When is a non-human primate (NHP) model required for AAV gene therapy toxicology?

An NHP model is indicated when the therapeutic target or mechanism of action exhibits low homology in lower species. It is also the preferred model when assessing potential immunogenicity against the vector capsid or transgene product, as the NHP immune system more closely resembles that of humans. Regulatory agencies often expect NHP data for AAV programs targeting novel pathways or demonstrating broad tissue tropism.

How do species-specific glycan profiles impact AAV biodistribution and toxicology?

Cell surface glycans function as primary attachment factors for many AAV serotypes. The type and density of these glycans can differ significantly across species, particularly in high-clearance organs like the liver. These differences can cause profound variations in vector biodistribution, leading to divergent on-target efficacy and off-target toxicity profiles between a rodent model and an NHP or human. Evaluating glycan-driven tropism is a key component of model selection.

Can rodent models still be relevant for advanced therapies with human-specific targets?

Yes. Rodent models are valuable for initial dose-range finding studies, evaluating the tolerability of the formulation buffer, and assessing toxicities related to manufacturing impurities. Genetically engineered rodent models that express a human target can also provide important early data on pharmacology and mechanism-based toxicity before committing to more complex NHP programs.

The selection of a pharmacologically relevant species is the foundation of any successful preclinical toxicology program for an advanced therapy. The selection between a rodent and an NHP model is a strategic determination based on mechanism of action, target homology, vector tropism, and potential immunogenicity. Species-specific biological factors, such as cell surface receptor or glycan expression for AAVs, directly dictate biodistribution and can invalidate safety data from an improperly selected model. A well-designed program often uses a multi-species approach to de-risk development and build a robust data package for an 18-24 month IND timeline.

The relevance of a preclinical toxicology program is defined by its ability to predict clinical outcomes. For cell and gene therapies, this begins with selecting a species that recapitulates human biology with respect to the therapeutic’s mechanism of action. This involves a rigorous assessment of cross-species target sequence homology, expression levels, and biological function. An irrelevant species, even if convenient, produces data that fails to inform the clinical risk assessment.

Our approach prioritizes this step, ensuring that investments in GxP toxicology studies yield meaningful, translatable safety data. This is a core principle of our work within our >100,000 sq ft facility, where we conduct studies that form the basis of successful regulatory submissions. Since 2019, programs we have supported have achieved a 100% IND success rate, with the Franklin Biolabs brand itself launching in 2024.

Rodent models provide an efficient system for initial safety assessments. They are instrumental in:

• Establishing a maximum tolerated dose (MTD) in early dose-range-finding studies.

• Evaluating acute toxicities associated with the formulation or manufacturing impurities.

• Providing foundational data in genetically engineered models that express the human target.

However, for many advanced therapies, the inherent biological divergence between rodents and humans limits their predictive power. Differences in immune response and a lack of target conservation can mask potential liabilities or fail to demonstrate on-target effects.

NHP models are often the most translationally relevant system for evaluating the safety of advanced therapies due to their high genetic and physiological homology to humans. They are particularly important for:

• Gene Therapies: Assessing biodistribution and shedding for viral vectors like AAVs, whose tropism is dictated by receptors with high NHP-human conservation.

• Cell Therapies: Evaluating cell engraftment, persistence, and potential immunogenicity or tumorigenicity in a host with a comparable immune system.

• Oligonucleotides: Characterizing complex pharmacology and identifying mechanism-based toxicities that may not manifest in lower-order species.

The importance of correct species selection is clearly illustrated by research into AAV vector engineering. As shown in studies such as PMID: 39001819, modifying an AAV capsid can significantly alter its tissue tropism. That work also revealed that inherent differences in glycan presentation between species can result in markedly different vector biodistribution profiles. A vector that shows favorable cardiac targeting and liver de-targeting in one species may behave differently in another due to these molecular-level incompatibilities. Relying on a single, potentially irrelevant species for non-target tissue biodistribution can lead to a misinterpretation of the safety profile.

A rigorous scientific approach to species selection aligns directly with our commitment to the highest standards of animal welfare. By choosing the most relevant model, we ensure that the minimum number of animals are used to generate the most meaningful data. Our programs are designed around the 3Rs (Reduction, Refinement, and Replacement), and our facility is registered with the USDA.

*

Choosing the appropriate species requires a balanced evaluation of scientific and logistical factors. In collaboration with our strategic partners in pathology, we help sponsors navigate this complex decision.