Abstract
Background
By June 30, 2022, 92·6% of the Chilean population older than 18 years had received a full primary SARS-CoV-2 vaccine series, mostly with CoronaVac (Sinovac Biotech), and 78·4% had received a booster dose, mostly heterologous with BNT162b2 (Pfizer–BioNTech) and ChAdOx1 (AstraZeneca). We previously reported national seroprevalence data from lateral flow testing of IgG SARS-CoV-2 antibodies up to 16 weeks after primary vaccination. Our aim here was to study IgG seropositivity dynamics up to 30 weeks after primary vaccination and, in CoronaVac recipients, up to 26 weeks after booster vaccination, and to establish the correlation between lateral flow tests and neutralising antibody titres.
Methods
In this cross-sectional study, testing stations for SARS-CoV-2 IgG detection were selected and installed from March 12, 2021, in hotspots in 24 large Chilean cities, and were maintained until March 31, 2022. Individuals voluntarily approaching the testing stations were invited to perform a rapid lateral flow test via a finger prick and complete a questionnaire. Between Aug 12, 2021, and April 1, 2022, volunteers seeking medical care in the Mutual de Seguridad de la Cámara Chilena de la Construcción provided blood samples for lateral flow testing and neutralising antibody studies; inclusion criteria were age at least 18 years, history of complete primary vaccination series with CoronaVac, BNT162b2, or ChAdOx1, or no vaccine, and no previous COVID-19 diagnosis. We tested the difference in IgG positivity across time, and between primary and booster doses, in all eligible participants with complete records, controlling for age, gender, and comorbidities. We also assessed the predictive power of neutralising antibody titres and sociodemographic characteristics on the probability of IgG positive results using multivariable logistic regression.
Findings
Of 107 220 individuals recruited at the testing stations, 101 070 were included in our analysis (59 862 [59·2%] women and 41 208 [40·8%] men). 65 902 (65·2%) received primary vaccination series with CoronaVac, 18 548 (18·4%) with BNT162b2, and 606 (0·6%) with ChAdOx1, and 16 014 (15·8%) received no vaccine. Among the 61 767 individuals with a complete primary vaccination series with CoronaVac, 608 (1·0%) received a CoronaVac booster, 10 095 (16·3%) received a BNT162b2 booster, and 5435 (8·8%) received a ChAdOx1 booster. After ChAdOx1 primary vaccination, seropositivity peaked at week 5 after the second dose, occurring in 13 (92·9%, 95% CI 79·4–100·0) of 14 individuals. In participants who received a complete CoronaVac primary series, the decline in seropositivity stabilised at week 18 after the second dose (86 [44·7%, 95% CI 41·8–47·7] of 1087 individuals), whereas after receiving BNT162b2, seropositivity declined slightly by week 25 after the second dose (161 [94·2%, 90·6–97·7] of 171). A lower proportion of individuals who received the CoronaVac primary series and a homologous booster were seropositive (279 [85·6%, 95% CI 81·8–89·4] of 326) by weeks 2–18 than those who received a BNT162b2 booster (7031 [98·6%, 98·4–98·9] of 7128) or a ChAdOx1 booster (2893 [98·0%, 97·5–98·5] of 2953). The correlation between IgG positivity and log of the infectious dose in 50% of neutralising antibodies was moderate, with a sensitivity of 81·4% (95% CI 76·3–86·6) and specificity of 92·5% (73·3–100·0).
Interpretation
Dynamic monitoring of IgG positivity to SARS-CoV-2 can characterise antibody waning over time in the absence or presence of booster doses, providing relevant data for the design of vaccination strategies. The correlation between lateral flow test IgG titres and neutralising antibody concentrations suggests that they could be a quick and effective surveillance tool to measure protection against SARS-CoV-2.
Funding
Instituto Sistemas Complejos de Ingeniería, Subsecretaría de Redes Asistenciales, Ministry of Health, Chile, and Mutual de Seguridad de la Cámara Chilena de la Construcción.
Research in context.
Evidence before this study
We searched PubMed for articles published from March 1, 2020, to Aug 4, 2022, using the search terms (“SARS-CoV-2 IgG lateral flow test”) and (“COVID vaccine” or “COVID-19 vaccination”), restricted to English language. To find more studies, we added the term “neutralizing antibodies” in the final search. We found 58 studies, of which seven were related to results of IgG lateral flow tests in a vaccinated population against COVID-19. A study of 160 health-care workers found that different lateral flow tests were useful in a qualitative assessment of vaccine responses after individuals received the BNT162b2 (Pfizer–BioNTech) vaccine. A similar study found that, depending on the assays used, sensitivities differed between serum panels containing post-infection samples and post-vaccination samples for mRNA and inactivated vaccines. In another study from Brazil, lateral flow tests showed good agreement with reference serological assays, suggesting their usefulness in assessing anti-SARS-CoV-2 IgG production induced by vaccination. A study from Northern Ireland assessed the spike protein IgG response following ChAdOx1 (AstraZeneca) COVID-19 vaccination using a lateral flow test. Samples were taken pre-vaccination until 6 months after the first dose, and strong responses were observed in almost all individuals who received two doses of the vaccine, which persisted for more than 6 months after the first vaccination. In a study performed by our group in Chile, SARS-CoV-2 IgG positivity, measured by lateral flow tests in more than 56 000 Chilean individuals, increased significantly 3 weeks after the second dose of primary vaccination with CoronaVac (Sinovac Biotech) or BNT162b2 (Pfizer–BioNTech), but steadily declined 16 weeks after the second dose with CoronaVac but not with BNT162b2. ChAdOx1 vaccine was not widely used between March, 2020, and July, 2021. By adding the term “neutralizing antibodies” we found two studies and one review. One study found an adequate correlation between IgG and neutralising antibody titres using a lateral flow test among vaccinated individuals, suggesting that the test might be useful in the short-term follow-up after vaccination for SARS-CoV-2. Finally, a 2022 study compared results of self-reported lateral flow tests with live virus neutralisation. Although lateral flow testing was less sensitive than quantitative antibody tests, the positivity in lateral flow tests correlated better with virus neutralisation.
Added value of this study
This study offers the largest sample to date (more than 100 000 individuals) and follow-up that provides evidence on the dynamic seropositivity after primary vaccination with three different vaccines: CoronaVac, BNT162b2, and ChAdOx1. The study population shows homologous and heterologous (BNT162b2 and ChAdOx1) booster vaccination after a CoronaVac primary scheme induces seropositivity. Furthermore, the correlation between seropositivity obtained by lateral flow tests and neutralising antibody concentrations can be assessed in this novel population vaccinated with different schemes.
Implications of all the available evidence
Our analysis confirms earlier findings on the dynamic effect of vaccination with primary series of CoronaVac and BNT162b2 vaccines. Additionally, it provides further evidence on the robust response from heterogeneous booster vaccination schemes. For recipients of CoronaVac, a homologous booster dose increased IgG positivity significantly, albeit in the short-term, whereas heterologous boosters with BNT162b2 or CHAdOX1 induced higher and more prolonged seropositivity levels. Our results on the correlation between the OnSite lateral flow tests for IgG positivity and neutralising antibody concentrations add to the evidence on the usefulness of some lateral flow tests in follow-up vaccination for SARS-CoV-2. In particular, our results support the hypothesis that lateral flow tests could be effective and convenient surveillance tools to implement as a proxy for protection against SARS-CoV-2.
Introduction
Of the nine COVID-19 vaccines distributed worldwide, the efficacy and effectiveness rates of mRNA vaccines are highest, followed by vector-based and inactivated virus vaccines.1, 2, 3 The Chilean vaccination plan includes all these vaccine platforms; however, the inactivated vaccine CoronaVac (Sinovac Biotech) is the most ubiquitous primary vaccination series. Waning immunity after primary vaccination occurs with all vaccine platforms, especially inactivated vaccines.4, 5, 6
On March 12, 2021, we initiated a national seroprevalence study in 29 strategically located testing stations throughout the largest cities in Chile. Participants agreed to a SARS-CoV-2 IgG lateral flow test (OnSite, CTK Biotech) and a structured web-based questionnaire performed by trained personnel. We reported results on July 6, 2021, and of 56 261 individuals, seropositivity steadily declined in participants who had received primary vaccination series with CoronaVac but not in those who received primary vaccination with BNT162b2 (Pfizer–BioNTech) 16 weeks after the second dose (35 696 [63%] vs 9552 [17%]).4 These results coincided with a mild but steady decrease in effectiveness7 and instigated the provision of a booster dose to individuals who had received CoronaVac, followed by booster doses to all recipients of any COVID-19 vaccine.
Chile is one of the world leaders in the uptake of COVID-19 vaccination, in terms of complete primary series and booster doses. According to the Ministry of Health Department of Statistics and Information, by June 30, 2022, 17·6 million (92·6%) of the Chilean population older than 18 years and 3·3 million (87·2%) of the population younger than 18 years had received a full primary SARS-CoV-2 vaccine series, with 14·9 million (78·4%) older than 18 years and 1·9 million (50·4%) younger than 18 years receiving the full series plus a booster dose.8 Although the two-dose primary vaccination series given to individuals older than 18 years was almost exclusively homologous (72·7% received two doses of CoronaVac, 23·2% received two doses of BNT162b2, and 0·8% received two doses of ChAdOx1 [AstraZeneca]; 3·3% received a single dose of Ad5-nCoV [CanSino Biologics, Tianjin, China]), the secondary series was mostly heterologous. The most used booster vaccine was BNT162b2 (76% of individuals aged >18 years), predominantly in individuals younger than 50 years, followed by ChAdOx1 (20% of individuals aged >18 years), predominantly in individuals older than 50 years (appendix p 2).
In the context of booster vaccination, we continued monitoring SARS-CoV-2 IgG positivity up to March 31, 2022, with the following aims: to extend surveillance up to 8 months after primary vaccination with CoronaVac or BNT162b2; include a new population vaccinated with the ChAdOx1 vaccine; and determine seropositivity over time among recipients of CoronaVac primary vaccination, after a heterologous booster with BNT162b2 or ChAdOx1 vaccines, or in a small homologous booster group. Additionally, we included a subset of volunteers for simultaneous evaluation of a SARS-CoV-2 lateral flow test and concentration of neutralising antibodies, as the neutralising antibody concentration is considered to be a better correlate of protective immunity than a positive lateral flow test.9, 10
Methods
IgG lateral flow test-based surveillance
In this cross-sectional observational study, from March 12, 2021, to July 2, 2021, 29 SARS-CoV-2 testing stations operated in public open spaces in 24 of the most populated cities in Chile. From July 2, 2021, testing stations began to be progressively closed, and by March 31, 2022, four stations remained operational in two cities, including Santiago. Site locations were determined weekly by use of an optimisation (mixed-integer) model provided with national mobile phone mobility data as previously described.4 The Comité de Ética de Investigación en Seres Humanos, University of Chile, provided ethics approval (protocol number 184–2020).
At the testing stations, participants older than 18 years responded to an online ad-hoc questionnaire and provided a finger prick blood sample for the IgG lateral flow test using the OnSite COVID-19 IgG/IgM Rapid Test Kit (CTK Biotech, Poway, CA, USA).4 The test targets, nucleocapsid (N) and spike (S) proteins, had an IgG sensitivity of 96·7% and specificity of 98·1%.11 Test results were read after 15 min by trained study personnel and entered onto an electronic platform. Results were classified as positive, negative, or invalid. If possible, invalid outcomes were retested on site. Personal selection and training were conducted by staff from the Subsecretaria de Redes Asistenciales (Ministry of Health, Chile). Participants gave written informed consent.
SARS-CoV-2 neutralisation assays and IgG lateral flow test subset
Between Aug 12, 2021, and April 1, 2022, participants older than 18 years seeking medical care in the Mutual de Seguridad de la Cámara Chilena de la Construcción (a private non-profit organisation devoted to labour risk prevention operating in more than 80 health-care centres across 61 cities) provided a finger prick sample for lateral flow testing of SARS-CoV-2 IgG and a blood sample for neutralising antibody studies. Participant demographics and clinical (comorbidities) data were obtained via an online questionnaire. The Comité de Ética Científica, Mutual de Seguridad de la Cámara Chilena de la Construcción provided ethics approval (protocol number 284–2021).
Inclusion criteria were being aged 18 years or older; having had a full primary vaccination series with CoronaVac, BNT162b2, or ChAdOx1 vaccines, or no vaccine; and having no previous diagnosis of COVID-19. Participants provided written consent.
Study personnel from the Mutual de Seguridad de la Cámara Chilena de la Construcción performed the IgG lateral flow test as described and obtained a 10 mL blood sample that was centrifuged immediately, transported to the Laboratory of Molecular and Cellular Virology at the University of Chile, and stored at 4°C. Neutralising antibody titres were measured with a validated HIV-1-based SARS-CoV-2 pseudotyped virus containing the firefly luciferase gene and expressing the S protein of the Wuhan reference strain (Wuhan-Hu-1) on the surface.12 Relative luminescence units (RLUs) of non-transduced HEK293T-ACE2 cells were averaged and used to define 100% neutralisation, whereas RLUs measured at the highest dilution of each sample were used to define 0% neutralisation. Thus, the percentage neutralisation of each of the eight sample dilutions was calculated as the complement of the division between the corresponding RLUs and the RLUs obtained at the higher dilution after subtracting the background (non-transduced HEK293T-ACE2 cells). 50% of the pseudovirus neutralising antibody titre (ID50) was calculated in GraphPad Prism (version 9.1.2) by modelling a four-parameter non-linear regression with variable slope, constraining top values to 100% and bottom values to 0%. Results are presented as log(ID50).
Data management
The web-based structured questionnaire recorded sociodemographic characteristics, comorbidities, and COVID-19 vaccine status, and was administered on site at both the lateral flow testing stations and the Mutual de Seguridad de la Cámara Chilena de la Construcción by trained personnel. Options available to report gender were male, female, and not declared. All data, except for IgG positivity, were self-reported. Results were anonymised and instantly uploaded to a centralised database and stored on the servers of Instituto Sistemas Complejos de Ingeniería. For this cross-sectional study, we only considered IgG results; individuals with a positive IgM result were advised to proceed to PCR testing.
Statistical analysis
In our analysis, data are presented in tables and figures as counts, percentages, and 95% CIs. Mean comparisons between groups were performed using standard two-proportion Z tests.
Data from all participants with complete records were included in the IgG positivity analysis, except for individuals providing incomplete self-reported information on vaccination status, those who were younger than 18 years, those who had not declared their gender to be male or female, those who had an invalid test result, or recipients of other vaccines. Individuals who were unvaccinated were not included in the analysis but were not considered as having incomplete records. Instead of pursuing the alternative of a multiple imputation strategy, we excluded participants with incomplete records because they were scarce (452 [4·2%] of 107 220 participants). Subsequently, participants with previous positive PCR test results were also excluded, as well as those who received a second booster dose by the time of the finger prick. Analyses (tables, figures, tests, regressions, 95% CIs, and p value computation) were performed using the open-source Julia programming language (version 1.6).13 Time periods are stated in terms of the number of weeks elapsed since vaccination: week 1 therefore included 0–6 days after vaccination, week 2 included 7–13 days, etc.
Data collected on the neutralisation assays and the IgG lateral flow test subset cohort were used to conduct a multivariable logistic regression analysis to explain IgG positivity results (dependent variable) as a function of ID50 count, vaccine (primary series) received, gender, and age (appendix pp 18–20).
Role of the funding source
The funders of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the report.
Results
Of 107 220 individuals recruited, 101 070 (94·3%) were included in the IgG positivity analysis. Of 6150 excluded participants, 4296 (69·9%) had incomplete self-reported information on vaccination status, 1245 (20·4%) were vaccinated with other vaccines, 273 (4·4%) were younger than 18 years, 88 (1·4%) reported gender other than male or female, 43 (0·7%) had an invalid test result, and 469 (7·6%) had a combination of these criteria (appendix p 3). The resulting samples were reasonably representative of the targeted populations (at the county level;4 appendix p 4); information on the contribution of each testing site to the overall study population and a comparison between the sample and national population can be found in the appendix (pp 5–6).
Of 101 070 total participants, 65 902 (65·2%) individuals had received a primary series with CoronaVac (one or two doses), 18 548 (18·4%) with BNT162b2, and 606 (0·6%) with ChAdOx1, and 16 014 (15·8%) individuals were not vaccinated. Among 61 767 individuals with a complete (two doses) primary series with CoronaVac (4135 participants had received only a first dose when tested), 608 (1·0%) had received a CoronaVac booster, 10 095 (16·3%) had a BNT162b2 booster, and 5435 (8·8%) had a ChAdOx1 booster. As previously reported,4 vaccinated individuals tended to be older, have more comorbidities (due to the national vaccination priorities), and be more likely to be female than unvaccinated individuals. Individuals who had received CoronaVac primary series were, on average, older than individuals who had received BNT162b2 primary series, as early massive vaccination efforts began in older individuals (aged >60 years) using mostly CoronaVac (table 1 ).
Table 1.
Baseline characteristics of study population by primary vaccination status
| Overall (n=101 070) | Unvaccinated (n=16 014) | CoronaVac (n=65 902) | BNT162b2 (n=18 548) | ChAdOx1 (n=606) | ||
|---|---|---|---|---|---|---|
| Age, years | ||||||
| 18–39 | 44 426 (44·0%) | 10 408 (65·0%) | 24 723 (37·5%) | 9046 (48·8%) | 249 (41·1%) | |
| 40–49 | 20 234 (20·0%) | 3203 (20·0%) | 12 659 (19·2%) | 4149 (22·4%) | 223 (36·8%) | |
| 50–59 | 18 197 (18·0%) | 1768 (11·0%) | 11 952 (18·1%) | 4395 (23·7%) | 82 (13·5%) | |
| ≥60 | 18 213 (18·0%) | 635 (4·0%) | 16 568 (25·1%) | 958 (5·2%) | 52 (8·6%) | |
| Gender | ||||||
| Male | 41 208 (40·8%) | 7177 (44·8%) | 26 172 (39·7%) | 7311 (39·4%) | 548 (90·4%) | |
| Female | 59 862 (59·2%) | 8837 (55·2%) | 39 730 (60·3%) | 11237 (60·6%) | 58 (9·6%) | |
| Booster dose | ||||||
| CoronaVac | 615 (0·6%) | 0 | 608 (0·9%) | 7 (0·0%) | 0 | |
| BNT162b2 | 13 082 (12·9%) | 0 | 10 095 (15·3%) | 2908 (15·7%) | 79 (13·0%) | |
| ChAdOx1 | 5470 (5·4%) | 0 | 5435 (8·2%) | 32 (0·2%) | 3 (0·5%) | |
| Previous positive COVID-19 PCR | 6678 (6·6%) | 1103 (6·9%) | 4237 (6·4%) | 1286 (6·9%) | 52 (8·6%) | |
| Previous positive COVID-19 IgG | 1872 (1·9%) | 94 (0·6%) | 1367 (2·1%) | 399 (2·2%) | 12 (2·0%) | |
| Comorbidities | ||||||
| Obesity | 4738 (4·7%) | 660 (4·1%) | 3136 (4·8%) | 912 (4·9%) | 30 (5·0%) | |
| High blood pressure | 16 108 (15·9%) | 965 (6·0%) | 12 877 (19·5%) | 2219 (12·0%) | 47 (7·8%) | |
| Diabetes | 7967 (7·9%) | 523 (3·3%) | 6213 (9·4%) | 1200 (6·5%) | 31 (5·1%) | |
| Cancer | 1241 (1·2%) | 82 (0·5%) | 969 (1·5%) | 185 (1·0%) | 5 (0·8%) | |
| Chronic pulmonary disease | 4619 (4·6%) | 594 (3·7%) | 3250 (4·9%) | 764 (4·1%) | 11 (1·8%) | |
| Chronic cardiovascular disease | 2599 (2·6%) | 235 (1·5%) | 2038 (3·1%) | 320 (1·7%) | 6 (1·0%) | |
Data are n (%).
For the analysis of IgG positivity, we additionally excluded individuals who self-reported a previous positive PCR result (n=6678), were unvaccinated (n=16 014), or both (n=1103). As previously reported,4 among participants without a vaccine booster dose, IgG positivity peaked 3 weeks after the second dose for both CoronaVac and BNT162b2, occurring in 1641 (78·5%, 95% CI 76·7–80·3) of 2090 individuals who had received the CoronaVac primary series and in 609 (97·0%, 95% CI 95·7–98·3) of 628 individuals who had received the BNT162b2 primary series. Additionally, after receiving the ChAdOx1 primary series, IgG positivity peaked 5 weeks after the second dose, occurring in 13 (92·9%, 95% CI 79·4–100·0) of 14 participants (figure 1 ). In participants who received the CoronaVac primary series, the decline in the proportion of participants with seropositivity stabilised at about 40% by week 22 after the first dose, whereas after receipt of BNT162b2 primary series the proportion of participants with seropositivity declined only slightly to about 90% by week 25 (for the overall evolution of IgG positivity throughout the study see appendix p 8).
Figure 1.
Seropositivity over time by vaccination status
Seropositivity over time among 59 601 individuals who received a primary vaccination series not reporting previous PCR or IgG positive results. Vaccinated participants included those who had received one or two doses of CoronaVac, BNT162b2, or ChAdOx1, but not a booster dose. The shaded region includes participants inoculated with a first dose (we excluded 1155 participants who received a second dose 5 weeks or more after receiving the first dose), and the non-shaded region includes those inoculated with two doses (we excluded 100 participants who received their second ChAdOx1 dose more than 5 weeks before being tested, and 731 who received the second dose 30 weeks or more before tested). Error bars represent 95% CIs. Horizontal lines represent weekly positivity. Absolute numbers including additional weeks are not presented but are included in the appendix (p 7).
Among 19 167 individuals who received a booster dose, 16 138 (84·2%) had a full primary series with CoronaVac and 2908 (15·2%) had a homologous booster dose with BNT162b2; the remaining 121 (0·6%) received another combination. As expected, due to the national vaccination plan, most individuals who had received the CoronaVac vaccine with a BNT162b2 booster dose were younger than 50 years, and most of those who had received the CoronaVac vaccine with a ChAdOx1 booster dose were aged 50 years or older (table 2 ).
Table 2.
Baseline characteristics of study population who had received a complete series with CoronaVac, according to the booster vaccine received
| Overall (n=16 138) | CoronaVac (n=608) | BNT162b2 (n=10 095) | ChAdOx1 (n=5435) | ||
|---|---|---|---|---|---|
| Age, years | |||||
| 18–39 | 4904 (30·4%) | 30 (4·9%) | 4859 (48·1%) | 15 (0·3%) | |
| 40–49 | 2682 (16·6%) | 38 (6·2%) | 2625 (26·0%) | 19 (0·3%) | |
| 50–59 | 2939 (18·2%) | 113 (18·6%) | 1703 (16·9%) | 1123 (20·7%) | |
| ≥60 | 5613 (34·8%) | 427 (70·2%) | 908 (9·0%) | 4278 (78·7%) | |
| Gender | |||||
| Male | 6090 (37·7%) | 211 (34·7%) | 3644 (36·1%) | 2235 (41·1%) | |
| Female | 10 048 (62·3%) | 397 (65·3%) | 6451 (63·9%) | 3200 (58·9%) | |
| Previous positive COVID-19 PCR | 1042 (6·5%) | 43 (7·1%) | 675 (6·7%) | 324 (6·0%) | |
| Previous positive COVID-19 IgG | 476 (2·9%) | 15 (2·5%) | 324 (3·2%) | 137 (2·5%) | |
| Comorbidities | |||||
| Obesity | 837 (5·2%) | 43 (7·1%) | 581 (5·8%) | 213 (3·9%) | |
| High blood pressure | 3667 (22·7%) | 253 (41·6%) | 1298 (12·9%) | 2116 (38·9%) | |
| Diabetes | 1775 (11·0%) | 137 (22·5%) | 640 (6·3%) | 998 (18·4%) | |
| Cancer | 297 (1·8%) | 26 (4·3%) | 169 (1·7%) | 102 (1·9%) | |
| Chronic pulmonary disease | 812 (5·0%) | 41 (6·7%) | 489 (4·8%) | 282 (5·2%) | |
| Chronic cardiovascular disease | 572 (3·5%) | 101 (16·6%) | 227 (2·2%) | 244 (4·5%) | |
Data are n (%)
Among individuals who received the CoronaVac primary series with a booster, excluding those who received a second booster dose (n=513), the proportion of participants with IgG positivity peaked at 44 (95·7%, 95% CI 89·8–100·0) of 46 participants 9–10 weeks after a homologous booster with CoronaVac, at 770 (99·7%, 95% CI 99·3–100·0) of 772 participants 7–8 weeks after receiving a heterologous booster with BNT162b2, and at 366 (99·5%, 95% CI 98·8–100·0) of 368 participants 3–4 weeks after receiving a ChAdOx1 booster (figure 2 ). In the individuals who received a ChAdOX1 booster, the proportion of participants with IgG positivity declined considerably to 216 (90·4%, 95% CI 86·7–94·1) of 239 individuals from week 19 onwards, whereas no decline was observed for those who received a BNT162b2 booster up to week 23. In individuals who received a CoronaVac booster, the proportion of participants with IgG positivity declined from week 9–10 onwards (to 20 [57·1%, 95% CI 40·7–73·5] of 35 participants by weeks 25–26).
Figure 2.
Seropositivity over time among individuals previously vaccinated with CoronaVac receiving a booster
Participants included 14 454 individuals who received CoronaVac, BNT162b2, or ChAdOx1 and who did not report previous positive PCR or IgG positive results; individuals who received a second booster dose are not shown (n=513). Horizontal lines represent weekly positivity. Error bars represent 95% CIs. Absolute numbers including additional weeks are not presented but are included in the appendix (p 10).
Overall IgG positivity among individuals who received the primary series with CoronaVac 2–18 weeks after a CoronaVac booster (279 [85·6%, 95% CI 81·8–89·4] of 326 participants) was lower than in individuals who received a BNT162b2 booster (7031 [98·6%, 98·4–98·9] of 7128) or a ChAdOx1 booster (2893 [98·0%, 97·5–98·5] of 2953; appendix pp 11–12). The homologous IgG booster response was lower than the heterologous booster response across all sociodemographic and clinical characteristics for which there was a large enough sample size (n>30; appendix p 11). IgG positivity did not differ between individuals who received BNT162b2 and ChAdOx1 booster doses across most of the various sociodemographic and clinical subgroups with sufficient sample size, but significant differences in IgG positivity were seen in female individuals (IgG positivity was higher with the BNT162b2 booster than with the ChAdOx1 booster; p=0·0009) and those with non-Chilean nationality (BNT162b2 booster had higher IgG positivity than ChAdOx1 booster; p=0·0048). Overall IgG positivity among individuals aged 60 years or older was lower than IgG positivity in younger individuals, across all booster vaccines.
In individuals who had received CoronaVac primary series, by 19–26 weeks after receiving a CoronaVac booster dose, seropositivity was lower (63·2%, 95% CI 56·3–70·2) than in individuals who had received booster doses of BNT162b2 (97·9%, 97·1–98·6) and ChAdOx1 (90·4%, 88·8–92·0; appendix pp 11–12). This lower seropositivity was observed in individuals who had received a CoronaVac booster dose across all sociodemographic and clinical characteristics for which there was a large enough sample size (n>30), except for participants with cancer when compared with individuals who received a ChAdOx1 booster dose (p=0·10), and for participants with chronic cardiovascular disease when compared with individuals who received a BNT162b2 booster (p=0·30; appendix pp 11–12). The proportion of individuals with IgG positivity among those who received a BNT162b2 booster was significantly higher than that among individuals who received a ChAdOx1 booster across all sociodemographic and clinical characteristics with sufficient sample size, except for patients with chronic cardiovascular disease (p=0·20). Of individuals who received a BNT162b2 booster dose, IgG positivity was significantly lower in individuals aged 60 years or older than in individuals younger than 60 years (p=0·0097; appendix pp 11–12). IgG positivity was not significantly different between male and female gender, for any vaccine at 19–26 weeks. A more detailed comparison of IgG positivity in the sample for the different boosters across gender and age can be found in the appendix (p 13). We conducted multivariable logistic regression analyses starting at week 3 after the booster dose in individuals who received CoronaVac primary series (excluding a second booster dose). Results showed a significant decrease in seropositivity over time for all booster vaccines, and no significant effects associated with gender, age, or comorbidities (appendix p 14).
Data from all 261 participants providing finger prick and venous blood samples for correlation between neutralisation and IgG positivity were included in the analysis. Participants had received a full primary vaccination series with CoronaVac (n=127), BNT162b2 (n=86), or ChAdOx1 (n=30), or were unvaccinated (n=18; appendix p 15). Among individuals who received ChAdOx1, four participants received a second dose of BNT162b2 due to safety guidelines issued by the government, and 26 had received a homologous booster dose.
The correlation between IgG positivity and the log(ID50) was moderate (0·57; figure 3 ). The sensitivity of the lateral flow test computed on the basis of the blood samples with ID50 more than the detection limits (log[ID50]>0) was 81·4% (95% CI 76·3–86·6; with 180 of 221 participants with detectable ID50 being identified as IgG positive). The specificity computed on the basis of samples with ID50 lower than the detection was 92·5% (95% CI 73·3–100·0; with 37 of 40 participants with no detectable ID50 being identified as not IgG positive). On average, individuals who received primary vaccination series with CoronaVac and individuals who were unvaccinated had lower ID50 counts than those who received primary vaccination series with BNT162b2 or ChAdOx1, uniformly across age and gender groups (appendix p 16). The ID50 count was also lower in participants with more than 150 days since the last inoculation (appendix p 17). Regarding IgG positivity, participants with a low (<4) ID50 count had predominantly negative IgG results, whereas positive results were associated with higher ID50 counts than negative results (figure 3). Participants who received a booster dose have higher neutralising antibody titres than those who had not received a booster (appendix p 16). All four participants who received a heterologous primary series had positive IgG results, with log(ID50) values of 7·42, 8·12, 8·48, and 8·68 in these four participants, respectively. Our multivariable logistic regression analysis showed that ID50 values were associated with IgG positivity (appendix p 18) and were a significant predictor of the probability of an IgG positive result, whereas age and gender were not. When the primary vaccination series was introduced as a dependent variable, the result stayed the same, although full homologous primary vaccination with BNT162b2 and ChAdOx1 significantly increased the probability of obtaining a positive result; however, it is difficult to separate the effect of the vaccine from the large ID50 counts (appendix p 19). A comparison between observed IgG positivity results and the fitted probability for positive IgG results as a function of ID50 count, and homologous vaccination with BNT162b2 or ChAdOx1, is shown in the appendix (p 20).
Figure 3.
Overall IgG positivity compared with the neutralising antibody response
The histograms depict the number of positive or negative IgG antibody results as a function of the log(ID50). Data are from 261 individuals in the Mutual de Seguridad Hospital subset. The numbers on top of the bars denote the number of positive (blue) or negative (red) results associated with participants who had received a booster dose.
Discussion
Following our first report of dynamic SARS-CoV-2 IgG positivity in Chile,4 in this cross-sectional study we now report a sustained decline in seropositivity to nearly 40% by 20–34 weeks after a full (two-dose) primary vaccination series with CoronaVac compared with a sustained seropositivity of more than 86% in individuals who received a primary vaccination series with BNT162b2. In a small number of individuals who received primary vaccination with ChAdOx1, seropositivity ranged from 81% to 93% 1–5 weeks after the second dose. Overall, our results for IgG positivity in individuals who had received a primary vaccination series are aligned with those previously reported.4
Heterologous booster vaccination with an mRNA or adenovirus vectored vaccine resulted in higher IgG seropositivity rates than homologous vaccination among individuals in Chile who received a primary vaccination series with CoronaVac. The benefit of a heterologous vaccination for individuals who received CoronaVac is probably due to the superior immunogenicity seen in mRNA and adenovirus vector vaccines compared with inactivated vaccines.14, 15 Notably, IgG positivity in individuals who received ChAdOx1 also began to decline 18 weeks after the booster, and began to decline in those who received BNT162b2 after 23 weeks, although considerably less so than in those who received CoronaVac.
Several studies have reported antibody responses after different booster dose regimens were used in individuals who received primary inactivated vaccines. All studies concluded that homologous and heterologous booster doses significantly increase concentrations of different anti-SARS-CoV-2 antibodies (and sometimes cell-mediated immunity).16, 17, 18 Studies comparing homologous versus heterologous booster responses among individuals receiving primary inactivated vaccines (mostly CoronaVac in China, Brazil, Turkey, and Thailand19, 20, 21, 22) report a more robust response with heterologous boosters. An antibody study from Brazil23 and one effectiveness study from Chile24 provided the most robust evidence for this conclusion. Individual studies suggest that the homologous booster provides a short-lived rise in antibodies23 and heterologous boosters provide higher responses to the omicron variant (although to a lesser extent than to other variants).25
Although these studies report similar findings, our study adds the comparison of three vaccines as primary and booster vaccinations over time, providing a longitudinal population-based perspective, pinpointing the curves of antibody prevalence in a continuous manner. Most importantly, continued epidemiological surveillance, undertaken by the Chilean Health Ministry and Instituto Sistemas Complejos de Ingeniería, reported substantial benefits of booster doses in individuals who had received a two-dose CoronaVac primary series by reducing rates of COVID-19, admissions to the intensive care unit, and deaths before the omicron surge.7 Our time-continuous curves of the proportion of tested individuals with IgG positivity after primary doses between the different vaccines and, more importantly, after the different booster regimens, are consistent with the results (ie, the data at specific timepoints rather than the trends over time) described above using other methodologies.
We found that SARS-CoV-2 IgG positivity determined by the OnSite lateral flow test was directly correlated with titres of neutralising antibodies (measured via pseudovirus neutralisation), which is a more widely accepted measure of the humoral immune response than lateral flow test positivity and a proposed correlate of protection.9, 10 Sensitivity and specificity, computed for the OnSite lateral flow test on the basis of ID50 count, are similar to those reported for other lateral flow tests that are computed on the basis of results from plaque reduction neutralisation.26 Therefore, until affordable and easy to implement point-of-care tests for SARS-CoV-2 neutralising antibodies are fully developed,27, 28 monitoring population IgG seroprevalence via lateral flow tests is a fast, viable, and valid means to infer changes in neutralising antibody levels in the population. For example, our data show that participants who received a booster dose have higher neutralising antibody titres than those who had not received a booster (appendix p 16), explaining the likelihood of their IgG positive results. We therefore conclude that heterologous boosters have a large and quick immunological benefit, shown by the increase in IgG seropositivity in the population. This effect should translate into an increase in protection for individuals receiving a heterologous CoronaVac booster dose and, to a lesser extent, those receiving a homologous CoronaVac booster dose.
Our study has several limitations. Although the lateral flow test used showed robust responses to inoculation with inactivated mRNA and vector-based vaccines, it is unclear how this test will respond to other vaccines, or more importantly, how other lateral flow tests will respond, considering the variability among them.29 The self-reported nature of our data could affect the representativeness and accuracy due to self-selection and recall biases. In this regard, self-reporting before testing does not allow the exclusion of infection-induced positive IgG results, which is particularly relevant for our seropositivity data in participants who received a booster; half of the data in this subgroup were collected during a surge in cases due to the omicron variant, from January to March, 2022. These data affect the observed seropositivity decline about 6 months after a booster dose with BNT162b2 and ChAdOx1 vaccines. Thus, if infection-induced seropositivity is higher in data collected in 2022 than in 2021, then the decline in seropositivity might be underestimated.
Additionally, ethnicity data were not recorded. A further limitation is the relatively small sample size of individuals who received homologous vaccines, and of young recipients (aged ≤45 years) of the ChAdOx1 booster. This is a structural limitation due to Chile's booster roll-out strategy. Importantly, SARS-CoV-2 IgG detected by the lateral flow test used in this study does not assure protection. The positive correlation between IgG positivity and neutralisation documented in our study supports the use of this lateral flow test as a proxy for protection after vaccination.
Our results show that dynamic monitoring of IgG positivity to SARS-CoV-2 can characterise antibody waning over time in the absence or presence of booster doses, providing data for the design of vaccination strategies.
Data sharing
Data cannot be shared directly by the authors because of data protection regulation. Data is accessible to authorised researchers after application to the Chilean Ministry of Health.
Declaration of interests
We declare no competing interests.
Acknowledgments
Acknowledgments
We thank Digital Entel Ocean, Empresa Nacional de Telecomunicaciones for processing and sharing fine granular mobility data. We thank the Servicio de Salud Metropolitano Centro, the Ministry of Sciences, and BHP for their support in setting up the pilot scheme for the dynamic COVID-19 IgG monitoring system. This study was partially funded by a grant from Instituto Sistemas Complejos de Ingeniería (ANID PIA/APOYO [AFB180003 and AFB220003]) that was used to hire research assistants for the information systems, data management, and training of field staff. The field work was funded by Subsecretaría de Redes Asistenciales, Ministry of Health, Chile, and Mutual de Seguridad de la Cámara Chilena de la Construcción.
Contributors
DS, MO'R, JPT, MZ, MR, and LJB conceived the study. LJB led and coordinated the project. DS and LJB directed the Instituto Sistemas Complejos de Ingeniería team that worked on the testing site optimisation module and the web-based platform for collecting data. MZ, CA, FA, and MR led the fieldwork teams. RS-R, FV-E, AG-A, VS, and MLA conducted neutralisation assays at the Laboratory of Molecular and Cellular Virology of the University of Chile. DS and IN were responsible for data cleaning, data analyses, and generating tables and figures. DS, MO'R, JPT, LJB, RS-R, and FV-E drafted the original manuscript. All authors contributed to the methods, revised the manuscript, and approved the final version. LJB, MZ, and MR led the funding acquisition efforts. LJB, DS, IN, MZ, and MR are the data guarantors with full access to the primary data, verifying that this manuscript is an honest, accurate, and transparent account of the study that has been conducted. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Supplementary Material
References
- 1.WHO COVID-19 vaccine tracker and landscape. 2022. https://www.who.int/publications/m/item/draft-landscape-of-covid-19-candidate-vaccines
- 2.Fiolet T, Kherabi Y, MacDonald CJ, Ghosn J, Peiffer-Smadja N. Comparing COVID-19 vaccines for their characteristics, efficacy and effectiveness against SARS-CoV-2 and variants of concern: a narrative review. Clin Microbiol Infect. 2022;28:202–221. doi: 10.1016/j.cmi.2021.10.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tregoning JS, Flight KE, Higham SL, Wang Z, Pierce BF. Progress of the COVID-19 vaccine effort: viruses, vaccines and variants versus efficacy, effectiveness, and escape. Nat Rev Immunol. 2021;21:626–636. doi: 10.1038/s41577-021-00592-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sauré D, O'Ryan M, Torres JP, Zuniga M, Santelices E, Basso LJ. Dynamic IgG seropositivity after rollout of CoronaVac and BNT162b2 COVID-19 vaccines in Chile: a sentinel surveillance study. Lancet Infect Dis. 2022;22:56–63. doi: 10.1016/S1473-3099(21)00479-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Altmann DM, Boyton RJ. Waning immunity to SARS-CoV-2: implications for vaccine booster strategies. Lancet Respir Med. 2021;9:1356–1358. doi: 10.1016/S2213-2600(21)00458-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Dolgin E. COVID vaccine immunity is waning—how much does that matter? Nature. 2021;597:606–607. doi: 10.1038/d41586-021-02532-4. [DOI] [PubMed] [Google Scholar]
- 7.Ministerio de Salud de Chile Immunization campaign against SARS-CoV-2: early estimates of the effectiveness of booster shots in Chile. Oct 2021. https://www.minsal.cl/wp-content/uploads/2021/10/2021-10-07-EFECTIVIDAD-DOSIS-DE-REFUERZO_ENG.pdf
- 9.Khoury DS, Cromer D, Reynaldi A, et al. Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection. Nat Med. 2021;27:1205–1211. doi: 10.1038/s41591-021-01377-8. [DOI] [PubMed] [Google Scholar]
- 10.Cromer D, Steain M, Reynaldi A, et al. Neutralising antibody titres as predictors of protection against SARS-CoV-2 variants and the impact of boosting: a meta-analysis. Lancet Microbe. 2022;3:e52–e61. doi: 10.1016/S2666-5247(21)00267-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.CTK Biotech OnSite COVID-19 IgG/IgM rapid test– (serum/plasma/whole blood) 2020. https://drive.google.com/file/d/1ATIHN4yKm_X9GPgJZER46e9YSJZfuO6R/view?usp=sharing
- 12.Beltrán-Pavez C, Riquelme-Barrios S, Oyarzún-Arrau A, et al. Insights into neutralizing antibody responses in individuals exposed to SARS-CoV-2 in Chile. Sci Adv. 2021;7 doi: 10.1126/sciadv.abe6855. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bezanson J, Edelman A, Karpinski S, Shah VB. Julia: a fresh approach to numerical computing. SIAM Rev. 2017;59:65–98. [Google Scholar]
- 14.Lim WW, Mak L, Leung GM, Cowling BJ, Peiris M. Comparative immunogenicity of mRNA and inactivated vaccines against COVID-19. Lancet Microbe. 2021;2:e423. doi: 10.1016/S2666-5247(21)00177-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Naranbhai V, Garcia-Beltran WF, Chang CC, et al. Comparative immunogenicity and effectiveness of mRNA-1273, BNT162b2, and Ad26.COV2.S COVID-19 vaccines. J Infect Dis. 2022;225:1141–1150. doi: 10.1093/infdis/jiab593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Schultz BM, Melo-González F, Duarte LF, et al. A booster dose of CoronaVac increases neutralizing antibodies and T cells that recognize delta and omicron variants of concern. MBio. 2022;13 doi: 10.1128/mbio.01423-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Leung NHL, Cheng SMS, Martín-Sánchez M, et al. Immunogenicity of a third dose of BNT162b2 to ancestral SARS-CoV-2 & omicron variant in adults who received two doses of inactivated vaccine. Clin Infect Dis. 2022 doi: 10.1093/cid/ciac458. published online June 8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zuo F, Abolhassani H, Du L, et al. Heterologous immunization with inactivated vaccine followed by mRNA-booster elicits strong immunity against SARS-CoV-2 omicron variant. Nat Commun. 2022;13 doi: 10.1038/s41467-022-30340-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Cheng H, Peng Z, Si S, et al. Immunogenicity and safety of homologous and heterologous prime-boost immunization with COVID-19 vaccine: systematic review and meta-analysis. Vaccines (Basel) 2022;10:798. doi: 10.3390/vaccines10050798. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Marra AR, Miraglia JL, Malheiros DT, et al. Effectiveness of heterologous COVID-19 vaccine booster dosing in Brazilian healthcare workers, 2021. Clin Infect Dis. 2022 doi: 10.1093/cid/ciac430. published online May 27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kuloğlu ZE, El R, Guney-Esken G, et al. Effect of BTN162b2 and CoronaVac boosters on humoral and cellular immunity of individuals previously fully vaccinated with CoronaVac against SARS-CoV-2: a longitudinal study. Allergy. 2022;77:2459–2467. doi: 10.1111/all.15316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Tawinprai K, Siripongboonsitti T, Porntharukchareon T, et al. Reactogenicity, immunogenicity, and humoral immune response dynamics after the third dose of heterologous COVID-19 vaccines in participants fully vaccinated with inactivated vaccine. Expert Rev Vaccines. 2022 doi: 10.1080/14760584.2022.2099380. published online July 13. [DOI] [PubMed] [Google Scholar]
- 23.Costa Clemens SA, Weckx L, Clemens R, et al. Heterologous versus homologous COVID-19 booster vaccination in previous recipients of two doses of CoronaVac COVID-19 vaccine in Brazil (RHH-001): a phase 4, non-inferiority, single blind, randomised study. Lancet. 2022;399:521–529. doi: 10.1016/S0140-6736(22)00094-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Jara A, Undurraga EA, Zubizarreta JR, et al. Effectiveness of homologous and heterologous booster doses for an inactivated SARS-CoV-2 vaccine: a large-scale prospective cohort study. Lancet Glob Health. 2022;10:e798–e806. doi: 10.1016/S2214-109X(22)00112-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Pérez-Then E, Lucas C, Monteiro VS, et al. Neutralizing antibodies against the SARS-CoV-2 delta and omicron variants following heterologous CoronaVac plus BNT162b2 booster vaccination. Nat Med. 2022;28:481–485. doi: 10.1038/s41591-022-01705-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Krone M, Gütling J, Wagener J, et al. Performance of three SARS-CoV-2 immunoassays, three rapid lateral flow tests, and a novel bead-based affinity surrogate test for the detection of SARS-CoV-2 antibodies in human serum. J Clin Microbiol. 2021;59 doi: 10.1128/JCM.00319-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Wang JJ, Zhang N, Richardson SA, Wu JV. Rapid lateral flow tests for the detection of SARS-CoV-2 neutralizing antibodies. Expert Rev Mol Diagn. 2021;21:363–370. doi: 10.1080/14737159.2021.1913123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Girl P, Zwirglmaier K, von Buttlar H, Wölfel R, Müller K. Evaluation of two rapid lateral flow tests and two surrogate ELISAs for the detection of SARS-CoV-2 specific neutralizing antibodies. Front Med (Lausanne) 2022;9 doi: 10.3389/fmed.2022.820151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Moshe M, Daunt A, Flower B, et al. SARS-CoV-2 lateral flow assays for possible use in national covid-19 seroprevalence surveys (React 2): diagnostic accuracy study. BMJ. 2021;372:n423. doi: 10.1136/bmj.n423. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data cannot be shared directly by the authors because of data protection regulation. Data is accessible to authorised researchers after application to the Chilean Ministry of Health.