/Yamagata and 1 or both of the A subtypes; the finding of lower immunogenicity to type B HA antigens has been shown by others.²⁹ 

The findings from the safety assessment of QIVc in children 6 through 47 months of age extend the favorable clinical profile reported for older children.^23,26  There were no clinically meaningful differences in the frequencies of solicited local and systemic AEs and unsolicited AEs for QIVc and US-licensed QIV, and the most frequent unsolicited AEs assessed as at least possibly related to study vaccination are common signs and symptoms of vaccine reactogenicity. Serious adverse events were reported for <1% of children in both vaccine groups and none were assessed as related to study vaccine. The 2 SAE reports of seizures in the QIVc group were not related to receipt of study vaccine, occurring at 17 and 168 days after vaccination; neither were 2 reports of medically attended febrile seizures that were distant in time from vaccination, occurring at 17 and 84 days. These safety data are consistent with findings from phase 3 studies with other quadrivalent inactivated influenza vaccines in this pediatric age range.^27–29 

Strengths of this randomized controlled trial included a large sample size that allowed for a meaningful comparison of safety as well as immunogenicity outcomes. The immunogenicity assessment was extensive with antibody responses evaluated in 2 types of assays (HI and MN) with both cell- and egg-derived target viruses. The study population included subjects with diverse racial and ethnic demographic characteristics, representative of the US population. Limitations of this study included the conduct of the study during a single influenza season (Northern Hemisphere 2019–2020). Although the onset of the COVID-19 pandemic coincided with the follow up period of the study, the impact was limited because safety assessments were ascertained from protocol-stipulated phone calls rather than clinical visits. Finally, the study had a surrogate immunogenicity endpoint and did not directly assess protection against disease.

A cell-based vaccine offers the advantage of avoiding antigenic changes that often occur when influenza vaccine strains are grown in eggs that may result in a mismatch to circulating strains. The current study demonstrated that a cell-based quadrivalent influenza vaccine elicited immune responses in children aged 6 through 47 months that were noninferior to those elicited from a licensed quadrivalent inactivated influenza vaccine. QIVc was well tolerated with a safety profile similar to that of a US-licensed QIV. These results indicate that QIVc is a suitable option for use in children, starting at 6 months of age for prevention of influenza. Future real-world evidence generated from data collected as part of routine patient care, for example, through test-negative case control studies, may be useful for comparative evaluation of relative effectiveness of cell- and egg-based vaccine platforms in the youngest pediatric population.^21,22,30 

We thank all parents and children who participated in the study and all study personnel who conducted the measurements, fieldwork, and data management. C. Gordon Beck and Amanda M. Justice provided editorial and medical writing support funded by Seqirus Ltd.

CMI

cell-mediated immunity

FAS

full analysis set

GMR

geometric mean ratio

GMT

geometric mean titer

HI

hemagglutination inhibition

MN

microneutralization

QIV

quadrivalent inactivated influenza vaccine

QIVc

cell-based quadrivalent inactivated influenza vaccine