FDA Draft Proposes Targeted Reductions in Oncology Toxicology Studies

A Focus on Three-Month General Toxicology

FDA announced draft guidance on streamlined nonclinical safety studies for oncology biologics and conjugated products on 1 June 2026. The document, dated May 2026, focuses primarily on three-month general toxicology studies and supplements the existing ICH S9 framework. Its recommendations concern specified product classes and their safety characteristics.

The draft draws on analyses of toxicology data and experience gained while reducing nonhuman-primate use. It asks developers to build a scientific risk assessment around the product, its target and prior knowledge. Impurity and excipient safety assessments are outside its scope.

Species Relevance and Antibody Classes

For biologics, a relevant test species should show target binding and the intended pharmacological effects. If no pharmacologically relevant species exists, the draft allows consideration of a weight-of-evidence assessment in place of animal toxicology studies. If rodent and nonrodent species show pharmacological activity similar to humans, one- and three-month general toxicology studies may use a single rodent species, supplemented by a risk assessment as appropriate.

For monospecific antibodies blocking PD-1 or PD-L1, a weight-of-evidence assessment may address chronic effects in place of a three-month toxicology study. With justification, the one-month study may use a nonsacrificial design and additional risk evidence. A similar weight-of-evidence approach is proposed for the chronic-effect assessment of CD3 bispecific T-cell engagers.

When the ADC Payload Drives Toxicity

The antibody–drug conjugate recommendation has two important conditions: the cytotoxic payload’s safety should be well characterized, and the payload should be the main driver of toxicity. Under those conditions, a three-month study may be conducted in rodents alone, regardless of whether the ADC binds the rodent target antigen. The study can use the ADC or its payload, as appropriate.

If the antibody recognizes a novel target and does not bind that target in the rodent species, the draft also calls for a weight-of-evidence risk assessment. The product’s target biology and the basis for interpreting payload toxicity therefore remain part of the justification for the streamlined program.

What the Integrated Assessment Contains

The proposed assessment can combine pharmacology, safety and pharmacokinetic information generated with the investigational product. Literature on target expression and physiological roles in healthy and diseased tissues adds another source of evidence. Published animal and human toxicity experience with the drug class can further inform the anticipated risks.

Fit-for-purpose NAMs can contribute to this package where appropriate. Their value depends on the safety question being addressed and the other information in the assessment. FDA oncology review divisions would judge adequacy from the totality of evidence.

The draft also invites scientifically justified alternatives for classes not specifically described. Examples include nonsacrificial designs, designs using fewer animals or risk assessments for well-understood targets. Developers are encouraged to discuss proposals early, and the document notes that conducting a three-month study earlier may avoid separate one- and three-month studies.

These are draft recommendations for consultation. They define conditional ways to assemble a safety program around the properties of a product, with the actual study plan and supporting evidence assessed in its development context.