Fractional versus standard BNT162b2 boosters after non-mRNA priming in Mongolia: 24-month immunogenicity and safety evidence from a randomised controlled trial
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Booster doses of COVID-19 vaccines help restore protection against waning immunity and emerging variants. Having found that fractional BNT162b2 boosters were non-inferior to standard boosters in eliciting anti-spike IgG responses at day 28 in Mongolian adults, we assessed long-term immunogenicity and safety in the same cohort.
Methods
In this randomised, controlled, non-inferiority trial, adults previously vaccinated with two doses of ChAdOx1-S, BBIBP-CorV, or Gam-COVID-Vac were randomly assigned (1:1) to receive a 15μg (fractional) or 30μg (standard) BNT162b2 booster. IgG concentrations and functional surrogate virus neutralisation test (sVNT) levels against Wuhan-Hu-1 and Omicron BA.1 were assessed over 24 months. A subset had IFN-γ cell-mediated-immunity (CMI) measured (Ag1/Ag2). Immune response GMRs (fractional:standard) were estimated from log-transformed values via multivariable linear regression. SARS-CoV-2 infections, adverse events, and serious adverse events (SAEs) were recorded. ClinicalTrials.gov Identifier: NCT05265065 .
Findings
Of 601 participants enrolled between May 27 and Sept 30, 2022, 520 (86.5%) completed 24-month follow-up. Although IgG levels declined from six to 24 months, relative responses remained similar between arms at 18 months (GMR 1·08 [95% CI 0·97-1·22]) and 24 months (GMR 1·06 [95% CI 0·95-1·18]). In the CMI subset, IFN-γ responses peaked at day 28, waned to 18 months, and returned to baseline by 24 months, with fractional and standard arms similar at 24-month GMRs (Ag1 GMR 1·17 [95% CI 0·82–1·66]; Ag2 GMR 1·06 [95% CI 0·73–1·54]). Median sVNT inhibition against both Wuhan-Hu-1 and Omicron BA.1 was high and comparable between groups at 18 and 24 months (both 88% [95% CI 86-90]). SARS-CoV-2 infection was confirmed in 28 participants, with an additional 386 suspected infections after day 28, inferred from a > 1·2-fold rise in IgG titres between study visits. No SARS-Cov-2 infections resulted in hospitalisation. Fifty-three SAEs were reported, evenly distributed between groups, with no vaccine-related events.
Interpretation
Fractional and standard BNT162b2 boosters showed comparable neutralising activity (sVNT) that persisted to 24 months, while binding IgG declined to baseline by 24 months. In the CMI subset, IFN-γ responses followed a similar trajectory, indicating alignment of humoral and cellular immunity over time. Despite widespread SARS-CoV-2 circulation, no COVID-19 hospitalisations or deaths occurred, and safety was reassuring. These data support fractional dosing as a pragmatic, cost-saving option.
Funding
Coalition for Epidemic Preparedness Innovations (CEPI). This study was supported by the Victorian Government’s Operational Infrastructure Support Programme.
Panel: Research in Context
Evidence before this study
We searched PubMed up to Aug 27, 2025, for English-language studies in adults assessing immune responses to fractional or reduced-dose COVID-19 boosters, focusing on mRNA vaccines and long-term follow-up. Search terms combined “fractional dose”, “low dose”, “reduced dose”, “booster”, “immunogenicity”, “antibodies”, “durability”, “persistence”, and related terms. Evidence is limited and largely confined to short- and medium-term follow-up. A trial in Thailand found half-dose AZD1222 after CoronaVac was non-inferior to a full dose at 14 and 90 days, with lower reactogenicity. 1 A subsequent Thai study showed half-dose BNT162b2 or AZD1222 after CoronaVac achieved non-inferior immunogenicity at 28 and 90 days, particularly with longer intervals. 2 In the UK, COV-BOOST reported similar anti-spike IgG responses with half- and full-dose BNT162b2 at three months, 3 and by eight months, decay patterns varied by platform, but half-dose BNT162b2 tracked closely with full doses. 4 In Belgium, the IMCOVAS trial showed low-dose mRNA-1273 and heterologous schedules were non-inferior to reference regimens up to one year, though intradermal vaccination was less effective. 5 In Brazil, the FRACT-COV trial followed 1451 adults for six months: full-dose BNT162b2 elicited the highest titres, but fractional BNT162b2 outperformed full-dose AZD1222 and Sinovac. 6 Across studies, fractional mRNA boosters generally induced robust short-term responses and, in some cases, non-inferior durability up to 12 months. However, no trial had reported a 24-month follow-up of fractional versus standard intramuscular BNT162b2 boosters.
Added value of this study
This randomised trial is the first to provide 24-month data comparing fractional (15 µg) and standard (30 µg) BNT162b2 boosters in adults primed with non-mRNA vaccines (ChAdOx1-S, BBIBP-CorV, Gam-COVID-Vac). Fractional dosing produced humoral and cellular responses equivalent to standard dosing: anti-spike IgG declined to baseline by 24 months, but neutralising activity and IFN-γ responses were preserved, with geometric mean ratios consistently close to unity. Age-stratified analyses showed higher early IgG responses in older adults that converged with younger participants by 18–24 months, indicating similar long-term durability across age groups. With 87% retention, detailed serological follow-up, and prespecified sensitivity analyses confirming that missingness did not bias results, this trial provides the first randomised long-term evidence from an LMIC setting. Our findings extend shorter-term results from Brazil, Thailand, and the UK by showing persistence of humoral and cellular immunity for two years.
Implications of all the available evidence
Fractional BNT162b2 boosters provide durable humoral and cellular immunity comparable to standard dosing, now shown up to 24 months. This supports dose-sparing as a feasible, cost-saving strategy to extend vaccine supply and improve equity, particularly in LMICs reliant on inactivated or adenoviral vector priming. While continued surveillance of emerging variants is needed, the available evidence indicates that fractional dosing can be used without compromising long-term protection.
