TYPE Original Research PUBLISHED 11 January 2023 DOI 10.3389/fimmu.2022.1079884 Temporal trends of COVID-19 OPEN ACCESS antibodies in vaccinated EDITED BY William Tolbert, Henry M Jackson Foundation for the healthcare workers undergoing Advancement of Military Medicine (HJF), United States repeated serological sampling: REVIEWED BY Luminița-Smaranda Iancu, Grigore T. Popa University of Medicine An individual-level analysis and Pharmacy, Romania Valentyn Oksenych, within 13 months in the University of Oslo, Norway *CORRESPONDENCE Paolo Boffetta ORCHESTRA cohort Paolo.boffetta@unibo.it SPECIALTY SECTION Giulia Collatuzzo 1, Giuseppe De Palma 2, This article was submitted to Francesco S. Violante 1,3, Stefano Porru 4, Viral Immunology, a section of the journal Francesca Larese Filon 5, Eleonora Fabianova 6, Frontiers in Immunology Concepción Violán 7,8,9,10, Luigi Vimercati 11, RECEIVED 25 October 2022 ACCEPTED 13 December 2022 Mihaela Leustean 12, Marta Maria Rodriguez-Suarez 13, PUBLISHED 11 January 2023 Emanuele Sansone 3, Emma Sala 14, Carlotta Zunarelli 1, CITATION Vittorio Lodi 3, Maria Grazia Lourdes Monaco 15, Collatuzzo G, De Palma G, Violante FS, Porru S, Larese Filon F, Fabianova E, Gianluca Spiteri 15, Corrado Negro 5, Jana Beresova 6, Violán C, Vimercati L, Leustean M, Rodriguez-Suarez MM, Sansone E, LucÌa A. Carrasco-Ribelles 7, Silvio Tafuri 11, Shuffield S. Asafo 1, Sala E, Zunarelli C, Lodi V, Giorgia Ditano 1, Mahsa Abedini 1 and Paolo Boffetta 1,16* Monaco MGL, Spiteri G, Negro C, Beresova J, Carrasco-Ribelles LA, 1 Department of Medical and Surgical Sciences, University of Bologna, Bologna, Italy, 2 Department Tafuri S, Asafo SS, Ditano G, Abedini M of Medical and Surgical Specialties, Radiological Sciences and Public Health, University of Brescia, and Boffetta P (2023) Temporal trends Brescia, Italy, 3 Occupational Medicine Unit, IRCCS Azienda Ospedaliero-Universitaria di Bologna, of COVID-19 antibodies in vaccinated Bologna, Italy, 4 Section of Occupational Medicine, Department of Diagnostics and Public Health, healthcare workers undergoing University of Verona, Verona, Italy, 5 Unit of Occupational Medicine, University of Trieste, Trieste, repeated serological sampling: An Italy, 6 Occupational Health Department, Regional Authority of Public Health, Banská Bystrica, individual-level analysis within 13 Slovakia, 7 Unitat de Suport a la Recerca Metropolitana Nord, Institut Universitari d’Investigació en months in the ORCHESTRA cohort. Atenció Primària Jordi Gol (IDIAP Jordi Gol), Mataró, Spain, 8 Direcció d’Atenció Primària Front. Immunol. 13:1079884. Metropolitana Nord Institut Català de Salut, Barcelona, Spain, 9 Germans Trias i Pujol Research doi: 10.3389/fimmu.2022.1079884 Institute (IGTP), Badalona, Spain, 10 Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Spain, 11 Interdisciplinary Department of Medicine, University of Bari, Bari, Italy, 12 National Institute of Public Health, Bucharest, Romania, 13 Instituto de Investigación Sanitaria del Principado de Asturias (ISPA) and Universitario Central de Asturias (HUCA), University of Oviedo, Oviedo, Spain, 14 Unit of Occupational Health, Hygiene, Toxicology and Prevention, ASST Ospedali Civili di Brescia, Brescia, Italy, 15 Occupational Medicine Unit, University Hospital of Verona, Verona, Italy, 16 Stony Brook Cancer Center, Stony Brook University, Stony Brook, NY, United States Short summary: We investigated changes in serologic measurements after COVID-19 vaccination in 19,422 subjects. An individual-level analysis was performed on standardized measurements. Age, infection, vaccine doses, time between doses and serologies, and vaccine type were associated with changes in serologic levels within 13 months. Background: Persistence of vaccine immunization is key for COVID-19 prevention. Frontiers in Immunology 01 frontiersin.org Collatuzzo et al. 10.3389/fimmu.2022.1079884 Methods: We investigated the difference between two serologic measurements of anti-COVID-19 S1 antibodies in an individual-level analysis on 19,422 vaccinated healthcare workers (HCW) from Italy, Spain, Romania, and Slovakia, tested within 13 months from first dose. Differences in serologic levels were divided by the standard error of the cohort-specific distribution, obtaining standardized measurements. We fitted multivariate linear regression models to identify predictors of difference between two measurements. Results: We observed a progressively decreasing difference in serologic levels from <30 days to 210–240 days. Age was associated with an increased difference in serologic levels. There was a greater difference between the two serologic measurements in infected HCW than in HCW who had never been infected; before the first measurement, infected HCW had a relative risk (RR) of 0.81 for one standard deviation in the difference [95% confidence interval (CI) 0.78–0.85]. The RRs for a 30-day increase in time between first dose and first serology, and between the two serologies, were 1.08 (95% CI 1.07–1.10) and 1.04 (95% CI 1.03–1.05), respectively. The first measurement was a strong predictor of subsequent antibody decrease (RR 1.60; 95% CI 1.56– 1.64). Compared with Comirnaty, Spikevax (RR 0.83, 95% CI 0.75–0.92) and mixed vaccines (RR 0.61, 95% CI 0.51–0.74) were smaller decrease in serological level (RR 0.46; 95% CI 0.40–0.54). Conclusions: Age, COVID-19 infection, number of doses, time between first dose and first serology, time between serologies, and type of vaccine were associated with differences between the two serologic measurements within a 13-month period. KEYWORDS serial serology, antibodies, temporal trends, COVID-19 vaccination, health care workers Introduction infection and vaccination; conversely, new virus strains to which people are immunologically naïve require adjuvanted vaccines Vaccines are of utmost importance for human health and are and remain a possible cause of new pandemics (2). As reviewed by one of the first medical interventions that everyone receives at Castellino and coworkers, the Highly Pathogenic Avian Influenza birth (as long as they are available in the specific setting). 1 (H5N1) influenza strain, which spread in 1997 after its first Whereas vaccinations in childhood are usually long lasting, occurrence in Hong Kong, was effectively contained by MF59 with Measles–Mumps–Rubella (MMR), polio, and yellow fever adjuvanted vaccines, which were found to provide high levels of vaccines conferring lifelong immunity, vaccination effects in protection against H5N1 6 months after the administration of just adults are generally short lived (1). one boost (3). Messenger RNA (mRNA) vaccines have been Immune memory depends on the type of vaccine, and extensively studied in recent decades and have been determined vaccine-specific immunity can vary in different subjects to be an effective recourse in times of extreme need, given their depending on individual and environmental factors (2). Live rapid manufacturing process (4). Indeed, mRNA technology— attenuated vaccines are highly effective in providing lifelong Comirnaty and Spikevax—became the protagonist of the COVID- protection, whereas glycoconjugate vaccines’ immunity duration 19 pandemic and was crucial in controlling the infection derives from the characteristics of their carrier. Remarkably, non- worldwide. Viral vector vaccines, e.g., Vaxzevria, were also put adjuvanted vaccines provide sufficient protection against seasonal on the market and represent a novel approach that could be influenza because the population is already primed by previous implemented during future pandemics (5). Frontiers in Immunology 02 frontiersin.org Collatuzzo et al. 10.3389/fimmu.2022.1079884 The introduction of vaccines against COVID-19 was pivotal (Bucharest, Iasi, Cluj, and Timisoara). In both cohorts, local in effecting a substantial reduction of cases of symptomatic teams followed a unique protocol for blood sampling and disease and the number of hospitalizations and deaths (5–8). preparation and shipment of samples to the central laboratory. However, the administration of additional boosters with the aim Data on sociodemographic characteristics, PCR testing, and of prolonging their protective effect has proven to be necessary. vaccination status (including date of vaccination, number of doses, This may be due to the suboptimal duration of immunization and type of vaccination administered) were either abstracted from conferred by available vaccines and the appearance of new medical surveillance records or collected through questionnaires. COVID-19 variants such as Delta and Omicron (9). The levels of anti-COVID-19 S1 antibodies were derived from ORCHESTRA, a multicenter prospective cohort of medical records or generated through ad hoc testing. Because the healthcare workers (HCW) from different South-Eastern cohorts are included in the European Commission-sponsored European countries, was assessed with the aim of investigating Orchestra project, their data have undergone extensive COVID-19 infections in hospital personnel. harmonization. Methods of measurement of the antibody levels Previous analyses in this cohort described the determinants varied between the included centers and in the time periods. of 3- and 6-months serology following COVID-19 vaccination Different analytical methods were used for different cohorts. We (10, 11). The objective of the present study is to delineate the therefore log transformed the results and then we divided them by trends of serologic levels in the first 13 months post vaccination, the standard errors (SE) of each cohort-specific mean. We used the both overall and by month, in a population of approximately same approach in previous analyses within ORCHESTRA (10, 11). 20,000 HCW based on nine European study populations (five Selected characteristics of the study population are described in from Italy, two from Spain, one multicentric from Romania, and Table 1. This study comprises 19,923 HCW from Italy (Bari, Bologna, one multicentric from Slovakia) included in ORCHESTRA. Brescia, Trieste, and Verona), Spain (Northern Metropolitan Area of Barcelona (Barcelona) and Oviedo), Romania (multicentric), and Slovakia (multicentric), with multiple serologies during a 13-month Methods time frame from first dose administration (between December 2020 and March 2021, depending on the cohort), defined as the interval ORCHESTRA is a prospective multicenter cohort of HCW 150–210 days, including the 13-months serology. After excluding 501 from several European countries [https://orchestra-cohort.eu HCW who received fewer than two vaccine doses, 19,422 subjects (12)]. This analysis includes HCW in nine cohorts from four were included in the analysis. countries; of these, Slovakia and Romania were multicenter The outcome of this analysis was the difference between the cohorts. In Slovakia, participants were HCW and workers last and first serologic measurement of antibodies in the period 1– from social care facilities in COVID-19 departments from four 13 months from administration of the first dose of the vaccine. To regions (Banská Bystrica, Bratislava, Ružomberok, and Martin). avoid combining both positive and negative differences, we In Romania, participants were HCW employed at the National subtracted the observed difference from that of the subject with Institute of Public Health and worked in four different locations the highest increase between the two measurements. The analyses TABLE 1 Distribution of healthcare workers (HCW) by selected characteristics and mean difference between first and second serologic measurements. Variable N (%) Mean difference* (SE) Cohort Italy, Bari 157 (0.8) 1.21 (0.07) Italy, Bologna 5,536 (28.6) –0.28 (0.01) Italy, Brescia 5,881 (30.4) –0.33 (0.00) Italy, Trieste 2,058 (10.6) –0.69 (0.04) Italy, Verona 4,569 (23.6) –0.19 (0.01) Romania, multicenter 67 (0.3) –0.85 (0.08) Slovakia, multicenter 583 (3.0) –0.84 (0.05) Spain, Barcelona 480 (2.5) 0.37 (0.08) Spain, Oviedo 33 (0.2) –0.78 (0.13) (Continued) Frontiers in Immunology 03 frontiersin.org Collatuzzo et al. 10.3389/fimmu.2022.1079884 TABLE 1 Continued Variable N (%) Mean difference* (SE) Sex Male 5,295 (27.3) –0.29 (0.01) Female 14,069 (72.7) –0.31 (0.01) Age ≤ 29 2,263 (11.7) –0.36 (0.01) 30–39 4,288 (22.1) –0.36 (0.01) 40–49 4,714 (24.3) –0.33 (0.01) ≥ 50 8,099 (41.8) –0.23 (0.01) Job title Administration 1,568 (8.1) –0.25 (0.03) Physician (including residents) 4,989 (25.8) –0.27 (0.01) Nurse 7,310 (37.7) –0.33 (0.01) Technician 1,622 (8.4) –0.25 (0.03) Other HCW (including auxiliary workers) 3,875 (20.0) –0.34 (0.02) Previous COVID-19 infection Never infected 16,634 (85.9) –0.29 (0.01) Infected before first serologic measurement 2,327 (12.0) –0.62 (0.02) Infected between first and second serologic measurement 376 (1.9) 1.01 (0.07) Infected at both times 27 (0.1) 0.96 (0.29) Number of vaccine doses 2 doses 18,196 (94.0) –0.38 (0.01) 3 doses 1,168 (6.0) 0.92 (0.04) Time between first and second serologic measurements < 30d 13 (0.0) 0.74 (0.77) 30d–60d 173 (0.9) –0.37 (0.06) 60d–90d 649 (3.3) –0.34 (0.03) 90d–120d 3,162 (16.3) –0.30 (0.01) 120d–150d 5,548 (28.6) –0.36 (0.01) 150d–180d 1,806 (9.3) –0.48 (0.02) 180d–210d 4,823 (24.9) –0.44 (0.02) 210d–240d 1,780 (9.2) –0.30 (0.03) 240d–270d 603 (3.1) 0.23 (0.06) 270d–300d 426 (2.2) 0.95 (0.07) 300d–330d 274 (1.4) 1.01 (0.08) 330d–360d 71 (0.4) 0.68 (0.15) 360d–390d 19 (0.1) 1.25 (0.21) (Continued) Collatuzzo et al. 10.3389/fimmu.2022.1079884 TABLE 1 Continued Variable N (%) Mean difference* (SE) 390d–410d 16 (0.1) 1.92 (0.38) 410d–440d 1 (0.0) 0.89 (–) Time between first serologic measurement and first dose of vaccine <90d 9,989 (51.6) –0.29 (0.01) ≥90d 9,375 (48.4) –0.32 (0.01) * Difference in cohort-specific standardized antibody level. SE, standard error. aimed at examining this difference according to HCW and vaccine COVID-19 infection; the vast majority (94.0%) had received characteristics, including age, sex, study center, previous COVID- two doses of COVID-19 vaccine, and 1,178 (6.0%) from the Bari, 19 infection, number of doses, and type of vaccine. Bologna, Romania, Slovakia, Barcelona, and Oviedo cohorts We first conducted descriptive analyses of the outcome and received a third dose. The time between two consecutive explanatory variables. The main analysis involved calculating the serologic measurements ranged from less than 30 days to 440 difference in serologic response within 13 months. Multiple linear days, with 16.3% of measurements being performed between 90 regression models were used to calculate the relative risks (RR) and and 120 days, 28.4% between 120 and 150 days, 9.4% between their 95% confidence intervals (CI) for the difference between the 150 and 180 days, 24.9% between 180 and 210 days, and 9.3% last and first serologic measurement. Results are expressed as RR for between 210 and 240 days. We stratified the measurements one logarithmic unit in the difference of one standard error of the based on whether they were collected more or less than 90 days cohort-specific distribution between the last and the first serologic from the first dose of the COVID-19 vaccine (52.0% vs. 48.0%). measurement. Therefore, RRs greater than one correspond to a Most subjects who received two doses (98.2%) were given larger decrease in serologic level between the last and the first Comirnaty; the remaining subjects were given Spikevax and measurement compared with the reference category, and vice versa. other or mixed vaccines (not shown in detail). We performed a secondary analysis that was restricted to HCW When calculating the mean serology difference between two who received only two vaccine doses, because most subjects with an measurements, we observed some difference by study center, increase in antibody levels between the last and first measurements with Bari and Barcelona showing positive differences. This is had received three doses of the vaccine, with the last measurement mainly explained by the fact that several blood samples in these taken after the third dose. cohorts were collected after the third vaccine dose for a second The models included terms for cohort, sex, age (10-year increase), time. Figure 1 illustrates the timing of serologic measurements job title, number of vaccine doses received, previous COVID-19 by study center. As the dots contain the average number of days infection, time difference between first serology measure and first between first and second serology, last serology appears to be dose of vaccine (30-day increase), and time difference between first infrequently performed after the booster vaccine dose, and blood serology measure and second serology measure (30-day increase). samples were collected at quite different time points in each Stata® software 16 (StataCorp LP, College Station, Texas, center. On the other hand, vaccinations were administered USA) was used for the statistical analysis. mostly with the same timing, with approximately 30 days The study was approved by the Italian Medicine Agency between the first and second doses, and the booster dose (AIFA) and the Ethics Committee of the Italian National administered approximately 300 days after first vaccination. Institute of Infectious Diseases (INMI) Lazzaro Spallanzani. According to the time between the two serologic samples, we calculated a progressively reduced difference from <30 days to 210–240 days (SE 0.74 to –0.30), and a subsequent increase in Results the difference until days 410–440 (SE 0.22 to 3.10). No Overall, our analysis included 19,422 HCW, with repeated substantial difference was observed in the time between first measurements performed within 13 months after first dose and serology and first vaccination (SE –0.30 and –0.32, respectively). with two or more vaccine doses. The distribution of Table 2 shows the results of the main analysis. Compared measurements by time since first dose and cohort is shown in with the cohort from Bologna, the cohorts from Bari, Trieste, Supplementary Figure 1. Most of the study population consisted Verona, Romania, Slovakia, and Barcelona and Oviedo of women (72.6%), and a large proportion of subjects were ≥ 50 experienced a larger decrease in their antibody levels, and no years old (42.5%). HCW were mostly nurses (37.7%) and cohorts experienced a smaller decrease. Age was associated with physicians (25.7%). Overall, 85.8% had had no previous a larger decrease in serologic levels. No difference was found by Frontiers in Immunology 05 frontiersin.org Collatuzzo et al. 10.3389/fimmu.2022.1079884 FIGURE 1 Timing of vaccination and serologic measurements by cohort. sex or job title, and HCW with a history of COVID-19 infection samples, was associated with the opposite; i.e. a larger decrease in experienced a smaller decrease in antibody levels. The RR for a antibody levels. While several of these factors were correlated 30-day increase in time between first vaccine dose and first with one another, the large sample size allowed us to conduct serology was 1.08 (95% CI 1.07–1.10) and for a 30-day increase multivariate analyses, whose stable results should exclude between the two samples was 1.04 (95% CI 1.03–1.05). The reciprocal confounding. antibody level at first measurement was a strong predictor of When considering previous COVID-19 infection, a subsequent decrease in the serologic level (RR 1.60, 95% CI 1.56– significantly smaller decrease was found in subjects reporting 1.64). Compared with Comirnaty, Spikevax was associated with previous infection compared with those never infected. In a smaller decrease in the serologic level (RR 0.83, 95% CI 0.75– particular, a one standard error decrease in serologic level 0.92), as was the administration of mixed vaccines (RR 0.61, 95% was 67% less likely to occur in HCW who became infected 0.51–0.74). Finally, administration of a third vaccine dose was between the first and second serologic measurements associated with a smaller decrease in the serologic level (RR 0.46, compared with those who were never infected. This can be 95% CI 0.40–0.54). explained by the renewal of the immunological peak driven by When the analysis was restricted to HCW who were the infection, leading to a stable antibody level—steadily higher administered two vaccine doses (N=18,347, Supplementary than average—while a difference can be appreciated in naïve Table 1), the results were comparable to those reported for the subjects, who experience only a decline in the antibody level whole study population. The only different result was observed with time. Among subjects infected before the first serologic for gender, with women exhibiting a smaller decrease in sampling. There was a larger difference in serologic level antibody levels than men (RR 0.97, 95% CI 0.95–0.99). among subjects infected before the first serological sample than among those infected in other time frames, which may be explained by the fact that infection confers a high degree of Discussion immunization, thus registering a higher antibody level at the first measurement compared with naïve subjects. Although the The present study analyzed several potential clinical, effectiveness of COVID-19 vaccines has been debated and individual, and work-related determinants of the difference in given the prevalence of breakthrough infection, when serologic level between multiple measurements after COVID-19 considering our data, the number of infections occurring in vaccination over 13 months in a large European cohort of HCW. vaccinated HCW is approximately one-sixth of the number of Although the effect of sex appeared to be modest, increasing age infections that occurred before vaccination, corroborating the was associated with a larger decrease in antibody levels. There fact that vaccination is the most important instrument in the were no differences according to job title. Previous COVID-19 prevention of communicable diseases (13–15). infection, having received Spikevax or a mixed vaccine, and Subjects administered with three doses of the vaccine administration of a third dose of the vaccine resulted in a smaller experienced a 54% lower probability of undergoing a one decrease in antibody levels, whereas an increase in time between standard error decrease in antibody level than subjects who had first vaccine dose and first serology, and that between the two received two doses. This may be due to the timing of the blood Frontiers in Immunology 06 frontiersin.org Collatuzzo et al. 10.3389/fimmu.2022.1079884 TABLE 2 Multivariate analysis of characteristics associated with changes in antibody levels between first and second serologic measurement. Characteristics RR 95% CI Cohort Italy, Bari 1.90 1.53–2.36 Italy, Bologna Ref Italy, Brescia 1.03 0.99–1.06 Italy, Trieste 2.49 2.34–2.64 Italy, Verona 1.68 1.59–1.78 Romania, multicenter 1.80 1.56–2.07 Slovakia, multicenter 2.77 2.55–3.00 Spain, Barcelona 1.81 1.62–2.02 Spain, Oviedo 0.89 0.59–1.33 Sex Male Ref Female 0.98 0.96–1.00 Age 10 years increase 1.03 1.02–1.04 Job title Administration Ref Physician (including residents) 1.01 0.97–1.05 Nurse 1.01 0.98–1.05 Technician 0.99 0.95–1.04 Other HCW (including auxiliary workers) 1.01 0.97–1.05 Previous COVID-19 infection Never infected Ref Infected before the first serologic measurement 0.81 0.78–0.85 Infected between the two serologic measurements 0.33 0.29–0.38 Infected at both times 0.70 0.50–0.99 Number of vaccine doses 2 doses Ref 3 doses 0.46 0.40–0.54 Time between first vaccine dose and first serology measurement 30 days increase 1.08 1.07–1.10 Time between first and second serology measurements 30 days increase 1.04 1.03–1.05 Antibody level at first serologic measurement 1 SD increase in ln(antibody level) 1.60 1.56–1.64 Type of vaccine – (Continued) Frontiers in Immunology 07 frontiersin.org Collatuzzo et al. 10.3389/fimmu.2022.1079884 TABLE 2 Continued Characteristics RR 95% CI Comirnaty Ref Spikevax 0.83 0.75–0.92 Vaxzevria 0.98 0.64–1.50 Mixed vaccines 0.61 0.51–0.74 Missing 0.44 0.37–0.52 RR, relative risk for decrease of one standard error of the cohort-specific distribution of ln(antibody level) between the last and the first serologic measurements, adjusted for all the variables in the table. CI, confidence interval; HCW, healthcare worker; Ref, reference category; SD, standard deviation. sampling, specifically, the overlap between the second The serologic trend was likely to decline with every 30 day measurement and the number of days following the booster increase in time between first vaccination and first serologic dose, which determines the immunological response. Indeed, (i) measurement, corroborating the evidence of a progressive most of the HCW from Bari underwent their last serology test decline in the level of antibodies and providing a precise after their third dose of the vaccine and (ii) on average, in quantification of the effect (9). When considering 30 day-time Barcelona, HCW underwent their first serology test at the same increases between the two serologic measurements, the declining time as their second dose of the vaccine (Figure 1). This is also trend was smaller. Moreover, the higher the serologic level captured by the variable related to time between the serologic detected at the first measurement, the larger the difference measurements: between 1 and 7 months the difference in serologic detected at the second blood sample. level increases, which implies a rapid decline in the number of We found a small difference in the serologic samplings by antibodies; an increase in serologic levels can be detected between type of vaccine, namely Comirnaty or Spikevax, with the former 8 and 13 months. This largely corresponds to the period when leading to a slightly smaller reduction in the antibody level than HCW were recommended to obtain the third dose of vaccine. the latter. However, Spikevax has been shown to confer greater Age was related to the progressive decline in trends of serologic protection in the long term than Comirnaty (23). Based on our levels, expressed through a 3% higher likelihood of undergoing a one analysis, Spikevax resulted in a 20% increase of the serologic standard error decrease in antibody level with every10-year increase level compared with Comirnaty, which is consistent with in age. Several studies have demonstrated the lower levels of previous evidence (23). This difference was attributed to a immunological responsiveness of older subjects (16). This is higher mRNA content in Spikevax that in Comirnaty and the consistent with our previous study that showed age as a longer interval between priming and boosting for Spikevax (4 determinant of lower level of serologic response in HCW (11). weeks for Spikevax vs. 3 weeks for Comirnaty) (23). Overall, When focusing on gender differences, a 2% smaller probability of large differences can be seen among viral vector vaccines and undergoing a one standard error decrease in antibody level was mRNA vaccines (24, 25), and our data did not allow us to suggested for women, consistent with previous literature (17–19). address this question in detail. Indeed, sexual dimorphism in immune responses has been described, The description of these trends, together with the evidence of which is more evident following infection or vaccination (20, 21). increased difference in antibody level in HCW who received a With regard to job title, no difference was detected. This is third vaccine dose, corroborates previous findings on the ability likely to be due to the adjustments for different potential of vaccines to stimulate immunological response (26). Assuming confounders which may mediate an effect on occupational a proportional capacity of the immune system to protect from categories, because of the different risk exposure among infection, this implies a greater ability of the vaccines to professional healthcare figures. Previous studies focused on guarantee protection against COVID-19 infection when HCW as a job category rather than distinguishing potential booster doses are administered; thus the greater effectiveness association with COVID-19 infection and vaccination outcomes. of a three-dose scheme of vaccination (27). Despite this, the For example, in a previous analysis we once again found no present analysis can only allow assumptions on the antibody relationship between job title and immunological response (11). levels as a sign of effective vaccination. Moreover, a previous study based on the Italian cohort of the Describing the trends of antibody levels after COVID-19 ORCHESTRA project found no difference in the risk of HCW vaccination is currently a major issue. Indeed, it enables to read contracting COVID-19, even when considering HCW in the pandemic and the public health interventions which have departments dedicated to COVID-19 (22). been introduced to control it in the light of quantitative data, and Frontiers in Immunology 08 frontiersin.org Collatuzzo et al. 10.3389/fimmu.2022.1079884 to understand if and to what extent vaccination schemes have same method was used in our previous analysis (11). This been useful and effective. Vaccines are usually administered in approach is particularly valuable given the global connotation multiple doses (28). Viral pathogens that have a short life cycle of COVID-19 infection, because it allows us to overcome and a high rate of replication are subject to higher mutagenicity, potential heterogeneity in data collection methods and to leading to continuous exposure to potential infectious risk for compare data from different populations. Indeed, all the both unvaccinated and vaccinated people. This is what differences found in the present analysis must be interpreted commonly happens with influenza viruses, requiring as representing real phenomena rather than being attributed to vaccination every year (29). Indeed, HCW in most countries differences in serologic testing across the cohorts. are strongly recommended to receive an influenza vaccination Besides this, our analysis was limited by the small number of every year as they are occupationally at a higher risk of infection; HCW who had three doses of the vaccine, leading to a low power both for their own protection and to prevent transmission in the of analysis of the effect of the third dose. Moreover, only a small hospital environment (30, 31). number of HCW had three or more measurements of antibody The introduction of COVID-19 vaccines into usual medical levels. Thus, we used only the first and the last measurements practice has been hypothesized based on the circumstances registered because most subjects had received only two vaccine doses. worldwide (32, 33). The main issue relates to the capacities of the Future analysis of this population will include more subjects COVID-19 vaccines to reduce the risk of symptomatic disease, and less with three or more measurements of antibody levels, allowing so to reduce the risk of becoming infected (34). for the assessment of non-linear trends in antibody levels. This study has some strengths and limitations. To our We could not distinguish individuals by health status, and knowledge, this is one of the largest analyses on changes in therefore could not analyze potential variability in antibody individual-level data regarding antibody level after COVID-19 levels due to conditions such as immunodepression. However, vaccination. Most previous studies were based on comparisons of this is a working population, representing an overall healthy groups of individuals tested at different times after vaccination, group, and HCW in particular have been reported to be healthier including a large prospective study conducted in England on more than the general population (40). In addition, given the paucity than 212,000 vaccinated subjects (35). Individual-level serologies of data on different types of vaccine, the possible comparisons were collected in other studies, which, however, included no more were limited in terms of viral vector vaccines. than a few hundred participants (36–39). The large sample size, Furthermore, subclinical infections were not systematically together with the collection of detailed information, allowed assessed (anti-nucleocapsid antibodies were available for only a assessment of multiple factors associated with trends in subset of participants). This may have diluted the estimated antibody levels within 13 months from vaccination, enabling effect of previous COVID-19 infection. reciprocal adjustment. Moreover, the prospective design of this study enables further follow-up with the participants, offering the possibility to expand this investigation and add interesting insight Conclusions in future analyses. In addition, this is one of the few studies covering a 13-month period when considering COVID-19 This analysis defines the trends of serologic levels in HCW vaccination-related immunity. who received multiple COVID-19 vaccines within the first 13 The results we obtained are consistent with previous months, by calculating the mean difference observed in nine literature (34), showing robustness, and are important in European cohorts. Positive trends were detected in those centers providing further confirmation of individual-level data on the where serologic samples were collected around the time of progressive waning of COVID-19 vaccine immunization by vaccination, as well as among previously infected HCW. enhancing the confidence of these observations. This gives Increasing age, time between first vaccine and first serology, additional strengths to previous data. Furthermore, we added and time between serologies determined a negative trend in useful information on time-trend serology from first vaccination antibody level. Spikevax was associated with a smaller decrease and in between serologic measurements, as well as comparisons in the serologic level than Comirnaty. The booster dose between naïve HCW and those previously infected with determines the renewal of the immunological response, COVID-19. We used a strong methodological approach to expressed as a smaller decrease in the serologic level. standardize the level of antibodies, overcoming the issue of These individual-level data support published studies different types of tests adopted in the various cohorts, making showing a progressive decline of immune response after the measurement from different populations comparable. COVID-19 vaccination. Analyses over longer time frames The main limitation of this study was the heterogeneity of would be of interest to better understand the longevity of testing methods used within the cohorts. However, as mentioned COVID-19 vaccine immunization. Such studies, combined above, we addressed this issue by using the standardized log with our cohort, would provide further information on the value and calculating RR per one SD increase, thus generating optimal time frame of administration of subsequent doses. harmonized results on measurements of antibody levels. The This information should be combined with a risk profile for Frontiers in Immunology 09 frontiersin.org Collatuzzo et al. 10.3389/fimmu.2022.1079884 COVID-19 infection to improve the quantitative information Verona is funded by the Regional Health Authority (Azienda needed to optimize the schedule of vaccine administration, and Zero), Veneto Region, Italy. the subgroups of population to be prioritized for boosters. Further studies are warranted to further describe temporal changes in serologic levels after COVID-19 vaccination and to Acknowledgments clarify the role of different types of vaccines and the timing of infection. To better address the pandemic and manage Bologna, Italy cohort: Laboratory of Microbiology of the vaccination strategy at both the occupational and population- University Hospital. Trieste, Italy cohort: All personnel of the based level, studies focused on the protectiveness of vaccination- Clinical Unit of Occupational Health, Laboratory of Virology driven antibodies are needed. and Laboratory of Microbiology of the University Hospital. Verona, Italy cohort: General Management, Medical Management, and all personnel of the Units of Occupational Data availability statement Health, Laboratory Medicine and Microbiology and of University Hospital of Verona and all personnel of the Unit of The raw data supporting the conclusions of this article will Epidemiology and Medical Statistics, University of Verona. be made available by the authors, without undue reservation Slovakia cohort: Alena Koš č á lová , Infectology Clinic, upon reasonable request. University Hospital, Bratislava; Oto Osina, Occupational Medicine Clinic, University Hospital, Martin; Zuzana Sirotná , Laboratory Dpt, Public Health Authority of the Slovak Republic, Ethics statement Bratislava, Jarmila Belá ková and co-workers, Occupational health Dpt., Regional Authority of Public Health, Banská The study was approved by the Italian Medicine Agency Bystrica; Alexandra Braž inová , Institute of Epidemiology, (AIFA) and the Ethics Committee of the Italian National Faculty of Medicine Comenius University, Bratislava. The Institute of Infectious Diseases (INMI) Lazzaro Spallanzani. Oviedo Cohort is funded by Health Research Institute of The patients/participants provided written informed consent Asturias (ISPA), the Epidemiology and Public Health Ciber to participate in this study as well as by local ethics committees. (CIBERESP), and the University of Oviedo, Asturias, Spain. Romania multicenter cohort: National Institute of Public Health (INSP), Bucharest. Author contributions GC and PB conceived and designed the study. PB coordinated Conflict of interest the international collaboration. PB, FV, VL, GC, and CZ coordinated and conducted the study in Bologna, Italy. GDP, The authors declare that the research was conducted in the ESan, and ESal coordinated and conducted the study in Brescia, absence of any commercial or financial relationships that could Italy. SP, MM, and GS coordinated and conducted the study in be construed as a potential conflict of interest. Verona, Italy. FLF and CN coordinated and conducted the study in Trieste, Italy. CV and LC-R coordinated and conducted the study in Barcelona, Spain. ML coordinated and conducted the Publisher’s note study in Bucarest, Romania. EF and JB coordinated the study in Slovakia. MR-S coordinated and conducted the study in Oviedo, All claims expressed in this article are solely those of the Spain. LV and ST coordinated and conducted the study in Bari, authors and do not necessarily represent those of their affiliated Italy. MA, GP, and SA performed the statistical analysis. GC and organizations, or those of the publisher, the editors and the PB supervised the statistical analysis. GC and PB drafted the reviewers. Any product that may be evaluated in this article, or manuscript. All authors reviewed and approved the final version claim that may be made by its manufacturer, is not guaranteed of the manuscript. or endorsed by the publisher. Funding Supplementary material This project has received funding from the EU Horizon 2020 The Supplementary Material for this article can be found research and innovation programme under the ORCHESTRA online at: https://www.frontiersin.org/articles/10.3389/ project Grant Agreement No. 101016167. The cohort from fimmu.2022.1079884/full#supplementary-material Frontiers in Immunology 10 frontiersin.org Collatuzzo et al. 10.3389/fimmu.2022.1079884 References 1. Amanna IJ, Slifka MK. Questions regarding the safety and duration of 22. Boffetta P, Violante F, Durando P, De Palma G, Pira E, Vimercati L, et al. immunity following live yellow fever vaccination. Expert Rev Vaccines (2016) 15 Determinants of SARS-CoV-2 infection in Italian healthcare workers: a multicenter (12):1519–33. doi: 10.1080/14760584.2016.1198259 study. Sci Rep (2021) 11(1):5788. doi: 10.1038/s41598-021-85215-4 2. Castellino F, Galli G, Del Giudice G, Rappuoli R. Generating memory with 23. Steensels D, Pierlet N, Penders J, Mesotten D, Heylen L. Comparison of vaccination. Eur J Immunol (2009) 39(8):2100–5. doi: 10.1002/eji.200939550 SARS-CoV-2 antibody response following vaccination with BNT162b2 and 3. Galli G, Hancock K, Hoschler K, DeVos J, Praus M, Bardelli M, et al. Fast rise mRNA-1273. JAMA (2021) 326(15):1533–5. doi: 10.1001/jama.2021.15125 of broadly cross-reactive antibodies after boosting long-lived human memory b 24. van Gils MJ, Lavell A, van der Straten K, Appelman B, Bontjer I, Poniman cells primed by an MF59 adjuvanted prepandemic vaccine. Proc Natl Acad Sci U S M, et al. Antibody responses against SARS-CoV-2 variants induced by four A (2009) 106(19):7962–7. doi: 10.1073/pnas.0903181106 different SARS-CoV-2 vaccines in health care workers in the Netherlands: A 4. Zhang C, Maruggi G, Shan H, Li J. Advances in mRNA vaccines for infectious prospective cohort study. PloS Med (2022) 19(5):e1003991. diseases. Front Immunol (2019) 10:594. doi: 10.3389/fimmu.2019.00594 25. Fan YJ, Chan KH, Hung IF. Safety and efficacy of COVID-19 vaccines: A 5. de Vries RD, Rimmelzwaan GF. Viral vector-based influenza vaccines. Hum systematic review and meta-analysis of different vaccines at phase 3. Vaccines Vaccin Immunother (2016) 12(11):2881–901. doi: 10.1080/21645515.2016.1210729 (Basel) (2021) 9(9):989. doi: 10.3390/vaccines9090989 6. Tenforde MW, Self WH, Adams K, Gaglani M, Ginde AA, McNeal T, et al. ; Influenza 26. Slifka MK, Amanna I. How advances in immunology provide insight into improving and other viruses in the acutely ill (IVY) network. association between mRNA vaccination and vaccine efficacy. Vaccine (2014) 32(25):2948–57. doi: 10.1016/j.vaccine.2014.03.078 COVID-19 hospitalization and disease severity. JAMA (2021) 326(20):2043–54 doi: 10.1001/ 27. Barda N, Dagan N, Cohen C, Herná n MA, Lipsitch M, Kohane IS, et al. jama.2021.19499. Effectiveness of a third dose of the BNT162b2 mRNA COVID-19 vaccine for 7. Cabezas C, Coma E, Mora-Fernandez N, Li X, Martinez-Marcos M, Fina F, et al. preventing severe outcomes in Israel: an observational study. Lancet (2021) 398 Associations of BNT162b2 vaccination with SARS-CoV-2 infection and hospital (10316):2093–100. doi: 10.1016/S0140-6736(21)02249-2 admission and death with covid-19 in nursing homes and healthcare workers in 28. Di Pietrantonj C, Rivetti A, Marchione P, Debalini MG, Demicheli V. Catalonia: prospective cohort study. BMJ (2021) 374:n1868. doi: 10.1136/bmj.n1868 Vaccines for measles, mumps, rubella, and varicella in children. Cochrane Database 8. Lopez Bernal J, Andrews N, Gower C, et al. Effectiveness of the pfizer- Syst Rev (2020) 4(4):CD004407. BioNTech and Oxford-AstraZeneca vaccines on covid-19 related symptoms, 29. Skeik N, Jabr FI. Influenza viruses and the evolution of avian influenza virus hospital admissions, and mortality in older adults in England: test negative case- H5N1. Int J Infect Dis (2008) 12(3):233–8. doi: 10.1016/j.ijid.2007.07.002 control study. BMJ (2021) 373:n1088. 30. Stokel-Walker C. Covid-19: The countries that have mandatory vaccination 9. Fiolet T, Kherabi Y, MacDonald CJ, Ghosn J, Peiffer-Smadja N. Comparing for health workers. BMJ (2021) 373:n1645. doi: 10.1136/bmj.n1645 COVID-19 vaccines for their characteristics, efficacy and effectiveness against 31. Dini G, Toletone A, Sticchi L, Orsi A, Bragazzi NL, Durando P. Influenza SARS-CoV-2 and variants of concern: a narrative review. Clin Microbiol Infect vaccination in healthcare workers: A comprehensive critical appraisal of the literature. (2022) 28(2):202–21. doi: 10.1016/j.cmi.2021.10.005 Hum Vaccin Immunother (2018) 14(3):772–89. doi: 10.1080/21645515.2017.1348442 10. Visci G, Zunarelli C, Mansour I, Porru S, De Palma G, Duval X, et al. 32. Gibelli F, Ricci G, Sirignano A, De Leo D. COVID-19 compulsory Serological response after SARS-CoV2 vaccination in healthcare workers: a vaccination: Legal and bioethical controversies. Front Med (Lausanne) (2022) multicenter study. Med Lav (2022) 113(2):e2022022 doi: 10.23749/mdl.v113i2.13017. 9:821522. doi: 10.3389/fmed.2022.821522 11. Collatuzzo G, Visci G, Violante FS, Porru S, Spiteri G, Monaco MGL, et al. 33. Opel DJ, Diekema DS, Ross LF. Should we mandate a COVID-19 vaccine for Orchestra WP5 Working Group. Determinants of anti-S immune response at 6 children? JAMA Pediatr (2021) 175(2):125–6. doi: 10.1001/jamapediatrics.2020.3019 months after COVID-19 vaccination in a multicentric European cohort of healthcare workers - ORCHESTRA project. Front Immunol (2022) 13:986085. 34. Hodgson SH, Mansatta K, Mallett G, Harris V, Emary KRW, Pollard AJ. doi: 10.3389/fimmu.2022.986085 What defines an efficacious COVID-19 vaccine? a review of the challenges assessing the clinical efficacy of vaccines against SARS-CoV-2. Lancet Infect Dis 12. Available at: https://orchestra-cohort.eu. (2021) 21(2):e26–35. doi: 10.1016/S1473-3099(20)30773-8 13. Fowlkes A, Gaglani M, Groover K, Thiese MS, Tyner H, Ellingson K, et al. 35. Ward H, Whitaker M, Flower B, Tang SN, Atchison C, Darzi A, et al. Effectiveness of COVID-19 vaccines in preventing SARS-CoV-2 infection among frontline Population antibody responses following COVID-19 vaccination in 212,102 workers before and during B.1.617.2 (Delta) variant predominance - eight U.S. locations, individuals. Nat Commun (2022) 13(1):907. doi: 10.1038/s41467-022-28527-x December 2020-august 2021. MMWR Morb Mortal Wkly Rep (2021) 70(34):1167–9 doi: 10.15585/mmwr.mm7034e4. 36. Wheeler SE, Shurin GV, Yost M, Anderson A, Pinto L, Wells A, et al. Differential antibody response to mRNA COVID-19 vaccines in healthy subjects. 14. Bajema KL, Dahl RM, Evener SL, Prill MM, Rodriguez-Barradas MC, Marconi Microbiol Spectr (2021) 9(1):e0034121. doi: 10.1128/Spectrum.00341-21 VC, et al. Comparative effectiveness and antibody responses to moderna and pfizer- BioNTech COVID-19 vaccines among hospitalized veterans - five veterans affairs 37. Ikezaki H, Nomura H, Shimono N. Dynamics of anti-spike IgG antibody medical centers, united states, February 1-September 30, 2021. MMWR Morb Mortal level after the second BNT162b2 COVID-19 vaccination in health care workers. J Wkly Rep (2021) 70(49):1700–5. doi: 10.15585/mmwr.mm7049a2 Infect Chemother (2022) 28(6):802–5. doi: 10.1016/j.jiac.2022.02.024 15. Doherty M, Buchy P, Standaert B, Giaquinto C, Prado-Cohrs D. Vaccine impact: 38. Choi JH, Kim YR, Heo ST, Oh H, Kim M, Lee HR, et al. Healthcare workers Benefits for human health. Vaccine (2016) 34(52):6707–14. doi: 10.1016/j.vaccine.2016.10.025 in south Korea maintain a SARS-CoV-2 antibody response six months after 16. Ssentongo P, Ssentongo AE, Voleti N, Groff D, Sun A, Ba DM, et al. SARS-CoV-2 receiving a second dose of the BNT162b2 mRNA vaccine. Front Immunol (2022) vaccine effectiveness against infection, symptomatic and severe COVID-19: a systematic 13:827306. doi: 10.3389/fimmu.2022.827306 review and meta-analysis. BMC Infect Dis (2022) 22(1):439. doi: 10.1186/s12879-022- 39. Zhang S, Xu K, Li C, Zhou L, Kong X, Peng J, et al. Long-term kinetics of SARS-CoV-2 07418-y antibodies and impact of inactivated vaccine on SARS-CoV-2 antibodies based on a COVID-19 patients cohort. Front Immunol (2022) 13:829665. doi: 10.3389/fimmu.2022.829665 17. He Q, Mao Q, Zhang J, Bian L, Gao F, Wang J, et al. COVID-19 vaccines: Current understanding on immunogenicity, safety, and further considerations. 40. Dayoub E, Jena AB. Chronic disease prevalence and healthy lifestyle Front Immunol (2021) 12:669339. doi: 10.3389/fimmu.2021.669339 behaviors among US health care professionals. Mayo Clin Proc (2015) 90 (12):1659–62. doi: 10.1016/j.mayocp.2015.08.002 18. Li Z, Xiang T, Liang B, Deng H, Wang H, Feng X, et al. Characterization of SARS-CoV-2-Specific humoral and cellular immune responses induced by COPYRIGHT inactivated COVID-19 vaccines in a real-world setting. Front Immunol (2021) 12:802858. doi: 10.3389/fimmu.2021.802858 © 2023 Collatuzzo, De Palma, Violante, Porru, Larese Filon, Fabianova, ́ , Vimercati, Leustean, Rodriguez-Suarez, Sansone, Sala, Zunarelli, Violan 19. Yalç ı n TY, Topç u DI, Doğ a n Ö, Aydın S, Sarı N, Erol Ç , et al. Immunogenicity after two doses of inactivated virus vaccine in healthcare Lodi, Monaco, Spiteri, Negro, Beresova, Carrasco-Ribelles, Tafuri, Asafo, workers with and without previous COVID-19 infection: Prospective Ditano, Abedini and Boffetta. This is an open-access article distributed observational study. J Med Virol (2022) 94(1):279–86. under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, 20. Pennell LM, Galligan CL, Fish EN. Sex affects immunity. J Autoimmun provided the original author(s) and the copyright owner(s) are credited (2012) 38(2-3):J282–91. doi: 10.1016/j.jaut.2011.11.013 and that the original publication in this journal is cited, in accordance 21. Klein SL, Flanagan KL. Sex differences in immune responses. Nat Rev with accepted academic practice. No use, distribution or reproduction Immunol (2016) 16(10):626–38. doi: 10.1038/nri.2016.90 is permitted which does not comply with these terms. Frontiers in Immunology 11 frontiersin.org