Advertisement

Susceptibility to Reinfection with SARS-CoV-2 Virus Relative to Existing Antibody Concentrations and T cell Response

Open AccessPublished:January 23, 2023DOI:https://doi.org/10.1016/j.ijid.2023.01.006

      Abstract

      Background: We have investigated the reinfection rate of vaccinated or convalescent immunized SARS-CoV-2 in 952 expatriate workers with SARS-CoV-2 serological antibody patterns and surrogate T cell memory at recruitment and follow up.
      Methods: Trimeric spike, nucleocapsid, and neutralizing antibodies were measured along with a T cell stimulation assay targeting SARS-CoV-2 memory in CD4+ and CD8+ T cells. The subjects were then followed up for reinfection for up to six months.
      Results: Seroprevalence positivity at enrollment was greater than 99%. T cell reactivity in this population was 38.2%. Of the 149 (15.9%) participants that were re-infected during the follow up period (74.3%) had nonreactive T cells at enrollment. Those who had greater than 100 BAU/mL increase from the median concentration of Anti-S IgG antibodies had a 6% reduction in the risk of infection. Those who were below the median concentration had a 78% greater risk of infection.
      Conclusions: Significant immune protection to reinfection was observed in those that retained T cell activation memory. Additional protection was observed when Anti-S was greater than the median value.
      • SARS-CoV-2 immunocompetence
      • seroprevalence
      • T cell response
      • neutralizing antibodies
      • spike and receptor binding protein antibodies
      • reinfection.

      INTRODUCTION

      Knowledge of the prevalence of active infections, and immune status from vaccinations and or convalescence is critical to manage health care resources efficiently. The Centers for Disease Control and World Health Organization [1, 2] recommend periodic SARS-CoV-2 serology surveys. Information obtained from serology and NAAT surveys are used to understand how SARS-CoV-2 passes through communities and how individual immune responses can confer protection against reinfection as many countries are in the midst of their fifth wave of the infection [1, 3]
      In July 2020, the United Arab Emirates (UAE) conducted the first national population-based seroprevalence study with over 13,000 participants. The study uncovered a striking difference in infectivity and seroprevalence between expatriate workers housed in company sponsored accommodation (See supplement) and individuals living in households [4]. Expatriate accommodation is required for certain worker categories per UAE regulations [5]. A significantly higher seroprevalence was found in the expatriate accommodations, reflecting the high efficiency with which SARS-CoV-2 spreads in dormitory housing [4]. The communal living conditions in the worker accommodations were likely the major contributing factor to this difference between prevalence in households at 10.4% compared to 68.6% in expatriate housing [4]. In the initial study workers were observed to have higher concentrations of the viral spike protein IgG binding antibodies than individuals living in households indicating immune boosting from more frequent exposure or reinfection. The UAE reacted quickly to the pandemic and began early vaccination with Sinopharm vaccination followed by boosters of the Pfizer vaccine approximately six months post completion of the initial course [6]. This naturally exposed and vaccinated worker population is ideal to study immune status and susceptibility to infection with SARS-CoV-2.
      Understanding susceptibility to infection with SARS-CoV-2 by the level of putative protective antibody concentrations and T cell reactivity, may help us understand the timing and need for re-vaccination and other preventive measures.
      This prospective study investigated, at enrollment, the prevalence of four SARS-CoV-2 immune markers; anti-spike (anti-S) and anti-nucleocapsid (anti-N) IgG antibodies (Abs), neutralizing IgG Abs (Nabs), and circulating T cell reactivity (CD8+ and CD4+/8+ cells) in expatriate workers housed in sponsored accommodations approximately two years after an initial infectivity and serology study. The aim of this study was to observe whether re-infection was common in this infected and vaccinated population and if specific antibody concentrations may have conferred a protective effect in a dose-dependent manner.

      MATERIALS and METHODS

      Please refer to the supplemental materials for the detailed materials and methods.
      Enrollment of expatriate workers who had tested positive for anti-S IgG and anti-N IgG antibodies at the onset of the pandemic, were recruited (N= 952). Nurses provided lay person study descriptions verbally and in writing in either English, Arabic or Urdu. Signed informed consent was obtained from all the participants. The Abu Dhabi Department of Health Institutional Review Board approved the protocol. No inducements or monetary compensation was offered to participate. A nasopharyngeal swab, blood specimens and a questionnaire were obtained at recruitment. Health information was obtained from the national healthcare database, Malaffi [7].
      Tests for neutralizing, Spike 1, Spike 2, receptor binding domain and nucleocapsid antibodies were performed by chemiluminescent techniques. T cell response to SARS CoV-2 was determined using an interferon gamma releasing assay. PCR for SARS CoV-2 nucleic acids was performed using the N gene and ORF1a targets at the time of collection. Follow up PCR testing was obtained from the Malaffi database.
      The Research Use Only version of the SARS-CoV-2 interferon gamma releasing assay (Qiagen, Germany) was used to detect T cell memory to SARS-CoV-2 antigens.
      To determine if specific antibodies, their concentrations as well as T cell mediated immunity had a protective effect on reinfection statistical analyses were performed using SPSS IBM Statistics (v26). P-values <0.05 were considered statistically significant. To correlate the level of anti-spike IgG levels to immunity and protection from SARS-CoV-2 infection, we followed up our study group for subsequent positive SARS-CoV-2 PCR result using data obtained from Malaffi [7] to detect reinfected cases after samples collection until February 2022.
      Ethical considerations
      The study was reviewed and approved by the UAE National COVID-19 Research Ethics Committee DOH/CVDC/2021/856 and amendment number: DOH/CVDC/2021/1703.

      RESULTS

      Please refer to the supplemental materials for the detailed results and tables.

      Characteristics of expatriate workers at enrollment

      Of the 3,585 male expatriate workers listed to be enrolled and followed up, between October 3 and December 15, 2021, 952 workers were available and enrolled (Figure 1). Mean age of our enrolled cohort was 35.5 years (SD 8.40), the majority (92.5%) were Asians and 11.6% had at least one existing chronic comorbidity, mainly hypertension (7%). Almost all (98.6%) had received at least one vaccine course and over three-quarters (79.1%) were boosted with at least one dose while 2.2% were boosted with two vaccine doses against SARS-CoV-2. The mean duration between the last received vaccine dose and blood collection was 89.2 days (SD 54.5, range 1–295 days). Regardless of the number of received doses, 91.8% were vaccinated with Sinopharm vaccine. 12.2% of the cohort reported testing PCR positive in the past 12 months prior to enrollment (Table 1).
      Figure 1
      Figure 1Flowchart of study population included in this study
      1 Either left the country, moved to different camps, or possibly were deceased or where on annual leave, offsite, or on-duty during the scheduled survey date (n= 63).
      2 Eleven cases had no information on their PCR testing results during the follow up, possibly they just left their workplace after the survey.
      Table 1Distribution of the surveyed labor workers by their measured sociodemographic, clinical characteristics, vaccination status against SARS-CoV-2 and history of testing PCR-positive at enrollment
      Measured characteristicsTotal

      n = 952 (valid %)
      Age, median, IQR (range, mean ± SD) – year35.0, 29 – 41 (20 – 65, 35.5 ± 8.40)
      Missing5
      Nationality
      Asian881 (92.6)
      African66 (6.9)
      Others5 (0.5)
      Education
      Primary schooling and below
      Started Sputnik boosted with Pfizer or Sinopharm (5) or first dose was Sinopharm and second dose Pfizer or Sputnik (2)
      502 (53.0)
      Secondary schooling352 (37.2)
      University and postgraduate level93 (9.8)
      Missing5
      Tobacco smoking
      Current smoker203 (21.4)
      Ex-smoker45 (4.8)
      Never smoke699 (73.8)
      Missing5
      Received flu shot
      Yes3 (0.3)
      No944 (99.7)
      Missing5
      BMI, median, IQR (Mean ± SD) Kg/m
      Started Sputnik boosted with Pfizer or Sinopharm (5) or first dose was Sinopharm and second dose Pfizer or Sputnik (2)
      25.2, 22.6 – 27.7 (25.3 ± 3.8)
      Underweight (< 18.5 Kg/m
      Started Sputnik boosted with Pfizer or Sinopharm (5) or first dose was Sinopharm and second dose Pfizer or Sputnik (2)
      )
      28 (3.2)
      Normal weight (18.5 – 24.9 Kg/m
      Started Sputnik boosted with Pfizer or Sinopharm (5) or first dose was Sinopharm and second dose Pfizer or Sputnik (2)
      )
      386 (44.2)
      Overweight (25.0 – 29.9 Kg/m
      Started Sputnik boosted with Pfizer or Sinopharm (5) or first dose was Sinopharm and second dose Pfizer or Sputnik (2)
      )
      357 (40.9)
      Obese (≥ 30 Kg/m
      Started Sputnik boosted with Pfizer or Sinopharm (5) or first dose was Sinopharm and second dose Pfizer or Sputnik (2)
      )
      102 (11.7)
      Missing79
      With at least one chronic condition
      No771 (88.4)
      Yes101 (11.6)
      High blood pressure66 (7.0)
      Diabetes mellitus39 (4.1)
      Hyperlipidemia17 (1.8)
      Heart problem2 (0.2)
      Asthma/COPD disease,2 (0.2)
      Cancer1 (0.1)
      Rheumatological diseases1 (0.1)
      Chronic liver disease, chronic neurological disorder, organ transplant, chronic blood/hematological disease, autoimmune disease, and anemia0 (0.0)
      Missing80
      Tested PCR positive in the past 12 months prior to enrollment
      No830 (87.8)
      Yes115 (12.2)
      Missing7
      COVID-19 vaccination status
      Not vaccinated13 (1.4)
      Only one dose9 (0.9)
      Two doses192 (20.2)
      One booster dose – three doses714 (75.2)
      Two booster dose – four doses21 (2.2)
      Missing3
      Vaccine-booster status
      Not boosted - two doses only (mean duration: 159.6 ± 71.8 days)
      Mean time duration post-last vaccine dose received prior to blood collection.
      192 (20.7)
      Boosted (mean duration: 70.2 ± 24.8 days)
      Mean time duration post-last vaccine dose received prior to blood collection.
      735 (79.3)
      Vaccine type
      Only Sinopharm (mean duration: 87.5 51.0 day)
      Mean time duration post-last vaccine dose received prior to blood collection.
      874 (92.1)
      Only Sputnik (mean duration: 210 ± 13.0 days)
      Mean time duration post-last vaccine dose received prior to blood collection.
      14 (1.5)
      Only Pfizer (mean duration: 106.7 ± 12.5 days)
      Mean time duration post-last vaccine dose received prior to blood collection.
      3 (0.3)
      Primary Sinopharm boosted with Pfizer (mean duration: 70.2 ± 69.2 days)
      Mean time duration post-last vaccine dose received prior to blood collection.
      38 (4.0)
      Mixed vaccine type
      Started Sputnik boosted with Pfizer or Sinopharm (5) or first dose was Sinopharm and second dose Pfizer or Sputnik (2)
      (mean duration: 154.7 ± 64.4 days)
      Mean time duration post-last vaccine dose received prior to blood collection.
      7 (0.7)
      Not vaccinated (8) or the first dose was after blood collection (5)13 (1.4)
      Missing3
      Duration: last dose to blood collection, median, IQR (mean ± SD)79.0, 56.0–96.0 (89.2 ± 54.5) - days
      1 – 14 days19 (2.0)
      15 – 30 days34 (3.6)
      31 – 60 days256 (27.0)
      61 – 295 days624 (65.8)
      Not vaccinated13
      T-cell reactivity
      Reactive353 (38.2%)
      Non-Reactive572 (61.8%)
      1 Mean time duration post-last vaccine dose received prior to blood collection.
      2 Started Sputnik boosted with Pfizer or Sinopharm (5) or first dose was Sinopharm and second dose Pfizer or Sputnik (2)
      At enrollment, all workers tested negative for SARS-CoV-2 PCR. Table 3 shows distribution of the measured characteristics at enrollment among the 940 workers with their PCR results retrieved from the Malaffi database during the follow up. The mean duration of follow-up was 88.9 (± 29.48) days. The earliest SARS-CoV-2 positive case occurred at 48 days of follow up. In total, there were 149 (15.9%) RT-PCR positive cases for SARS-CoV-2 during the follow up. Figure 3 shows probability of not contracting COVID-19 infection during the follow up. There was no significant difference in the mean age of those tested PCR positive and negative (P-value = 0.812). Most PCR-positive cases were boosted with only one booster dose (79.7%) or vaccinated with only Sinopharm (93.2%). There were more PCR positive cases among those who had been vaccinated since ≥ 61 days (55.2%) compared to those who had been vaccinated since ≤ 30 days (5.6%) (Table 3).
      Overall, 940 workers were available during the follow up period. Of this cohort, 149 (15.9%) were tested RT-PCR positive for SARS-CoV-2. Table 6 shows the mean anti-S IgG Abs correlated to the neutralizing IgG Abs protection levels in all populations and in populations tested positive and negative for SARS-CoV-2 during the follow up. This period correlated with the end of Delta as the predominant variant and the establishment of the Omicron variant BA.2 in the UAE.
      During the study no fatalities due to covid occurred in the study group and only one hospitalization took place. We were unable to determine if this was related to SARS-CoV-2.

      Prevalence of SARS-CoV-2 anti-S IgG, anti-N IgG, neutralizing IgG, and SARS CoV-2 specific reactive T cells at enrollment

      All the 952 collected blood samples were serologically tested for anti-S IgG, anti-N IgG, and surrogate neutralizing IgG antibodies. T cell reactivity was tested for each of the 952 enrolled individuals. Table 2 shows the laboratory findings and prevalence of the four markers. Nearly, all were sero-positive to anti-S IgG (99.7%), anti-N IgG (99.9%), and neutralizing IgG (99.3%) antibodies. The median interquartile range (IQR) concentration was 357.5 BAU/mL (173–930.5), 146.5 COI (92.0–205.5), and 172.0 AU/mL (51–greater than 800), for the anti-S IgG, anti-N IgG, and neutralizing IgG Abs, respectively. T cell reactivity was observed in 38.2% of the tested workers (Table 2).
      Table 2Prevalence of the four immune-response biomarkers measured against SARS-CoV-2 at enrollment.
      Anti-S IgG antibodies – AU/mL

      n = 952 (%)
      Anti-N IgG antibodies – COI

      n = 952 (%)
      Neutralizing IgG antibodies – AU/mL

      n = 952 (%)
      T-cell reactivity

      n = 952 (%)
      Mean ± SD (median, IQR)648.1 ± 641.7

      (357.5, 173–930.5)
      145.0 ± 74.8

      (146.5, 92.0 – 205.5)
      363.1 ± 339.4

      (172.0, 51 – 800)
      Range25 – 20800.0 – 320.03 – 810
      Positive/reactive949 (99.7)951 (99.9)945 (99.3)353 (38.2)
      Negativen-reactive3 (0.3)1 (0.1)7 (0.7)572 (61.8)
      Equivocal/invalid0 (0.0)0 (0.0)0 (0.0)7
      Missing/not tested20

      Correlation between the measured anti-SARS-CoV-2 antibodies at enrollment

      There was an observed positive correlation between the three anti-SARS-CoV-2 antibodies. The strongest correlation (r = 0.85, p<0.001) was between the combined anti-S IgG and neutralizing IgG Abs (Figure 2). Anti-N antibodies provided the weakest correlation between Anti-S and neutralizing antibodies. Figures 2b and 2c distinguishes the correlations between antibodies in the groups that did and did not have SARS-CoV-2 specific T cell responses. T cell reactivity did not impact the correlation of the neutralizing antibodies with either anti-N or anti-S.
      Figures 2
      Figures 2a, b, c. Matrix scatterplot - correlation between the three measured sero-biomarkers (anti-S IgG, anti-N IgG, and neutralizing IgG Abs) (2a) and by T-cells reactivity (2b: with reactive T-cells, 2c: with no reactive T-cells) at enrollment.

      Association between anti-SARS-CoV-2 antibodies at enrollment and infection with SARS-CoV-2

      The baseline median and mean concentration of anti-S IgG (381.0 and 678.4 vs 245.0 and 472.5 BAU/mL, p-value <0.001), anti-N IgG (143.0 and 150.3 vs 127.0 and 136.4 COI, p-value = 0.019), and neutralization IgG Abs (223.0 and 385.5 vs 85.0 and 236.2 AU/mL, p-value <0.00) was higher in PCR-negative compared to PCR-positive cases. Nearly, three-quarters (74.3%) of infections with SARS-CoV-2 during follow up had non-reactive T cells at enrollment (Table 3). Figure 3 shows the probability for not contracting SARS-CoV-2 in the follow up period. The hazard ratio was adjusted for an increase of 100 BAU/mL in anti-S IgG, 100 AU/mL increase in neutralizing antibodies and a 100 COI increase in anti-N IgG. By the end of follow up at 110 days the probability of reinfection increased to greater than 20% in the population tested. Figure 4 shows the distribution of anti-S IgG, anti-N IgG, and neutralizing IgG Abs concentration across the groups tested PCR positive and PCR negative for SARS-CoV-2 during the follow up.
      Table 3Distribution of the measured characteristics and levels of the measured immune-response biomarkers at enrollment by infection with SARS-CoV-2 (testing RT-PCR positive) during the follow up (n = 940).
      Total

      (valid %)
      Tested RT-PCR positive for SARS-CoV-2 during the follow upP - value
      No, n (valid %)Yes, n (valid %)
      Total population940 (100)791 (84.1%)149 (15.9%)
      Duration of follow up (mean ± SD) – day88.9 ± 29.4891.78 ± 30.3373.45 ± 18.08<0.001
      Age (mean ± SD) – year(35.5 ± 8.40)35.56 ± 8.4735.47 ± 8.080.812
      Nationality<0.001
      Asian874 (93.0)749 (94.7)125 (83.9)
      African66 (7.0)42 (5.3)24 (16.1)
      Education0.002
      Primary schooling and below2496 (52.8)437 (55.3)59 (39.6)
      Secondary schooling351 (37.3)278 (35.1)73 (49.0)
      University and postgraduate level93 (9.9)76 (9.6)17 (11.4)
      Tobacco smoking0.248
      Current smoker or Ex-smoker244 (26.0)211 (26.7)33 (22.1)
      Never smoke696 (74.0)580 (73.3)116 (77.9)
      Received flu shot0.067
      Yes3 (0.3)1 (0.1)2 (1.3)
      No937 (99.7)790 (99.9)147 (98.7)
      BMI (mean ± SD) – Kg/m225.26 ± 3.8425.21 ± 3.8425.59 ± 3.810.245
      With at least one chronic condition0.783
      No836 (89.2)703 (89.1)133 (89.9)
      Yes101 (10.8)86 (10.9)15 (10.1)
      Tested PCR positive in the past 12 months prior to enrollment<0.001
      No824 (87.8)681(86.1)143 (96.6)
      Yes115 (12.2)110 (13.9)5 (3.4)
      COVID-19 vaccination status0.652
      Not vaccinated prior to blood collection12 (1.3)9 (1.1)3 (2.0)
      Only one dose9 (1.0)7 (0.9)2 (1.4)
      Two doses189 (20.1)164 (20.8)25 (16.9)
      One booster dose – three doses709 (75.5)591 (74.7)118 (79.7)
      Two booster doses – four doses20 (2.1)20 (2.5)0 (0.0)
      Vaccine type0.111
      Only Sinopharm866 (92.2)728 (92.0)138 (93.2)
      Only Sputnik14 (1.5)12 (1.5)2 (1.4)
      Only Pfizer3 (0.3)1 (0.1)2 (1.4)
      Started Sinopharm boosted with Pfizer37 (3.9)34 (4.3)3 (2.0)
      Mixed vaccine type27 (0.8)7 (0.9)0 (0.0)
      Not vaccinated at enrollment12 (1.3)9 (1.1)3 (2.0)
      Missing1
      Time duration since last vaccine dose (mean ± SD)89.24 ± 54.5489.65 ± 54.0987.62 ± 57.610.168
      1 – 14 days18 (1.9)14 (1.8)4 (2.8)0.004
      15 – 30 days34 (3.7)30 (3.8)4 (2.8)
      31 – 60 days255 (27.5)198 (25.3)57 (39.3)
      61 – 295 days620 (66.9)540 (69.1)80 (55.2)
      Not vaccinated1394
      Anti-S IgG, min-max – AU/ml25 – 208025 – 2,08032 – 2,080
      Median (IQR)

      Mean ± SD
      357.5 (173–930.5)

      648.36 ± 642.03
      381.0 (177.0-997.0)

      678.4 ± 649.8
      245.0 (128.0-50.8)

      472.5 ± 564.1
      MD = 205.9

      (p<0.001)
      Anti-N IgG, min – max – COI0 – 3203 – 2980 – 286
      Median (IQR)

      Mean ± SD
      147 (92.0–205.5), 148.06 ± 74.76143.0 (95.0-107.0)

      150.3 ± 74.4
      127.0 (79.5-127.0)

      136.4 ± 76.1
      MD = 13.9

      (p =0.019)
      Neutralizing IgG, min-max, AU/ml3 – 93912 – 8583 – 800
      Median (IQR)

      Mean ± SD
      172.0 (51-800)

      362.67 ± 339.24
      223.0 (56.0-800.0)

      385.5 ± 343.1
      85.0 (37.0-360.5)

      236.2 ± 286.4
      MD = 138

      (p<0.001)
      T-cell reactivity<0.001
      Reactive348 (38.0)310 (40.4)38 (25.7)
      Non-reactive568 (62.0)458 (59.6)110 (74.3)
      Categorical variables were compared using the Chi-squared or Fisher's exact tests, and continuous variables were compared using the unpaired t-test or the Non-parametric Mann-Whitney U test.
      SD: Standard deviation.
      MD: mean difference
      Figure 3
      Figure 3Survival probability for not contracting SARS-CoV-2 during the follow up.
      Abs: antibodies.
      Figure 4
      Figure 4Distribution of anti-S IgG, anti-N IgG, and neutralizing IgG antibodies at enrollment by infection with SARS-CoV-2 during the follow up. P-values retrieved from the non-parametric Mann-Whitney U test.
      In the multivariable Cox-proportional hazard model fitted to investigate the association between antibody levels and incidence of SARS-CoV-2 infections, Table 4 shows that an increase in the anti-S IgG Abs concentration by 100 BAU/mL was associated with a 6% reduction in the risk of infection (HRa: 0.94, 95% CI: 0.91–0.98, p-value 0.003). Subjects who had an anti-S IgG Ab concentration below the median (357.5 BAU/mL) were at 78% increased risk of infection with SARS-CoV-2 compared with those who had ≥ median concentration of anti-S IgG. Compared to subjects who had > 8-fold increase in anti-S IgG concentration, the risk of infection with SARS-CoV-2 increased to 67% and 177% in those who had 4–8 fold and <2-fold increase, respectively. Those with anti-S IgG concentration that ranged from 25 to 225 and from 225.1 to 654.0 BAU/mL were 160% and 85% at increased risk of infection with SARS-CoV-2 respectively compared with those who had > 654.1 BAU/mL anti-S IgG concentration, An increase in the anti-S IgG concentration by 100 BAU/mL was associated with a similar magnitude of decreased risk of infection (6%) in workers who were boosted with at least one booster dose and those who were not boosted, but this decreased risk was significant (p-value = 0.006) for boosted but not for non-boosted (p-value = 0.132) (Table 4).
      Table 4Crude (HRc) and adjusted hazard ratio (HRa) for the risk of infection with SARS-CoV-2 by the level of anti-S IgG and anti-N IgG antibodies at enrollment.
      Anti-S IgG AbsAnti-N IgG Abs
      HRc

      (95% CI)
      HRa

      (95% CI)
      HRc

      (95% CI)
      HRa

      (95% CI)
      Increase by 100 unit in the Abs concentration (AU/mL in anti-S Abs / 100 COI in anti-N Abs concentration)0.95 (0.92–0.98)

      P < 0.001
      0.94 (0.91–0.98)

      P = 0.003
      0.78 (0.63–0.97)

      P = 0.028
      0.80 (0.63–1.02) P = 0.075
      ≥ Median (anti-S: 357.5 AU/mL / anti-N: 146.5 COI)1.001.001.001.00
      < Median (anti-S: 357.5 AU/mL / anti-N: 146.5 COI)1.89 (1.35–2.64) P = <0.0011.78 (1.22–2.53)

      P = 0.002
      1.41 (1.02–1.95) P = 0.0391.32 (0.94–1.87) P = 0.114
      Folds of increase relative to the cut off value

      (33.8 AU/mL for anti-S IgG Abs and 1 COI for anti-N IgG Abs)
      > 8 folds increase1.001.00
      4 – 8 folds increase1.67 (1.13–2.46) P = 0.0101.59 (1.06–2.40)

      P = 0.026
      2 – <4 folds increase1.84 (1.20–2.85) P = 0.0061.59 (0.99-2.54)

      P = 0.054
      < 2–fold increase2.77 (1.43–5.37) P = 0.0032.30 (1.13–4.66)

      P = 0.021
      > 200 folds increase1.001.00
      100 – 200 Folds increase1.50 (0.97–2.29)

      P = 0.064
      1.47 (0.93–2.30)

      P = 0.099
      < 100 folds increase1.03 (0.68–1.56) P = 0.8880.98 (0.63–1.53)

      P = 0.944
      Tertile
      Third tertile (anti-S: 654.1–2,080 AU/mL, anti-N: 0–110 COI)1.001.001.001.00
      Second tertile (anti-S: 225.1–654.0 AU/mL, anti-N: 111–185 COI)1.85 (1.17–2.92) P = 0.0091.96 (1.16–3.28)

      P = 0.011
      0.94 (0.62–1.43) P = 0.7820.86 (0.55–1.34)

      P = 0.501
      First tertile (anti-S: 25.0–225.0 AU/mL, anti-N: 186–320 COI)2.60 (1.68–4.04) P <0.0012.55 (1.54–4.21) P <0.0011.40 (0.95–2.07) P = 0.0861.39 (0.92–2.10)

      P = 0.115
      By vaccine-booster status - increase by 100 AU/mL in anti-S/100 COI in anti-N antibodies
      Boosted with at least one dose0.93 (0.90-0.97)

      P = <0.001
      0.94 (0.89-0.98)

      P = 0.006
      0.90 (0.69–1.16) P = 0.420.95 (0.72–1.25) P = 0.71
      Not-boosted – only two doses0.94 (0.87–1.01) P = 0.1050.94 (0.87–1.02)

      P = 0.132
      0.40 (0.22–0.73) P = 0.0030.42 (0.22–0.79) P = 0.008
      Adjusted hazard ratio for age, BMI, number of vaccine doses, type of vaccine, time interval between last vaccine dose and blood collection, and history of previous infection
      Table 4 also shows that an increase in anti-N IgG Abs concentration by 100 COI was associated with a 20% decreased risk of infection with SARS-CoV-2 (HRa: 0.80, 95% CI: 0.63–1.02), but this decreased risk did not reach to the statistical significance level (p-value = 0.075). Although workers who had ≥ median (146.5 COI), > 200-fold increase in their anti-N IgG Abs concentration, or with an anti-N IgG Abs concentration ranged from 654.1 to greater than 2,080 COI were at decreased risk of infection, but this decreased risk did not reach to a statistically significant association in both crude and adjusted models. Stratification by vaccine-booster status revealed that an increase in anti-N IgG Abs concentration by 100 COI at enrollment was associated with 58% decreased risk of infection with SARS-CoV-2 in those who were not-boosted (HRa: 0.42, 95% CI: 0.22–0.79, p-value = 0.008) with at least one booster dose (Table 4).
      Table 5 shows that an increase in the neutralizing IgG Abs by 100 AU/mL was significantly associated with a 13% decreased risk of infection with SARS-CoV-2 (HRa: 0.87, 95% CI: 0.82–0.93, p-value <0.001). Workers who had negative/weakly (<32.9 AU/mL) or moderate (33.3–82.9 AU/mL) level versus high level (≥ 83.3 AU/mL) of neutralizing IgG Abs were 110% and 80% at increased risk of infection Workers with a neutralizing IgG Abs concentration of 3–73 AU/mL (first tertile) and those with 74–799 AU/mL (second tertile) were at 228% and 162% increased risk of infection with SARS-CoV-2 compared to those with 800–939 AU/mL (third tertile) Ab concentration. Reduction in the risk of infection with SARS-CoV-2 by the increased concentration of neutralizing IgG Ab by 100 AU/mL was similar (overlapping 95% CIs) in those boosted and not-boosted workers but the significance level was more apparent in those who had at least one booster dose. Having had SARS-CoV-2 antigen non-reactive T cells at enrollment was associated with 2-fold increased risk of reinfection compared to those with reactive T cells (HRa: 1.99, 95% CI: 1.34–2.95, p-value <0.001).
      Table 5Crude (HRc) and adjusted hazard ratio (HRa) for the risk of infection with SARS-CoV-2 by the level of neutralizing IgG antibodies and T-cell reactivity at enrollment.
      HRc

      (95 % CI)
      HRa

      (95 % CI)
      Increase by 100 AU/mL in neutralizing IgG Abs concentration0.87 (0.82–0.92), P <0.0010.87 (0.82–0.93), P <0.001
      Level of protection
      Strongly protective (≥ 83.3 AU/ml)1.001.00
      Moderately protective (33.3–82.9 AU/ml)1.80 (1.23–2.63), P <0.0011.64 (1.09–2.46), P = 0.018
      Negative (8) or weakly protective (<32.9 AU/ml)2.10 (1.39-3.16), P = 0.0021.72 (1.10–2.70), P = 0.017
      Folds of increase relative to the cut off value of 10 AU/ml0.99 (0.98–0.99), P <0.0010.99 (0.98–0.99), P <0.001
      > 50 folds1.001.00
      ≤ 50 folds2.60 (1.75–3.86), P <0.0012.47 (1.58–3.65), P <0.001
      Tertile
      Third tertile (800 – 939 AU/ml)1.001.00
      Second tertile (74 – 799 AU/ml)2.62 (1.63–4.23), P <0.0012.67 (1.57–4.55), P <0.001
      First tertile (3 – 73 AU/ml)3.28 (2.06–5.22), P <0.0013.05 (1.81–5.17), P <0.001
      Effect of vaccine-booster status - Increase by 100 AU/mL
      Boosted with at least one dose (n = 734)0.87 (0.82–0.92), P <0.0010.88 (0.82–0.95) P <0.001
      Not-boosted – only two doses (n = 192)0.82 (0.71–0.95), P = 0.0080.84 (0.72–0.97), P = 0.017
      T-cell reactivity
      Reactive1.001.00
      Non-reactive1.97 (1.36-2.85), P <0.0011.99 (1.34–2.95), P <0.001
      Adjusted hazard ratio for age, BMI, number of vaccine doses, type of vaccine, time interval between last vaccine dose and blood collection, and history of previous infection.
      Table 6Mean anti-S IgG Abs by protection level of the neutralizing IgG Abs at enrollment in all population and in populations tested positive and negative for SARS-CoV-2 during the follow up.
      Neutralizing IgG antibodies – level of protectionAnti-S IgG antibodies
      All population

      n (%)

      mean ± SD

      median (IQR)
      Tested negative for SARS-CoV-2

      n (%)

      mean ± SD

      median (IQR)
      Tested positive for SARS-CoV-2

      n (%)

      mean ± SD

      median (IQR)
      P-value
      Obtained from the non-parametric Mann-Whitney U test.
      Strongly protective (≥ 83.3 AU/ml)601 (63.1)

      921.96 ± 655.5

      669.0 (364.0–1,390)
      518 (65.5)

      941.7 ± 652.4

      720.5 (375.8–1,402.5)
      75 (50.3)

      762.8 ± 654.4

      457.0 (305.0–923.0)
      0.005
      Moderately protective (33.3–82.9 AU/ml)212 (22.4)

      228.9 ± 216.4

      180.5 (141.3–238.8)
      170 (21.5)

      223.4 ± 211.6

      173.5 (137.0–246.0)
      41 (27.5)

      249.3 ± 238.6

      189.0 (159.0–236.5)
      0.179
      Negative or weakly protective (<32.9 AU/ml)138 (14.5)

      101.1 ± 45.3

      94.5 (71.0–123.0)
      103 (13.0)

      104.8 ± 44.9

      102.0 (71.0–126.0)
      33 (22.1)

      73.5 ± 46.9

      84.0 (58.5–101.5)
      0.031
      Between groups P-value
      Obtained from ANOVA test.
      P < 0.001P < 0.001P < 0.001
      1 Obtained from the non-parametric Mann-Whitney U test.
      2 Obtained from ANOVA test.

      Discussion

      After all the SARS-CoV-2 studies conducted globally, we still do not have a solid understanding of the relationship between measured immunity and clinical protection from SARS-CoV-2 infection. This is urgently needed to plan future COVID-19 vaccine programs, to validate if more vaccine boosters are necessary and ascertain if the boosters need to be modified to confer some protection from new variants of interest. Moreover, there is a need to supply vaccine to low income countries especially those with large populations of immunocompromised individuals. In this prospective cohort study, we assessed the level of three different SARS-CoV-2 antibodies and the reactivity of circulating T CD4+/CD8+ in a naturally, continuously exposed, and vaccinated cohort of expatriate workers against SARS-CoV-2.
      We also investigated the correlation between the measured antibodies and assessed the risk of infection with SARS-CoV-2 in this cohort. Our study enrollment and follow up period correlated with the transition of the Delta variant to the initial Omicron variant in the region based on the study period of September 2021 and February 2022.
      Although our study showed that the increase in neutralizing IgG was significantly associated with a decreased risk of infection with SARS-CoV-2, this did not conclusively demonstrate that neutralizing antibodies confer protection from re-infection. This is likely due in part, to the effect of immune waning and changes of immune recognition sites on subsequent SARS-CoV-2 variants. This finding is in accordance with other studies [8, 9] and it illustrates the importance of the complete immune response including T cell responses and central plasma cell memory to the viral antigens as conjunctive additional mechanisms of protection. This protection is further hindered by the different variants of interest that arise. We also note that the current vaccinations are designed to reduce morbidity and mortality and not necessarily to prevent infection. Interestingly, the commercial assays employ enough different conserved epitopes that there does not appears to be impact on the performance of the assays with the variants (unpublished communications, Diasorin and YHLO)
      Not surprisingly, the increase in anti-S IgG and neutralizing IgG antibodies were strongly associated with a significant reduction in the risk of infection with SARS-CoV-2, particularly with the increase in neutralizing IgG antibodies. Increased anti-N antibodies did not show any significant reduction in the risk of infection except for among the non-boosted individuals.
      As expected, successive variants of interest demonstrate increased infectivity. More concerning is that this occurs even in convalescent or vaccinated populations. Some are demonstrating immune evasion to prior infection and vaccination [10]. This is apparent with the Omicron variants A.2.12.1, BA.4 and BA.5 which all share key changes in the receptor binding protein amino acid 452, L452Q with A.2.12.1 or L452R with BA.4 and BA.5. These mutations appear to have independently appeared in all three variants and seem to reduce reactivity in those previously infected with the first Omicron variant. It is thought that this effect may be an evolutionary immunologic pressure response to the global immune surge caused by the initial Omicron variant [11].
      Our study revealed that having reactive T cells with memory to SARS-CoV-2 antigen at enrollment was associated with a protective effect against future infection with SARS-CoV-2. This is possibly due to the current population's antibody specificities having less effect on the emerging variants with the L452Q or R mutations. T cell responses to spike proteins following infection or vaccination are distributed across the spike epitopes and populations show a wide distribution of T cell responses due to HLA diversity. The immune response of T-helper and cytotoxic T cells may increasingly become the important determinant of reduced infectivity or morbidity and mortality as the general population's humoral immunity responds to counter further receptor binding protein mutations, either through infection or vaccination, and may play a lesser role in reduction of infectivity or virulence.
      Based on our findings, a value of 669 BAU/mL was a peak immunity cutoff with respect to spike protein antibodies. This value is quite high and may be sustainable for anything more than short periods with the current vaccines. While this titer provided the greatest protection, it wasn't 100% protective suggesting that we may not be able to define complete protective effect based solely on antibody titer. A similar study was performed by Raymond and his colleagues [12] who attempted to evaluate the Anti-S IgG antibodies cutoff using different testing platform for neutralizing activity and suggested a values of 25.5 COI and 6.8 S/C for Roche and Abbott platforms respectively.
      Our results are consistent with other studies [13, 14] that showed the increases in anti-S IgG antibodies concentration were associated with a reduction in the risk of infection in those who were boosted with at least one dose when compared to those who were not boosted. This could be explained by the model of protection provided by Wei [15] and estimated anti-spike IgG mean half-lives after second dose vaccination by 81-52 days according to vaccine type. However, correlating time-updated antibody measurements to protection from infection is likely important to inform the timing of boosters and other control measures. Existing studies combining correlates of protection and longitudinal data using the same assay are limited.
      In addition, we should take into consideration that multiple doses of COVID-19 vaccines are likely to increase the number and quality of antibody production and memory B cells should be more efficacious in preventing reinfection when compared with a single dose of vaccine irrespective of most variant changes and likewise, a similar increase in memory retention of helper and killer T cells should reduce severe disease and hospitalization.
      Early throughout the COVID-19 pandemic, the public has widely been led to believe that antibodies provide the bulk of protective immunity. This is mainly because antibodies are easy to detect, whereas T cell testing is complex and involves advanced technology. The knowledge of all factors contributing to immunity and protection through vaccination is still lacking and requires further studies.
      Although almost all our cases in this study were serologically reactive to SARS-CoV-2, T cell reactivity was detected only in 38.2% of the cases. Because of the complexity of T cell assays they are often omitted in studies or performed in smaller cohorts. We employed a commercial assay that is relatively simple to use, which allowed us to test at a scale rarely seen in the SARS-CoV-2 pandemic. Some other studies have demonstrated a response greater than 90% in subjects six months post receipt of two vaccine doses. [16-18] while other studies [19, 20] using the Qiagen IGRA assay demonstrated a substantial waning effect on T cell memory to varying degrees post 6 months. The technologies of the commercial IGRA assays are very different and likely have significant differences in sensitivity. Given that any of the commercial IGRA assays are not based on clinical concordance, clinical preference of one assay over the other cannot be determined without further studies.
      Nevertheless, our use of the QuantiFERON assay enabled a dichotomous categorization of study participants into those with or without a detectable T cell response and enabled us to demonstrate a predictive role for T cell reactivity. Similar observations have been made by [16] who reported similar decrease in T cell activity. Other vaccine studies have also demonstrated that the T cell response following a booster vaccine dose is consistently higher than the response observed after initial vaccination [21, 22].
      Our study is unique in that the workers residing in company accommodations served as an accelerated model for SARS-CoV-2 transmission and re-transmission in the general population. Infectivity in the general population is more gradual and possibly not as extensive as that observed in this cohort. One limitation of the study is that the participants are all the male sex and of Asian origin. SARS-CoV-2 has not shown any racial or sex bias for infectivity and this should not impact the study results. While it is possible that variant selection bias may occur in the cohort relative to general population, we note that during this study, variant penetration reflects that of the general population. Also, recruitment at a total of 24 housing sites in the Emirate of Abu Dhabi negates the potential for variant selection bias that could potentially occur with a smaller cohort. Other limitations include that we did not study mucosal immunity and our antibody studies focused on the Spike, RBP, NAbs and nucleocapsid antibodies.

      Conclusions

      SARS-CoV-2 continues to infect individuals across the globe. Recent immunity pressures have driven selection to variants that can escape humoral responses. In this study, we were able to estimate a serological cut off value that provides some, but not a complete, degree of protection from reinfection for a defined period. We have also determined that T cell antigen stimulation response is closely linked to immunocompetence to SARS-CoV-2 infection in our study population.

      Funding

      Funding for the study was provided by the Abu Dhabi Public Health. Center Grant number DOH/CVD/2021/856. The funding source was independent of study design, execution, and data analysis.

      Ethical Approval

      The study was reviewed and approved by the UAE National COVID-19 Research Ethics Committee DOH/CVDC/2021/856 and amendment number: DOH/CVDC/2021/1703.
      We thank James Paul Lund, Manar Al Ramahi, Abigail, Baguio, Mohammed Ahmed Hassan, Stuart McCallum and Matthew Richard Jones for valuable technical support.

      References

      [1] Center for Disease Control and Prevention, Covid-19- Large-scale Geographic Seroprevalence Surveys, https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/geographic-seroprevalence-surveys.html (Accessed on May, 31,2022)
      [2] World Health Organization. Population-based Age-Stratified Seroepidemiological Investigation Protocol for Coronavirus 2019 (COVID-19) Infection. 2020. https://apps.who.int/iris/handle/10665/332188 (31) (Accessed on January 2021,).
      [3] Grant R, Dub T, Andrianou X, Nohynek H, Wilder-Smith A, Pezzotti P, Fontanet A. SARS-CoV-2 population-based seroprevalence studies in Europe: a scoping review BMJ Open 2021;11:e045425. doi: 10.1136/bmjopen-2020-045425
      [4] Ahmed R Alsuwaidi, Farida I Al Hosani, Shammah Al Memari, Hassib Narchi, Laila Abdel Wareth et al., Seroprevalence of COVID-19 infection in the Emirate of Abu Dhabi, United Arab Emirates: a population-based cross-sectional study, International Journal of Epidemiology, Volume 50, Issue 4, August 2021, Pages 1077–1090, https://doi.org/10.1093/ije/dyab077
      [5] Labour Housing Regulations. Government of the United Arab Emirates Available at: https://u.ae/en/information-and-services/jobs/labour-accomodation Official Portal of the Government. Accessed 16, August, 2022
      [6] Alsuwaidi, A.R., Al Hosani, F.I., ElGhazali, G. Al-Ramadi B. The COVID-19 response in the United Arab Emirates: challenges and opportunities. Nat Immunol 22, 1066–1067 (2021). https://doi.org/10.1038/s41590-021-01000-5.
      [7] Abu Dhabi Health Data Services- SP LLC. (“Malaffi”) 2022, Page Title. (Accessed 1 June 2022 from https://malaffi.ae/).
      [8] Dan, J. M, Mateus J, Kato Y, Hastie, K, Yu E, Faliti C, et al. Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection. Science 371, (2021). https://doi.org/10.1126/science.abf4063
      [9] Khoury, D.S., Cromer, D., Reynaldi, A., Schlub, T.E., Wheatley A.K., Juno J.A. et al. Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection. Nat Med 27, 1205–1211. https://doi.org/10.1038/s41591-021-01377-8
      [10] Vogel G., New versions of Omicron are masters of immune evasion. Science 13 May 2022 , Vol 376 Issue 6594 . doi: 10.1126/science.adc9473
      [11] Tchesnokova V, Kulakesara H, Larson L, Bowers V. Rechkina E. et al, Acquisition of the L452R mutation in the ACE2-binding interface of Spike protein triggers recent massive expansion of SARS-Cov-2 variants. bioRxiv. 2021 Mar 11:2021.02.22.432189. doi: 10.1101/2021.02.22.432189.
      [12] Suhandynata, R.T., Hoffman M.A., Huang D. Tran J.T. Kelner M.J., et al. Commercial Serology Assays Predict Neutralization Activity against SARS-CoV-2. Clinical Chemistry 67(2) October 2020. https://doi.org/10.1093/clinchem/hvaa262
      [13] Gilbert, P.B., Montefiori, D.C., McDermott A.B., Fong Y., Benkeser D., et al. Immune correlates analysis of the mRNA-1273 COVID-19 vaccine efficacy clinical trial. Science 375, 43–50 (2021). https://doi.org/10.1126/science.abm3425
      [14] Feng, S., Phillips D.J., White T., Sayal H., Aley P.K., et al. Correlates of protection against symptomatic and asymptomatic SARS-CoV-2 infection. Nat. Med. 27, 2032–2040 (2021). https://doi.org/10.1038/s41591-021-01540-1
      [15] Wei, J., Pouwels, K.B., Stoesser, N., Matthews P.C., Diamond I., et al. Antibody responses and correlates of protection in the general population after two doses of the ChAdOx1 or BNT162b2 vaccines. Nat Med 28, 1072–1082 (2022). https://doi.org/10.1038/s41591-022-01721-6
      [16] Demaret J., Corroyer-Simovic B., Alidjinou E.K., Goffard A., Trauet J., et al., (2021) Impaired Functional T-Cell Response to SARS-CoV-2 After Two Doses of BNT162b2 mRNA Vaccine in Older People. Front. Immunol. 12:778679. doi: 10.3389/fimmu.2021.778679
      [17] Mateus, J., Dan, M., Zhang, Z., Rydyznski Moderbacher, C., Lammers, M., Goodwin, B., et al., Low-dose mRNA-1273 COVID-19 vaccine generates durable memory enhanced by cross-reactive T cells. Science374. https://doi.org/10.1126/science.abj9853
      [18] Zhang, Z., Mateus, J., Coelho. C.H., Dan, J.M., Rydyznski Morderbacher, C., Galvez R.I., et al., Humoral and cellular immune memory to four COVID-19 vaccines. Cell DOI:https://doi.org/10.1016/j.cell.2022.05.022
      [19] Banki, Z., Seekircher, L., Falkensammer, B., Bante, D., Shafer, H., et al. Six-month follow up of immune responses after rapid mass vaccination against SARS-CoV-2 with BNT162b2 in the district of Schwaz/Austria. Viruses 202, 14, 1642. https://doi.org/10.3390/v14081642
      [20]. www.nature.com/scientificreports. (Accessed July 11, 2022). Takeuchi, J.S., Fukunaga, A., Yamamoto, S., Tanaka, A., Matsuda, K., SARS-CoV-2 specific T cell and humoral immune responses upon vaccination with BNT162b2: a nine months longitudinal study. https://doi.org/10.1038/s41598-022-19581-y
      [21] Moore, S.C., Kronsteiner, B., Longet, S., Adele, S., Deeks, A.S., et al., Evolution of long-term hybrid immunity in healthcare workers after different COVID-19 vaccination regimens: a longitudinal observational cohort study. Preprint. https://doi.org/10.1101/2022.06.06.22275865
      [22] Jackson L.A., Anderson E.J., Rouphael N.G., Roberts P.C., Makhene M., et al. An mRNA vaccine against SARS-CoV-2—preliminary report. N Engl J Med. 2020;383(20):1920–1931. https://doi.org/10.1056/nejmoa2022483

      Competing Interests

      The authors have no competing interests in this study or the publication of this manuscript. Acknowledgments