Immunogenicity and safety of quadrivalent versus trivalent inactivated influenza vaccine: a randomized, controlled trial in adults

Seasonal influenza affects 5 to 15% of the worldwide population in all age groups every year, and causes high morbidity and mortality, as well as substantial socio-economic disruption [1]. Currently, the influenza viruses in circulation in the human population are A(H1N1), A(H3N2) and two distinct lineages of B-strain: the B/Victoria/2/87 and B/Yamagata/16/88 lineages [2]. These have coexisted since 1983 in the human population [3], and both strains have circulated broadly worldwide since 1988 [1, 2, 4, 5].

The B influenza strain is the major cause of an epidemic every 2 to 4 years, with infection occurring in all age groups [6]. Influenza B causes a substantial number of hospitalizations and deaths and, although seasonal mortality estimates are highest for the H3N2 strain, the impact of B-strains is greater than that of H1N1 [7, 8]. From 1976 to 1999, 16% of the influenza-associated deaths in the US were caused by influenza B and occurred mostly in the elderly population, although they also represent 46% of all influenza related death in children below the age of 5 years [8].

Current seasonal influenza vaccines include the two A-strains (H1N1 and H3N2) but only one B-strain [9]. The prediction for which B-strain to include in the vaccine is not always accurate [6], giving rise to a variable level of mismatch between the B-strain contained in the vaccine and the B-strain causing infection and disease each season. In 6 of the past 11 years, in the US, the main circulating B-virus did not match the strain recommended and included in the vaccine [10]. In the EU, mismatch occurred in 4 out of 8 seasons between the 2003/2004 and 2010/2011 Northern Hemisphere influenza seasons [11]. During seasons with a mismatch of B-strains, protection is reduced in adults and even more so in children [12‐15] because cross-reactivity between the two B-strain lineages is limited [2, 16, 17]. In vaccine-naïve children, live attenuated influenza vaccines have a high vaccine efficacy against well-matched B-viruses, but provide only poor or no cross-protection against strains of the opposite B-lineage [13]. In order to ensure increased vaccine coverage against viruses of both B-lineages, one option may be to include one strain of each lineage in the influenza vaccine, and in doing so potentially offer a better protection against influenza disease [16, 18]. In recognition of the value of a quadrivalent influenza vaccine, the World Health Organization for the first time in February 2012 issued a recommendation for a strain of the alternate B lineage to be included in quadrivalent influenza vaccines made for the 2012–2013 Northern Hemisphere immunization campaigns [19].

Adjuvanted vaccines have been shown to be highly immunogenic and well tolerated in children and adults, providing an important antigen-sparing strategy [20, 21] and cross immunogenicity between clades of the same lineage [21]. Thus, the use of adjuvants may be a way to address possible concerns regarding the manufacturing capability to handle a fourth strain within the short timeframe of seasonal production.

This study was the first conducted by GlaxoSmithKline Vaccines to assess, in adults aged 18 to 60 years, the immunogenicity and safety of an influenza vaccine containing four strains (two A-strains: H1N1 and H3N2 and two B-strains: Victoria and Yamagata lineages), with and without adjuvant, and using 2 different doses of hemagglutinin (15 μg and 5 μg hemagglutinin in the unadjuvanted and adjuvanted formulations, respectively, for each strain).

Trivalent influenza vaccines recommended for immunization against influenza contain the two A-strains (H1N1 and H3N2) but only one B-strain selected each year in the Yamagata-like lineage group or the Victoria-like one [9, 29]. The efficacy and effectiveness of influenza vaccines depend on the degree of similarity between the circulating viruses and those included in the vaccine [12].

The present study was conducted to examine how the addition of another strain would affect the immune response and reactogenicity observed with a TIV. The addition of a fourth strain in the QIV vaccine formulations did not impact the immune response elicited in healthy adults aged 18 to 60 years to the three strains contained in the TIV vaccines, as demonstrated by the statistical non-inferiority of the immunological response elicited by QIV and LD QIV-AS over TIV and LD TIV-AS 21 days after vaccination. Furthermore, the response elicited against the B/Jiangsu strain contained in the QIV formulations, but not in the trivalent vaccines was shown to be superior in terms of levels of antibodies reached 21 days after vaccination. As shown in other studies [30], some levels of cross reactive antibodies were elicited in this adult population against the heterologuous B strains by the TIV formulations. However, levels of antibodies indicative of seroprotection were only reached with the QIV formulations in the present study. The criteria as defined in the CHMP and CBER guidelines for seasonal influenza vaccines in adults were met for the A/Solomon Islands, A/Wisconsin and B/Malaysia strains for all four vaccine groups. Nevertheless, only the quadrivalent formulations were able to elicit the level of immunological response required to fulfill all CHMP and CBER criteria against the B/Jiangsu strain 21 days after vaccination. The micro-neutralization assay results confirmed the data obtained with HI assay. Similar results were previously reported with a quadrivalent live attenuated vaccine containing two influenza subtype A strains (A/H1N1 and A/H3N2) and two B strains (B/Yamagata and B/Victoria). This live attenuated vaccine was shown to be non-inferior to trivalent live attenuated vaccines both in adults and in children [31, 32].

As a clinical impact on vaccine efficacy has been suggested or described in several studies when the circulating strain belonged to the other B lineage [13, 14, 30, 33‐35], the immunological improvement elicited by QIV formulations is likely to translate into a direct clinical benefit. Additionally, the use of quadrivalent vaccines including an influenza B strain representing both lineages could result in substantial cost-savings. Recently, two modeling papers [36, 37] evaluated the potential benefits associated with the use of a quadrivalent vaccine compared with a TIV for the time period 1999–2009 in the US. Based on the clinical burden of influenza illness and viral surveillance data, the Center for Disease Control and Prevention estimated that the use of a QIV rather than a TIV between 1999 and 2009 could have resulted in additional reductions in the number of cases of influenza illnesses, influenza-associated respiratory hospitalizations, and influenza-related respiratory deaths [36]. Furthermore, Lee et al. calculated that the use of QIV during the same time period may have been associated with substantial cost-savings for society and third party payers (median savings: $3.1 billion and $292 million, respectively) [37].

Another objective of the present study was to see whether an adjuvant would trigger cross reactive antibodies against the 2 B lineages. GlaxoSmithKline Vaccines proprietary α-tocopherol, oil-in-water emulsion-based Adjuvant System, AS03, has been shown to elicit cross-reactive response against drifted influenza strains [21]. In the present study adjuvant alone was not sufficient to induce a cross-reactive immune response to the B/Jiangsu strain, which may be expected as the two B-strains are genetically too distinct [3] and so may not be simply considered as drift variants. However, a dose-sparing effect has been demonstrated with the use of 5 μg HA per strain in the adjuvanted formulations (20 μg HA/dose) instead of 15 μg in the non-adjuvanted vaccines (60 μg HA/dose), as illustrated by the immunological non-inferiority of LD QIV-AS vs. QIV. This is in line with other studies using AS03 adjuvant [21, 38].

In terms of reactogenicity, no clinically relevant impact was found with the addition of a fourth strain in the vaccines. The safety profile of the study vaccines was comparable to that seen in other studies conducted in adults [39‐41]. The non adjuvanted QIV formulation reactogenicity profile was generally similar to the TIV vaccine. As expected, the adjuvanted vaccines had a higher reactogenicity profile compared to the non-adjuvanted formulations, being associated with a higher incidence of solicited injection site and general symptoms. Similar observations were reported in previous clinical trials of influenza A/H1N1pdm09 [38, 42‐44] and H5N1 [45] vaccines comparing safety outcomes between adjuvanted and non-adjuvanted vaccines. Local and general adverse events tended to be more frequent with adjuvanted vaccines compared with non-adjuvanted vaccines, in both adults and children. No potentially related serious adverse events were reported during the conduct of the study. However, our study with 105 vaccines per vaccine group was not designed to assess rare events.

Another limitation of the current study is the absence of further follow-up of the participants. Additional information (beyond 21 days post vaccination) on the immune responses to the additional B-strain, including data on the persistence of antibodies against B/Jiangsu strain during the influenza season, would be of interest. Another limitation is that the study enrolled relatively young participants (<60 years of age). Whether our findings can be extended to other populations such as the elderly, with decreased immune response to influenza vaccines [46] remains to be evaluated.

As observed by others [1, 2], this study confirms that adults vaccinated against one B-lineage type have reduced post-vaccination GMTs and a lower percentage of titers ≥40 to viruses from the alternative lineage. Consequently, adults vaccinated with influenza vaccine against one type B-lineage would have reduced protection against infection with the heterologuous lineage if it were circulating. The use of a quadrivalent influenza vaccine, containing B-strains from both circulating lineages, could therefore provide a broader protection against influenza infection, while maintaining a similar safety profile.

This study, conducted at a single site in Czech Republic, provides evidence of the benefit of QIV vaccination to the participants of this study aged between 18 and 60 years. Other clinical studies have been conducted to confirm this finding in a broader adult population, as well as in other populations, including children and elderly persons (registered at http://​www.​ClinicalTrials.​gov: NCT01204671 and NCT01196988).