key: cord-0752309-o5te0ldx
authors: Seekircher, L.; Banki, Z.; Kimpel, J.; Roessler, A.; Schaefer, H.; Falkensammer, B.; Bante, D.; Forer, L.; Schoenherr, S.; Harthaller, T.; Sacher, M.; Ower, C.; Tschiderer, L.; Ulmer, H.; Krammer, F.; von Laer, D.; Borena, W.; Willeit, P.
title: Immune response to 2-dose BNT162b2 vaccination and risk of SARS-CoV-2 breakthrough infection: The Shieldvacc-2 study
date: 2022-04-21
journal: nan
DOI: 10.1101/2022.04.19.22273872
sha: 83d24d441cb013491d999328fea16748bcb53bc6
doc_id: 752309
cord_uid: o5te0ldx

It is uncertain to which extent antibody and T-cell responses after vaccination against SARS-CoV-2 are associated with reduced risk of breakthrough infection and whether their measurement enhances risk prediction. We conducted a phase-4 open-label clinical trial in the pre-omicron era, enrolling 2,760 individuals aged [≥]16 years 35{+/-}8 days after having received the second dose of BNT162b2 (baseline 15-21 May 2021). Over a median 5.9-month of follow-up, we identified incident SARS-CoV-2 breakthrough infections using weekly antigen tests, a confirmatory PCR test, and/or serological evidence for incident infection. We quantified relative risks adjusted for age, sex, and prior SARS-CoV-2 infection for different immunological parameters and assessed improvements in risk discrimination. In contrast to the T-cell response, higher plasma levels of binding antibodies and antibodies in a surrogate neutralization assay were associated with reduced risk of breakthrough infection. Furthermore, assessment of anti-spike IgG levels enhanced prediction of breakthrough infection and may therefore be a suitable measurable correlate of protection in practice.

Measurable correlates of protection that help assess to which extent a person is protected from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection after vaccination . CC-BY 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 preprint this version posted April 21, 2022. ;  https://doi.org/10.1101/2022.04. 19 .22273872 doi: medRxiv preprint

Study enrolment took place between 15 Baseline immunological parameters are also described in Table 1 . Except for two immunosuppressed participants, all individuals were seropositive for IgG antibodies targeting the receptor binding domain (RBD) of the spike protein (anti-S IgG) and 86.0% had seropositive titers of neutralizing antibodies in a surrogate SARS-CoV-2 virus neutralization test (sVNT). Among the 712 participants with prior SARS-CoV-2 infection, 655 were seropositive for anti-N Ig (92.0%). Titers of anti-S IgG, neutralizing antibodies in the sVNT, and anti-N Ig were all significantly lower in participants who had a SARS-CoV-2 infection . CC-BY 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)

The copyright holder for this preprint this version posted April 21, 2022 . CC-BY 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)

The copyright holder for this preprint this version posted April 21, 2022 Fifty-three (77.9%) of the breakthrough infections were symptomatic. The frequencies of reported symptoms are shown in Supplementary Fig. 3 . The most common symptoms were cough (57.4%), loss of taste or smell (45.6%), muscle or limb pain (44.1%), and fever or chills (36.8%). One case of SARS-CoV-2 infection required hospitalization; none were fatal.

No study participant died from other causes.

The relative risks for breakthrough infection per two-fold higher level were 0.72 (95% Secondary analyses restricted to symptomatic SARS-CoV-2 infections yielded broadly similar results (Fig. 2) . Additional secondary analyses quantified relative risks across different categories of levels of immunological parameters (Fig.3) . In all analyses, P-values for trend were 0.001 or lower and shapes of associations were log-linear, with no evidence of any thresholds that would clearly delineate population groups at high vs. low risk. For example, we observed a significantly lower relative risk for individuals with anti-S IgG levels ≥3,000

Binding Antibody Units per milliliter (BAU/mL) compared to those with <500 BAU/mL (hazard ratio: 0.20 [0.06-0.64]; P=0.007) and for individuals with anti-N Ig levels ≥25 Cut-. CC-BY 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)

The copyright holder for this preprint this version posted April 21, 2022 P=0.019). Cumulative incidence plots according to categories of immunological parameters are depicted in Supplementary Fig. 4 .

To assess the incremental value of immunological parameters for predicting SARS-CoV-2 infection, we quantified improvements in the C-index when adding their measurements to a common base model (Fig. 4) . The base model incorporated information on age and sex and 

In the present study involving 2,760 participants aged 16 years or older, we evaluated humoral and cellular immunological parameters after 2-dose BNT162b2 vaccination as potential correlates of protection against SARS-CoV-2 infection over a six-month follow-up period. We observed strong inverse log-linear associations with risk of incident SARS-CoV-2 breakthrough infections for a number of immunological parameters, i.e., anti-S IgG, titers of neutralizing antibodies in a sVNT, andin people that had a SARS-CoV-2 infection before . CC-BY 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. 2 In comparison to these trials, effect sizes for anti-S IgG in our study were stronger and were robust in analysis of the pre-specified primary outcome of "any infections" as well as in a sensitivity analysis restricted to symptomatic infections.

Our results are also in agreement with previous observational studies that were conducted in vaccinated community samples, [3] [4] [5] health-care workers, 6 patients with autoimmune rheumatic diseases, 7 and patients undergoing dialysis [8] [9] [10] . However, while the majority of studies compared risk across categories (e.g., dichotomizing the study population at arbitrary cut-offs for anti-S IgG levels), our study revealed associations compatible with a log-linear shape, . CC-BY 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 preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.19.22273872 doi: medRxiv preprint thereby suggesting the higher the level of immune response the lower the risk of breakthrough infection without evidence for a threshold or saturation effect.

For T-cell responses, we did not detect measurable differences in post-vaccination levels between people with or without breakthrough infection, which is in concordance with the findings of studies among BNT162b2-vaccinated individuals from Italy 11 and Singapore 12 . This is may be explained by the main function of T-cells, which is to facilitate early viral clearance 13 , and hence circumvent severe clinical course, rather than prevent a primary infection per se. The shorter incubation time of the Delta variant (2-3 days) 14 . CC-BY 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 preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04. 19.22273872 doi: medRxiv preprint No prior study has investigated the usefulness of these immunological parameters in SARS-CoV-2 risk prediction and, therefore, our study provides entirely novel evidence on this topic. Our study has several strengths. It has a prospective study design, is adequately sized, covers a ~6 months follow-up post second dose, and compared immunological parameters for humoral and cellular immunity measured with validated assays. To the best of our knowledge, our study is the first to analyze associations between T-cells and incident SARS-CoV-2 infection using time-to-event analysis. Furthermore, all incident SARS-CoV-2 infections and related symptoms were validated rigorously in structured telephone interviews. Our study also has limitations. First, cellular immune parameters were available only for a (random) subgroup of participants, thereby limiting statistical power. Second, the QuantiFERON . CC-BY 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 preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.19.22273872 doi: medRxiv preprint SARS-CoV-2 assay was limited to measure the Interferon-Gamma (IFN-γ) production after stimulation with CD4 and combined CD4 and CD8 peptide pools, therefore a detailed characterization of T-cell response with respect of the source of IFN-γ (CD4 or CD8 T-cells), phenotypical and further functional analysis of T-cells is missing. Third, the proportion of participants with a prior SARS-CoV-2 infection was relatively high (25.8%) and is likely due to high interest to participate in the study. Ascertainment of prior infection was of high quality as it was performed by trained study staff and was confirmed by a positive anti-N Ig measurement in 92% of cases. Fourth, we conducted our study during a phase in which Delta was the predominant variant of concern and associations of immunological parameters with Omicron may be weaker. Finally, these analyses have been conducted on samples taken after two doses of BNT162b2 and might not apply to other coronavirus disease 2019 (COVID- 19) vaccines or to people that received a third doses of BNT162b2.

In conclusion, in contrast to the T-cell response, higher plasma levels of binding and neutralizing antibodies in a surrogate neutralization assay were associated with reduced risk of breakthrough infection. Assessment of anti-S IgG levels enhances prediction of incident SARS-CoV-2 infection and may therefore be a suitable measurable correlate of protection in practice.

The Shieldvacc-2 study is a phase-4 open-label clinical trial conducted at two centers (Jenbach, Zell am Ziller) in the district of Schwaz, Tyrol, Austria. Individuals were eligible for inclusion if they (i) were aged 16 years or older; (ii) had received two 30 µg doses of . CC-BY 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 preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.19.22273872 doi: medRxiv preprint BNT162b2 delivered by intramuscular injection, with the second dose having been administered 35±8 days prior to enrolment into the study; (iii) understood and agreed to comply with the study procedures; and (iv) were willing to be contacted by telephone or to complete a diary throughout the study in an online participant portal. Exclusion criteria were: . CC-BY 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 preprint this version posted April 21, 2022. 

RUO IFN-γ release assay (Qiagen, Hilden, Germany) in response to CD4 and combined CD4 and CD8 peptide pools derived from SARS-CoV-2 spike antigen (S1 S2 RDB). The ratios of IFN-γ values from SARS-CoV-2 specific stimulation and the unstimulated control (Nil) was . CC-BY 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 preprint this version posted April 21, 2022 

The primary outcome was defined as the occurrence of a SARS-CoV-2 breakthrough infection during a follow-up period of six months, identified by a positive PCR test (n=50), a seroconversion of anti-N Ig from baseline to follow-up (n=14), or a threefold increase of a positive anti-N Ig level during the study period (n=4). To preclude underascertainment of asymptomatic or pauci-symptomatic events, participants were asked to undergo SARS-CoV-2 antigen testing every 7 (±3) days throughout the course of the study and to record test results and related symptoms via the online participant portal. A secondary analysis focused on symptomatic SARS-CoV-2 infections defined as in the BNT162b2 phase-2/3 trial as having one or more symptoms of fever or chills, cough, breathing difficulties, muscle or limb pain, lost sense of smell or taste, sore throat, diarrhea, or vomiting. 36 All recorded SARS-CoV-2 infections, including dates of infections, symptoms, and clinical course, underwent rigorous validation in structured telephone interviews. For events detected through serological tests only, the date of infection was estimated using the dates of plausible contagions (e.g., symptoms or close contact with infected individuals) (n=12) or otherwise using the median date of all SARS-CoV-2 events that were recorded in the study (n=6). Prior SARS-CoV-2 infection was based on self-report or seropositivity of anti-N Ig at the time of enrolment.

Because the distributions of immunological parameters were skewed, we log-transformed their values for all analyses. We used t-tests for continuous variables and χ 2 -tests for . CC-BY 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 preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.19.22273872 doi: medRxiv preprint categorical variables to compare baseline characteristics of participants with and without incident SARS-CoV-2 infection. We calculated Pearson correlation coefficients to assess the cross-sectional correlation of immunological parameters at study baseline.

To quantify the associations between immunological parameters and the risk of a SARS-CoV-2 breakthrough infection, we estimated relative risks per 2-fold higher level adjusted for age, sex, and prior SARS-CoV-2 infection. For the majority of parameters (i.e., anti-S IgG, neutralizing antibodies in a sVNT, T-cell response, and anti-N Ig), we analyzed time-to-event data using Cox regression. In these analyses, participants were censored at the time of a SARS-CoV-2 infection, end of the follow-up period, or loss to follow-up, whichever came first. Participants were considered lost to follow-up if they withdrew from the study or had more than one consecutive missing antigen test result, no positive PCR test result, and did not provide anti-N Ig test results at the beginning and the end of the study. The proportionalhazards assumption was tested on the basis of Schoenfeld's residuals and was met. For the titers of neutralizing antibodies in a pVNT, which was measured only in a nested case-control sample, we used conditional logistic regression to estimate odds ratios for breakthrough infection adjusted for age, sex, and prior SARS-CoV-2 infection. As expected, SARS-CoV-2 events were relatively rare in our study and we therefore describe odds ratios and hazard ratios collectively as measures of relative risk ("rare disease assumption"). We conducted secondary analyses that (i) focused on symptomatic SARS-CoV-2 infections and (ii) quantified relative risks across different categories of baseline antibody levels.

To assess the incremental predictive values of measuring different immunological parameters, we quantified improvements in the C-index when adding them to a model containing information on age and sex. 37 The C-index is the preferred measure of risk discrimination for time-to-event data and assesses whether the model correctly predicts the order of failure of . CC-BY 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 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 preprint this version posted April 21, 2022 

Data on COVID-19 cases in the district of Schwaz, Tyrol, in Austria is publicly available (data.gv.at -Open Data Österreich). Tabular data on the Shieldvacc-2 cohort can be requested from the corresponding authors by researchers who submit a methodologically sound proposal (including a statistical analysis plan); participant-level data on the Shieldvacc-2 cohort cannot be shared due to regulatory restrictions.

. CC-BY 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 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) Fig. 1 : Correlation coefficients and scatter plots of the baseline levels of the immunological parameters assessed in the study.

*The analyses of pVNT values were restricted to the nested case-control sample. † The analyses of anti-N Ig values were restricted to participants with prior SARS-CoV-2 infection. For all analyses log-transformed immunological parameters were used. The upper part of the matrix indicates unadjusted Pearson correlation coefficients with 95% confidence intervals and number of participants. Areas are shadowed according to the magnitude of the point estimates of Pearson correlation coefficients, i.e., the darker the closer to one, the lighter the closer to zero. The lower part of the matrix depicts scatter plots of different immunological parameters, with both axes presented on a log scale. Abbreviations: BAU, Binding Antibody Units; COI, Cut-off-Index ; IU, international unit; mL, milliliter; pp, peptide pool; pVNT, pseudotyped virus neutralization test; SI, stimulation index; sVNT, surrogate virus neutralization test.

. CC-BY 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. Table 1 provides additional information on participants with prior SARS-CoV-2 infection and with incident SARS-CoV-2 breakthrough infection. Symptomatic SARS-CoV-2 infection was defined as having one or more symptoms including fever or chills, cough, breathing difficulties, muscle or limb pain, lost sense of smell or taste, sore throat, diarrhea, or vomiting. Cox regression was applied for anti-S IgG, sVNT, CD4 pp, CD4 & CD8 pp, and anti-N Ig and conditional logistic regression for pVNT. Immunological parameters entered as log2-transformed continuous terms. Abbreviations: pp, peptide pool; pVNT, pseudotyped SARS-CoV-2 virus neutralization test; sVNT, surrogate SARS-CoV-2 virus neutralization test.

*pVNT was measured in a subset of 68 cases and 204 individual-matched controls. Cox regression was applied for anti-S IgG, sVNT, and anti-N Ig and conditional logistic regression for pVNT. For anti-N Ig, the regression model was adjusted for age and sex (and not for prior SARS-CoV-2 infection due to collinearity) and for anti-S IgG, sVNT, and pVNT additionally for prior SARS-CoV-2 detected by seropositivity of anti-N Ig at the time of enrolment or self-report. Ptrend indicates the p-value of the likelihood ratio test comparing regression models including categories of antibody levels as a continuous variable and without antibody information. Supplementary Table 1 provides additional information on participants with prior SARS-CoV-2 infection and with incident SARS-CoV-2 breakthrough infection. Abbreviations: BAU, Binding Antibody Units; CI, confidence interval; COI, Cut-Off-Index; IR, incidence rate; IU, international unit; NA, not available; mL, milliliter; PD, person-days; pVNT, pseudotyped SARS-CoV-2 virus neutralization test; sVNT, surrogate SARS-CoV-2 virus neutralization test.

. CC-BY 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)

The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.19.22273872 doi: medRxiv preprint

*Refers to a SARS-CoV-2 infection prior to study entry detected by self-report. Participants with complete data on all variables are included in analyses (2,760 participants; 68 incident SARS-CoV-2 events). We quantified improvements in the C-index when adding them to a model containing information on age and sex. Immunological parameters entered as log2-transformed continuous terms. Abbreviations: CI, confidence interval; sVNT, surrogate SARS-CoV-2 virus neutralization test.

. CC-BY 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)

The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.19.22273872 doi: medRxiv preprint Tables   Table 1: Baseline characteristics of the entire study population and separately for participants with and without incident SARS-CoV-2 infection during follow-up.

Correlates of protection against symptomatic and asymptomatic SARS-CoV-2 infection

Immune correlates analysis of the mRNA-1273 COVID-19 vaccine efficacy clinical trial

Waning of SARS-CoV-2 antibodies targeting the Spike protein in individuals post second dose of ChAdOx1 and BNT162b2 COVID-19 vaccines and risk of breakthrough infections: analysis of the Virus Watch community cohort

Antibody responses and correlates of protection in the general population after two doses of the ChAdOx1 or BNT162b2 vaccines

Levels of SARS-CoV-2 Antibodies Among Fully-Vaccinated Individuals With Delta or Omicron Variant Breakthrough Infections: A Prospective Cohort Study

An observational study of breakthrough SARS-CoV-2 Delta variant infections among vaccinated healthcare workers in Vietnam

Postvaccination antibody titres predict protection against COVID-19 in patients with autoimmune diseases: survival analysis in a prospective cohort

SARS-CoV-2 Vaccine Antibody Response and Breakthrough Infection in Patients Receiving Dialysis

COVID-19 outbreak in vaccinated patients from a hemodialysis unit: antibody titers as a marker of protection from infection

Humoral response and breakthrough infections with SARS-CoV-2

.617.2 variant in vaccinated maintenance hemodialysis patients

SARS-CoV-2 vaccine breakthrough infections with the alpha variant are asymptomatic or mildly symptomatic among health care workers

Decreased memory B cell frequencies in COVID-19 delta variant vaccine breakthrough infection

The T cell immune response against SARS-CoV-2

Transmission, viral kinetics and clinical characteristics of the emergent SARS-CoV-2 Delta VOC in Guangzhou

SARS-CoV-2 Omicron Variant Infection of Individuals with High Titer Neutralizing Antibodies Post-3 rd mRNA Vaccine Dose

Omicron escapes the majority of existing SARS-CoV-2 neutralizing antibodies

A serological assay to detect SARS-CoV-2 seroconversion in humans

Development and characterization of a quantitative ELISA to detect anti-SARS-CoV-2 spike antibodies

Robust neutralizing antibodies to SARS-CoV-2 infection persist for months

BNT162b2 COVID-19 vaccine and correlates of humoral immune responses and dynamics: a prospective, single-centre, longitudinal cohort study in healthcare workers

Waning Immune Humoral Response to BNT162b2 Covid-19 Vaccine over 6 Months

Correlation between a quantitative anti-SARS-CoV-2 IgG ELISA and neutralization activity

Evolution of antibody responses up to 13 months after SARS-CoV-2 infection and risk of reinfection

B and T cell response to SARS-CoV-2 vaccination in health care professionals with and without previous COVID-19

Commercialized kits to assess T-cell responses against SARS-CoV-2 S peptides. A pilot study in health care workers

Severity of adverse reactions is associated with T-cell response in mRNA-1273 vaccinated health care workers

Evaluation of the QuantiFERON SARS-CoV-2 interferon-ɣ release assay in mRNA-1273 vaccinated health care workers

Lasting antibody and T cell responses to SARS-CoV-2 in COVID-19 patients three months after infection

Dynamics of SARS-CoV-2 neutralising antibody responses and duration of immunity: a longitudinal study

Discordant neutralizing antibody and T cell responses in asymptomatic and mild SARS-CoV-2 infection

Comparison of Four SARS-CoV-2 Neutralization Assays. Vaccines (Basel)

Pseudotype Neutralization Assays: From Laboratory Bench to Data Analysis

Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

Comparing the Predictive Powers of Survival Models Using Harrell's C or Somers' D

Overall (n=2,760) Incident SARS-CoV-2 infection

Yes (n=68) No (n=2,692)

Absolute level in BAU/mL

Surrogate virus neutralization No. with measurement 2,760 68 2

*Refers to a SARS-CoV-2 infection prior to study entry detected by seropositivity of anti-N Ig at the time of enrolment or self-report. † Nested case-control sample. ‡ Restricted to participants with prior SARS-CoV-2 infection. Values are considered as positive if BAU/mL ≥7.1 for anti-S IgG, IU/mL ≥200 for surrogate virus neutralization, reciprocal titer >16 for pseudotyped virus neutralization, COI ≥1.0 for anti-N Ig, and SI ≥3 for CD4 and combined CD4 and CD8 peptide stimulation

We thank Bianca Neurauter, Eva Hochmuth, and Luiza Hoch for excellent assistance when organizing the study and Brigitte Müllauer, Evelyn Peer, Lisa-Maria Raschbichler, Evelyn Peer, and Albert Falch for excellent technical support. We are also grateful to all BMAs in the diagnostic section of Institute of Virology at the Medical University of Innsbruckespecially