key: cord-0293222-c752tmka authors: Barbeau, D.; Martin, J.; Carney, E.; Dougherty, E.; Doyle, J.; Dermody, T.; Hoberman, A.; Williams, J.; Michaels, M.; Alcorn, J.; Duprex, P.; McElroy, A. K. title: Comparative analysis of human immune responses following SARS-CoV-2 vaccination with BNT162b2, mRNA-1273, or Ad26.COV2.S date: 2021-09-23 journal: nan DOI: 10.1101/2021.09.21.21262927 sha: 5d24eb3ec9b1cfe98f4c1cdfca62b4d25f034608 doc_id: 293222 cord_uid: c752tmka Background: Three SARS-CoV-2 vaccines, two based on mRNA, BNT162b2 and mRNA-1273, and one based on an adenovirus platform, Ad26.COV2.S, received emergency use authorization by the U.S. Food and Drug Administration in 2020/2021. These vaccines displayed clinical efficacy in initial studies against confirmed COVID-19 of 95.0%, 94.1%, and 66.9%, respectively. Methods: Individuals receiving one of these vaccines were invited to participate in a prospective longitudinal comparative study of immune responses elicited by the three vaccines. In this observational cohort study, humoral responses were evaluated using a SARS-CoV-2 receptor-binding domain (RBD) ELISA and a SARS-CoV-2 virus neutralization assay at 21-32 days and again at 47-64 days following each initial vaccination. Results: The two mRNA-based platforms elicited similar RBD ELISA responses, but significantly higher neutralizing antibody responses were achieved by mRNA-1273. The adenovirus-based vaccine elicited significantly lower RBD ELISA and SARS-CoV-2 virus neutralization activity. IFN-gamma ELISPOT assays were conducted with peripheral blood mononuclear cells obtained 47-64 days after each initial vaccination. The mRNA-1273 vaccine elicited significantly higher spike glycoprotein-specific T cell responses than either the BNT162b2 or the Ad26.COV2.S vaccines. Conclusions: These findings are consistent with published efficacy data for the three vaccines and support the use of neutralizing antibody titers as a correlate of protection against symptomatic COVID-19. 3 Conclusions: Both mRNA based vaccines elicited higher magnitude humoral responses 47 than Ad26.COV2.S and mRNA1273 elicited the highest magnitude of T cell response. 48 Neutralizing antibody titers correlated with reported estimates of vaccine efficacy. 49 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 23, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 Two mRNA-based vaccines (BNT162b2 and mRNA-1273) and one adenovirus-based 51 vaccine (Ad26.COV2.S) have been used in the US since EUA was granted for each 52 (December 2020 for the mRNA vaccines and February 2021 for the adenovirus 53 vaccine). While data have been published for each of these vaccines confirming safety, 54 immunogenicity, and efficacy [1] [2] [3] [4] [5] [6] [7] , only limited data are available that compare vaccine-55 induced immune responses amongst the three vaccines using identical immunological 56 assays [8] . In this study, we assessed a cohort of SARS-CoV-2 naive individuals who 57 received BNT162b2, mRNA-1273, or Ad26.COV2.S vaccines for antigen-specific 58 humoral and T cell immunity using identical immunologic assays to allow direct 59 comparisons of the elicited responses. To date, an immune correlate of protection has 60 not been established, however, neutralizing antibody titers may be a possible correlate 61 [9]. Mathematical modelling combined with analysis of published immunogenicity data 62 has been used to demonstrate that the ratio of neutralization titer following vaccination 63 to that during convalescence following naturally-acquired disease correlates with the 64 reported clinical efficacy for each vaccine [10] . Our study demonstrates a correlation 65 between neutralizing antibody titer and reported vaccine efficacy, but we did not 66 observe this correlation with IFN-γ ELISPOT responses, supporting the use of 67 neutralizing antibody titer as a correlate of protection. 68 A convenience sample of participants 18 years of age and older were prospectively 71 enrolled if they were planning to receive two doses of mRNA-1273 or BNT162b2, or a 72 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint ELISA and FRNT assays were conducted as previously described [11] . Samples with 83 FRNT 50 titers below the limit of detection (LOD=10) were assigned a value of 5 for 84 graphical depiction and statistical analysis on the figures. 85 PMBCs were incubated for 24 hours with 2 μg/mL of a SARS-CoV-2 complete spike 87 glycoprotein mega pool consisting of 15-mer peptides overlapping by 11 residues based 88 upon the original Wuhan-Hu-1 strain (GenBank MN908947.3) (Miltenyi) with 1 x Cell 89 Activation Cocktail (without Brefeldin A, BioLegend) or media alone. Human IFN-γ 90 ELISPOT assays were conducted according to the manufacturer's instructions 91 (Mabtech). 92 Graphpad Prism was used for statistical analysis and to prepare figures. Four samples 94 were missing from visit 2 for the mRNA-1273 participants. Samples were available for 95 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint compare ELISPOT data between vaccine groups. Sex, age, and racial/ethnic 100 distributions of participants were compared with those of the catchment area using one-101 way chi-squared goodness-of-fit tests. 102 103 Participant age, sex, and race were provided at the time of enrollment (Table 1) . 105 Information about co-morbidities was collected from the medical record or directly from 106 study participants. 107 An RBD ELISA was conducted using plasma samples from all participants at all three 108 time points (enrollment/date of first vaccination and at subsequent visits, Table 2 ). 109 Participants with higher-than-average baseline RBD titers were tested by SARS-CoV-2 110 nucleoprotein (N) ELISA, and 5 individuals were found to have N-specific ELISA titers 111 greater than or equal to 900 [11] . As this finding is consistent with prior SARS-CoV-2 112 infection, data from these participants were excluded from further analysis. One mRNA-113 1273 vaccine recipient relocated after enrollment precluding collection of additional 114 data. After these exclusions, 25 BNT162b2 recipients, 24 mRNA-1273 recipients, and 115 24 Ad26.COV2.S recipients were included in the study. On the date of first vaccination 116 (pre-bleed), 6 individuals had detectable RBD ELISA titers relative to a pre-pandemic 117 normal human control group. There were no statistically significant differences in 118 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 23, 2021. ; https://doi.org/10. 1101 baseline titers between the vaccine groups ( Figure 1 ). At visit 2 (median 21-31 days 119 post initial vaccination), most study participants had notable antibody titers with the 120 exception of two BNT162b2 recipients, one of whom was immunocompromised 121 secondary to ongoing therapy for breast cancer with capecitabine. At visit 2, RBD 122 ELISA titers were significantly higher in mRNA-1273 recipients compared with 123 Ad26.COV2.S recipients (p=0.0008) At visit 3 (median 45-63 days post initial 124 vaccination), following booster doses for both of the mRNA-based vaccines, the 125 BNT162b2 titers were equivalent to the mRNA-1273 titers, and both were significantly 126 higher than those achieved by the Ad26.COV2.S (single-dose regimen) (p=0.0059 for 127 BNT162b2 vs. Ad26.COV2.S and p=0.0315 for mRNA-1273 vs. Ad26.COV2.S). RBD 128 ELISA titers observed in these cohorts correlated with those previously reported in 129 immunogenicity studies of the various vaccines with Ad26.COV2.S having 2-3 log GMT 130 [6], and mRNA-1273 and BNT162b2 having 4-5 log GMT after completion of the two 131 dose series [3, 12] . 132 Plasma samples from visit 2 and visit 3 were tested in a neutralization assay using 133 SARS-CoV-2. FRNT 50 titers were at or below the limit of detection (LOD=10) for a small 134 subset of participants after a single immunization and did not differ significantly between 135 vaccines ( Figure 2A ). Following a second dose of either of the mRNA vaccines, FRNT 50 136 titers were significantly higher than those achieved with single dose Ad26.COV2.S 137 vaccination (p<0.0001 for both BNT162b2 and mRNA-1273 vs. Ad26.COV2.S). A trend 138 towards higher mean FRNT 50 titer was observed following the vaccination series with 139 mRNA-1273 than with BNT162b2, but this was not significant. Given the relatively low 140 neutralization titers observed in some subjects, several observations did not meet the 141 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 23, 2021. ; https://doi.org/10.1101/2021.09.21.21262927 doi: medRxiv preprint 50% focus-reduction threshold. Therefore, the data were analyzed as a function of 142 percent neutralization of input virus at a constant 1:20 dilution of plasma ( Figure 2B ). 143 This strategy allowed a more nuanced assessment of the capacity of individual plasma 144 specimens to neutralize virus. Participants who received a single dose of mRNA-1273 145 achieved higher percent neutralization at 1:20 dilution than those receiving 146 Ad26.COV2.S (p<.0001). Following the booster dose, BNT162b2 and mRNA-1273 147 vaccination elicited similar neutralizing capacity, and percent neutralization at 1:20 148 dilution elicited by both mRNA vaccines was higher in magnitude than that elicited by 149 the single dose Ad26.COV2.S at visit 3 (p<0.0001 for both BNT162b2 and mRNA-1273 150 versus Ad26.COV2.S). Interpretation of these data in the context of previous 151 independently published reports of each vaccine should take into consideration the 152 differences in methods used in these studies. In some studies, pseudovirion assays are 153 used, in others live virus is used but the readouts vary based upon the assay-luciferase 154 reporters in some, foci in others. Additionally, if the inoculum (virus combined with 155 serum) is left in place following adsorption, this could alter the assay to assess not only 156 antibodies that block binding but also those that block cell to cell spread or post-binding 157 steps in fusion as has been shown for antibodies that target the spike N-terminal 158 domain [13, 14] . Slightly lower FRNT 50 titers noted in our study could be because we 159 remove the inoculum after a 1-hour incubation period, hence our FRNT assay is specific 160 for antibodies that block binding to the cellular ACE2 receptor. For this reason it is 161 important to have a comparator control group that can be used in all studies to permit 162 normalizing between assays as has been proposed by WHO [15] . 163 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. platforms [3, 6, 7, 12] . However, as with all vaccines, the goal is to achieve a balance 186 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted September 23, 2021. ; https://doi.org/10.1101/2021.09.21.21262927 doi: medRxiv preprint vaccine group (approximately 50%) was slightly lower than that observed in the Phase 210 III clinical trial for Ad26.COV2.S (66.9%) [5] , which may be attributable to the fact that 211 our Ad26.COV2.S cohort is mostly older men or it could be attributable to the difficulty of 212 comparing data using different immunological assays and the lack of a clear definition of 213 "convalescent" in the calculation of GMT from convalescent patients. Further studies are 214 necessary to confirm a precise serum neutralizing titer that correlates with protection 215 and to identify and standardize assays for this determination. ranging from 0 to 100 per 1x 10 5 PBMC. This assay is another area in which 227 standardization is needed for cross study comparison. 228 The finding of significantly higher levels of spike-specific T cells following vaccination 229 with mRNA-1273 than with either BNT162b2 or Ad26.COV2.S could be attributable to 230 higher levels of antigen or to the dosing regimen. A notable feature of all three vaccines 231 is the intracellular expression of viral antigen, which allows intracellular processing and 232 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. indicated that BNT162b2 has an efficacy of 95% after the two-dose series, while 239 Ad26.COV2.S efficacy was 66.9%. Since these two vaccines elicit similar T cell 240 responses in IFN-γ ELISPOT assays, these data suggest that the T cell component of 241 the immune response is probably not responsible for the enhanced protection afforded 242 by BNT162b2 versus Ad26.COV2.S. However, the IFN-γ ELISPOT assay did not 243 discriminate between CD4+ and CD8+ T cells, and thus, we cannot exclude the 244 possibility that a specific T cell subset or function plays a role in protection, or that T cell 245 mediated protection plays a role in the context of reduced humoral immunity. 246 concern? One of the most pressing problems in the pandemic is the emergence of viral 248 variants of concern (VOC). Serum from vaccine recipients appears to neutralize the 249 variants reported thus far, despite a reduction in viral neutralizing capacity in some 250 cases [18] [19] [20] [21] . Additionally, vaccinated individuals appear to be protected from severe 251 disease and hospitalization following infection with VOC [22] . Studies of influenza 252 vaccination suggest that heterologous protection is provided by virus-specific T cell 253 immunity [23] . Most SARS-CoV-2 T cell epitopes are largely conserved between the 254 variants [24], and even unexposed individuals have some virus-specific T cells due to 255 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 23, 2021. ; https://doi.org/10.1101/2021.09.21.21262927 doi: medRxiv preprint cross-reactivity with epitopes conserved among common human coronaviruses [25] . 256 Given the existence of virus-specific, cross-reacting T cells, we hypothesize that 257 increased protection from disease caused by the variants could be provided by vaccines 258 that also stimulate strong T cell responses in addition to strong humoral responses. If 259 this were the case, mRNA-1273 could potentially provide better protection from variants 260 than either BNT162b2 or Ad26.COV2.S. Further research will be required to test this 261 Limitations of our study include that it was an observational cohort study without 263 randomization, and participants receiving the various vaccines were not matched for 264 demographics. As a result, there were sex and age differences and different racial 265 distributions between the groups. The cohorts were of insufficient size to permit sub- is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 23, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 Interval between vaccine 2 and visit 3 36 (32-41) 24 (22-25) N/A . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 23, 2021. ; https://doi.org/10.1101/2021.09.21.21262927 doi: medRxiv preprint 397 398 399 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 23, 2021. ; https://doi.org/10.1101/2021.09.21.21262927 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 23, 2021. ; https://doi.org/10. 1101 Immunogenicity of Ad26.COV2.S vaccine against 318 SARS-CoV-2 variants in humans Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine An mRNA Vaccine against 322 SARS-CoV-2 -Preliminary Report Safety and Efficacy of the BNT162b2 324 mRNA Covid-19 Vaccine Vaccine against Covid-19 Interim Results of a Phase 1-2a Trial of 329 Covid-19 Vaccine Immunogenicity of the Ad26.COV2.S 331 Vaccine for COVID-19 Potential therapeutic targets 333 and promising drugs for combating SARS-CoV-2 Evidence for antibody as a 336 protective correlate for COVID-19 vaccines Neutralizing antibody levels are highly 338 predictive of immune protection from symptomatic SARS-CoV-2 infection A Cross-Sectional Study of SARS-CoV-2 341 Seroprevalence between Fall Safety and Immunogenicity of Two 344 RNA-Based Covid-19 Vaccine Candidates N-terminal domain antigenic 347 mapping reveals a site of vulnerability for SARS-CoV-2 Neutralizing and protective human 350 monoclonal antibodies recognizing the N-terminal domain of the SARS-CoV-2 351 spike protein WHO International Standard for 353 anti-SARS-CoV-2 immunoglobulin Comparison of SARS-355 CoV-2 Antibody Response Following Vaccination With BNT162b2 and mRNA-356 1273 SARS-CoV-2 mRNA vaccines 358 induce broad CD4+ T cell responses that recognize SARS-CoV-2 variants and 359 HCoV-NL63 Neutralising antibody activity against SARS-361 CoV-2 VOCs B.1.617.2 and B.1.351 by BNT162b2 vaccination Neutralization of SARS-CoV-2 lineage 364 7 pseudovirus by BNT162b2 vaccine-elicited human sera Infection and mRNA-1273 vaccine antibodies 367 neutralize SARS-CoV-2 UK variant. medRxiv 2021 Durable Humoral and Cellular 369 Immune Responses Following Ad26.COV2.S Vaccination for SARS-CoV-2 Delta VOC in Scotland: demographics, risk of hospital admission, 373 and vaccine effectiveness Negligible impact of SARS-CoV-2 variants on 377 CD4 (+) and CD8 (+) T cell reactivity in COVID-19 exposed donors and 378 vaccinees. bioRxiv 2021 Targets of T Cell Responses to SARS-380 CoV-2 Coronavirus in Humans with COVID-19 Disease and Unexposed 381 Individuals