Differences in the immune response elicited by two immunization schedules with an inactivated SARS-CoV-2 vaccine in a randomized phase 3 clinical trial

Abstract

Background: The development of vaccines to control the coronavirus disease 2019 (COVID-19) pandemic progression is a worldwide priority. CoronaVac is an inactivated severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccine approved for emergency use with robust efficacy and immunogenicity data reported in trials in China, Brazil, Indonesia, Turkey, and Chile.

Methods: This study is a randomized, multicenter, and controlled phase 3 trial in healthy Chilean adults aged ≥18 years. Volunteers received two doses of CoronaVac separated by 2 (0-14 schedule) or 4 weeks (0-28 schedule); 2302 volunteers were enrolled, 440 were part of the immunogenicity arm, and blood samples were obtained at different times. Samples from a single center are reported. Humoral immune responses were evaluated by measuring the neutralizing capacities of circulating antibodies. Cellular immune responses were assessed by ELISPOT and flow cytometry. Correlation matrixes were performed to evaluate correlations in the data measured.

Results: Both schedules exhibited robust neutralizing capacities with the response induced by the 0-28 schedule being better. No differences were found in the concentration of antibodies against the virus and different variants of concern (VOCs) between schedules. Stimulation of peripheral blood mononuclear cells (PBMCs) with Mega pools of Peptides (MPs) induced the secretion of interferon (IFN)-γ and the expression of activation induced markers in CD4+ T cells for both schedules. Correlation matrixes showed strong correlations between neutralizing antibodies and IFN-γ secretion.

Conclusions: Immunization with CoronaVac in Chilean adults promotes robust cellular and humoral immune responses. The 0-28 schedule induced a stronger humoral immune response than the 0-14 schedule.

Funding: Ministry of Health, Government of Chile, Confederation of Production and Commerce & Millennium Institute on Immunology and Immunotherapy, Chile.

Clinical trial number: NCT04651790.

Keywords: COVID-19; CoronaVac; SARS-CoV-2; human; immunization; immunology; inflammation; medicine; phase 3 clinical trial; vaccines.

Conflict of interest statement

NG, GP, BS, FM, JS, LD, LG, DR, MR, RB, YV, DM, OV, CA, GH, CI, MU, MN, ÁR, RF, JF, JM, ER, AG, MA, FV, JG, MJ, PG, AK No competing interests declared, RS has received funding from ANID - ICM, ICN 2021_045. The author has no other competing interests to declare, DW has received funding support from the NIH under contract number 75N93019C00065. The La Jolla Institute for Immunology (LJI) has filed for patent protection for various aspects of T cell epitope and vaccine design work. The author has no other competing interests to declare, AG The La Jolla Institute for Immunology (LJI) has filed for patent protection for various aspects of T cell epitope and vaccine design work. The author has no other competing interests to declare, AS is a consultant for Gritstone Bio, Flow Pharma, Arcturus, Immunoscape, CellCarta, Moderna, AstraZeneca, Fortress, Repertoire, Gilead, Gerson Lehrman Group, RiverVest, MedaCorp, Guggenheim, OxfordImmunotech, and Avalia. The author has received funding support from the NIH under contract 75N93021C00016 and 75N93019C00065. The La Jolla Institute for Immunology (LJI) has filed for patent protection for various aspects of T cell epitope and vaccine design work. The author has no other competing interests to declare, GZ, WM is a SINOVAC employee and contributed to the conceptualization of the study (clinical protocol and eCRF design), DG acts as the Executive Director of the clinical trials PedCoronaVac03CL clinical study (ClinicalTrials.gov NCT04992260) and CoronaVac03CL (ClinicalTrials.govNCT04651790) (funds to the institution), and receives research support from Millennium Institute on Immunology and Immunotherapy. The author received funding from Agencia Nacional de Investigación y Desarrollo, Fondo de Fomento al Desarrollo Científico y tecnológico. The author has no other competing interests to declare, KA acts as the Scientific Director of clinical trials PedCoronaVac03CL clinical study (ClinicalTrials.gov NCT04992260) and CoronaVac03CL (ClinicalTrials.govNCT04651790) (funds to the institution), and receives research support from Millennium Institute on Immunology and Immunotherapy. The author has received funding from Agencia Nacional de Invetsigación y Desarrollo, Fondo de Fomento al Desarrollo Cientí fico y tecnológico ID20I10082. The author has no other competing interests to declare, SB acts as the General Director of clinical trials PedCoronaVac03CL clinical study (ClinicalTrials.gov NCT04992260) and CoronaVac03CL (ClinicalTrials.govNCT04651790). The author has received funding from Agencia Nacional de Investigació n y Desarrollo (ANID) - Millennium Science Initiative Program - ICN09_016 / ICN 2021_045: Millennium Institute on Immunology and Immunotherapy (ICN09_016 / ICN 2021_045; former P09/016-F) and Agencia Nacional de Investigación y Desarrollo [FONDECYT grant numbers 1190830]. The author has no other competing interests to declare

© 2022, Gálvez, Pacheco, Schultz et al.

Figures

Figure 1.. Circulating neutralizing antibodies against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) measured by surrogate virus neutralization test (sVNT) and conventional virus neutralization test (cVNT) for the Ancestral strain in immunized volunteers.

Neutralizing antibody titers were evaluated with an sVNT, which quantifies the interaction between S1-RBD and human ACE2 (hACE2) pre-coated on ELISA plates (A,C) and with a cVNT, which quantifies the cytopathic effect (CPE) induced in Vero cells as plaques formation (B, D). n=372 volunteers for cVNT (Ancestral) and n=130 volunteers for sVNT (for both schedules). Data is represented as the reciprocal antibody titer of neutralizing antibody versus the different times evaluated. Numbers above the bars show either the arbitrary international units (IU) (A) or the geometric mean titer (GMT) (B), and the error bars indicate the 95% CI. Seropositivity rates are also displayed (C, D). Data from IU and GMT values were analyzed by a two-tailed unpaired t-test of the base 2 logarithms of data to compare immunization schedules. Data from seropositivity rates were analyzed by a two-tailed Fisher’s exact test. Numbers above each bracket represent calculated p values comparing both immunization schedules. Statistical significance was set at p<0.05 and highlighted numbers indicate statistical significance.

Figure 1—figure supplement 1.. Study design for this phase 3 clinical trial comparing two different immunization schedules as of August 2021.

This study aims to characterize the differential immune response elicited by two immunization schedules with CoronaVac, with each dose separated by either 2 or 4 weeks.

Figure 1—figure supplement 2.. Geometric mean titer (GMT) values and seropositivity rates of circulating neutralizing antibodies against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) measured through pseudotyped virus neutralization test (pVNT) (ID80).

Neutralizing antibody titers were evaluated with a pVNT. n=94 volunteers for both schedules. In (A), data is represented as the reciprocal antibody titer of neutralizing antibody versus the different times evaluated. Numbers above the bars show the geometric mean titer (GMT), and the error bars indicate the 95% CI. A two-tailed one-way ANOVA for repeated measures was performed over the Log2 of antibody titers, followed by Bonferroni’s multiple comparisons test to compare geometric mean units (GMUs). (B) Data from seroconversion rates were analyzed by a two-tailed Fisher’s exact test. Values above the lines indicate p values. Significant p values are shown in red. Statistical significance was set at p<0.05.

Figure 1—figure supplement 3.. Circulating neutralizing antibodies against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) measured through surrogate virus neutralization test (sVNT), conventional virus neutralization test (cVNT), and pseudotyped virus neutralization test (pVNT) (ID80) in volunteers immunized with CoronaVac aged 18–59 and ≥60 years.

Neutralization assays were performed with an sVNT, which quantifies the interaction between S1-RBD and human ACE2 (hACE2) pre-coated on ELISA plates (A); with a cVNT, which quantifies the cytopathic effect induce in Vero cells as plaques formation for the Ancestral and D614G strains (B); and with a pVNT (C). Results were obtained from volunteers aged 18–59 years (blue circles for the 0–14 schedule and blue triangles for the 0–28 schedule) and ≥60 years (red circles for the 0–14 schedule and red triangles for the 0–28 schedule) before immunization (0 days), 2 weeks after second dose, and 4 weeks after the second dose. Data is represented as the reciprocal antibody titer versus time after the second dose. Numbers above the bars show the international units (IU) or geometric mean titer (GMT), and the error bars indicate the 95% CI. A two-tailed one-way ANOVA for repeated measures was performed over the Log2 of antibody titers, followed by Bonferroni’s multiple comparisons test to compare between schedules and age groups in each schedule. Values above the lines indicate p values. Significant p values are shown in red. Statistical significance was set at p<0.05.

Figure 2.. Total anti-S1 and anti-RBD antibodies circulating in immunized volunteers.

Concentrations of IgG antibodies after two doses of CoronaVac were evaluated for immunized volunteers before the first (preimmune) and second dose and 2 and 4 weeks after the second. Specific IgG against the S1-RBD and the spike protein of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were measured. n=162 volunteers for ELISA assays (A) and n=44 volunteers for meso-scale discovery (MSD) assays (B). Data are expressed as the reciprocal antibody titer in arbitrary WHO international unit versus the different times evaluated. Error bars indicate the 95% CI. Spots represent individual values of each volunteer, with the numbers above each set of spots showing the geometric mean unit (GMU) estimates. Data were analyzed using a two-tailed unpaired t-test of the Log2 of data to compare immunization schedules. Numbers above each bracket represent calculated p values comparing both immunization schedules. Statistical significance was set at p<0.05 . Dotted line on A is showing a value of 4, which is the threshold established for the seroconversion rate of each volunteer. Therefore, every spot over the dotted line represents volunteers that were considered positive for seroconversion relative to their preimmune sample.

Figure 3.. Similar levels of interferon (IFN)-γ-secreting cells and expression of activation-induced markers (AIM) on T cells are found upon stimulation with Mega Pools of peptides derived from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) for both immunization schedules with CoronaVac.

Total number of IFN-γ+ spot forming cells (SFCs) were determined by ELISPOT. Data were obtained upon stimulation of peripheral blood mononuclear cells (PBMCs) for 48 hr with MP-S and -R (A) or with MP-CD8A and -B (B). The percentage of activated CD4+ (AIM+ [OX40+, CD137+]) and CD8+ (AIM+ [CD69+, CD137+]) T cells was determined by flow cytometry, upon stimulation for 24 hr with MP-S and -R (C), or with MP-CD8A and -B (D) in samples obtained before the first (preimmune) and second dose, and 2 and 4 weeks after the second dose. n=124 samples stimulated with MP-S and -R for ELISPOT (A). n=117 samples stimulated with MP-CD8A and -B for ELISPOT (B). n=116 stimulated with MP-S and -R for flow cytometry (C). n=110 samples stimulated with MP-CD8A and -B for flow cytometry (D) (for both schedules). Numbers above the bars show the mean and the error bars correspond to the 95% CI. Data were analyzed by a mixed-effect two-way ANOVA, followed by a Bonferroni’s post hoc test to compare immunization schedules. Numbers above each bracket represent calculated p values comparing both immunization schedules. Statistical significance was set at p<0.05 and highlighted numbers indicate statistical significance.

Figure 3—figure supplement 1.. Total number of interferon (IFN)-γ+ spot forming cells (SFCs) induced upon stimulation with Mega Pools (MPs) of peptides derived from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) proteome in volunteers immunized with CoronaVac aged 18–59 and ≥60 years.

Changes in the secretion of IFN-γ were measured, determined as the number of SFCs per 3×105 peripheral blood mononuclear cell (PBMC). Data was obtained upon stimulation of PBMC with MP-S+R (A), and upon stimulation of PBMC with MP-CD8A+B (B), for 48 hr in samples obtained before immunization, 2 weeks after the second dose, and 4 weeks after the second dose. All data were normalized for DMSO unspecific stimulation. Data are presented as geometric means and error bars represent the 95% CI. A two-tailed one-way ANOVA for repeated measures was performed over the Log10 of SFCs, followed by Bonferroni’s multiple comparisons test to compare between schedules. Values above the lines indicate p values. Significant p values are shown in red. Statistical significance was set at p<0.05.

Figure 3—figure supplement 2.. Percentage of activation-induced markers (AIM+) T cells induced upon stimulation with Mega Pools (MPs) of peptides derived from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) proteome in volunteers immunized with CoronaVac aged 18–59 and ≥60 years.

The percentage of activated CD4+ (AIM+ [OX40+, CD137+]) and CD8+ (AIM+ [CD69+, CD137+]) T cells was determined by flow cytometry, upon stimulation for 24 hr with MP-S and -R (A–B), and with MP-CD8A and -B (C–D) in samples obtained before the first (preimmune) and second dose, and 2 and 4 weeks after the second dose. Data are presented as means and error bars represent the 95% CI. A two-tailed one-way ANOVA for repeated measures was performed, followed by Bonferroni’s multiple comparisons test to compare between schedules. Values above the lines indicate p values. Significant p values are shown in red. Statistical significance was set at p<0.05.

Figure 3—figure supplement 3.. Immunization with CoronaVac in a 0–28 schedule does not induce major IL-4 responses in peripheral blood mononuclear cells (PBMCs).

Changes in the secretion of IL-4 were measured, determined as the number of spot forming cells (SFCs) per 3×105 PBMC. Data was obtained upon stimulation of PBMC with MP-S+R (A), and upon stimulation of PBMC with MP-CD8A+B (B), for 48 hr in samples obtained before immunization, 2 weeks after the second dose, and 4 weeks after the second dose. All data were normalized for DMSO unspecific stimulation. Data are presented as geometric means and error bars represent the 95% CI. A two-tailed one-way ANOVA for repeated measures was performed over the Log10 of SFCs, followed by Bonferroni’s multiple comparisons test to compare between schedules. Values above the lines indicate p values. Significant p values are shown in red. Statistical significance was set at p<0.05.

Figure 4.. Antibodies against the spike (S) and receptor-binding domain (RBD) from variant of concern (VOC) of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are similar between schedules, while breadth index varies between schedules.

Antibodies concentrations against the S (A) and the RBD (B) of different VOCs of SARS were evaluated through meso-scale discovery (MSD). n=44 volunteers for the 0–14 schedule. n=40 volunteers for the 0–28 schedule. Samples evaluated were obtained at 4 weeks after the second dose. Data is represented as the reciprocal antibody titer of neutralizing antibody versus the different VOCs evaluated. With these values, a breadth index was calculated for each VOC for anti-S (C) and anti-RBD (D) antibodies. Numbers above the bars show either the international units (IU) (A, B) or the breadth index (C, D), and the error bars indicate the 95% CI. Data were analyzed by a mixed-effect two-way ANOVA, followed by a Bonferroni’s post hoc test to compare immunization schedules. Numbers above each bracket represent calculated p values comparing both immunization schedules. Statistical significance was set at p<0.05 and highlighted numbers indicate statistical significance.

Figure 4—figure supplement 1.. Antibodies against the spike (S) and receptor-binding domain (RBD) from variant of concern (VOC) of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and their breadth indexes are reduced relative to the Ancestral strain, except in the receptor-binding domain (RBD)-related parameters for the Delta strain.

Antibodies concentrations against the S (A) and the RBD (B) of different VOCs of SARS were evaluated through meso-scale discovery (MSD). Results were obtained from 44 volunteers for the 0–14 schedule and 40 volunteers for the 0–28 schedule, from samples obtained at 4 weeks after the second dose. Data is represented as the reciprocal antibody titer of neutralizing antibody versus the different VOCs evaluated. With these values, a breadth index was calculated for each VOC for anti-S (C) and anti-RBD (D) antibodies. Numbers above the bars show either the international units (IU) (A, B) or the breadth index (C, D), and the error bars indicate the 95% CI. Data were analyzed by a repeated measures two-way ANOVA, followed by a Bonferroni’s post hoc test to compare the level of antibodies against each VOC relative to the Ancestral strain. Numbers above each bracket represent calculated p values comparing both immunization schedules. Statistical significance was set at p<0.05 and red numbers indicate statistical significance.

Figure 4—figure supplement 2.. Antibodies against the spike (S) and receptor-binding domain (RBD) from variant of concern (VOC) of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are similar between schedules, while breadth index varies between schedules for the 18–59 years age group.

Antibodies concentrations against the S (A) and the RBD (B) of different VOCs of SARS were evaluated through meso-scale discovery (MSD). Results were obtained from 20 volunteers for the 0–14 schedule and 19 volunteers for the 0–28 schedule, from samples obtained at 4 weeks after the second dose. Data is represented as the reciprocal antibody titer of neutralizing antibody versus the different VOCs evaluated. With these values, a breadth index was calculated for each VOC for anti-S (C) and anti-RBD (D) antibodies. Numbers above the bars show either the international units (IU) (A, B) or the breadh index (C, D), and the error bars indicate the 95% CI. Data were analyzed by a mixed-effect two-way ANOVA, followed by a Bonferroni’s post hoc test to compare immunization schedules. Numbers above each bracket represent calculated p values comparing both immunization schedules. Statistical significance was set at p<0.05 and red numbers indicate statistical significance.

Figure 4—figure supplement 3.. Antibodies against the spike (S) from variant of concern (VOC) of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and breadth indexes vary between schedules for the >60 years age group, while these parameters for anti-receptor-binding domain (RBD) antibodies remain similar.

Antibodies concentrations against the S (A) and the RBD (B) of different VOCs of SARS were evaluated through meso-scale discovery (MSD). Results were obtained from 24 volunteers for the 0–14 schedule and 21 volunteers for the 0–28 schedule, from samples obtained at 4 weeks after the second dose. Data is represented as the reciprocal antibody titer of neutralizing antibody versus the different VOCs evaluated. With these values, a breadth index was calculated for each VOC for anti-S (C) and anti-RBD (D) antibodies. Numbers above the bars show either the international units (IU) (A, B) or the breadh index (C, D), and the error bars indicate the 95% CI. Data were analyzed by a mixed-effect two-way ANOVA, followed by a Bonferroni’s post hoc test to compare immunization schedules. Numbers above each bracket represent calculated p values comparing both immunization schedules. Statistical significance was set at p<0.05 and red numbers indicate statistical significance.

Figure 4—figure supplement 4.. Neutralizing antibodies against the receptor-binding domain (RBD) from variant of concern (VOC) of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are similar between schedules, while breadth index varies between schedules.

Neutralization assays were performed with a conventional virus neutralization test (cVNT), which quantifies the cytopathic effect induce in Vero cells as plaques formation for the strains Alpha, Gamma, and Delta (A) (n=31 volunteers for the 0–14 schedule, n=15 volunteers for the 0–28 schedule); and with cVNT for the Omicron strain (B) (n=22 volunteers for the 0–14 schedule, n=20 volunteers for the 0–28 schedule). Samples evaluated were obtained at 4 weeks after the second dose. Data is represented as the reciprocal antibody titer of neutralizing antibody versus the different VOCs evaluated. With these values, a breadth index was calculated for each VOC for each technique (C, D). Numbers above the bars show either the reciprocal antibody titers (A, B) or the breadth index (C, D), and the error bars indicate the 95% CI. Data were analyzed with a repeated measures two-way ANOVA, followed by a Bonferroni’s post hoc test to compare between VOCs. Numbers above each bracket represent calculated p values comparing both immunization schedules. Statistical significance was set at p<0.05. Data analyses for panels A and C were performed with data previously published (Melo-González et al., 2021).

Figure 5.. Multivariate analyses show correlated humoral and cellular immune responses.

Pearson correlation matrixes were generated independently for the 0–14 (A) and 0–28 (D) immunization schedules, including humoral and cellular immune response variables. Colors indicate r values, and the scale is shown next to each matrix. Individual selected Pearson correlations for the 0–14 (B–C) and 0–28 (E–F) immunization schedules are shown, indicating n, r, and p values. Statistical significance was set at p<0.05. Shaded gray areas show the 95% CI of the correlations.