Efficacy and safety of a universal influenza A vaccine (MVA-NP+M1) in adults when given after seasonal quadrivalent influenza vaccine immunisation (FLU009): a phase 2b, randomised, double-blind trial

Summary

Background

In animal, epidemiological, and human challenge studies, a pre-existing T-cell response to internal proteins of influenza A has been associated with improved virological and disease outcomes. The aim of this study was to assess whether inducing additional responses to conserved CD4 and CD8 T-cell antigens provides added benefit to standard influenza vaccination.

Methods

We designed a phase 2b, randomised, placebo-controlled, double-blind trial of a recombinant viral-vectored vaccine (modified vaccinia Ankara expressing virus nucleoprotein and matrix protein 1; MVA-NP+M1), which has been shown to induce both CD4 and CD8 T cells, at eight outpatient clinical trial sites in Australia over two consecutive influenza seasons. We recruited non-immunosuppressed adults (≥18 years) who had received the 2019 quadrivalent influenza vaccine (QIV) vaccine within 28 days before study enrolment and randomisation (day 0). Participants were randomly assigned (1:1) according to a computer-generated random sequence to receive one dose of 1·5 × 10⁸ plaque-forming units of MVA-NP+M1 or saline (placebo) intramuscularly. Randomisation was stratified by age (<65 years or ≥65 years). The patients and trial assessors were masked to treatment assignment. During the subsequent influenza seasons, participants with symptoms related to respiratory illness or influenza-like illness were to attend the clinic within 72 h of symptom onset for two nasal swabs for influenza testing by quantitative RT-PCR. The primary endpoint was the incidence rate of laboratory-confirmed influenza in the intention-to-treat (ITT) population. Safety (solicited adverse events within 7 days and unsolicited adverse events within 28 days after study vaccination, and serious adverse events for the study duration) was assessed in all randomly assigned participants who received at least one vaccination (according to the treatment received). The trial is registered with ClinicalTrials.gov, NCT03880474.

Findings

Between April 2 and June 14, 2019, 2152 adults were randomly allocated and received MVA-NP+M1 (n=1077) or placebo (n=1075), comprising the efficacy (ITT) analysis set. Participants were followed up throughout the 2019 Australia influenza season (May 1 to Oct 15, 2019). 419 (19·5%) of 2152 participants were aged 65 years or older. The incidence of laboratory-confirmed influenza did not differ between the MVA-NP+M1 group (35 of 1077 participants; 3·25% [95% CI 2·31–4·44]) and the placebo group (23 of 1075; 2·14% [1·39–3·14]; Fisher's exact p=0·14). 23 severe solicited local injection site reactions were reported in 13 (0·6%) of 2152 participants, 22 of which were reported in the MVA-NP + M1 group (in 12 [1·1%] participants). 100 severe systemic events were reported in 45 (4·2%) MVA-NP + M1 recipients, and 20 were reported in 14 (1·3%) placebo recipients. Three unsolicited grade 3 events in three participants (two headache and one nausea, all in the MVA-NP+M1 group) were deemed vaccine related. 21 serious adverse events were reported in 18 (1·7%) of 1077 participants in the MVA-NP+M1 group and 25 serious adverse events were reported in 22 (2·0%) of 1075 participants in the placebo group; none were considered vaccine related. The trial was stopped after one season for futility on the recommendation of the data monitoring committee.

Interpretation

MVA-NP+M1 was well tolerated with no vaccine-associated serious adverse events. A vaccine designed to induce moderate T-cell responses to the cross-reactive internal proteins of influenza A did not lead to improved incidence when given within 28 days after standard QIV immunisation. A greater magnitude of T-cell response with a different vaccine or regimen, or localisation in the lungs via alternative delivery, such as intranasal or aerosol, might be successful and require further investigation.

Funding

Vaccitech.

Introduction

Seasonal influenza has a substantial annual global impact, accounting for an estimated 1 billion illnesses worldwide and more than 600 000 life-years lost in the USA alone.¹ Present vaccine approaches have limited efficacy, with the most recent reports estimating protection against illness of 40–60% in adults.2, 3 In situations in which animals are exposed to a newly arisen influenza strain against which they have no protective neutralising antibodies, cross-reactive T cells against conserved internal antigens of influenza have been shown to be associated with limiting viral shedding and reduced duration of symptoms.4, 5, 6, 7, 8, 9 In human influenza challenge studies, protection against symptoms and viral replication has been associated with the magnitude of the influenza-specific CD4 T-cell response.¹⁰ During the 2009 influenza A H1N1 pandemic, individuals with documented influenza in the previous season were less likely to become ill, despite an absence of cross-reactive antibodies.¹¹ In the prospective Flu Watch study and in previous trials, the two influenza proteins most commonly recognised by T cells were virus nucleoprotein (NP) and matrix protein 1 (M1), and the quartile with the highest previous T-cell responses was less likely to develop influenza.¹² NP and M1 have little polymorphism between influenza A strains, which appears to allow strong T-cell cross-reactivity.¹³ Thus, a vaccine against influenza that induces protective T-cell responses against conserved internal antigens might provide lasting immunity against human seasonal influenza and other subtypes with pandemic potential.6, 9, 14, 15, 16

Recombinant viral-vectored vaccines have been used in humans to induce a high magnitude of CD4 and CD8 T-cell responses to a wide range of antigens in large numbers of patients. One such vector is modified vaccinia Ankara (MVA), which has been used to generate strong T-cell responses to HIV-1,17, 18 tuberculosis,¹⁹ hepatitis C virus,²⁰ influenza,¹³ and melanoma.²¹ MVA-NP+M1 is a replication-deficient vector expressing the NP and M1 antigens from an H3N2 strain.¹³ Studies have shown that vaccination with MVA-NP+M1 results in an increase in influenza-specific cross-reactive effector T cells that secrete interferon γ (IFNγ), which were maintained at higher levels for longer than a year.13, 22 These increases were similar to that associated with decreased influenza infection in a natural history study, although different assay methods were used.¹² In a small phase 2a human challenge study, MVA-NP+M1-vaccinated individuals experimentally challenged with live influenza tended to show less severe symptoms and shorter duration of viral shedding than control participants.²³ These and other studies demonstrated the safety of MVA-NP+M1 in both young and older adults when used alone and in combination with licensed inactivated seasonal influenza vaccine.24, 25, 26, 27

On the basis of these findings, we designed a trial to test the hypothesis that MVA-NP+M1 would provide added benefit to seasonal vaccination.