A single-center, randomized, double-blind, active-controlled, phase III immuno-bridging clinical trial was conducted, aiming to assess the immunogenicity and safety of Omicron vaccine as compared with the authorized prototype vaccine CoronaVac in Jiangsu Province, China. Healthy adults aged 18 years and older, without a history of or current SARS-CoV-2 infection, were eligible for enrolment. Participants should have previously received either 2 or 3 doses of CoronaVac at least 6 months before (2C or 3C cohorts), for which the time intervals were 21-60 days between the first and second doses and at least 6 months between the second and third doses. Within each cohort, enrolled eligible participants were randomly assigned, in a ratio of 2:1, to receive one booster dose of Omicron vaccine (the trial group) or CoronaVac (the control group), and the first 80 participants in the trial group and the first 40 participants in the control group were assigned to the immunogenicity subgroups. Randomization codes were generated by the statistician using stratified block randomization, according to age stratification (18 to 59 years, or ≥60 years; 4:1 ratio). The randomization code was allocated to each participant in the sequence of enrolment order, and then the participants received the investigational products labeled with the same code. Concealed random group allocations and blinding codes were kept in sealed envelopes, and investigators, participants, and laboratory staff were all blinded to the group allocation. Detailed information on the inclusion/exclusion criteria for participants is listed in the Appendix.

This trial was approved by the ethics committee of Jiangsu Provincial Centre for Disease Control and Prevention. Written informed consent was obtained from all participants before screening, and the trial was conducted in accordance with the principles of the International Council for Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use, the standards of Good Clinical Practice, and Chinese regulatory requirements. The trial was registered in the ClinicalTrials.gov registration database with an identifier of NCT05381350.

Except for the SARS-CoV-2 antigen, Omicron vaccine was following the same manufacturing process and facilities for vaccine production of the licensed prototype vaccine CoronaVac as has been described previously (21–23). In brief, the inactivated Omicron and CoronaVac vaccines were created from African green monkey kidney cells (Vero cells), which had been inoculated with SARS-CoV-2 Omicron BA.1 strain or ancestral strain, respectively. The two SARS-CoV-2 strains were then harvested, inactivated with β-propiolactone, concentrated, purified, and finally adsorbed onto aluminum hydroxide. Both vaccines were prepared in a Good Manufacturing Practice-accredited facility of Sinovac, which was periodically inspected by the National Medical Products Administration (NMPA) committee for compliance. Vaccines were prefilled in ready-to-use syringes and were designed to be administered intramuscularly at a dose of 1200SOU/0.5ml (6μg antigen) for Omicron vaccine and 600SU/0.5ml (3μg antigen) for CoronaVac. In this trial, the schedules of both vaccines were one booster dose for all participants.

The safety objectives were to evaluate the safety and reactogenicity of one booster dose of Omicron vaccine as compared with CoronaVac. Participants were observed in the trial center for at least 30 min after vaccination. For the first 7 days, participants were required to record any local adverse events at the injection site and systemic adverse events on diary cards. Solicited local adverse events included injection site pain, induration, swelling, erythema, rash, and pruritus. Solicited systemic adverse events included fever, acute hypersensitive reaction, diarrhea, nausea, myalgia, headache, cough, and fatigue. Investigators then conducted a face-to-face interview to confirm safety profiles on Day 7. From Day 8 to 28, adverse events were collected by spontaneous records or reports from participants combined with regular visits on Day 14 (only for immunogenicity subgroup participants) and Day 28. Serious adverse events (SAE) and adverse events of special interest (AESI) were recorded throughout the trial until 12 months after vaccination. The reported adverse events were graded according to the NMPA guidelines (24). The causal relationship between adverse events and vaccination was established by investigators.

Safety assessments in this article included solicited local and systemic adverse reactions and unsolicited adverse reactions, as well as SAE and AESI within 28 days after vaccination. The follow-up for SAE and AESI is continuing.

For all participants, serum samples were collected on the day of booster dose (before vaccination) and at 28 days after vaccination. For immunogenicity subgroup participants, extra blood samples at 7 and 14 days after vaccination were collected to explore whether booster vaccination could induce earlier immune response. Previous back-up serum samples at 28 days after two initial doses of CoronaVac were also collected (the historical control group), in adults aged 26-45 years old from the previous lot-to-lot consistency study (25). Neutralizing antibody titers against ancestral strain, Delta, and Omicron BA.1 and BA.5 were then tested using a micro cytopathogenic effect assay, which was done by the National Institute for Food and Drug Control (NIFDC). Details on immunological assessment methods and related testing procedures are described in the Appendix.

There were two primary objectives. One of them was to evaluate the superiority of neutralizing antibody response against Omicron BA.1 at 28 days after Omicron vaccine booster as compared with CoronaVac booster, based on the geometric mean titers’ (GMT) ratio and on the seroconversion rate difference. It is worth noting that the 3C control group was used for either 2C or 3C trial group according to the requirement of NMPA (2/3C+1O vs. 3C+1C). The other one was to evaluate the non-inferiority of neutralizing antibody response against Omicron BA.1 at 28 days after one booster dose of Omicron vaccine, compared with the immune response against ancestral strain at 28 days after two initial doses of CoronaVac, based on GMT ratio (2/3C+1O vs. 2C+0).

The statistical analyses were done separately for two different cohorts (2C and 3C cohorts). Safety profiles were assessed in the safety set (SS) of participants who received one booster dose. Immunological endpoints were assessed in the per-protocol set (PPS), which included all participants who completed one booster dose and had available neutralizing antibody results according to the protocol, as well as all back-up serum samples with available neutralizing antibody results. According to the timepoints postvaccination, PPS was divided into PPS1 (Day 7), PPS2 (Day 14) and PPS3 (Day 28). The demographics of participants in the full analysis set (FAS) were summarized by cohorts and vaccination groups. Pearson chi-squared test or Fisher’s exact test were applied to analyze categorical variables for group comparison, and t-tests or analysis of variance (ANOVA) were used for continuous variables.

For the primary objectives, superiority is considered to be fulfilled when the lower boundary of 95% confidence intervals (CIs) for the GMT ratio is >1, which is equivalent to the GMT difference after log₁₀ transformation greater than 0, and the lower boundary of 95%CI for the seroconversion rate difference is >0%. Noninferiority is considered to be fulfilled when the lower boundary of 95%CI for the GMT ratio is ≥2/3, which is equivalent to the GMT difference after log₁₀ transformation at least -0.176. The immunogenicity was assessed in 1,500 enrolled participants, of which 750 participants were in each cohort (500 and 250 in the trial and control groups, respectively), as well as 250 back-up serum samples in the historical control group. This sample size would allow for a statistical power of 92% to detect superiority in terms of an expected GMT difference of 0.2 (on a log₁₀ scale, with a standard deviation [SD] of 0.55) and a statistical power of 98% to detect superiority in terms of an expected seroconversion rate difference of 20%, with an expected drop-out rate of 20% and a two-sided significance level of 5%. This sample size would also allow for a statistical power of 90% to detect noninferiority in terms of an expected GMT difference of 0 (on a log₁₀ scale, with a SD of 0.60), with an expected drop-out rate of 20% and a two-sided significance level of 5%. The sample size was calculated using PASS 2022.

Observed GMTs and corresponding 95%CIs were calculated based on a standard normal distribution of log-transformed neutralizing antibody titers, of which titers lower than the lower limit of quantitation (LLOQ, 1:4) were presented as half of LLOQ. Analysis of covariance (ANCOVA) was used to compare the difference between groups, which used log-transformed antibody titers after vaccination as the dependent variable, study vaccines and age (18-59 or ≥60 years) as the fixed effects, and log-transformed antibody titers before vaccination as the covariate. GMTs and 95% CIs were then estimated by the geometric least-square mean from the ANCOVA model, and the differences in antibody responses between groups were estimated by the ratio of geometric least-square mean and 95% CIs. Seroconversion refers to GMT changes from <LLOQ to ≥4-fold LLOQ (1:16) after vaccination, or at least 4-fold increases after vaccination if baseline GMT is ≥LLOQ. The corresponding 95%CIs of seroconversions were calculated according to Clopper-Pearson analysis, and seroconversion rate differences and 95%CIs between groups were estimated by the age stratified (18-59 or ≥60 years) Cochran-Mantel-Haenszel test.

Subgroup analyses were performed by age (18-59 and ≥60 years) for neutralizing antibodies. Exploratory analyses were applied to evaluate the antibodies response at 7 and 14 days after vaccination in the immunogenicity subgroup participants. Hypothesis testing was two-sided, and p values of less than 0.05 were considered to be significant. All analyses were conducted with SAS (Version 9.4, SAS Institute Inc., Cary, USA).

Between June 1^st and July 21^st, 2022, a total of 1500 eligible participants were enrolled ( Figure 1 ). Given that one participant was wrongly distributed, 751 (499 and 252) and 749 (500 and 249) participants were finally allocated into 2C and 3C cohorts to get vaccine, respectively, and were included in the SSs. Among these participants who were vaccinated, some of them were not included in the PPSs, due to failed of blood sampling, lost to follow-up, out of window visit, and so on. In 2C cohort, 657 out of 751 participants (87.5%; 437 and 220) were included in the PPS3, and 114 (95.0%; 78 and 36) participants were in the PPS1 and PPS2. In 3C cohort, 730 out of 749 participants (97.5%; 481 and 249) were included in the PPS3, and 114 (95.0%; 76 and 38) and 115 (95.8%; 77 and 38) participants were in the PPS1 and PPS2, respectively. In addition, 250 serum samples were selected from the previous lot-to-lot trial, of which 238 (95.2%) were included in the PPS ( Figure 1 ).

The baseline characteristics of participants were similar between the trial and control groups within 2C or 3C cohort, respectively ( Table 1 ). In 2C cohort, the average ages were 43.5 years in the trial group and 42.7 years in the control group, and 60.44% and 66.14% were male, respectively; the time intervals to the latest dose were similar between groups, with median values of 342 days (range from 183 to 407) and 349 days (208 to 402), respectively. Comparative baseline characteristics were also observed within 3C cohort. Among participants aged 18-59 or ≥60 years, similar baseline characteristics were also observed between groups within each cohort, as shown in Supplementary Table 1 .

Given that the 2C and 3C trial groups shared the same control group (the 3C control group) in the primary immunogenicity analysis, participants’ demographic characteristics across cohorts were also shown ( Table 1 ). There were significant differences in age (p<0.0001) and gender (p=0.0244) distributions between the 2C trial group and the 3C control group, and a longer time interval to the latest dose was observed in the 2C trial group (median: 342 vs. 204 days; p<0.0001). Compared with the historical control group, the average ages in the 2C and 3C trial groups were higher (43.5 and 48.9 vs. 34.8 years).

A total of 183 out of 1499 participants (12.21%) reported adverse reactions within 28 days after booster dose ( Table 2 and Supplementary Table 2 ). Occurrences of adverse reactions were similar between the trial (114 out of 998 participants, 11.42%) and control groups (69 out of 501 participants, 13.77%), and the majority of them were solicited (proportions of 97.37% and 100%, respectively). The incidences of solicited local adverse reactions were 6.71% for Omicron vaccine and 9.38% for CoronaVac, of which injection-site pain was the most reported event (5.61% and 7.98%), followed by induration (1.20% and 2.59%) and pruritus (1.00% and 2.40%). The most frequent solicited systemic adverse reactions were fever, with incidence rates of 3.51% and 4.39%, respectively. Unsolicited adverse reactions occurred in 3 participants (0.30%) in the trial group and 1 participant (0.20%) in the control group. Most of adverse reactions were mild (grades 1) for both vaccine boosters in two groups, with proportions of 92.81% and 97.46%. Two adverse reactions in grade 3 occurred, which were solicited events of fever and injection-site redness in the trial group. No grade 4 events, SAEs and AESIs related to vaccines were observed.

For the age subgroups of 18-59 and ≥60 years, the overall incidence rates of adverse reactions within 28 days were 12.68% (152 out of 1199 participants) and 10.33% (31 out of 300 participants), respectively, with no differences between the trial and control groups ( Supplementary Table 2 ). For the trial group with Omicron vaccine, the frequency of adverse reactions was 7.43% in participants aged ≥60 years, which was relatively lower than that in participants aged 18-59 years (12.44%). In addition, no unsolicited adverse reactions and grade 3 adverse reactions were observed for participants aged ≥60 years.

The immunogenicity results are shown in Table 3 and Supplementary Table 3 . In participants within 3C cohort, the observed GMTs and 95%CIs of BA.1 neutralizing antibody at 28 days after vaccination were 10.36 (9.39, 11.44) for Omicron vaccine and 6.52 (5.73, 7.43) for CoronaVac ( Table 3 ). The estimated GMTs were 9.53 (8.65, 10.50) and 5.81 (5.11, 6.61), respectively, resulting in a GMT ratio of 1.64 (1.42, 1.89). Similarly, a GMT ratio of 1.84 (1.57, 2.16) was observed when comparing the estimated GMT in the 2C trial group [8.87 (7.98, 9.87)] with that in the 3C control group [4.81 (4.20, 5.52)]. In addition, the superiority criteria for seroconversion rate differences between immuno-comparative groups were also fulfilled, which were 17.73% (11.46, 24.00) between the 3C trial and control groups, and 19.04% (12.62, 25.45) between the 2C trial group and the 3C control group. In subgroup analyses by age, as shown in Supplementary Table 3 , the superiority criteria were all met for participants either aged 18-59 or ≥60 years, except for the seroconversion rate difference between the 2C trial group and the 3C control group in participants aged ≥60 years [9.13% (-4.72, 20.92)].

Comparing with the historical control group, not all pre-specified non-inferiority criteria were met ( Table 3 ). Specifically, the estimated GMT ratios for the 2C and 3C trial groups against Omicron BA.1 as compared with the historical control group against ancestral strain were 0.20 (0.17, 0.24) and 0.23 (0.19, 0.26), respectively. When restricting to participants aged 18-59 years, the non-inferiority criteria were still not fulfilled, with GMT ratios of 0.22 (0.19, 0.26) and 0.24 (0.20, 0.28), respectively ( Supplementary Table 3 ).

Detailed observed neutralizing antibody levels against Omicron BA.1, BA.5, Delta and ancestral strains are shown in Figure 2 and Supplementary Figure 1 . Pre-vaccination GMTs in 3C cohort were higher than those in 2C cohort for all strains in all age groups. For the trial groups 28 days after vaccination, GMTs against Omicron BA.1 in 3C cohort were slightly higher than those in 2C cohort, whereas for the control groups GMTs against ancestral strain in 3C cohort were lower than those in 2C cohort in all age groups. Additionally, in either 2C or 3C cohort 28 days after vaccination, GMTs against Omicron BA.5 were relatively lower than those against Omicron BA.1, and GMTs against ancestral strain were much higher than those against Omicron BA.5 and BA.1. The neutralizing antibody levels at 7 and 14 days after vaccination are shown in Supplementary Figure 2 and Supplementary Table 4 . Results showed that GMTs for all strains appeared to sharply increase at the first 7 days and then reached the highest levels at 14 days.