The nonclinical package FDA expects when a veterinary or ex-US compound enters human development

A molecule that already carries a veterinary approval, or that is marketed for humans outside the United States, does not enter US human development with a discounted nonclinical burden. FDA and the ICH framework assess the compound against the same first-in-human (FIH) standard as any new investigational drug: the sponsor must be able to justify the safe starting dose, dose range, and duration of the proposed clinical trial from an adequate, well-documented nonclinical dataset. Prior animal or foreign human experience is relevant, and it can reduce or replace individual studies, but only where the existing data are scientifically adequate, generated to acceptable quality standards, and demonstrably relevant to the proposed human use. In practice the work shifts from "generate everything de novo" to "run a rigorous gap analysis, then bridge the gaps."

The baseline: what a first-in-human package has to contain

ICH M3(R2) is the anchor. It sets the nonclinical studies expected before and during clinical trials and ties study scope to the intended clinical exposure. The general principle is that nonclinical assessment should use all available data to support the starting dose and dose range, with the program extended as clinical development advances; particular studies may be "abbreviated, deferred, omitted, or added" on a case-by-case basis with scientific rationale ^1229828. The core elements are:

• Repeat-dose toxicity in two mammalian species, one non-rodent. Study duration should generally be at least as long as the proposed clinical trial, up to the maxima in M3(R2) Table 1: a 2-week trial is supported by 2-week studies; trials of 2 weeks to 6 months by studies of matched duration; trials longer than 6 months by 6-month rodent and 9-month non-rodent studies ¹⁰⁶. For marketing, the recommended durations are longer (Table 2), rising to 6-month rodent and 9-month non-rodent studies for chronic use ¹⁰⁷.
• Single-dose / acute toxicity. Dedicated single-dose studies are generally not needed when acute toxicity information is captured within other studies ⁷¹⁰. For short (under 14-day) or single-dose human trials, an extended single-dose toxicity study can serve as the toxicology support ¹⁰⁶.
• Safety pharmacology core battery (ICH S7A). Assessment of effects on the three vital systems: central nervous (motor activity, behavior, coordination, body temperature; e.g., functional observational battery or modified Irwin's), cardiovascular (blood pressure, heart rate, ECG, repolarization/conductance), and respiratory (rate, tidal volume, oxygen saturation). The battery may be supplemented or omitted with justification ³⁷.
• Genotoxicity battery (ICH S2(R1)). A bacterial reverse mutation (Ames) test, an in vitro mammalian test for chromosomal damage (chromosome aberration, micronucleus, or mouse lymphoma Tk assay), and an in vivo test, assembled as one of two equivalent standard battery options ³⁴⁴³. Negative results in appropriate in vivo assays with adequate exposure generally suffice to show absence of significant genotoxic risk ⁴⁵.
• Toxicokinetics / PK. Exposure data underpin dose selection; all relevant nonclinical PK is considered when setting the human starting dose ¹⁰⁷²¹.
• Local tolerance. Evaluated for the route/formulation; ICH S6(R1) notes the marketing formulation should be tested, though representative formulations may be acceptable with justification, and findings within general toxicity studies can sometimes obviate a separate study ⁴⁶.
• Reproductive toxicity (ICH S5), timed to the population. The standard genotoxicity battery must be complete before enrolling women of childbearing potential (WOCBP) not on highly effective contraception ⁴⁷. Preliminary reproduction data from two species can support up to 150 WOCBP for up to three months before definitive testing; in the US, embryo-fetal development assessment may be deferred to before Phase III with pregnancy precautions, whereas the EU and Japan generally expect it earlier ⁴⁷. Male fertility studies should be complete before large or long Phase III trials, though men can enter Phase I/II earlier because male reproductive organs are assessed within repeat-dose studies ¹⁰⁹. Embryo-fetal data may be used to support WOCBP enrollment (ICH S5(R3)) ¹⁵.
• Carcinogenicity. A marketing-stage requirement rather than an FIH gate; combination carcinogenicity studies are generally not recommended where the individual agents have been tested to current standards ¹⁰⁸.

Starting dose is derived from all relevant nonclinical data, especially the NOAEL in the most appropriate species, using exposure margins rather than dose alone. For example, exploratory approaches target a human AUC near 1/50 of the AUC at the NOAEL in the more sensitive species and cap the maximum clinical exposure using AUC-at-NOAEL limits ^1074951.

For biologics, the species question is decisive. ICH S6(R1) requires a pharmacologically relevant species selected on more than sequence homology (target binding, occupancy, and functional activity data), typically both a rodent and non-rodent for short-term studies and one species long-term, with the rodent preferred absent a rationale; where no relevant species exists, homologous proteins or transgenic models expressing the human target are the fallback ⁶¹⁰⁵. A veterinary product's toxicology generated in a non-relevant species may therefore carry little weight for a human biologic program.

Why a prior veterinary or ex-US approval does not shortcut the package

Three FDA expectations explain why an existing approval elsewhere is treated as supporting information, not as a substitute for the FIH dataset.

GLP compliance. FDA expects nonclinical safety studies supporting an IND to be conducted under Good Laboratory Practice (21 CFR Part 58). The IND regulations state that toxicology studies intended to support the safety of a clinical investigation are subject to GLP, and that non-GLP studies require a statement of the reason for noncompliance in the final report ^819596101. Non-GLP data can be considered only if the sponsor shows the study was rigorous and adequately controlled to preserve reliability and integrity; FDA guidance does not endorse inadequately documented non-GLP studies as pivotal safety studies ^81787984. Legacy veterinary or foreign development-era studies frequently fail this documentation bar.

Relevance and adequacy. The existing studies must actually address the human risk questions, in a relevant species, at exposures that cover the proposed clinical range. Where the available toxicity information does not support the exposure or exposure pattern from the intended human formulation or route, additional studies are recommended ¹⁰¹¹¹⁹.

FDA's ability to validate the data. For foreign studies not conducted under an IND, acceptance under 21 CFR 312.120 requires that the study was well-designed and well-conducted in accordance with GCP, including independent ethics committee review and informed consent, with supporting information demonstrating GCP conformance ^5556596769. FDA must be able to validate the data, potentially through on-site inspection or access to source records; if the records are unavailable, FDA may decline to accept the data ^575962635866. Where a marketing approval would rest solely on foreign data (21 CFR 314.106), the data must also be applicable to the US population and US medical practice, considering demographics, disease manifestations, and standards of care, and FDA may still request US-relevant studies ^5456646572.

How prior data can legitimately reduce the workload

The same framework that refuses an automatic pass also provides structured routes to leverage what already exists.

Use all available data (M3(R2) and the 3Rs). Adequate existing toxicology can support clinical trials and replace some dedicated studies: acute toxicity information captured elsewhere removes the need for separate single-dose studies, and general toxicity studies can support or replace separate safety pharmacology studies when they address the endpoints ⁷¹⁰. M3(R2) and related guidelines actively promote reducing animal use consistent with reduce/refine/replace principles, so duplicating a sound existing study is discouraged ²⁴¹⁸²¹.

Reliance on prior findings and published literature (505(b)(2)). A sponsor may rely on the published literature and/or FDA's previous finding of safety and effectiveness for a listed drug, but only to the extent the proposed product shares the active ingredient, dosage form, strength, route, indication, and conditions of use; any differences must be supported by additional data, clinical or animal as appropriate ². Literature alone is usually insufficient, and sponsors should access and independently review primary data ². The governing concept is bridging: establishing that the earlier information is scientifically relevant to the current product and use before it can be leveraged ¹¹⁰⁴.

Bridging triggers and impurity qualification. Additional nonclinical studies become more likely as the new product diverges from the prior one, even when the composition is unchanged, notably a new route of administration or a markedly different duration of use ¹⁰¹¹¹⁹. A change in route can raise new safety or efficacy questions that make bridging difficult or impossible, particularly for complex biologics ¹⁰⁴. Impurities are a common trigger: new impurities above the relevant ICH Q3A/Q3B thresholds should be identified and qualified, and qualification may require safety testing where the impurity cannot be reduced or qualified from existing data ¹¹⁸¹²⁶. (In the oncology setting, ICH S9 allows a risk-assessment approach in which a dedicated NOAEL qualification study is often not warranted ¹⁵¹.)

A practical way to frame it

The efficient path for a compound with veterinary or ex-US history is a documented gap analysis against the M3(R2)/S-series package for the specific trial being proposed, asking three questions of every existing study: Is it in a relevant species at relevant exposures? Was it conducted and documented to GLP (or defensibly rigorous) standards FDA can validate? Does the proposed human use (route, formulation, impurity profile, duration, population) introduce questions the old data cannot answer? Studies that clear all three can support the IND; the remainder define the new nonclinical work.

Where the intended early clinical exposure is genuinely minimal, the exploratory IND / phase 0 route offers a reduced package scaled to exposure: microdose studies (for example, a total dose under 100 µg, below 1/100 of the NOAEL and of the pharmacologically active dose) can be supported by an extended single-dose toxicity study in a single, usually rodent, species, with routine genotoxicity and safety pharmacology not recommended; the package then expands to repeated-dose, two-species, and longer-duration studies as human exposure increases ^{133135136134140142}. This can be an attractive first step for a repurposed veterinary molecule where the sponsor wants human PK or target-engagement data before committing to the full toxicology program.