The breadth of this guide reflects a real divergence: while the strategic framework is shared, the scientific execution differs sharply between small molecules and biologics. Differences in molecular structure, pharmacology, metabolism, species relevance, and toxicity mechanisms require distinct nonclinical development strategies and regulatory approaches.

Small molecules are typically low-molecular-weight compounds, often less than 1,000 Daltons, chemically synthesized, and frequently active across standard toxicology species. Their safety programs follow ICH M3(R2) and typically include pharmacology, DMPK, safety pharmacology, repeat-dose toxicology, and the standard battery of genetic toxicology studies. Genotoxicity testing is generally required because reactive metabolites, direct DNA interactions, and unintended off-target effects may contribute to safety risk. Predictive, mechanism-based toxicology strategies are especially valuable for small-molecule oncology programs where the intended patient population may tolerate greater risk than healthy volunteers. ICH S9 permits a more flexible package appropriate to the severity of the disease, unmet medical need, and anticipated clinical benefit.

Large molecules (biologics—monoclonal antibodies, recombinant proteins, antibody-drug conjugates, and beyond) present fundamentally different development challenges. Unlike small molecules, biologics tend to cause toxicity through exaggerated pharmacology, target-mediated effects, immune modulation, or unintended interactions with biological pathways rather than through nonspecific chemical toxicity. Under ICH S6(R1), species selection is driven by pharmacological relevance. Species selection depends on target expression, binding affinity, functional activity, and biological response. In some cases, only non-human primates (NHPs) provide meaningful pharmacological relevance. In others, transgenic animals, surrogate molecules, or alternative experimental systems may be required when no naturally relevant species exists. Genotoxicity studies are generally not required for proteins; instead, immunogenicity, cross-reactivity, cytokine-release, and immune activation risks are assessed where appropriate, and local tolerance become central concern, and human proteins are often immunogenic in animals in ways that complicate exposure and interpretation. High-dose selection for biologics should be scientifically justified using pharmacology, exposure, target engagement, toxicity, and the anticipated clinical exposure, consistent with ICH S6(R1), with route and schedule matched to the clinical plan. Biologics are a large and growing share of the pipeline—more than 250 have been approved since recombinant insulin in 1982, and they were projected to reach roughly a third of all therapeutics—making fluency in this second strategy essential.

Specialized modalities—oligonucleotides, peptides, and vaccines—sit between these poles and are handled case-by-case; early FDA alignment on species selection and study design is especially valuable for them.