Toxicologic Pathology Forum: Tissue Evaluation in Nonclinical Toxicity Studies for Prophylactic Vaccines

Abstract

Nonclinical toxicity testing (GLP) of prophylactic vaccines to support human clinical trials is outlined in the World Health Organization nonclinical vaccine-development guidelines, which are followed by most regulatory agencies globally. Vaccine GLP toxicity studies include at least two groups: a buffer control (often phosphate-buffered saline) group and a highest anticipated clinical dose formulation group. However, studies may include additional groups, including lower-dose formulation groups and adjuvant-containing formulation control groups. World Health Organization guidelines touch upon expectations for dose group and tissue selection for microscopic evaluation, but there is variation in the interpretation of this aspect of these guidelines between vaccine developers. This opinion piece proposes a scientifically based approach for defining appropriate groups to evaluate in the dosing and recovery phases in nonclinical vaccine toxicity studies, as well as suggestions on selecting tissues for microscopic evaluation at the recovery phase of studies to promote alignment between vaccine manufacturers.

This is an opinion article submitted to the Toxicologic Pathology Forum. It represents the views of the author. It does not constitute an official position of the Society of Toxicologic Pathology, British Society of Toxicological Pathology, or European Society of Toxicologic Pathology, and the views expressed might not reflect the best practices recommended by these societies. This article should not be construed to represent the policies, positions, or opinions of their respective organizations, employers, or regulatory agencies.

Background

Vaccines against infectious diseases, termed prophylactic vaccines (hereafter referred to as vaccines), have had a tremendous impact on world health, eliminating or significantly reducing diseases or disease severity and preventing death, most recently with COVID-19. While we think of vaccines as a relatively recent phenomenon, exposure-based protection has been practiced for centuries. It was not until 1986, however, that the first vaccine manufactured using recombinant DNA technology, Recombivax, was approved for human use to protect against the hepatitis B virus. Prior to that, viruses and proteins used in vaccines were produced in mammalian or bacterial cells and isolated either from those cells or their supernatants to generate the vaccine antigenic components (i.e., proteins, polysaccharides, live or inactivated viruses). Often, these vaccines were effective without the addition of adjuvants such as aluminum because of the assortment of immune activators (pathogen-associated molecular patterns and damage-associated molecular patterns) that were retained in the antigen extract (e.g., flagella, microbial DNA/RNA, lipopolysaccharide [LPS], lipoproteins). ¹ Advances in molecular and protein technologies, and good laboratory practices (GLP) and good manufacturing practices (GMP), have resulted in the generation of vaccines that do not have these residual materials and are safer and more consistently manufactured than at any time in history. ²

Governance around prophylactic vaccine nonclinical testing was relatively limited until the late 20th and early 21st century; in 1997, the European Medicines Agency (EMA) published Note for Guidance on Pharmacological and Toxicological Testing of Vaccines, ³ which was followed by the publication of the World Health Organization (WHO) guidelines on nonclinical studies to support vaccine clinical trials in 2005. ⁴ The 2005 WHO guidelines were followed in 2014 with the WHO publication Guidelines on the Non-Clinical Evaluation of Vaccine Adjuvants and Adjuvanted Vaccines. ⁵ In addition, there is a 2005 EMA Guideline on Adjuvants in Vaccine for Human Use ⁶ ; although this guideline is still active, deference to the 2014 WHO guideline seems to be the norm. Prior to that, nonclinical toxicity studies supporting vaccine clinical trials either did not occur or were limited and usually included limited or no microscopic evaluation of tissues. Many of the historical vaccines, such as the measles, mumps, and rubella vaccines; pertussis vaccines; and tetanus vaccines, did not undergo rigorous nonclinical toxicity evaluation by today’s standards prior to testing in human clinical trials (e.g., in-life and immunogenicity nonclinical studies only). While this may seem extraordinary, prophylactic vaccine active components (modified live or inactivated viruses, pathogen-specific proteins/polysaccharides) are foreign antigens and are designed to not engage endogenous targets. ⁷ Because vaccines function through activation of the immune system against a foreign antigen, concerns around direct toxicity and off-target adaptive immune responses are limited, although molecular mimicry could result in immune responses against self-antigens. For this reason, the assortment of required regulatory studies to support clinical trials for vaccines is limited compared to small molecules. Most countries defer to the WHO vaccine guidelines, but for historical vaccine-related Food and Drug Administration regulatory guidance, see https://www.fda.gov/vaccines-blood-biologics/biologics-guidances/vaccine-and-related-biological-product-guidances.

Although the antigen is considered the active component of the vaccine, it may be formulated with other components (adjuvants or stabilizers) that may enhance the immune response to the antigen(s). The highest-dose-containing vaccine formulations intended for clinical administration must be evaluated in nonclinical repeat-dose toxicity studies. Unlike small-molecule studies, only one dose level of the vaccine may be tested, as immune responses to antigens are not anticipated to follow a dose-response relationship. ⁷ A buffer control group (often phosphate-buffered saline) is expected to be included in the study; formulation controls (particularly those containing adjuvants) are recommended to “aid in the interpretation of safety data from the adjuvanted vaccine.” ⁵ If an adjuvant is well characterized (e.g., Al[OH]₃), inclusion in nonclinical repeat-dose toxicity studies may not be necessary. Therefore, nonclinical toxicity studies may include as few as 2 groups or as many as deemed necessary.

Microscopic Assessments in Vaccine Studies

World Health Organization recommendations nonclinical vaccine repeat-dose toxicity studies are intentionally nonproscriptive. On microscopic tissue evaluation, they suggest either evaluation of all collected tissues (as defined by the WHO 2014 guideline in Appendix 2) ⁵ at the end of dosing (including target tissues) or simply pivotal organs (e.g., brain, lung, heart, kidneys, liver, reproductive organs) and anticipated target tissues (e.g., injection sites, draining lymph nodes, spleen/lymphoid organs). Other than suggesting microscopic evaluation of all tissues from vaccine formulations with novel adjuvants, it is unclear when one approach should be selected over the other (i.e., full tissue list vs selected tissues). Additionally, other than the buffer control group and the group administered the highest-anticipated vaccine formulation, the microscopic evaluation of tissues from other treatment groups is generally not specified. Finally, the WHO guidance documents do not define whether expectations for microscopic evaluation are the same at the end of recovery as they are at the end of dosing.

Because of this regulatory ambiguity, there is wide variation between vaccine manufacturers regarding which groups and which tissues are microscopically evaluated at the end of the dosing and recovery phases. Very conservative interpretation of the guidelines results in the microscopic evaluation of all tissues in all groups (including formulation controls) from both dosing and recovery phase animals. Less conservative interpretation may result in the microscopic evaluation of only selected dose groups and, particularly at the end of the recovery phase, limited tissues.

In the author’s experience and from personal communications, microscopic findings outside of target tissues (i.e., injection site/regional tissues, draining lymph node, and other lymphoid tissues) are rarely identified in prophylactic vaccine nonclinical toxicity studies at the end of either the dosing or recovery phase. One report of lacrimal gland degeneration in rabbits was reported, related to administration of a specific adjuvant ⁸ (hence its inclusion in the tissue list in Appendix 2 of the 2014 WHO guideline). ⁵ It is, therefore, reasonable to evaluate all tissues only from the buffer control group and the vaccine formulation reflecting the highest anticipated human clinical dose end of the dosing phase. However, microscopic evaluation of formulation controls or lower vaccine dose groups at the end of dosing is likely not needed unless there were unexpected findings or findings of concern in the vaccine group such that evaluation of other groups would elucidate the cause (e.g., adjuvant, antigen, or a combination of both). When vaccine formulations contain adjuvants not in any licensed vaccines or with components not previously evaluated in vaccine repeat-dose GLP toxicity studies, evaluation of formulation controls may be warranted or expected by regulators. However, specific understanding of novel adjuvants should be assessed outside of nonclinical GLP toxicity studies, as appropriate, to understand immune effects and absorption, distribution, metabolism, and excretion (ADME).

The WHO guidelines^4,5 indicate that the rationale for evaluating recovery tissues microscopically is “to investigate the reversibility of any adverse effects observed during the treatment period and to screen for possible delayed adverse effects.” ⁴ Specifications on microscopic tissue evaluation at the end of recovery are difficult to discern from the guidelines, and there is some variation in the extent of tissues evaluated at the end of this phase between vaccine manufacturers (personal communication).

While nonclinical repeat-dose toxicity study findings are consistent with clinical reactogenicity (i.e., manifestations of innate immune responses), they are not predictive of delayed immune-mediated adverse events in humans (i.e., adaptive immune responses). Most delayed adverse events in humans associated with vaccines are exceptionally rare and commonly the result of autoimmunity, often against nerve tissue (e.g., Guillain-Barre syndrome).^9,10 However, other immune-related adverse events have been reported after vaccination in humans, including myocardial/pericardial inflammation with the mRNA-lipid nanoparticle (LNP) vaccines against severe acute respiratory syndrome coronavirus 2, smallpox, and anthrax.^11-13 To the author’s knowledge, no delayed toxicity has been identified or reported in nonclinical prophylactic vaccine repeat-dose toxicity studies. Rare, delayed vaccine adverse events in humans often only become evident after hundreds of thousands or even millions of doses have been administered (i.e., in the post-marketing phase). In large part, this may be due to species-specific differences and previous exposure to pathogens between nonclinical species and humans. Furthermore, nonclinical toxicity studies cannot be powered adequately to identify very rare immunological events.

While there may be exceptions (e.g., live viral vaccines), the likelihood of identifying delayed toxicity during the recovery phase in vaccine repeat-dose toxicity studies is negligible. Therefore, microscopic evaluation could be limited to target tissues identified at the end of the dosing phase, unless there is a scientific or toxicologic rationale for assessing more tissues at the end of the recovery phase. Such an approach would be consistent with small- and large-molecule nonclinical toxicity studies. Inclusion of pivotal organs for microscopic evaluation at the end of the recovery phase (e.g., brain, lung, heart, kidneys, liver, reproductive organs) may lend comfort to vaccine developers, but it is unlikely they will add additional understanding of human safety risk. The addition of selected tissues should be considered if there is a historical association of the microbe with immune-mediated lesions in specific tissues (e.g., staphylococcus-associated myocarditis) or if data from the clinical trials or post-marketing surveillance identify risk associated with similar/same antigens or adjuvants.

One theoretical caveat to this proposal to limit the microscopic evaluation of tissues at the end of the recovery phase is related to vaccines intended for single-dose administration clinically. In such cases, the repeat-dose toxicity study typically has 2 doses administered 2 to 3 weeks apart. For such studies, the dosing phase necropsy will occur ~2 days after the second dose. Therefore, adaptive immune responses may not have had adequate time to fully manifest following the second dose and may not be evident until the end of the recovery period 2 to 4 weeks later. Microscopic findings at the end of the dosing phase may better reflect innate immune responses to the vaccine, with potentially less insight into the adaptive immune responses to the vaccine. In such cases, historical experience with the vaccine antigen and formulation (i.e., adjuvant), tissue distribution, and exposure should be considered when selecting tissues for evaluation in the recovery phase. In two-dose vaccine toxicity studies, microscopic evaluation of major organs (e.g., heart, lung, kidney, liver, brain, reproductive organs) in addition to target organs may be appropriate at the end of the recovery phase.

Conclusion

World Health Organization guidance documents are intentionally nonproscriptive regarding the microscopic evaluation of tissues at the end of dosing and recovery phases in vaccine repeat-dose toxicity studies. This has led to variability between vaccine manufacturers regarding which treatment groups and which tissues are microscopically evaluated at the end of dosing and recovery phases. Vaccine developers are encouraged to define a common and science-based approach to the microscopic assessment of study groups and tissues in nonclinical repeat-dose vaccine toxicity studies.

A reasonable approach is to perform microscopic evaluation only from the buffer control groups and the groups administered the highest-anticipated vaccine formulations intended for clinical trials. Microscopic evaluation of non–antigen-containing formulation control groups and lower-dose vaccine groups should be performed only on a case-by-case basis if unanticipated or adverse findings have been identified in the vaccine formulation group (e.g., to determine if the finding was antigen- or adjuvant-related or the result of their combination). Microscopic evaluation of formulation control groups for adjuvants not previously assessed in GLP toxicity studies may be valuable and/or expected by regulators.

Regarding the selection of tissues for microscopic evaluation, it is reasonable to perform microscopic evaluation on all tissues at the end of the dosing phase from the buffer control group and the group administered the highest-anticipated vaccine formulation intended for clinical trials. At the end of the recovery phase, it would be reasonable to perform microscopic evaluation on only the target tissues identified at the end of the dosing phase, as identifying delayed toxicity in nonclinical repeat-dose toxicity studies is unlikely. However, if there are clinical signs or historical data that suggest delayed toxicity is possible, microscopic evaluation of additional tissues or test groups may be prudent. A potential exception of evaluating only target tissues at the end of the recovery phase is in studies in which only two doses of vaccine have been administered, as adaptive immune responses may not be apparent until the end of the recovery phase. Microscopic evaluation of tissues from formulation controls or lower-dose formulation groups at the end of the recovery phase could be performed on a case-by-case basis to better understand toxicity. The proposals for microscopic assessments in vaccine toxicity studies included in this opinion piece are both scientifically justifiable and could improve the efficiency of vaccine development.