Background: Protection afforded by inactivated influenza vaccines can theoretically be improved by inducing T-cell responses to conserved internal influenza A antigens. We assessed whether, in an influenza controlled human infection challenge, susceptible individuals receiving a vaccine boosting T-cell responses would exhibit lower viral load and decreased symptoms compared with placebo recipients.

Methods: In this single centre, randomised, double-blind phase 2 study, healthy adult (aged 18-55 years) volunteers with microneutralisation titres of less than 20 to the influenza A(H3N2) challenge strain were enrolled at an SGS quarantine facility in Antwerp, Belgium. Participants were randomly assigned double-blind using a permuted-block list with a 3:2 allocation ratio to receive 0·5 mL intramuscular injections of modified vaccinia Ankara (MVA) expressing H3N2 nucleoprotein (NP) and matrix protein 1 (M1) at 1·5 × 108 plaque forming units (4·3 × 108 50% tissue culture infectious dose [TCID50]; MVA-NP+M1 group) or saline placebo (placebo group). At least 6 weeks later, participants were challenged intranasally with 0·5 mL of a 1 × 106 TCID50/mL dose of influenza A/Belgium/4217/2015 (H3N2). Nasal swabs were collected twice daily from day 2 until day 11 for viral PCR, and symptoms of influenza were recorded from day 2 until day 11. The primary outcome was to determine the efficacy of MVA-NP+M1 vaccine to reduce the degree of nasopharyngeal viral shedding as measured by the cumulative viral area under the curve using a log-transformed quantitative PCR. This study is registered with ClinicalTrials.gov, NCT03883113.

Findings: Between May 2 and Oct 24, 2019, 145 volunteers were enrolled and randomly assigned to the MVA-NP+M1 group (n=87) or the placebo group (n=58). Of these, 118 volunteers entered the challenge period (71 in the MVA-NP+M1 group and 47 in the placebo group) and 117 participants completed the study (71 in the MVA-NP+M1 group and 46 in the placebo group). 78 (54%) of the 145 volunteers were female and 67 (46%) were male. The primary outcome, overall viral load as determined by quantitative PCR, did not show a statistically significant difference between the MVA-NP+M1 (mean 649·7 [95% CI 552·7-746·7) and placebo groups (mean 726·1 [604·0-848·2]; p=0·17). All reported treatment emergent adverse events (TEAEs; 11 in the vaccination phase and 51 in the challenge phase) were grade 1 and 2, except for two grade 3 TEAEs in the placebo group in the challenge phase. A grade 4 second trimester fetal death, considered possibly related to the MVA-NP+M1 vaccination, and an acute psychosis reported in a placebo participant during the challenge phase were reported.

Interpretation: The use of an MVA vaccine to expand CD4+ or CD8+ T cells to conserved influenza A antigens in peripheral blood did not affect nasopharyngeal viral load in an influenza H3N2 challenge model in seronegative, healthy adults.