Immunogenicity and efficacy of vaccine boosters against SARS-CoV-2 Omicron subvariant BA.5 in male Syrian hamsters

Omicron subvariant BA.5, emerging in early 2022, carries additional spike mutations (e.g., 69-70del, L452R, F486V, Q493) relative to BA.2 and is associated with increased reinfections, breakthroughs, and hospitalizations. Current vaccines, based on adenoviral vectors, mRNA, and protein subunits, have reduced severe COVID-19 outcomes but face challenges from immune-evasive variants. To inform booster strategies, this study investigates the durability of humoral responses after primary vaccination with Ad26.COV2.S or BNT162b2 in male Syrian hamsters and evaluates homologous and heterologous boosters (mRNA-1273 or NVX-CoV2373) for neutralizing antibody responses and protection against BA.5 challenge over a 9-month period.

The paper contextualizes the emergence of VOCs (Alpha, Delta, Omicron) with spike mutations compromising neutralization, prompting WHO-recommended boosters. It reviews vaccine platforms in use (Ad26.COV2.S, AZD1222, BNT162b2, mRNA-1273, NVX-CoV2373) and prior observations that mRNA vaccine-induced antibodies wane by 6–8 months, whereas Ad26.COV2.S responses are more stable over time. Prior preclinical and clinical studies demonstrated NVX-CoV2373 efficacy and enhanced neutralization against Omicron sublineages after boosting, and improved protection in animals with variant-adapted Ad26 boosters. However, comprehensive long-term kinetics and BA.5 challenge data in hamsters across prime/boost combinations were limited.

Design: Four-to-five-week-old male Syrian (golden) hamsters (HsdHan: AURA) were assigned to primary vaccination with either a single intramuscular dose of Ad26.COV2.S at 1×10^10 or 1×10^8 viral particles (100 µL) or two intramuscular doses of BNT162b2 (5 µg each, 3 weeks apart; 100 µL/dose). Sham-vaccinated controls received PBS. Animals were followed up to 252 days post-vaccination (dpv). Sampling and assays: Blood was collected every 28 days for immunogenicity. Neutralizing antibodies against USA-WA1/2020 were measured by PRNT50; binding anti-S1 IgG by ELISA (OD490 and AUC). At day 168 (6 months), all animals were boosted: homologous (Ad26.COV2.S 1×10^8 vp for Ad26-primed; BNT162b2 0.5 µg for BNT-primed) or heterologous (mRNA-1273 0.5 µg; NVX-CoV2373 1 µg recombinant spike with 15 µg Matrix-M). Sham boosts (PBS) were included. Neutralization against Omicron BA.5 was assessed by FRNT50 at pre-boost (day 168), 1 month post-boost (day 196), and 3 months post-boost (day 252). Challenge: Three months post-boost, hamsters were intranasally challenged with Omicron BA.5; mock controls received PBS. Infectious virus titers were measured at 2 and 4 days post-infection (dpi) in nasal washes (PFU/mL), trachea (PFU/g), and lungs (PFU/g) by plaque assay on VeroE6-TMPRSS2 cells. Body weight and clinical signs were recorded daily to 4 dpi. Histopathology: Left lungs at 4 dpi were fixed, paraffin-embedded, sectioned, and H&E-stained. Blinded pathology scoring (0–4) assessed interstitial pneumonia, perivascular/peribronchiolar inflammation, hemorrhage, edema, and other lesions. Laboratory details: Virus stocks (USA-WA1/2020 and Omicron BA.5.5) were passaged once in VeroE6-TMPRSS2 and deep-sequenced (no spike mutations >5% frequency). PRNT and FRNT protocols used standard heat-inactivated sera, serial dilutions, and defined LOD of 1:20. ELISA used recombinant S1 antigen; positivity was defined as OD490 above negative control + 3 SD in ≥2 dilutions. Statistical analyses employed ANOVA with appropriate post-tests, Mann–Whitney, Wilcoxon, and Spearman correlations; p<0.05 considered significant. All procedures followed UTMB IACUC and BSL-3 approvals.

- Antibody kinetics: Ad26.COV2.S induced higher and more durable neutralizing and binding antibody responses than two-dose BNT162b2. Neutralizing GMTs peaked at day 56 for Ad26 1×10^10 vp (GMT 368) and 1×10^8 vp (GMT 139), and at day 28 after BNT162b2 series (GMT 65). At 252 dpv, Ad26 1×10^10 vp maintained higher titers than Ad26 1×10^8 vp (binding 2.6-fold, p<0.0001; neutralizing 1.8-fold, p=0.0005) and than BNT162b2 (binding 14.0-fold; neutralizing 9.3-fold; both p<0.0001). - Booster immunogenicity against BA.5: Heterologous NVX-CoV2373 boosters produced the largest increases in BA.5 FRNT50 at 1 month post-boost: 2.6-fold (Ad26 1×10^10 vp prime; n=22, GMT 80, p=0.006), 3.8-fold (Ad26 1×10^8 vp prime; n=22, GMT 50, p=0.013), and 2.5-fold (BNT162b2 prime; n=22, GMT 32, p=0.024). Titers remained stable to 3 months post-boost. Responder rates (FRNT50≥20) at 3 months post-boost were higher with NVX-CoV2373: in Ad26-primed cohorts, measurable nAb responses occurred in 75% with NVX-CoV2373 vs 50% with mRNA-1273 and 58% with homologous Ad26 boosting; in BNT162b2-primed cohorts, 67% with NVX-CoV2373 vs 33% with mRNA-1273 and 25% with homologous BNT162b2. - Clinical outcomes: Vaccinated groups exhibited reduced weight loss vs sham after BA.5 challenge. In Ad26 1×10^10 vp-primed animals, boosters (mRNA-1273, NVX-CoV2373, or homologous Ad26) reduced weight loss at 3 dpi (p=0.016, 0.006, 0.002) and 4 dpi (p=0.006, 0.002, 0.008). Similar effects were seen for Ad26 1×10^8 vp primes with heterologous boosters; homologous boost showed limited effect at 3 dpi. - Virologic protection: NVX-CoV2373 boosters yielded the greatest reductions in infectious virus. For Ad26 1×10^10 vp prime + NVX-CoV2373, reductions vs sham were 9-fold (nasal), 3-fold (trachea), 77-fold (lungs) at 2 dpi, and 32-fold (nasal), 86,000-fold (trachea), 690,000-fold (lungs) at 4 dpi. mRNA-1273 and homologous Ad26 boosters also reduced viral loads, with mRNA-1273 showing greater late reductions in lower respiratory tract at 4 dpi. Notably, unboosted Ad26-primed animals still showed ~10-fold reductions across samples, similar to homologous boost. - In BNT162b2-primed animals, both homologous BNT162b2 and heterologous mRNA-1273 boosters reduced viral loads at 4 dpi; at 2 dpi, mRNA-1273 conferred greater reductions (4-, 69-, and 20-fold in nasal, trachea, lungs) than homologous boosting (3-, 25-, 7-fold). - Histopathology: Sham-vaccinated, BA.5-challenged lungs had moderate to severe interstitial pneumonia and airway/vascular inflammation. Ad26-primed animals (with or without boosts) had only mild lesions. BNT162b2-primed animals boosted with mRNA-1273 or NVX-CoV2373 showed reduced lesion severity; sham-boosted or homologous-boosted BNT162b2 groups resembled sham pathology. - Correlates: Pre-challenge BA.5 FRNT50 inversely correlated with 4 dpi viral loads, strongest in lungs: Spearman r = -0.86 (Ad26 1×10^10 vp), -0.80 (Ad26 1×10^8 vp), -0.77 (BNT162b2); all p<0.0001. FRNT50≥20 predicted absence of lung infection at 4 dpi with NPV 91% (95% CI 80–96%) and PPV 98% (95% CI 87–100%). NVX-CoV2373 groups had the lowest breakthrough proportions.

This study addressed whether homologous or heterologous boosters enhance protection against the immune-evasive Omicron BA.5 after different primary vaccine platforms. Results show that single-dose Ad26.COV2.S, at both high and low doses, generates more durable humoral immunity than two-dose BNT162b2 over nine months. Critically, heterologous boosting, especially with NVX-CoV2373, substantially augments BA.5 neutralization, reduces viral replication—most notably in the lower respiratory tract—and mitigates disease indicators such as weight loss and lung pathology. The strong inverse correlation between BA.5 neutralizing titers and lung viral loads underscores nAb titers as a useful correlate of protection in this model. These findings support flexible heterologous booster strategies, highlight the durability advantages of Ad26-based priming (valuable where cold-chain and repeat visits are challenging), and suggest that protein-adjuvanted boosters can meaningfully improve protection against Omicron subvariants. Nonetheless, breakthrough infections persisted, indicating vaccination alone may not fully block infection or transmission against highly immune-evasive variants.

Across 9 months in male Syrian hamsters, Ad26.COV2.S priming yielded more durable antibody responses than BNT162b2, and heterologous NVX-CoV2373 boosting provided the strongest enhancement of BA.5 neutralization and protection, with large reductions in viral loads and improved clinical and pathological outcomes. Neutralizing antibody titers correlated with protection, particularly in the lungs. These data inform booster choices and support heterologous strategies to counter Omicron subvariants. Future research should evaluate bivalent/original-Omicron-adapted boosters and updated Ad26 constructs, include both sexes and additional immune readouts (binding breadth, T-cell responses), assess dose optimization across platforms, and extend challenges to newer BA.5 descendants (e.g., BQ.1/BQ.1.1).

- Omicron variants, including BA.4/BA.5, are less pathogenic in hamsters than earlier strains, potentially limiting disease readouts and cross-strain comparisons. - Only male hamsters were used to ensure power across cohorts; sex differences were not assessed. - Immunogenicity assessments focused on neutralizing antibodies; cross-reactive binding antibodies and T-cell responses were not measured and may contribute to protection. - Only one concentration of the NVX-CoV2373 booster and a relatively low mRNA-1273 dose were tested, possibly affecting comparative efficacy; cross-platform dose equivalence is imperfect in rodent models. - Due to cohort sizes and timepoints, only BA.5 was used for challenge; findings may not generalize to all Omicron sublineages. - Some methodological constraints inherent to high-dose rodent vaccinations and schedule choices may affect translatability.