key: cord-0801796-kh2ehuxj
authors: Pannus, P.; Neven, K. Y.; De Craeye, S.; Heyndrickx, L.; Vande Kerckhove, S.; Georges, D.; Michiels, J.; Francotte, A.; Van Den Bulcke, M.; Zrein, M.; Van Gucht, S.; Schmickler, M.-N.; Verbrugghe, M.; Matagne, A.; Thomas, I.; Dierick, K.; Weiner, J. A.; Ackerman, M. E.; Goriely, S.; Goossens, M. E.; Ariën, K. K.; Desombere, I.; Marchant, A.
title: Poor antibody response to BioNTech/Pfizer COVID-19 vaccination in SARS-CoV-2 naïve residents of nursing homes
date: 2021-06-09
journal: nan
DOI: 10.1101/2021.06.08.21258366
sha: 669a95975c23b8a083c8081e5e98acf2396b0ade
doc_id: 801796
cord_uid: kh2ehuxj

Background Residents of nursing homes (NH) are at high risk of COVID-19 related morbidity and death and may respond poorly to vaccination because of old age and frequent comorbidities. Methods Forty residents and forty staff members either naive or previously infected with SARS-CoV-2 were recruited in two NH in Belgium before immunization with two doses of 30g BNT162b2 mRNA vaccine at day 0 and day 21. Binding antibodies (Ab) to SARS-CoV-2 receptor binding domain (RBD), spike domains S1 and S2, RBD Ab avidity, and neutralizing Ab against SARS-CoV-2 wild type and B.1.351 variant were assessed at days 0, 21, 28, and 49. Results SARS-CoV-2 naive residents had lower Ab responses to BNT162b2 mRNA vaccination than naive staff. These poor responses involved lower levels of IgG to all domains of the vaccine antigen, lower avidity of RBD IgG, and lower levels of Ab neutralizing the vaccine strain. No naive resident had detectable neutralizing Ab to the B.1.351 variant. High and comparable Ab responses were observed in residents and staff previously infected with SARS-CoV-2. Clustering analysis revealed that poor vaccine responders not only included naive residents but also naive staff, emphasizing the heterogeneity of responses to mRNA vaccination in the general population. Conclusions The poor Ab responses to mRNA vaccination observed in infection naive residents and in some naive staff members of NH suggest suboptimal protection against breakthrough infection, especially with variants of concern. Adapted vaccination regimens may be needed to provide optimal protection against COVID-19 to vulnerable populations.

Introduction microplates using a sciFLEXARRAYER printing system (Scienion, Germany). Individual SARS-CoV-2 135 antigens, including Spike 1 domain (S1, encompassing AA16-685 of S), Spike 2 domain (S2, 136 encompassing AA686-1213 of S), and Receptor Binding Domain (RBD), were printed in duplicate 137 (GenBank YP009724390.1). Serial dilutions of test samples as well as positive and negative control 138 sera were incubated in microarray plates for 1h at room temperature (RT) and washed with 139 phosphate-buffered saline with 0.05% Tween 20 (PBST). Next, plates were incubated (1h, RT) with 140 horseradish peroxidase-conjugated goat anti-human IgG and washed with PBST before adding a 141 precipitating TMB solution for 20min (RT, dark). Then, TMB was removed and plates were dried at 142 37°C for 10min. Microplates were imaged and analyzed using a microplate reader (SciReader CL, 143

Scienion, Germany). The average pixel intensity for each spot was calculated for each antigen/dilution 144 and reported as net intensity. The dynamic range of each antigen measurement was defined using 145 serial dilutions of positive sera. Only antigen measurements within the dynamic range were 146 considered and were multiplied by the dilution factor. For each serum, quantitative results were eligible 147 if at least 2 dilutions report comparable results (%CV<28%). Results are reported as arbitrary pixel 148 units per milliliter (AU/ml). ROC-analyses using an independent population for validation generated 149 cutoff concentrations of 21.0 AU/ml, 19.5 AU/ml and 19.5 AU/ml for RBD, S1 and S2, respectively 150 (Supplementary methods). 151 152

Serial dilutions of heat-inactivated serum (1/50-1/25600 in EMEM supplemented with 2mM L-154 glutamine, 100U/ml -100μg/ml of Penicillin-Streptomycin and 2% fetal bovine serum) were incubated 155 during 1h (37°C, 7% CO 2 ) with 3xTCID100 of (i) a wild type (WT) Wuhan strain (2019-nCoV-Italy-156 INMI1, reference 008V-03893) and (ii) the B.1.351 variant of SARS-CoV-2, in parallel. Sample-virus Bio-layer interferometry measurements were performed with an Octet HTX instrument (Fortébio) using 165 AR2G biosensors. Data analyses were performed using FortéBio Data Analysis 9.0 software. Kinetic 166 assays were performed at 25-30˚C at a sample plate agitation speed of 1000rpm. Sensors were first 167 activated by immersion in a solution containing 20mM EDC and 10mM s-NHS. Then, 0.05mg/ml of 168 RBD antigen in 10mM sodium acetate pH 6 was loaded for 600sec. After antigen loading, the 169 biosensors were immersed in a solution of 1M ethanolamine pH8.5 to prevent non-specific 170 interactions. Antigen loaded AR2G sensors were first dipped in PBS to establish a baseline time 171 curve, and then immersed for 10min in wells containing purified serum IgG at three different dilutions 172 package "umap" for dimensionality reduction of the following outcomes at day 49: anti-RBD/S1/S2 187 IgG, anti-RBD IgG avidity, and WT NT50. To achieve normality, avidity was log 10 and neutralization 188 log 2 transformed. The optimal number of clusters was tested via the k-means (range 1:10) and visually 189 identified with an "elbow" in a plot of variance versus number of clusters. DBSCAN ("dbscan" package) 190 identified clusters within the UMAP reduced dimensions. 191

The study included 40 residents and 40 members of staff who were either naïve or previously infected 194 with SARS-CoV-2 before they received 2 x 30µg BNT162b2 mRNA vaccine at their respective NH. In 195 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258366 doi: medRxiv preprint previously infected subjects, SARS-CoV-2 infection occurred between 269 and 315 days before 196 vaccination. Complete cohort and demographic information is provided in Table 1. Although residents  197 with the poorest health status were excluded, most enrolled residents were frail and many suffered 198 multiple co-morbidities requiring medication. 199 200 Levels of binding Ab to SARS-CoV-2 spike receptor binding domain (RBD), spike subunit domains 201 one (S1) and two (S2) were measured in longitudinal serum samples using a multiplex immunoassay. 202

Detailed numerical data are presented in Tab.S1. At baseline, naïve staff and residents had 203 undetectable levels of SARS-CoV-2-specific IgG whereas high levels of Ab were detected in 204 previously infected subjects (Fig.1a, Fig.S1 ). Primary vaccination induced a significant increase in 205 SARS-CoV-2 Ab in naïve as well as previously infected staff and residents, and Ab levels were further 206 boosted following secondary vaccination at day 21 (Fig.1a) . Vaccine-induced Ab levels to RBD and 207 S1 were about six-fold lower in naïve residents as compared to naïve staff following primary 208 vaccination and two-fold lower after booster vaccination (Fig.1b) . In comparison to naïve subjects, Ab 209 levels were strongly increased in both residents and staff previously infected with SARS-CoV-2 210 ( Fig.1b and Fig.S1 ). Among previously infected subjects, residents had higher Ab responses to RBD 211 and S1 as compared to staff. Ab responses to S2 were lower than responses to RBD and S1, 212 especially in naïve subjects. 213

The avidity of RBD-specific Ab was measured using bio-layer interferometry. Rapid avidity maturation 215 was observed after primary vaccination in naïve and previously infected staff (Fig.2a) . High RBD IgG 216 avidity was also observed in previously infected residents at day 21, whereas avidity could only be 217 assessed in few naïve residents who had sufficiently high levels of RBD Ab to be characterized 218 ( Fig.2a) . Following booster vaccination, RBD IgG avidity further increased in naïve staff and residents, 219 but remained stable in previously infected subjects (Fig.2a) . Four weeks after booster vaccination (day 220 49), Ab avidity was significantly higher in naïve staff as compared to naïve residents, and was higher 221 in previously infected subjects as compared to naïve subjects (Fig.2b) . (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258366 doi: medRxiv preprint had detectable neutralizing Ab to the Wuhan strain at baseline and these titers further increased by 227 primary and booster vaccinations (Fig.2c) . Potent neutralizing Ab responses were also induced by 228 vaccination of naïve staff, although the proportion of subjects with detectable responses decreased 229 between day 28 (18/19) and day 49 (14/19). In contrast, only 6/20 naïve residents had detectable 230 neutralizing Ab at day 28 and this proportion increased to 9/20 at day 49 ( Fig.2c) . At day 49, naïve 231 residents had significantly lower neutralizing Ab responses as compared to naïve staff, whereas 232 higher responses were detected in previously infected subjects as compared to naïve subjects 233 ( Fig.2d ). Compared to the wild type strain, neutralizing titers against the B.1.351 variant were reduced 234 five to ten-fold across study groups (Fig.2e) The consistent differences in Ab responses observed between the four study groups suggested a 239 coordinated response to mRNA vaccination across the measured immunological parameters. Indeed, 240 titers of neutralizing Ab against the wild type strain strongly correlated with RBD, S1 and S2 binding 241

Ab, RBD IgG avidity, and neutralizing Ab to the B.1.351 variant (Fig.2f) . The clustering analysis therefore revealed a group of poor Ab responders that not only included naïve 254 residents but also naïve staff. 255 256 Discussion 257 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258366 doi: medRxiv preprint Reports on lower Ab responses to COVID-19 mRNA vaccination in older people and in people with 258 chronic comorbidities raise concern about the susceptibility of NH residents to severe breakthrough 259 infections, especially with SARS-CoV-2 variants of concern [10-13]. In this study, NH residents Another important finding of this study is that poor vaccine responders were not limited to naïve 287 residents, but also included healthy naïve staff. This observation emphasizes the heterogeneity of Ab 288 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. Identifying vulnerable populations who may benefit less from current mRNA vaccination regimens is 296 essential for the control of the COVID-19 pandemic. Adapted mRNA vaccination regimens may be 297 required to protect SARS-CoV-2 naïve residents of NH and younger poor vaccine responders against 298 breakthrough infections, especially with variants of concern. 299 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

The authors declare no conflicts of interest. 318 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258366 doi: medRxiv preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258366 doi: medRxiv preprint RBD S1

S2 Spike binding IgG, AU/mL x 1,000 *** *** *** *** *** *** * *** *** *** *** *** *** *** ** * *** *** *** *** *** *** ** * *** *** *** *** *** *** 

Long-Term Care Policy after Covid-19 -Solving the 320 Nursing Home Crisis

Correlates of protection against SARS-CoV-2 in 402 rhesus macaques

BNT162b2 Vaccination in People Over 80 Years of

Age Induces Strong Humoral Immune Responses with Cross Neutralisation of P

Clinical and immunological assessment of 408 asymptomatic SARS-CoV-2 infections

SARS-CoV-2 Infection after Vaccination in

Health Care Workers in California

Evidence for increased breakthrough rates of SARS

CoV-2 variants of concern in BNT162b2 mRNA vaccinated individuals

Accessed 5

Improving Vaccine-Induced Immunity: Can 418

Baseline Predict Outcome?

Broad immune activation underlies shared set 420 point signatures for vaccine responsiveness in healthy individuals and disease activity 421 in patients with lupus

No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity

Binding RBD-IgG, AU/mL 50% neutralisation (WT) Binding S1-IgG, AU/mL Binding S2-IgG, AU/mL * *** *** * *** * * *** ** *** ** *** *** *** *** *** NA