Antibody Responses and Reactogenicity of a Heterologous, Full-Dose Messenger RNA-1273 Booster in Heavily SARS-CoV-2–Exposed CoronaVac-Vaccinated Health-Care Workers in Indonesia: A Real-World Observational Study

ABSTRACT. Real-world data on heterologous boosting with messenger RNA (mRNA)-1273 (Moderna) after inactivated COVID-19 vaccination are limited. We report mRNA-1273 boosting in heavily SARS-CoV-2–exposed Indonesian health-care workers who received a two-dose CoronaVac 6 months prior. Between August and November 2021, we measured SARS-CoV-2 spike-specific IgG binding antibody (Bab) titers in all 304 participants, and neutralizing antibody titers in a random subset of 71 participants, on stored paired serum samples taken before and 28 days after a full-dose (100-μg) mRNA-1273 booster. At the time of the mRNA-1273 boost, 107 participants (35.2%) were not previously infected (naive vaccinated), 42 (13.8%) were infected before CoronaVac (infected vaccinated), and 155 (51.0%) were infected after CoronaVac (mostly during the Delta wave; vaccinated infected). At time of the mRNA-1273 boost, neutralizing antibodies could still be detected in 83% of participants (59 of 71) overall, 60% of naive-vaccinated participants (15 of 25), 95.7% of vaccinated-infected participants (22 of 23), and 95.7% of infected vaccinated participants (22 of 23). After the mRNA-1273 boost, 100% of participants (71 of 71) had neutralizing antibody activity, with increases in median Bab and neutralizing antibody serum titers of 9.3- and 27.0-fold overall, 89.1- and 2,803.4-fold in naive-vaccinated participants, 15.9- and 19.9-fold in infected-vaccinated participants, and 2.2- and 18.4-fold in vaccinated-infected participants. In the multivariable analysis, Bab titers after the mRNA-1273 boost were greatest in individuals who had a previous virus breakthrough post-CoronaVac, and when a longer time period (> 4 months) had elapsed since the most recent prior “spike antigen exposure” (either second CoronaVac or virus breakthrough). Overall, adverse reactions were mild and short-lived. In conclusion, a full-dose mRNA-1273 booster after CoronaVac was well tolerated and immunogenic after 28 days, including in those with very low antibody levels.


INTRODUCTION
Vaccination is a key global strategy to control the COVID-19 pandemic. To date, eight COVID-19 vaccines have received an Emergency Use Listing by the WHO, 1 most of which require a two-dose regimen. Accumulating evidence shows a progressive increase in breakthrough infections over time since the second dose, which is associated with diminishing humoral immunity. [2][3][4][5] Neutralization and vaccine effectiveness after two-dose vaccine schedules are particularly reduced for the recent Omicron variant (B.1.1.529), 6 but with significant restoration after a third vaccine booster dose. [7][8][9][10][11] Recent trials suggested that heterologous (or "mix and match") virus-vectored or messenger RNA (mRNA) booster strategies were more immunogenic than a homologous schedule, [12][13][14][15][16] albeit with increased reactogenicity in some combinations. 17 CoronaVac (SinoVac Life Sciences Co. Ltd., Beijing, China), an inactivated whole-virus vaccine that received WHO Emergency Use Listing on June 1, 2021, 18 is among the most widely administered COVID-19 vaccines worldwide, with more than 2 billion doses in 54 countries as of March 2022. Vaccine efficacy against symptomatic illness after two-dose CoronaVac was 50.5% to 83.5% in randomized trials. [18][19][20] Vaccine effectiveness against hospitalization, intensive care unit admission, and COVID-19-related death was reported at 87.5%, 90.3%, and 86.3%, respectively, based on real-world data from Chile. 21 However, recent studies reported Corona-Vac to have lower immunogenicity than virus vector or mRNA vaccines, 22,23 a pronounced decrease of neutralizing antibody titers within a few months, [23][24][25] and reduced effectiveness in the elderly. 26,27 CoronaVac has been the predominant vaccine rolled out in Indonesia's countrywide mass vaccination campaign, which has seen the second highest number of COVID-19 cases (5,910,000) and deaths (153,000) in Asia (as of March 15, 2022). 28 Frequent breakthrough infections during the second, Delta-driven wave in July 2021 prompted the Ministry of Health to offer a heterologous booster of full-dose (100-mg) mRNA-1273 vaccine (Moderna Inc., Cambridge, MA) to frontline health-care workers. Recent studies in individuals with no history of laboratory-confirmed COVID-19 found that an mRNA booster of BNT162b2 (Pfizer-BioNTech) after primary vaccination with CoronaVac was well tolerated and induced elevated, virus-specific antibody levels and potent neutralization activity, including against the Omicron variant, albeit to a lesser extent. [8][9][10][11] However, real-world data are limited on mRNA-1273 as the booster as well as on "hybrid immunity" induced by previous SARS-CoV-2 infection. 29 We therefore established the Indonesia Vaccine Immunity & Infection Evaluation (INVITE) study, which included an assessment of a full-dose (100-mg) mRNA-1273 booster dose given 6 months after two-dose CoronaVac among heavily SARS-CoV-2-exposed health-care workers in Jakarta, Indonesia. We documented the antispike (anti-S) IgG binding antibody (Bab) and neutralizing antibody (Nab) responses and reactogenicity of this booster, and examined the effects of previous SARS-CoV-2 infections occurring before or after primary CoronaVac vaccination.

MATERIALS AND METHODS
Design and population. The INVITE study is a longitudinal observational cohort that includes health-care workers (doctors, nurses, and clinical laboratory analysts) who are 18 years or older at St Carolus and Pasar Minggu hospitals in Jakarta. We enrolled participants consecutively on the day they received their government-provided 100-mg mRNA-1273 booster dose at their own hospital, between August 5 and October 15, 2021. Eligible staff for the booster were those who had completed their two-dose primary vaccination at least 6 months prior, had no known recent (, 3 months) SARS-CoV-2 infection, and had no other contraindication for vaccination. Our analysis included all study participants for whom a paired pre-and postbooster serum sample was available. All participants provided written informed consent.
Procedures. Basic demographic and clinical data were captured on an online case report form. Solicited signs of local or systemic reactogenicity during the 7 days after the booster dose were recorded in a daily patient diary and, after study physician verification, were graded for severity (mild, moderate, severe, or disrupting activities of daily living [ADLs]). Previous COVID-19 was defined as a documented polymerase chain reaction-confirmed SARS-CoV-2 infection, and participants were thus categorized as not previously SARS-CoV 2-infected (naïve vaccinated), SARS-CoV-2-infected before the Corona-Vac primary vaccination (infected vaccinated), or infected after the CoronaVac primary vaccination (vaccinated infected, also known as Delta breakthrough). Venous blood samples were drawn on the day of the booster dose before administration and 28 (1 10) days thereafter (Supplemental Figure 1).
Serum was stored at -80 C and titers of IgG Babs against the SARS-CoV-2 spike receptor-binding domain were determined using the chemiluminescent microparticle immunoassay SARS-CoV-2 IgG II Quant assay (Abbott Laboratories, Abbott Park, IL) on the Architect i2000sr platform in accordance with the manufacturer's instructions, and expressed in WHO International Standard binding antibody units (BAU) per milliliter using the manufacturer's conversion factor (1 BAU/mL 5 0.142 3 arbitrary units/milliliter) and seropositivity defined as $ 7.1 BAU/mL. 30 Samples with results greater than the upper limit of quantification were tested again after dilution. To validate the Bab measurements, we also measured Nabs using the SARS-CoV-2 Surrogate Virus Neutralization Test (cPass; GenScript, Piscataway, NJ) according to the manufacturer's instructions, converted to WHO international units per milliliter for SARS-CoV-2 neutralization assays, 31 in a random subset of the three participant groups. The cutoff value for Nabs was defined as 28 IU/mL or 30% signal inhibition. 31,32 Statistical analysis. Paired comparisons of Bab and Nab titers before and after the mRNA-1273 booster dose were performed using the Wilcoxon matched pairs signed-rank test. Correlations of log 10 titers were expressed using the Spearman coefficient (r s ). Multivariable ordinal (proportional odds) logistic regression was performed to assess factors associated with pre-and postboost log 10 -transformed Bab titers (categorized in quartiles). Multivariable logistic regression was performed to assess factors associated with occurrence of any severe or disrupting adverse reactions. The independent variables included in the analysis were age, gender, previous SARS-CoV-2 infection, timing of previous SARS-CoV-2 infection before or after CoronaVac primary vaccination, any comorbidity (cardiovascular disease including hypertension, obesity [body mass index . 30 kg/m 2 ], diabetes mellitus, and asthma), preboost Bab titer, time interval between the first and second CoronaVac dose, and time interval between the second CoronaVac and mRNA-1273 booster dose. A two-sided P , 0.05 was considered significant. Statistical analyses were performed with Stata/IC version 15.1 (StataCorp, College Station, TX).
Factors associated with pre-and postbooster Bab titers In multivariable analysis, previous SARS-CoV-2 infection either before or after the CoronaVac primary vaccination was strongly associated with a greater preboost Bab titer (P 5 0.001 and P , 0.0001, respectively) and, as expected, pre-mRNA-1273 boost Bab titers were lower when a longer period of time (. 3 months) had elapsed since the most recent prior spike antigen exposure (either the second Coro-naVac dose or a breakthrough infection) compared with a more recent exposure.
Post-mRNA-1273 boost Bab titers were greater in participants who had a previous SARS-CoV-2 infection after the CoronaVac primary vaccination (i.e., virus breakthrough), and when a longer period of time (. 4 months) had elapsed since the most recent prior spike antigen exposure (either the second CoronaVac dose or a breakthrough infection) compared with more recent exposure. Age, gender, comorbidity, and time interval between the first and second CoronaVac dose were not associated with the pre-or post-mRNA-1273 boost Bab titer; and pre-mRNA-1273 boost Bab titer and previous SARS-CoV-2 infection before Coro-naVac were not associated with the post-mRNA-1273-boost Bab titer ( Table 2). Reactogenicity of mRNA-1273 booster. The 304 participants reported a total of 300 adverse reactions within 7 days after receiving the mRNA-1273 booster. The percentage of participants with any solicited adverse reaction was 98.7% (n 5 300), 96.4% (n 5 293) for any local reaction (injection site pain and swelling most frequent), and 90.8% (n 5276) for any systemic reaction (myalgia, fever/chills, and headache being most frequent). Most solicited local and systemic adverse reactions were mild (n 5 86, 28.3%) or moderate (n 5 131, 43.1%), whereas severe (n 5 59, 19.4%) or those disrupting ADLs (n 5 24, 7.9%) were less common (Figure 4, Supplemental Table 2). All adverse reactions were short-lived and none required hospitalization. Severe or ADL-disrupting adverse reactions were associated with a longer interval between the second dose and the mRNA-1273 booster (OR, 1.37/month increase; IQR, 1.09-1.72; P 5 0.001) and were inversely associated with older age (OR, 0.71/age decade increase; IQR, 0.54-0.92; P 5 0.011). Gender, comorbidity, previous SARS-CoV-2 infection, time interval between the first and second dose, and pre-and postboost Bab titers were not associated with any adverse reactions (Supplemental Table 3).

DISCUSSION
This real-world study in Indonesian health-care workers who received two-dose CoronaVac primary vaccination 6 months prior found that humoral immunity diminished substantially over time, as reported previously. [23][24][25] As expected, this was more pronounced in those who were not previously infected with SARS-CoV-2 than in those who had hybrid immunity resulting from previous infection. The heterologous mRNA-1273 vaccine booster dose increased Bab and Nab titers significantly in all tested individuals, including those who had undetectable antibody levels before the booster. In adjusted multivariable analysis, Bab titers after receiving the mRNA-1273 booster were found to be greatest in individuals who had a history of a virus breakthrough after the primary CoronaVac vaccination, and in those for whom a longer time period (. 4 months) had elapsed since the most recent prior spike antigen exposure (either the second CoronaVac dose or a virus breakthrough) compared with more recent exposure. This finding of a longer interval between vaccine doses being beneficial for antibody production builds on existing literature for doses 1 and 2 of Oxford/AstraZeneca and mRNA vaccines. [33][34][35] Our observations confirm intense SARS-CoV-2 exposure among frontline health workers in Jakarta during the Delta wave in July 2021. 36 Our findings concur with the notion that Delta breakthrough infections postvaccination are associated with subsequent more potent and durable immune responses than those seen in individuals who were infected only or received only two doses of the COVID-19 vaccine. 37 From an immunological perspective, we noted a lower pre-mRNA-1273 booster correlation between Bab and Nab titers in the vaccinated-infected group, wherein the most recent exposure to the spike protein was from natural infection rather than vaccination. This finding is consistent with natural infection generating diverse anti-S Bab to nonneutralizing epitopes-a notion consistent with a recent study 38 in which plasma from individuals who had been infected and subsequently received mRNA vaccination could better neutralize an engineered SARS-CoV-2 polymutant spike protein or diverse sarbecovirus spike proteins. Moreover, we surmise the observed low post-mRNA-1273 booster Bab and Nab titer correlations could be attributed to antibody production ceiling effects, as reported previously 39 for the association between anti-S1 IgG titers and the inhibition of the binding between S1 and angiotensin converting enzyme 2 after the second mRNA vaccine dose.
The high-dose heterologous mRNA-1273 booster was generally well tolerated, with common reports of injection site pain and swelling; as well as muscle pain, fever/chills, and headache; and 7.9% experienced a short-lived impairment in performing ADLs. The mRNA-1273 booster was better tolerated with increasing age and for a shorter time period since the second CoronaVac dose. These findings largely concur with Moderna's phase II/III trial of the highdose (100-mg) booster dose, which was generally safe and well tolerated, with only slightly more frequent adverse reactions compared with the 50-mg dose. 40 Our study findings add to current evidence on the use of mRNA booster vaccines in three main ways. First, it established that heterologous mRNA vaccine boosting with mRNA-1273 (Moderna) after CoronaVac is immunogenic and well tolerated in a highly virus-exposed population reflecting real-world conditions. 29 This concurs with other recent trials in which heterologous boosting with the other widely used mRNA vaccine BNT162b2 (Pfizer-BioNTech) after CoronaVac demonstrated better immunogenicity than a homologous CoronaVac booster against wild-type and Delta and Omicron variants. [8][9][10][11] Heterologous boosting has been shown to result in an enhanced quantitative profile of antibody and T-cell responses, and expands the ability to recognize multiple regions of spike compared with a homologous booster. 41,42 Second, our study evaluated a 100-mg full dose of the mRNA-1273 booster rather than the authorized 50-mg booster dose. On December 12,, 2021, Moderna announced 40 preliminary data that a mRNA-1273 booster of 50 mg increased Nabs 37-fold, and that of 100 mg increased Nabs 83-fold, compared with preboost levels, in person without previous infection. Third, the findings contribute to informing policy on flexible options in deploying COVID-19 vaccines in mix-and-match schedules, with particular relevance for FIGURE 2. Antispike binding antibody (Bab) titers before messenger RNA (mRNA)-1273 booster and time since most recent "spike antigen exposure." Scatterplot of antispike Bab titers before mRNA-1273 booster and time since most recent spike antigen exposure (either by the CoronaVac second dose or SARS-CoV-2 infection, whichever was the most recent). Pre-mRNA-1273 boost Bab titers were higher overall in previously infected-vaccinated than in naive-vaccinated participants. Pre-mRNA-1273 boost Bab titers correlated inversely with the time since the most recent documented spike antigen exposure. Participants with a breakthrough infection post-CoronaVac (vaccinated-infected) had the shortest time since most recent spike antigen exposure, and had a higher pre-mRNA-1273 boost Bab titer than those infected before CoronaVac (infected-vaccinated) and naive-vaccinated participants. BAU 5 binding antibody units.      countries that are largely dependent on inactivated vaccines. As policymakers in several countries have started implementing third or periodic boosting to protect the most vulnerable populations, and mitigate health-care and economic impacts, real-world data will be critical to guide decisions regarding when, which populations, and which boosters should be administered. There are some limitations to our study. First, we did not assess anti-N IgG, Nabs against Delta and Omicron variants of concern, or cellular immunity. Although accumulating evidence suggests that the anti-S IgG Bab and Nab responses correlate with protection against infection, 43,44 it is important to recognize there is not yet an established or well-defined correlate of long-term vaccine protection, 45 and protection against hospitalization and death is likely to involve Bab and T-cell responses, which is supported by global data showing decreased risk of severe disease after Omicron infection compared with previous waves in well-vaccinated populations. 46 Second, we could not fully confirm all who had been infected by SARS-CoV-2 because routine surveillance may have been incomplete, and the use of anti-N IgG as a proxy for previous infection was not valid because of the initial immunization with a whole-virus inactivated vaccine. Third, the sample size and follow-up period were insufficient to identify less common or late adverse events after booster vaccination, and the immunogenicity data were limited to immune responses through study day 28. Fourth, only immunological data were collected and therefore this investigation lacks information regarding the effectiveness of a heterologous mRNA-1273 booster vaccination, in particular against infection with the emergent BA.1 and BA.2 subvariants, and future variants. Last, demographics of the volunteers were not representative of the Indonesian population, and elderly individuals were not represented.
In conclusion, our real-world data indicate that, after a primary course of two-dose CoronaVac inactivated vaccine, a heterologous, full-dose mRNA-1273 booster was immunogenic after 28 days of follow-up and was well tolerated among heavily SARS-CoV-2-exposed health-care workers, even in those with very low preboost antibody titers. Further trials on clinical end points of optimized booster doses and variant-specific or multivalent vaccines in response to decreased susceptibility to neutralization of emerging SARS-CoV-2 variants of concern are warranted. 47 the open-access fee for this article due to the ongoing COVID-19 pandemic.
Financial support: This work was funded by the Wellcome Trust, UK (106680/Z/14/Z and 222574/Z/21/Z). S. D. is funded by a National Institute for Health and Care Research Global Research Professorship (NIHR300791). The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.
Disclosure: S. D. is a scientific advisor to the Scottish Parliament on COVID-19 immunology, for which she receives a fee. The study was approved by the research ethics committee of the Faculty of Medicine Universitas Indonesia (841/UN2.F1.ETIK/PPM.00.02/2021) and the Oxford Tropical Research Ethics Committee .