Safety and immunogenicity of a measles-vectored SARS-CoV-2 vaccine candidate, V591 / TMV-083, in healthy adults: results of a randomized, placebo-controlled Phase I study

Abstract

Background: V591 (TMV-083) is a live recombinant measles vector-based vaccine candidate expressing a pre-fusion stabilized SARS-CoV-2 spike protein.

Methods: We performed a randomized, placebo-controlled Phase I trial with an unblinded dose escalation and a double-blind treatment phase at 2 sites in France and Belgium to evaluate the safety and immunogenicity of V591. Ninety healthy SARS-CoV-2 sero-negative adults (18-55 years of age) were randomized into 3 cohorts, each comprising 24 vaccinees and 6 placebo recipients. Participants received two intramuscular injections of a low dose vaccine (1 × 105 median Tissue Culture Infectious Dose [TCID50]), one or two injections of a high dose vaccine (1 × 106 TCID50), or placebo with a 28 day interval. Safety was assessed by solicited and unsolicited adverse events. Immunogenicity was measured by SARS-CoV-2 spike protein-binding antibodies, neutralizing antibodies, spike-specific T cell responses, and anti-measles antibodies. ClinicalTrials.gov, NCT04497298.

Findings: Between Aug 10 and Oct 13, 2020, 148 volunteers were screened of whom 90 were randomized. V591 showed a good safety profile at both dose levels. No serious adverse events were reported. At least one treatment-related adverse event was reported by 15 (20.8%) participants receiving V591 vs. 6 (33.3%) of participants receiving placebo. Eighty-one percent of participants receiving two injections of V591 developed spike-binding antibodies after the second injection. However, neutralizing antibodies were detectable on day 56 only in 17% of participants receiving the low dose and 61% receiving the high dose (2 injections). Spike-specific T cell responses were not detected. Pre-existing anti-measles immunity had a statistically significant impact on the immune response to V591, which was in contrast to previous results with the measles vector-based chikungunya vaccine.

Interpretation: While V591 was generally well tolerated, the immunogenicity was not sufficient to support further development.

Funding: Themis Bioscience GmbH, a subsidiary of Merck & Co. Inc., Kenilworth, NJ, USA; Coalition for Epidemic Preparedness Innovations (CEPI).

Keywords: COVID-19; SARS-CoV-2; measles vector; vaccine.

Conflict of interest statement

Declaration of interests KR, RT, YT, AG, MM are employees of Themis Bioscience GmbH, a subsidiary of Merck & Co. Inc., Kenilworth, NJ, USA. KR, RT, MM possess stock options of Merck & Co. A patent application including the design of V591 has been filed by the Institut Pasteur and is part of a licensing agreement between the Institut Pasteur and Themis/MSD, NE and CG are inventors. CIC Cochin-Pasteur, SGS, INSERM, Bioaster received payment to conduct the study. All other authors declare no conflict of interest.

Copyright © 2022 Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Kenilworth, NJ USA, The Author(s). Published by Elsevier B.V. All rights reserved.

Figures

Figure 1

Trial Profile. Participants received 2 intramuscular injections on day 0 and day 28 of a low dose vaccine (low dose), 2 injections of a high dose vaccine (high dose [2 inj.]), 1 injection of the high dose vaccine and 1 injection of placebo (high dose [1 inj.]), or 2 injections of placebo (placebo). * all 90 participants were included in the safety analysis set and in the mITT analysis set, ** The first 3 participants in these groups were assigned as sentinel participants to the unblinded dose-escalation phase, ***12 participants were excluded from the PP analysis set: i) 10 participants due to major protocol deviations, ii) 1 participant did not receive the second injection, iii) 1 participant was discontinued further to investigator decision.

Figure 2

Spike-binding IgG antibody response (a) and neutralizing antibody response (b) to SARS-CoV-2 in trial participants per treatment group and study day. Spike-binding IgG was measured by ELISA (Elisa units/mL [ELU/mL]), neutralizing antibodies (50% neutralizing titer [NT50]) were measured using a pseudoneutralization assay. Results of the mITT set are shown. Bars show median and interquartile ranges. The number of individuals with detectable antibody levels, defined as seroconversion (Seroconv.), and the total number of samples per timepoint (Total) are indicated above the plots. Black arrows indicate administration of vaccine, grey arrows administration of placebo. The dotted lines indicate the lower limits of quantification (a: 50.3 ELU/mL, b: 10 NT50). One participant in the placebo group seroconverted after PCR-confirmed SARS-CoV-2 infection on day 21 (no sample on day 28). One participant in the high dose [1 inj.] group was anti-N positive from day 14. One participant in the high dose [2inj.] group was discontinued before day 28.

Figure 3

Anti-Measles IgG levels by participant and timepoint. Anti-measles IgG ELISA results (mITT set) are indicated in international units/liter [IU/L], results below 200 IU/L (dotted line) are considered negative. Solid lines connect samples from the same participant. Black arrows indicate administration of vaccine, grey arrows administration of placebo. One participant in the high dose [2 inj.] group was discontinued before the second injection, the GMT of this group at baseline considering only the 23 participants who completed the study is 595 IU/L.

Figure 4

Impact of pre-existing anti-measles antibody levels on V591 immunogenicity. All participants (mITT set) having received V591 across the different groups (low dose, high dose [2 inj.], high dose [1 inj.]) were stratified into quartiles based on anti-measles antibody levels on day 0. SARS-CoV-2 spike-binding IgG levels on day 28 and day 56 are plotted for each V591 recipient according to quartile allocation. In quartile 1, results of 17 individuals are reported as one participant who was stratified into this quartile was discontinued before day 28. The other three quartiles show results from 18 individuals. Quartiles were compared using an LS means (i.e. GMT) comparison in the MMRM model. Statistically significant differences between quartiles are indicated. Bars indicate the median.