Immunogenicity of a single dose of BNT162b2, ChAdOx1 nCoV-19, or CoronaVac against SARS-CoV-2 delta and omicron variants among previously infected adults: A randomized trial

To the Editor,

In this journal, Westrop et al. (2022) reported that heterologous COVID-19 vaccine schedules involving adenoviral-vector and mRNA vaccines were highly immunogenic, and that the immune responses were higher among previously infected adults compared to infection-naïve adults.¹ These findings were consistent with a number of studies that demonstrated that two-dose mRNA COVID-19 vaccine regimens in individuals previously infected with SARS-CoV-2 were more immunogenic compared to natural infection or vaccination alone.2, 3, 4 However, this evidence was largely based on mRNA vaccine regimens and was conducted before the introduction and global spread of the omicron variant.

Prior to the emergence of omicron, a retrospective study in Israel suggested that a single dose of BNT162b2 after natural infection provided high vaccine effectiveness (82%) against SARS-CoV-2 reinfection compared to participants who were infected but unvaccinated; this was the same for two doses.⁵ However, it remained unclear whether natural immunity against previous SARS-CoV-2 variants (i.e., Wuhan or alpha), followed by a single dose of COVID-19 vaccine, was sufficient to protect individuals from omicron. As the immunogenicity of inactivated and adenoviral vectored vaccines among individuals previously infected with SARS-CoV-2 had not been described, we investigated the immunogenicity of a single dose of BNT162b2, ChAdOx1 nCoV-19 (ChAdOx1), or CoronaVac against SARS-CoV-2 delta and omicron variants in individuals who recovered from COVID-19.

Enrolled participants provided written informed consent, were ≥18 years old, and diagnosed with COVID-19 between 6-24 weeks prior to study recruitment. Exclusion criteria included individuals: with RT-PCR confirmed SARS-CoV-2 reinfection, that received two doses of any registered COVID-19 vaccine before diagnosis of COVID-19, that received prophylactic treatment or investigational agents against COVID-19, that had a history of vaccine hypersensitivity, that were immunocompromised or received immunosuppressive agents, or that had unstable underlying diseases. This randomized study was performed between May and August 2021, when D614G ancestral and alpha SARS-CoV-2 variants were circulating in Thailand. Participants were randomly assigned to receive a single dose of CoronaVac (Sinovac Life Science), ChAdOx1 (Oxford-AstraZeneca), or BNT162b2 (Pfizer-BioNTech), with blood samples collected before and 14 days after vaccination. SARS-CoV-2 specific anti-receptor binding domain IgG (Abbott SARS-CoV-2 IgG II Quant assay), pseudovirus-based neutralizing antibodies against delta and omicron variants, as well as SARS-CoV-2-specific memory T-Cell responses against spike and nucleoprotein-membrane protein-open reading frame (NMO) proteins of the ancestral strain (IFN-γ ELISpot, Mabtech, Sweden) were measured. This study was approved by the Siriraj Institutional Review Board (COA no. Si 546/2021), and its protocol registered at the Thai Clinical Trial Registry (TCTR20210720005).

Anti-RBD IgG and neutralizing antibodies were reported as geometric mean concentration (GMC) and geometric mean titers (GMT) with 95% confidence intervals (CI), respectively. For multiple comparisons, the analysis of variance (ANOVA) for parametric data and Wilcoxon Rank Sum test for non-parametric data were used to assess the differences among groups. Paired and unpaired t tests were used to compare GMC and GMT within group and between groups, respectively, using GraphPad Prism 9 version 9.2.0 (283) (GraphPad Software, CA, USA), respectively. Correlation analysis was performed using the Spearman correlation analysis. Other statistical analyses were conducted using STATA version 17 (Stata Corp, LP, College Station, TX, USA). Adverse events (AEs) and secondary outcome measures were descriptively analyzed.

Of the 166 participants screened, 23 were excluded from the study per aforementioned criteria, and 29 were excluded from the analysis due to receiving doses of COVID-19 vaccines prior to SARS-CoV-2 infection (Supplementary Fig. 1). A total of 114 adults were included in the final analysis , 52 (45.2%) participants were male, and the median age (IQR) was 38.5 (29-45) years (Table 1). There was no statistical difference in baseline characteristics and intervals between COVID-19 diagnosis and vaccination across the three vaccination groups (Table 1 ).

Table 1.

Baseline characteristics of vaccine-naïve participants enrolled by type of vaccines.

Baseline characteristics	Type of vaccinations following SARS-CoV-2 detection
	Total	BNT162b2	ChAdOx1	CoronaVac	P value
Number of participants, n (%)	114	31 (27.2)	45 (39.5)	38 (33.3)
Male, n (%)	52	14 (45.2)	19 (42.2)	19 (50.0)	0.78
Age, median (IQR)	38.5 (29, 45)	37 (28, 44)	39 (29, 48)	39 (29, 45)	0.90
Body mass index, median (IQR)	24.40 (21.0, 27.4)	23.3 (20.3, 26.4)	24.5 (21.8, 27.4)	25.4 (21.1, 28.6)	0.17
Days from diagnosis to enrollment, median (IQR)	85 (61, 119)	67 (54, 126)	111(61, 120)	88 (62, 114)	0.20

A single dose of either study vaccine significantly increased the SARS-CoV-2 IgG concentration (Supplementary Fig. 2) and neutralizing antibody titers (PVNT₅₀) against delta and omicron variants (Fig. 1 A-B) from pre-vaccination. Both BNT162b2 and ChAdOx1 groups induced significantly higher antibody responses than the CoronaVac group: the geometric mean PVNT₅₀ (95% CI) for BNT162b2, ChAdOx1, and CoronaVac groups for delta were 2,503 (1865, 3359), 1,093 (808,1480), and 340 (228,507); and for omicron were 372 (193,717), 176 (84.8, 364), and 6.9 (2.2, 22.1), respectively (Fig. 1A-B). Importantly, a single dose of BNT162b2 or ChAdOx1 but not CoronaVac induced similar PVNT₅₀ against delta and omicron as to those induced by three doses of COVID-19 vaccines (two primary doses of ChAdOx1 plus a BNT162b2b booster) previously reported in healthy Thai adults who never had SARS-CoV-2 infection (5) (Fig. 1A-B). A single dose of CoronaVac in previously infected individuals produced similar PVNT₅₀ as those who received two doses of BNT162b2 primary series (7) (Fig. 1B).

Fig. 1.

A) Neutralization titers against delta variant determined using the pseudovirus neutralization assay. B) Neutralization titers against omicron variant determined using the pseudovirus neutralization assay. C) IFN-γ T-Cell responses against spike protein of ancestral strain determined using ELISpot assay. D) IFN-γ T-Cell responses against NMO protein of ancestral strain determined using ELISpot assay. ELISpot indicates enzyme-linked immunospot; PVNT₅₀, 50% reduction of infectivity; IFN, interferon; SFU, spot-forming unit; pre-vac, pre-vaccination; post-vac, post-vaccination. In panel A and B, data were presented as geometric mean titers with 95% confidence interval. Paired pre- and post-vaccination PVNT₅₀ titers were log-transformed and compared using paired Student's t-test. PVNT₅₀ titers between vaccine groups were log-transformed and compared using unpaired Student's t-test. In panel C and D, data were presented as geometric mean SFU with 95% confidential interval. Paired pre- and post-vaccination SFU were log-transformed and compared using paired Student's t-test. SFU between vaccine groups were log-transformed and compared using unpaired Student's t-test. The reference bar in panel A and B were the results from studies conducted in the same setting as this study (6,7).

Similarly, all three vaccines significantly increased T-Cell responses to spike protein (Fig. 1C), while only CoronaVac significantly increased T-Cell responses to NMO proteins (Fig. 1D). Both BNT162b2 and ChAdOx1 induced significantly higher T-Cell responses than CoronaVac. There was a strong correlation between SARS-CoV-2 IgG and spike-specific T-Cell responses (Supplementary Fig. 3).

Our results suggested that among individuals who were previously infected with SARS-CoV-2 (ancestral strain, with D614G mutation, and alpha strain), a single dose of BNT162b2 or ChAdOx1, but not CoronaVac, may provide similar protection against delta and omicron variants as individuals vaccinated with three doses of COVID-19 vaccines. These findings are particularly relevant for countries where there have been with limited access to vaccines such as BNT162b2 and ChAdOx1 but have a high level of natural immunity to SARS-CoV-2. Limitations of the study include the small sample size, and generalization of our findings to other COVID-19 vaccines and previous infection with other SARS-CoV-2 variants.

Funding

This work was supported by the Faculty Research Fund, Faculty of Medicine Siriraj Hospital (grant number: R016436005); and the Petroleum Authority of Thailand.

Declaration of Competing Interests

The authors declare no competing interests.