Selecting Translationally Relevant Preclinical Systems
The predictive value of any nonclinical toxicology study hinges on the relevance of the chosen biological system. For targeted cytotoxic conjugates, assuming that a single model will be universally informative is a significant program risk. Predictive accuracy requires a model that faithfully recapitulates the three key biological components of the therapeutic: the antibody, the linker, and the payload.
Key sources of inter-species variability that must be characterized include:
-
Target Expression: The antigen target may have different expression levels or patterns on hematopoietic cells in nonclinical species compared to humans.
-
Payload Sensitivity: Hematopoietic progenitors from different species can exhibit varied intrinsic sensitivity to the cytotoxic payload.
-
Metabolism and Pharmacokinetics: Differences in linker stability and payload metabolism can significantly alter the exposure profile and resulting toxicity.
A systematic approach to model selection, integrating in silico analysis with in vitro functional data, is necessary to justify the use of a particular species for pivotal safety assessment.
| Consideration |
Rodent Models |
Non-Human Primate (NHP) Models |
| Target Homology |
Variable; requires sequence analysis and expression confirmation. |
High degree of homology with human hematopoietic system. |
| Payload Metabolism |
May differ significantly from human metabolic pathways. |
Generally more predictive of human pharmacokinetics and metabolism. |
| Reagents & Tooling |
Widely available commercial and custom reagents. |
More limited and specialized reagent availability. |
| Operational Scale |
Suitable for early screening and dose-range finding. |
Reserved for pivotal, IND-enabling GxP safety studies. |