The central safety-evaluation problem of vaccine adjuvants is structural: the same innate-immune mechanisms that produce adjuvanticity (an adjuvant's ability to strengthen the protective immune response) produce reactogenicity, and the field lacks validated, reproducible criteria for separating the local and systemic inflammation that is the expected consequence of immune activation from reactions that are toxicologically relevant. The open question, raised by Erik B. Lindblad's 2007 review 'Safety Evaluation of Vaccine Adjuvants', is how to establish a rigorous yet practical, broadly accepted international framework for adjuvant safety testing across human and veterinary applications. Literature from 2008-2026 documents real but partial progress - in vitro cytokine-release assays with provisional safety thresholds, genomic-biomarker screening methods, tiered industry evaluation programs for licensed combination adjuvants (Adjuvant Systems), large integrated post-licensure safety databases, and curated adjuvant knowledge bases - while explicitly reaffirming that safety findings cannot be extrapolated across species, formulations, or populations, that regulatory approaches still vary between authorities, and that no validated decision algorithm or shared evaluation-grade safety database exists. The problem remains open.

Background

Most modern vaccines are built from purified recombinant or synthetic antigens (subunit vaccines). Purification strips away the microbial molecules that make whole pathogens visible to the innate immune system - the fast, non-specific arm of immunity whose activation, through pattern-recognition receptors such as Toll-like receptors (TLRs) and inflammatory cytokine release, is the prerequisite for a strong adaptive antibody and T-cell response. An adjuvant is a substance formulated with the antigen to restore that immunogenicity. By 2015, five adjuvant classes were licensed in human vaccines - aluminum salts (alum), the oil-in-water emulsions MF59 and AS03, the immunostimulant-on-alum combination AS04 (monophosphoryl lipid A, a detoxified bacterial lipid that activates innate immunity through Toll-like receptor 4, adsorbed onto aluminum salts), and virosomes (liposome-like particles assembled from viral envelope components) (Rambe et al. 2015; Di Pasquale et al. 2015) - and dozens of experimental classes (TLR agonists, saponins such as QS-21 (plant-derived immunostimulatory molecules), emulsions, particulates) were in development. Since then, further classes have entered human-vaccine use, among them the CpG 1018 oligonucleotide (a short synthetic DNA motif that activates Toll-like receptor 9, used in a licensed hepatitis B vaccine) and the saponin-based Matrix-M (used in authorized COVID-19 vaccines), widening the same safety-evaluation problem.

The safety problem is not incidental but mechanistic. The processes that make adjuvants work - local innate immune activation, recruitment of inflammatory cells to the injection site, cytokine release - are the same processes that produce reactogenicity: the expected local reactions (pain, redness, swelling, induration, and in stronger cases granulomas, i.e. persistent nodules of immune cells and fibrous tissue at the injection site) and systemic reactions (fever, malaise) that follow vaccination. Immunotoxicity denotes the toxicologically relevant end of this spectrum: persistent lesions, tissue destruction, immune hyperstimulation, hypersensitivity, or autoimmunity. Because vaccines are given predominantly to healthy people - infants in pediatric programs above all - the acceptable margin between immune activation and harm is far narrower than for therapeutic drugs, and only minor expected reactions are tolerable.

Lindblad's 2007 review 'Safety Evaluation of Vaccine Adjuvants' systematized this tension and framed the open evaluation problem. Preclinical safety testing deliberately uses overdosing - high, repeated doses - to force rare adverse events to appear in small animal cohorts, yet dose escalation can itself convert normal local inflammation into clinically significant toxicity, so overdose findings blur exactly the distinction, expected inflammation versus toxic reactogenicity, that safety decisions depend on. The chapter documents that no reproducible biomarker threshold, defined time window, or validated decision algorithm existed for drawing that boundary; that findings in one animal species, strain, dose regimen, or antigen-adjuvant combination could not be assumed to extrapolate to others, including humans, infants, or therapeutic-vaccine recipients; that persistent local lesions and cytokine measurements had no agreed clinical safety correlates; and that the historical safety experience of licensed adjuvants was not organized into a shared evidence base that could guide new candidates. Regulatory guidance from European and WHO authorities existed, but a 2015 industry review of that landscape (Tavares Da Silva et al. 2015) records that approaches varied between authorities and remained under active discussion.

The open question is therefore not whether the licensed adjuvants are acceptably safe - accumulating post-licensure surveillance supports favorable benefit-risk profiles for the licensed classes - but whether the field can replace fragmented, product-by-product testing with an evaluation framework that is simultaneously rigorous (validated against real outcomes) and practical (feasible in animal studies and early clinical trials), and that is accepted internationally for both human and veterinary adjuvanted vaccines. This matters because potent adjuvant classes remain shelved principally over reactogenicity concerns: Zaitseva et al. (2012) note that of the many adjuvant formulations under development, very few are licensed mainly due to concerns about adverse side effects, and Sasaki et al. (2018) cite poly(I:C) - a synthetic double-stranded RNA that mimics viral genetic material and signals through Toll-like receptor 3 - as a highly effective adjuvant not clinically applied because of its human toxicity risk. The published record already contains evidence bearing on the framework's separation criteria and extrapolation claims.

Problem Statement

The problem is to develop and validate a safety-evaluation framework for vaccine adjuvants that is rigorous enough to support regulatory decisions and practical enough for routine preclinical and early clinical use, across human (pediatric, adult, and therapeutic) and veterinary applications. Such a framework must, jointly and in mutually supporting fashion: (1) provide reproducible criteria - validated biomarkers or assay endpoints with defined thresholds and time windows - that separate the local and systemic inflammation expected from successful immune activation from reactogenicity that is toxicologically relevant, including under the deliberate overdosing regimens used to reveal rare adverse events; (2) establish, with validation evidence, the extent to which adjuvant safety findings in one animal species, strain, dose regimen, or antigen-adjuvant combination predict outcomes in others, including extrapolation to humans, infants, and therapeutic vaccination settings; (3) anchor the interpretation of persistent local lesions and cytokine readouts to clinically meaningful long-term safety outcomes; and (4) organize the historical safety experience of licensed adjuvants into a shared, curated evidence base capable of guiding the evaluation of new adjuvants and delivery systems. All four requirements are mandatory; satisfying only a subset of them does not resolve the problem. The overarching objective is a framework meeting these requirements well enough to be adoptable as a broadly accepted international standard for adjuvant safety testing. The framework's separation criteria and extrapolation claims must themselves be validated against real outcomes rather than assumed; a claim of international consensus without such validation evidence does not by itself establish the framework. Where the published record cannot yet support a requirement, that requirement remains open and unresolved.

References

  1. Lindblad EB. Safety Evaluation of Vaccine Adjuvants. In: Singh M, ed. Vaccine Adjuvants and Delivery Systems. Wiley; 2007:421-444. DOI: 10.1002/9780470134931.ch18. https://doi.org/10.1002/9780470134931.ch18
  2. Ahmed SS, Plotkin SA, Black S, Coffman RL. Assessing the Safety of Adjuvanted Vaccines. Science Translational Medicine. 2011;3(93):93rv2. DOI: 10.1126/scitranslmed.3002302. https://doi.org/10.1126/scitranslmed.3002302
  3. Garçon N, Segal L, Tavares F, Van Mechelen M. The safety evaluation of adjuvants during vaccine development: the AS04 experience. Vaccine. 2011;29(27):4453-4459. DOI: 10.1016/j.vaccine.2011.04.046. https://doi.org/10.1016/j.vaccine.2011.04.046
  4. Batista-Duharte A, Lindblad EB, Oviedo-Orta E. Progress in understanding adjuvant immunotoxicity mechanisms. Toxicology Letters. 2011;203(2):97-105. DOI: 10.1016/j.toxlet.2011.03.001. https://doi.org/10.1016/j.toxlet.2011.03.001
  5. Zaitseva M, Romantseva T, Blinova K, Beren J, Sirota L, Drane D, Golding H. Use of human MonoMac6 cells for development of in vitro assay predictive of adjuvant safety in vivo. Vaccine. 2012;30(32):4859-4865. DOI: 10.1016/j.vaccine.2012.05.002. https://doi.org/10.1016/j.vaccine.2012.05.002
  6. Sayers S, Ulysse G, Xiang Z, He Y. Vaxjo: a web-based vaccine adjuvant database and its application for analysis of vaccine adjuvants and their uses in vaccine development. Journal of Biomedicine and Biotechnology. 2012;2012:831486. DOI: 10.1155/2012/831486. https://doi.org/10.1155/2012/831486
  7. Petrovsky N. Vaccine adjuvant safety: the elephant in the room. Expert Review of Vaccines. 2013;12(7):715-717. DOI: 10.1586/14760584.2013.811198. https://doi.org/10.1586/14760584.2013.811198
  8. Tavares Da Silva F, Di Pasquale A, Yarzabal JP, Garçon N. Safety assessment of adjuvanted vaccines: Methodological considerations. Human Vaccines & Immunotherapeutics. 2015;11(7):1814-1824. DOI: 10.1080/21645515.2015.1043501. https://doi.org/10.1080/21645515.2015.1043501
  9. Di Pasquale A, Preiss S, Tavares Da Silva F, Garçon N. Vaccine Adjuvants: from 1920 to 2015 and Beyond. Vaccines. 2015;3(2):320-343. DOI: 10.3390/vaccines3020320. https://doi.org/10.3390/vaccines3020320
  10. Rambe DS, Del Giudice G, Rossi S, Sanicas M. Safety and Mechanism of Action of Licensed Vaccine Adjuvants. International Current Pharmaceutical Journal. 2015;4(8):420-431. DOI: 10.3329/icpj.v4i8.24024. https://doi.org/10.3329/icpj.v4i8.24024
  11. Batista-Duharte A, Martínez DT, Carlos IZ. Efficacy and safety of immunological adjuvants. Where is the cut-off? Biomedicine & Pharmacotherapy. 2018;105:616-624. DOI: 10.1016/j.biopha.2018.06.026. https://doi.org/10.1016/j.biopha.2018.06.026
  12. Sasaki E, Momose H, Hiradate Y, Mizukami T, Hamaguchi I. Establishment of a novel safety assessment method for vaccine adjuvant development. Vaccine. 2018;36(46):7112-7118. DOI: 10.1016/j.vaccine.2018.10.009. https://doi.org/10.1016/j.vaccine.2018.10.009
  13. Laupèze B, Hervé C, Di Pasquale A, Tavares Da Silva F. Adjuvant Systems for vaccines: 13 years of post-licensure experience in diverse populations have progressed the way adjuvanted vaccine safety is investigated and understood. Vaccine. 2019;37(38):5670-5680. DOI: 10.1016/j.vaccine.2019.07.098. https://doi.org/10.1016/j.vaccine.2019.07.098
  14. Verstraeten T, Descamps D, David MP, Zahaf T, Hardt K, Izurieta P, Dubin G, Breuer T. Analysis of adverse events of potential autoimmune aetiology in a large integrated safety database of AS04 adjuvanted vaccines. Vaccine. 2008;26(51):6630-6638. DOI: 10.1016/j.vaccine.2008.09.049. https://doi.org/10.1016/j.vaccine.2008.09.049
  15. Canbolat F. Vaccine Adjuvants in Immunotoxicology. In: Synthetic Peptide Vaccine Models. CRC Press; 2021:202-237. DOI: 10.1201/9781003144533-14. https://doi.org/10.1201/9781003144533-14
  16. Bhattacharya S, Singh J, Gautam RK, Farooqui N, Manocha N, Malhotra H. Adjuvant and Vaccine Safety. In: Emerging Pathways of Vaccine Adjuvants. Wiley; 2025:235-253. DOI: 10.1002/9781394237654.ch9. https://doi.org/10.1002/9781394237654.ch9
  17. Monickaraj J, Rehana H, Bernardi A, Zheng Y, Lin T, Liu L, Madireddi A, Yeh L, Zheng J, Hur J, He Y. Vaxjo 2.0: An ontology- and large language model-powered knowledge base of vaccine adjuvants and mechanisms. Frontiers in Cellular and Infection Microbiology. 2026;16:1763384. DOI: 10.3389/fcimb.2026.1763384. https://doi.org/10.3389/fcimb.2026.1763384
  18. Yoshioka Y, Nishinaka-Yoshioka A, Kobiyama K, Hayashi T, Kidani Y, Yanagida Y, Kasahara J, Tsujii K, Asaoka Y, Kuroda N, Kugimiya A, Osawa H, Yoshimura A, Onishi M, Nakagawa T, Ishida S, Omoto S, Nagira M, Coban C, Ishii KJ. IL-1 delineates squalene-based adjuvant efficacy and reactogenicity in a cell-type-specific manner. npj Vaccines. 2026;11(1):67. DOI: 10.1038/s41541-026-01420-0. https://doi.org/10.1038/s41541-026-01420-0