Safety and immunogenicity of a heterologous booster with an RBD virus-like particle vaccine following two- or three-dose inactivated COVID-19 vaccine

ABSTRACT

LYB001 is an innovative recombinant SARS-CoV-2 vaccine that displays a repetitive array of the spike glycoprotein’s receptor-binding domain (RBD) on a virus-like particle (VLP) vector to boost the immune system, produced using Covalink plug-and-display protein binding technology. LYB001’s safety and immunogenicity were assessed in 119 participants receiving a booster with (1) 30 μg LYB001 (I-I-30 L) or CoronaVac (I-I-C), (2) 60 μg LYB001 (I-I-60 L) or CoronaVac in a ratio of 2:1 after two-dose primary series of inactivated COVID-19 vaccine, and (3) 30 μg LYB001 (I-I-I-30 L) after three-dose inactivated COVID-19 vaccine. A well-tolerated reactogenicity profile was observed for LYB001 as a heterologous booster, with adverse reactions being predominantly mild in severity and transient. LYB001 elicited a substantial increase in terms of the neutralizing antibody response against prototype SARS-CoV-2 28 days after booster, with GMT (95%CI) of 1237.8 (747.2, 2050.6), 554.3 (374.6, 820.2), 181.9 (107.6, 307.6), and 1200.2 (831.5, 1732.3) in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups, respectively. LYB001 also elicited a cross-neutralizing antibody response against the BA.4/5 strain, dominant during the study period, with GMT of 201.1 (102.7, 393.7), 63.0 (35.1, 113.1), 29.2 (16.9, 50.3), and 115.3 (63.9, 208.1) in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups, respectively, at 28 days after booster. Additionally, RBD-specific IFN-γ, IL-2, IL-4 secreting T cells dramatically increased at 14 days after a single LYB001 booster. Our data confirmed the favorable safety and immunogenicity profile of LYB001 and supported the continued clinical development of this promising candidate that utilizes the VLP platform to provide protection against COVID-19.

KEYWORDS:

• SARS-CoV-2
• safety
• immunogenicity
• virus-like particle (VLP)
• booster

Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) continues to evolve with the emergence of Omicron and its sublineages, outcompeting other variants of concern (VOCs), resulting in several variant-driven waves of breakthrough infections. Vaccination is the most cost-effective tool for tackling the coronavirus disease 2019 (COVID-19) pandemic. To date, most approved vaccines against COVID-19 target the prototype SARS-CoV-2 sequence, including those based on the mRNA, adenovirus vectors, protein/adjuvant subunits, and inactivated virus platforms. These vaccines have demonstrated reduced protective effectiveness against COVID-19 over time due to waning immunity and the emergence of immune-evading SARS-CoV-2 VOCs.^1–4 In the absence of Omicron-adapted vaccines, optimizing the delivery of first-generation vaccines using a heterologous booster strategy (mix and match) appears to induce a better immune response than a homologous booster strategy.^5–10 Two inactivated COVID-19 vaccines (ICVs), developed by Sinovac and Sinopharm in China, accounted for approximately 45% of the global delivered doses in 2021 and notably contributed to worldwide vaccine coverage.¹¹ These ICVs have also proven to be highly effective against severe COVID-19 disease outcomes.^12,13 However, they exhibited poor or even absent neutralizing antibody (NAb) activity and effectiveness against infection with Omicron sublineages after the two-dose primary series, a primary booster, or even a secondary booster.^14–17

The receptor-binding domain (RBD) on the spike glycoprotein of SARS-CoV-2 is an immunodominant antigen that contains epitopes for most neutralizing antibodies¹⁸ and LYB001 is an innovative recombinant vaccine displaying repetitive RBDs on the surface of a virus-like particle (VLP) vector.¹⁹ The array of RBD on the VLP was achieved using Covalink plug-and-display protein binding technology (isopeptide bond 4T/4C conjunction in Figure 1), similar to platforms described in other studies.²⁰ Because VLP and RBD can be expressed separately, the modular production of VLP in Escherichia coli and RBD in CHO cells is highly scalable. This platform also offers a shortened research and development cycle for a variant-adapted vaccine against rapidly evolving pathogens. Thus, vaccine adaptation can be easily accomplished, offering a major advantage in tackling the major global health challenges associated with human infectious diseases. Additionally, the highly repetitive antigen array (mimicking an actual virus) and relatively large particle size can enhance B-cell receptor cross-linking and antigen-presenting cell uptake and presentation, leading to strong stimulation of immune cells in the draining lymph nodes and overcoming the insufficient immunogenicity that can occur with soluble or monomeric recombinant subunit vaccines. Furthermore, the optimal orientation of the neutralizing epitope displayed on the VLP surface can result in a higher proportion of neutralizing antibodies.²¹

Figure 1. Design principle of LYB001.

RBD, receptor binding domain; VLP, virus-like particle.

The majority of the Chinese population had completed two- or three-dose ICV vaccinations at the beginning of the study, which made it challenging to obtain dose-response data in a naïve population. Additionally, the cross-NAb response against emerging variants, generated by the primary series of ICV targeting the prototype, waned substantially 6 months after the primary vaccination. A booster vaccination strategy was rolled out to restore immunity against emerging variants. In the absence of Omicron-adapted vaccines, we first explored the effects of a heterologous LYB001 booster at different doses compared with a homologous ICV booster among participants who completed two-dose ICV as dose-finding studies. Subsequently, we evaluated the safety and immunogenicity profiles of a heterologous LYB001 booster at a specific dose among participants who completed the three-dose ICV regimen. Our objective was to investigate whether alterations in reactogenicity and peak immunity occurred in relation to different primary ICV vaccinations. Herein, we present the safety and immunogenicity results of LYB001 as a booster vaccine at interval of 6–12 months in two- or three-dose ICV recipients.

Methods

Ethical approval

The study was registered with ClinicalTrials.gov (NCT05928455) and approved by the Institutional Review Boards of Chengdu Xinhua Hospital and Chongqing Red Cross Hospital. The trial was conducted in accordance with Good Clinical Practice guidelines and the principles stated in the Declaration of Helsinki. Written informed consent was obtained from each participant prior to any study-related procedures.

Study design and participants

This study was conducted at Chengdu Xinhua Hospital Affiliated to North Sichuan Medical College and Chongqing Red Cross Hospital (People’s Hospital of Jiangbei District). This study aimed to evaluate the safety and immunogenicity of a heterologous LYB001 booster at interval of 6–12 months following two or three doses of ICV in healthy participants aged 18–59 years. The study was performed in two parts: In part 1, a randomized, open-label, positive-controlled design was utilized to evaluate the safety and immunogenicity profile following different heterologous booster doses (30 and 60 μg) of LYB001 using a dose-escalation design. This was compared to a homologous booster dose of CoronaVac in adults aged 18–59 years who had completed a two-dose primary series of ICV 6–12 months earlier. Ninety participants (n = 45 per dose group) were planned to be enrolled in part 1, randomized 2:1 to receive LYB001 (30 and 60 μg) or CoronaVac. In part 2, a designated dose (30 μg) of LYB001 based on the preliminary results from part 1 was used as booster in 30 adults aged 18–59 years who had completed a three-dose primary series of ICV 6–12 months earlier. Participants with a known COVID-19 vaccination history other than ICV, history of SARS-CoV-2 infection, history of severe, uncontrolled chronic disease, or other conditions that, per the judgment of the investigator, might interfere with safety and immunogenicity assessment or pose possible risks to participants, were excluded from the study. For the full list of screening criteria, see the inclusion and exclusion criteria in the Supplementary Materials.

Randomization and masking

In part 1, participants were recruited using a dose-escalation study design. Participants who had completed two-dose primary series of ICV were randomly assigned in a ratio of 2:1 either to receive 30 μg LYB001 or a CoronaVac booster. After confirmation of an acceptable 7-day safety profile in this cohort, the study proceeded to the cohort of two-dose ICV recipients randomly assigned in a ratio of 2:1 either to receive 60 μg LYB001 or CoronaVac booster. Randomization of participants and vaccines was performed by an independent statistician using SAS statistical software version 9.4 or higher via the block randomization method, with a block of five and block size of nine. Randomization numbers were allocated to eligible participants in order of enrollment. Participants were randomly allocated to each group according to the randomization table. In part 2, randomization was not applicable because this was a single-arm study.

Blinding and masking were not applicable in this open-label study because the CoronaVac booster information of each participant had to be mandatorily recorded in the national vaccination system. However, all laboratory staff responsible for immunogenicity assessment and laboratory safety measures were blinded to the group allocation.

Procedures

The design of the investigational vaccine is shown in Figure 1. Briefly, LYB001 is a recombinant vaccine made using a procedure that expresses the VLP vector (NPM-4C) in Escherichia coli and the RBD (RBD-4T from the spike glycoprotein of the SARS-CoV-2 prototype strain) in CHO cells. Purified stock solutions of RBD-4T and NPM-4C were mixed to enable conjugation via isopeptide binding to produce final VLP (RBDM). Following the final purification of VLP, it was adsorbed onto the aluminum hydroxide adjuvant. The LYB001 (0.5 mL, doses of 30 or 60 μg) and CoronaVac (0.5 mL) vaccines were administered through intramuscular injection.

Safety assessments

In this trial, participants were required to stay at the trial site for a 30 min safety observation period for the potential development of immediate adverse events (AEs) after the vaccine booster. During the observation period, participants were instructed to fill out the diary card and were given a thermometer and a measurement scale for recording the AEs experienced within 7 days after the booster, including solicited local/systemic and unsolicited AEs. Solicited local AEs included injection-site pain, induration, redness, swelling, rash, and pruritus, whereas solicited systemic AEs included fever, diarrhea, nausea, vomiting, headache, myalgia (non-injection site), cough, fatigue, and acute allergic reaction. On day 8 after the booster, participants returned to the trial site to submit diary cards that were reviewed by the investigator, and contact cards were dispensed to participants to record unsolicited AEs within 8–28 days after the booster. The intensity of AEs was graded using appropriate guidelines issued by the National Medical Products Administration (NMPA) of China, and causality was assessed by the investigators.

Immunogenicity assessments

Blood samples for humoral immunogenicity assessment were drawn from participants at baseline (day 0 before vaccination) and at days 14, 28, and 90 after the booster, and used to determine: (1) spike glycoprotein binding IgG levels, and (2) NAb titers against prototype SARS-CoV-2 and circulating VOCs. Blood samples for cellular immunity were drawn from the participants at baseline and 14 days after the booster. Spike glycoprotein-binding IgGs were measured using ELISA, and the values were reported as binding antibody units (BAUs), in accordance with the manufacturer’s recommendations and the World Health Organization (WHO) International Standard and International Reference Panel for anti-SARS-CoV-2 immunoglobulin. NAb titers against the prototype SARS-CoV-2 and dominant VOCs were determined using vesicular stomatitis virus (VSV)-based pseudovirus-neutralizing assays by Chongqing Medleader Bio-Pharm Co., Ltd. Seroconversion was defined as either a four-fold increase in post-boost antibody levels from a seropositive (≥ cutoff value) baseline, or a seropositive conversion from a seronegative (< cutoff value) baseline. The cellular immune response was detected using enzyme-linked immunospot (ELISpot) assay, and was presented as the counts of spot forming cells (SFCs) per 3 × 10⁵ peripheral blood mononuclear cells (PBMCs) secreting interferon (IFN)-γ, interleukin (IL)-2, and IL-4 when stimulated by the RBD peptide pool ex vivo. Further details of the methodology used for the immunogenicity assays are provided in Supplementary Methods.

Outcomes

The primary objective of this study was to assess the safety and immunogenicity of LYB001 following a heterologous booster in adults aged 18–59 years who had previously completed two- or three-dose primary courses of ICV vaccination. The primary immunogenicity outcomes were the geometric mean titer (GMT), geometric mean fold rise (GMFR), and seroconversion rate (SCR) of spike glycoprotein-binding IgGs, NAb titers against the prototype SARS-CoV-2, and circulating VOCs at baseline and 14 and 28 days after the booster. The primary safety outcomes were immediate AEs within 30 min after the booster, solicited local/systemic AEs within 7 days, and unsolicited AEs within 28 days after the booster.

The secondary objective was to assess the immune response durability, which included the GMT and SCR of spike glycoprotein-binding IgGs, NAb titers against prototype SARS-CoV-2, and circulating VOCs measured 90 days after the booster. Secondary safety outcomes were severe adverse events (SAEs), adverse events of special interest (AESIs) within 90 days after the booster, and safety laboratory measures 3 days after the booster. The exploratory outcome was to assess the cellular immune response following a heterologous booster dose of LYB001, and the corresponding exploratory outcome was the RBD-specific IFN-γ, IL-2, and IL-4 secreting T cell response as measured by ELISpot assay at baseline and 14 days after the booster.

Statistical analysis

The sample size of this trial was not based on formal statistical hypotheses. Safety analyses were evaluated using the safety set (SS), which included all participants who received the booster dose. Immunogenicity analysis was performed in the per-protocol set for immunogenicity (I-PPS) following an intention-to-treat principle, including participants who had completed the booster immunization with immunogenicity results at day 0 before vaccination and at least one available post-boost immunogenicity result with no major protocol deviations. The counts and percentages of participants who experienced AEs were presented in the safety analyses, including solicited local/systemic AEs, unsolicited AEs, AEs graded as grade 3 or worse, AEs leading to participant withdrawal, SAEs, and AESIs. The NAb GMTs against the prototype SARS-CoV-2 and circulating VOCs at different time points after the booster were calculated with Clopper-Pearson 95% confidence intervals (CIs), and the t-test was used to compare log-transformed antibody titers between groups. Additionally, GMFRs and SCRs at different time points after the booster, relative to baseline, were calculated along with their Clopper-Pearson 95% CIs. The cellular immune responses (cytokine-secreting T cells by ELISpot assay) and their changes from baseline were statistically analyzed for each group at 14 days after the booster, and the differences were statistically tested using the Wilcoxon rank-sum test. The χ² test or Fisher’s exact test was used to analyze other categorical data. Given the exploratory nature of this study, no adjustments were made for multiple comparisons or multiplicity. Statistical analysis was performed using GraphPad Prism 9.0 (GraphPad Software Inc., California, USA).

Results

Participants

Between May 14 and June 30, 2022, 215 individuals were screened, and 120 eligible participants aged 18–59 years who had competed a two- or three-dose ICV 6–12 months earlier were enrolled in this study (Figure 2). One participant withdrew before the booster vaccination due to prior receipt of a two-dose recombinant protein subunit vaccine other than the ICV against COVID-19 (ZF2001). A total of 119 participants aged 18–59 years received the booster vaccination and were included in the analysis set. The mean (SD) participants’ age was 29.4 (8.2), 29.6 (7.8), 30.1 (9.1), and 30.3 (9.4) years, with a consistent sex distribution in the groups of participants receiving a booster with 30 μg LYB001 (I-I-30 L), 60 μg LYB001 (I-I-60 L), or CoronaVac (I-I-C) after two-dose primary series of ICV, or 30 μg LYB001 (I-I-I-30 L) after three-dose primary series of ICV, respectively. The mean (SD) prime-boost intervals were 290.2 (37.0), 267.8 (31.9), 276.9 (25.0), and 216.4 (37.3) days in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups, respectively. In each group, there was no medication (or vaccination) or medical (or allergy) history, which, in the opinion of the investigator, might compromise the participants’ well-being or confound protocol-specified assessments. The NAb titers against the prototype SARS-CoV-2 and Omicron BA.4/5 variants were low or absent at baseline; these were comparable across the I-I-30 L, I-I-60 L, and I-I-C groups and were lower than that of the I-I-I-30 L group (Table 1).

Figure 2. Study profile.

*One participant withdrew prior to booster vaccination due to prior receipt of a two-dose recombinant protein subunit vaccine against COVID-19 (ZF2001) other than ICV. ^†One participant was absent for blood collection at 90 days after the CoronaVac booster following two-dose ICV due to quarantine for COVID-19. ^‡One participant was absent for blood collection at 90 days after the CoronaVac booster following two-dose ICV due to quarantine for COVID-19. ^§One participant was absent for blood collection at 14 days after the 30 μg LYB001 booster following three-dose ICV due to quarantine for COVID-19. ICV, inactivated COVID-19 vaccine; SS, safety set; I-FAS, full analysis set for immunogenicity; I-PPS, per protocol set for immunogenicity (I-PPS1: I-PPS of 14 days after the booster; PPS2: I-PPS of 28 days after the booster), IPS, immunogenicity persistence set (IPS1: IPS of 90 days after the booster).

Table 1. Baseline characteristics of the participants.

	Booster after two-dose primary series of ICV		Booster after three-doses of ICV
	30 μg LYB001 (N = 30)	60 μg LYB001 (N = 30)	CoronaVac (N = 29)	30 μg LYB001 (N = 30)
Age (years)
Mean (SD)	29.4 (8.2)	29.6 (7.8)	30.1 (9.1)	30.3 (9.4)
Median (Range)	29.0 (18, 49)	27.5 (18, 54)	27.0 (19, 54)	27.0 (19, 52)
Sex
Male, n (%)	17 (56.67)	20 (66.67)	19 (65.52)	18 (60.0)
Female, n (%)	13 (43.33)	10 (33.33)	10 (34.48)	12 (40.0)
Ethnicity
Han Chinese, n (%)	29 (96.67)	30 (100.00)	27 (93.10)	26 (86.7)
Others, n (%)	1 (3.33)	0 (0.00)	2 (6.90)	4 (13.3)
History of prior infection (self-reported), n (%)	30 (100.0)	30 (100.0)	29 (100.0)	30 (100.0)
Prime-boost intervals
Mean (SD)	267.8 (31.9)	276.9 (25.0)	290.2 (37.0)	216.4 (37.3)
Median (Range)	264.0 (195, 330)	274.0 (231, 338)	288.0 (232, 362)	202.5 (183, 373)
Pre-booster immunity
GMCs of spike protein binding IgGs (95% CI), mIU/mL	17.3 (13.1, 22.9)	14.6 (11.8, 18.1)	17.2 (12.4, 24.0)	164.1 (113.3, 237.6)
GMTs of pseudovirus neutralizing antibody against prototype (95% CI)	8.5 (6.3, 11.6)	7.0 (5.2, 9.2)	9.0 (6.3, 12.7)	83.9 (55.5, 126.8)
GMTs of pseudovirus neutralizing antibody against BA.4/5 (95% CI)	5.0 (5.0, 5.0)	5.0 (5.0, 5.0)	5.0 (5.0, 5.0)	9.4 (6.2, 14.0)

SD, standard deviation; GMC, geometric mean concentration; GMT, geometric mean titer.

Safety

Overall adverse events/reactions

LYB001 as a heterologous booster after a two- or three-dose ICV was safe and well tolerated, with adverse reactions (vaccination-related AEs) being predominantly mild in severity (only two participants reported grade 2 adverse reactions) (Table 2). The majority of these adverse reactions were those commonly anticipated for intramuscularly administered vaccines and spontaneously resolved/recovered within a median duration of 2 days after symptom onset. The overall incidence rates of adverse reactions were 76.7% (23/30), 66.7% (20/30), 31.0% (9/29), and 63.3% (19/30) in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups, respectively, which were predominantly contributed by solicited local adverse reactions.

Table 2. Overall adverse events or reactions after booster administration.

Participants with at least one		Booster after two-dose primary series of ICV			Booster after three-dose ICV
		30 μg LYB001 (N = 30) n (%)	60 μg LYB001 (N = 30) n (%)	CoronaVac (N = 29) n (%)	30 μg LYB001 (N = 30) n (%)
Primary endpoints
AEs		24 (80.0)	22 (73.3)	10 (34.5)	20 (66.7)
Adverse reactions		23 (76.7)	20 (66.7)	9 (31.0)	19 (63.3)
Grade 3 or worse AEs		0 (0.0)	1 (3.3)	0 (0.0)	0 (0.0)
Grade 3 or worse adverse reactions		0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)
Solicited local adverse reactions	overall	21 (70.0)	17 (56.7)	5 (17.2)	18 (60.0)
	mild	20 (66.7)	17 (56.7)	5 (17.2)	18 (60)
	moderate	1 (3.3)	0 (0.0)	0 (0.0)	0 (0.0)
	severe	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)
Pain	incidence	21 (70.0)	17 (56.7)	5 (17.2)	18 (60.0)
	median duration	2.0	2.0	2.0	2.0
Swelling	incidence	2 (6.7)	1 (3.3)	0 (0.0)	0 (0.0)
	median duration	2.0	2.0	2.0	2.0
Redness	incidence	1 (3.3)	0 (0.0)	0 (0.0)	0 (0.0)
	median duration	2.5	1.0	/	/
Pruritus	incidence	0 (0.0)	1 (3.3)	0 (0.0)	0 (0.0)
	median duration	/	1.0	/	/
Solicited systemic adverse reactions	overall	6 (20.0)	5 (16.7)	1 (3.4)	2 (6.7)
	mild	6 (20.0)	5 (16.7)	1 (3.4)	2 (6.7)
	moderate	0 (0.0)	1 (3.3)	0 (0.0)	0 (0.0)
	severe	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)
Fatigue	incidence	5 (16.7)	3 (10.0)	1 (3.4)	2 (6.7)
	median duration	1.0	1.0	1.0	3.0
Nausea	incidence	1 (3.3)	0 (0.0)	0 (0.0)	0 (0.0)
	median duration	1.0	/	/	/
Myalgia	incidence	0 (0.0)	2 (6.7)	0 (0.0)	0 (0.0)
	median duration	/	1.5	/	/
Headache	incidence	2 (6.7)	0 (0.0)	0 (0.0)	1 (3.3)
	median duration	1.0	/	/	1.0
Cough	incidence	0 (0.0)	1 (3.3)	0 (0.0)	0 (0.0)
	median duration	/	4.0	/	/
Diarrhea	incidence	0 (0.0)	1 (3.3)	0 (0.0)	0 (0.0)
	median duration	/	1.0	/	/
Acute allergic reaction (rash)	incidence	0 (0.0)	1 (3.3)	0 (0.0)	0 (0.0)
	median duration	/	5.0	/	/
Unsolicited adverse reactions		7 (23.3)	5 (16.7)	4 (13.8)	8 (26.7)
Secondary endpoints
SAE		0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)
AESI		0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)
AEs leading to withdrawal		0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)
Clinically significant safety laboratory measures
Hematology	White blood cell count increased	2 (6.7)	0 (0.0)	0 (0.0)	0 (0.0)
	Lymphocyte count increased	1 (3.3)	0 (0.0)	0 (0.0)	0 (0.0)
Blood chemistry	Alanine aminotransferase increased	2 (6.7)	0 (0.0)	0 (0.0)	0 (0.0)
	Blood urea increased	2 (6.7)	0 (0.0)	0 (0.0)	1 (3.3)
	Blood glucose increased	1 (3.3)	0 (0.0)	0 (0.0)	1 (3.3)
	Blood creatinine increased	1 (3.3)	1 (3.3)	1 (3.4)	0 (0.0)
Coagulation function		0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)
Urinalysis	Red blood cells urine positive	1 (3.3)	0 (0.0)	0 (0.0)	3 (10.0)*

The incidence of adverse events or reactions is presented as n (%). *Two participants reported positive urine red blood cells at baseline without clinical significance. AEs, adverse events; ICV, inactivated COVID-19 vaccine; SAE, serious adverse event; AESI, adverse event of special interest.

Solicited local/systemic reactions and unsolicited reactions

Solicited local adverse reactions were reported by 70.0% (21/30), 56.7% (17/30), 17.2% (5/29), and 60% (18/30) of the participants and the most common (reported in ≥ 10% of the population in at least one group) solicited local adverse reaction was injection-site pain (predominantly mild in severity), which was reported by 70.0% (21/30), 56.7% (17/30), 17.2% (5/29), and 60% (18/30) of the participants in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups, respectively. The solicited systemic adverse reactions were reported by 20.0% (5/30), 16.7% (5/30), 3.4% (1/29), and 6.7% (2/30) of the participants and the most common solicited systemic adverse reaction was fatigue, which was reported by 16.7% (5/30), 10.0% (3/30), 3.4% (1/29), and 6.7% (2/30) of the participants in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups, respectively. The unsolicited adverse reactions within 28 days after the booster were reported by 7 (23.3%), 5 (16.7%), 4 (13.8%), and 8 (26.7%) participants in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups, respectively, which were comparable across groups.

Safety laboratory measures

Regarding the safety laboratory measures, the changes at day 3 after the booster from baseline did not indicate a particular trend concerning hematology, blood chemistry, urinalysis, and coagulation function parameters, with only a few significantly abnormal safety laboratory parameters reported (Table 2). All abnormal values spontaneously returned to normal at a subsequent visit without any clinical consequences.

SAEs, AESIs, and other significant AEs

No SAEs, AESIs, deaths, or AEs that led to withdrawal were reported within 90 days of booster administration. Only one participant in the 60 μg LYB001 booster group experienced a grade 3 or worse AE (preferred terms: pyrexia); however, this was judged as unrelated to the investigational vaccine.

Neutralization antibodies

As shown in Figure 3 and Table S1, the heterologous LYB001 booster elicited a potent NAb response against the prototype SARS-CoV-2 and cross-NAb response against the Omicron BA.4/5 strain, which was dominant during the study period, in participants previously immunized with two- or three-dose ICV. The NAb response was low or undetectable at baseline, increased considerably at day 14 or 28 after the booster, and moderately declined by day 90 after the booster. In every time points after the booster vaccination, the NAb responses against the prototype SARS-COV-2 were higher in the heterologous LYB001 booster groups (I-I-30 L, I-I-60 L, I-I-I-30 L) compared to the homologous CoronaVac booster group (I-I-C). The peak pseudovirus NAb GMTs (95% CI) against the prototype SARS-COV-2 were 1237.8 (747.2, 2050.6), 554.3 (374.6, 820.2), 181.9 (107.6, 307.6), and 1200.2 (831.5, 1732.3) at 28 days after the booster in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups, respectively. The peak pseudovirus NAb GMTs (95% CI) against the Omicron BA.4/5 were 109.6 (52.3, 229.6), 67.3 (36.5, 124.0), 29.9 (17.4, 51.4), and 135.7 (75.6, 243.5) at 14 days; 201.1 (102.7, 393.7), 63.0 (35.1, 113.1), 29.2 (16.9, 50.3), and 115.3 (63.9, 208.1) at 28 days after the booster in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups, respectively. Given the same two-dose primary course of ICV vaccination, the 30 μg LYB001 booster elicited the highest NAb response against prototype and BA.4/5, followed by the 60 μg LYB001, and then the CoronaVac, from a comparable (against prototype) or identical (against BA.4/5) baseline immunity. At 28 days after the booster, the pseudovirus NAb titers of the I-I-30 L group were 6.8 times (1237.8 vs 181.9) higher than that of the I-I-C group against prototype, and 6.9 times (201.1 vs 29.2) higher than that of the I-I-C group against BA.4/5. When evaluating the impact of different primary dose regimens (two-dose ICV vs three-dose ICV), the NAb responses of the I-I-30 L group were comparable or even higher than that of the I-I-I-30 L groups: (1237.8 vs 1200.2 against prototype, and 201.1 vs 115.3 against BA.4/5 at 28 days after booster), despite the higher baseline NAb titers in the I-I-I-30 L group. This indicated that a single 30 μg LYB001 booster could substantially restore the NAb response within a prime-boost interval of 6–12 months, irrespective of undetectable or partially detectable baseline antibody levels. Despite the decline in the NAb response at 90 days after the booster, the SCRs of NAb titers in the I-I-30 L group still remained 100% against prototype, and 80.0% against BA.4/5.

Figure 3. VSV-based neutralizing antibody titers against prototype SARS-CoV-2 and Omicron BA.4/5 strain.

Antibody values less than the lower limit of detection (LOD = 10) were replaced by 0.5 × LOD. The individual data in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups are indicated by Δ, ▽, ○, ▲, respectively. VSV, vesicular stomatitis virus; I-I-30 L, 30 μg LYB001 booster after two-dose inactivated COVID-19 vaccine; I-I-60 L, 60 μg LYB001 booster after two-dose inactivated COVID-19 vaccine; I-I-C, CoronaVac booster after two-dose inactivated COVID-19 vaccine; I-I-I-30 L, 30 μg LYB001 booster after three-dose inactivated COVID-19 vaccine.

Spike protein-binding antibodies

The spike glycoprotein-binding IgGs exhibited a similar trend to those seen for NAb responses, but with a slower waning 90 days after the booster (Figure 4). Specifically, the peak spike protein-binding IgGs (95% CI) were 1374.8 (892.2, 2118.3), 1161.1 (863.6, 1561.2), 455.9 (297.5, 698.7), and 1220.9 (862.5, 1728.2) at 28 days; and slowly declined to 1072.1 (686.5, 1674.3), 894.8 (606.0, 1321.3), 403.7 (245.6, 663.7), and 1011.2 (762.6, 1341.0) at 90 days after the booster in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups, respectively. The spike protein-binding IgG titers were comparable across the LYB001 booster groups (I-I-30 L, I-I-60 L, and I-I-I-30 L) regardless of different LYB001 dosages and primary vaccination schedules, all of which surpassed the concentrations observed in the I-I-C group.

Figure 4. The spike protein binding IgGs at baseline and 14, 28, and 90 days after booster administration.

Antibody values less than the lower limit of detection (LOD = 20) were replaced by 0.5 × LOD. The data in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups are indicated by Δ, ▽, ○, ▲, respectively. I-I-30 L, 30 μg LYB001 booster after two-dose inactivated COVID-19 vaccine; I-I-60 L, 60 μg LYB001 booster after two-dose inactivated COVID-19 vaccine; I-I-C, CoronaVac booster after two-dose inactivated COVID-19 vaccine; I-I-I-30 L, 30 μg LYB001 booster after three-dose inactivated COVID-19 vaccine.

Cellular immunity

T cell responses detected in the LYB001 groups were compared with those in the CoronaVac control group. The majority of participants exhibited low preexisting responses (Figure 5 and Table S2). Notably, the LYB001 booster induced significantly higher cytokine responses to the SARS-CoV-2 RBD peptide pool both in the 30 and 60 μg LYB001 groups as compared to the CoronaVac group, irrespective of administered dosages and primary vaccination regimens. SFCs per 3 × 10⁵ PBMCs as indicated by median (Q1, Q3) were 23.0 (8.0, 68.0), 23.0 (10.0, 42.0), and 18.0 (5.0, 46.0) for IFN-γ; 48.0 (26.0, 145.0), 39.0 (21.0, 70.0), and 54.0 (30.0, 99.0) for IL-2; 12.0 (4.0, 36.0), 8.0 (4.0, 33.0), and 18.0 (7.0, 43.0) for IL-4 at 14 days after the booster in the I-I-30 L, I-I-60 L, and I-I-I-30 L groups, respectively. In contrast, the SFCs per 3 × 10⁵ PBMCs were 2.0 (0.0, 4.0), 5.0 (3.0, 8.0) and 1.0 (0.0, 1.0) for IL-4, IL-2 and IL-4, respectively, at 14 days after the booster in the I-I-C group. While the cellular response measured in the I-I-C group was almost absent, the 30 and 60 μg LYB001 booster induced both Th1 (IFN-γ) and Th2 (IL-2 and IL-4) type cellular response, exhibiting over 10-fold increases versus the respective median SFCs at baseline.

Figure 5. RBD-specific IFN-γ, IL-2, or IL-4 secreting T-cells measured by ELISpot assay.

Bars and numbers in the figure indicate group medians, and error bars indicate interquartile range. The individual data in the I-I-30 L, I-I-60 L, I-I-C, and I-I-I-30 L groups are indicated by Δ, ▽, ○, ▲, respectively. RBD, receptor binding domain; I-I-30 L, 30 μg LYB001 booster after two-dose inactivated COVID-19 vaccine; I-I-60 L, 60 μg LYB001 booster after two-dose inactivated COVID-19 vaccine; I-I-C, CoronaVac booster after two-dose inactivated COVID-19 vaccine; I-I-I-30 L, 30 μg LYB001 booster after three-dose inactivated COVID-19 vaccine.

Discussion

In this study, we investigated the safety and immunogenicity of LYB001 as a heterologous booster following two or three doses of ICV in participants aged 18–59 years with a prime-boost interval of approximately 6–12 months. To the best of our knowledge, this is the first clinical trial conducted in China to report the preliminary safety, reactogenicity, and immunogenicity of a VLP-based vaccine against COVID-19. Overall, despite the higher incidence of reactogenicity events compared with the I-I-C group, AEs were mostly mild in severity and transient after a heterologous booster with LYB001 across the I-I-30 L, I-I-60 L, and I-I-I-30 L groups. The reported solicited adverse reactions represented reactogenicity events anticipated for intramuscularly administered vaccines, with local AEs including injection-site pain, swelling, redness, and pruritus, and systemic AEs including fatigue. In addition to mild reactogenicity, most AE symptoms resolve spontaneously, mostly within 48 h of onset. Those requiring treatment were managed using simple measures and widely available medications. No vaccination-related SAEs, AESIs, or AEs leading to participant withdrawal were reported within 90 days of booster administration. Although some participants reported abnormal safety laboratory measures with clinical significance 3 days after the booster, these abnormal values spontaneously returned to normal at the subsequent visit without any clinical consequences. In addition, these differences in laboratory results from baseline (calculated as values on day 3 after booster minus baseline values) for each participant in all groups did not show a particular trend. Furthermore, we found that the higher incidence rate of solicited local/systemic reaction for the heterologous LYB001 booster (in the I-I-30 L, I-I-60 L, and I-I-I-30 L groups) compared to the CoronaVac booster (in the I-I-C group) was predominantly due to injection-site pain. Injection-site pain accounted for 70.0% (n = 21), 56.7% (n = 17), 60.0 (n = 18), and 17.2% (n = 5) of the participants in the I-I-30 L, I-I-60 L, I-I-I-30 L, and I-I-C groups, respectively. The possible explanations for the increased incidence of injection-site pain are as follows: (1) the aluminum adjuvant content in LYB001 was higher than that of the inactivated vaccine (higher aluminum adjuvant was previously reported to correlate with a higher risk of pain),²² or (2) LYB001, with the display of repetitive RBD on the VLP vector, has a larger particle size, which may result in a longer local recruitment time of relevant molecules of the innate immune system and activation of antigen-presenting cells.²¹ Similarly, mRNA vaccines elicit transient increases in C-reactive protein (CRP) levels, an indicator of vaccine adjuvant activity.²³ Our results were consistent with a self-assembling, two-component nanoparticle vaccine (approved in South Korea) displaying the RBD of the SARS-CoV-2 spike glycoprotein in a highly immunogenic array. This vaccine demonstrated injection-site pain in 88.1% and 92.3% of participants receiving 10 μg GBP 510 and 25 μg GBP 510 in AS03 adjuvant, respectively.²⁴ In a phase III trial of another coronavirus-like particle (CoVLP) vaccine, injection site pain was reported in 85.0% of participants after the first dose of CoVLP with AS01 adjuvant versus 29.4% of participants in the placebo group.²⁵

The results of our study indicated that one heterologous booster dose of LYB001 could profoundly restore the NAb response, irrespective of the baseline antibody levels. LYB001 elicited a comparable NAb response (1237.8 vs. 1200.2) against the prototype 28 days after the booster in the I-I-30 L group versus the I-I-I-30 L group, despite a significantly different baseline (8.5 vs. 83.9). The heterologous LYB001 booster also induced superior humoral immune responses compared to the homologous booster with CoronaVac. When compared with a recombinant fusion protein vaccine V-01 (IFN-PADRE-RBD-Fc dimer, approved in China), the peak GMR (heterologous investigational vaccine booster versus homologous ICV booster) for VSV-based pseudovirus NAb GMT was 6.8 (1237.8 vs. 181.9) against prototype in the I-I-30 L group, higher than 3.3 GMR (893 vs. 268) after the 10 μg V-01 booster. The peak GMR (heterologous booster versus homologous ICV booster) was 6.9 (201.1 vs. 29.2) for the VSV-based pseudovirus NAb GMT against Omicron BA.4/5, higher than 3.8 GMR (211 vs. 56) against BA.1 after the V-01 booster.²⁶ Additionally, for a recombinant protein vaccine ZF2001 (approved in China, Colombia, Indonesia, Uzbekistan), the GMR (25 μg ZF2001 booster versus homologous ICV booster) was 2.4 (537 vs. 225) for the VSV-based pseudovirus NAb GMTs against prototype SARS-CoV-2, 1.7 (108 vs. 63) times higher than the ICV booster for the pseudovirus NAb GMTs against Omicron BA.1.²⁷ Taking into consideration a similar NAb detecting technique (VSV-based pseudovirus NAb assay), trial population (participants aged 18–59 years who completed a two-dose primary series of ICV), and comparison based on the active comparator of ICV, the 30 μg LYB001 booster can elicit a favorable NAb response against the circulating variant during the study period.

Although LYB001 was designed using the RBD from the prototype SARS-CoV-2, it demonstrated satisfactory immunogenicity against the prototype SARS-CoV-2 and robust cross-neutralizing activity against the Omicron BA.4/5 - SARS-CoV-2 variants that showed extensive immune escape. The conserved neutralizing epitopes between Omicron BA.4/5 and prototype SARS-CoV-2 may contribute to cross-neutralization. Immunogenicity could also be augmented by an innovative platform using RBD-VLP protein binding technology, which enhanced B cell activation, APC uptake and presentation, and efficient draining to lymph nodes. The optimal orientation for maximizing neutralizing epitope display, leading to a higher proportion of functional antibodies, was also reflected by a higher fold increase in the ratio of NAb titer versus spike glycoprotein-binding antibody concentration, comparing the LYB001 booster with the CoronaVac booster. Thus, this innovative design enables LYB001 to elicit robust NAb responses against the prototype SARS-CoV-2 and cross-neutralizing activity against circulating Omicron BA.4/5. Although a correlate of protection has not been established for predicting the individual-level risk of SARS-CoV-2 infection, a spike glycoprotein-binding IgG concentration of 1148 BAU/mL (also reported as BAUs in accordance with the WHO standard) may provide 75% protection against symptomatic infection with BA.5.²⁸ This indicates promising efficacy against BA.5 after LYB001 booster.

T-cell responses are also important for controlling disease development in patients with COVID-19. Targeted T-cell epitopes are broadly conserved between the prototype SARS-CoV-2 variant and Omicron.^29,30 Generally, the cellular immune response is absent or at least weak after booster administration in healthy adults receiving two-dose ICV.^31,32 The results from this trial indicated that LYB001 booster induced robust cellular responses to the SARS-CoV-2 RBD-specific peptide pool in the I-I-30 L, I-I-60 L, and I-I-I-30 L groups, as compared to the weak cellular responses in the I-I-C group. The RBD-specific IFN-γ secreting T-cells measured by ELISpot assay were dramatically increased (>10 times versus baseline) after a single LYB001 booster, comparable to one shot of adenovirus type-5-vectored COVID-19 vaccine (a median of approximately 10 IFN-γ secreting SFCs per 1 × 10⁵ PBMCs) which generally elicited robust cellular immune response.³³ Some participants in this study appeared to have preexisting cellular responses to the RBD-specific peptide pool used for PBMC restimulation. Such cross-reactive T-cell memory is possibly due to previous doses of ICV vaccination and/or previous exposure to common human coronaviruses.^34,35

This study has some limitations. The safety and immunogenicity findings were based on a small sample size (approximately 30 participants/group); therefore, the results should be interpreted with caution. This is an exploratory study without multiplicity adjustment to control type I error; therefore, the statistical power may be diminished. Additionally, this is an open-label study (participants receiving the CoronaVac booster must be registered with the vaccination system of China, and the fourth dose of the CoronaVac booster, which was used as an active comparator in part 1, was not approved when the study began), so the evidence provided in this trial is weaker than that in a blinded randomized control trial (RCT). To attain a more comprehensive and precise validation of the outcomes, minimize sample bias, and enhance statistical robustness, blinded RCTs of LYB001 booster with larger sample size and age stratification (including older participants aged ≥60 years) have now been conducted. Additionally, we only detected NAb titers of emerging Omicron sublineages during our study period in China, such as BA.5. Other emerging VOCs, such as XBB and its sublineages, will be included in future studies.

Conclusions

Taken together, the LYB001 as a heterologous booster, after inactivated COVID-19 vaccine, elicited a strong humoral immune response, demonstrated cross-neutralizing activity against dominantly circulating BA.4/5 variants, and induced robust RBD-specific cytokine secreting T-cell responses, without compromising vaccine’s safety. The reactogenicity events were predominantly mild (grade 1) in severity and transient and were anticipated for intramuscularly administered vaccines. Thus, our study provides valuable evidence that the LYB001 vaccine, developed using RBD-VLP protein-binding technology, is a promising candidate for preventing COVID-19.

Author contribution

All authors have read and approved this version of manuscript. Yan Liu, Xiaolan Yong, Jun Liu, Ying Zeng, Xuelian Cui and Yilin Wang conceived and designed the clinical study. Xiaolan Yong, Jun Liu and Jing Nie performed the clinical trial. Tao Wang performed the immunological (humoral and cellular) assays. Yiyong Chen, Wei Kang and Yan Liu drafted and revised the manuscript. Zhonghua Yang reviewed data analysis and interpretated the data. All authors agreed to be accountable for all aspects of the work.

Acknowledgments

The authors thank all the participants who volunteered to participate in this first-in-human study of LYB001.

Disclosure statement

Ying Zeng, Wei Kang, and Zhonghua Yang are employees of the Yantai Patronus Biotech Co., Ltd. The other authors declare no conflicts of interest.

Data availability statement

The data used in this study are available from the corresponding author upon reasonable request.

Supplementary data

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/21645515.2023.2267869.

Additional information

Funding

This study was partially supported by a grant from the Key Program of the Yantai Science & Technology Bureau [No. 2022CXG010508]. This study was funded by the Patronus Biotech Co. Ltd.