The Regulatory Framework
Regulatory Framework / Soft Law
The Soft Law Layer
International guidelines from ICH, OECD, and EPA that carry no force of law yet function as the single most powerful mechanism locking beagles into pharmaceutical and chemical testing worldwide.
What “Soft Law” Means in This Context
In legal theory, “soft law” refers to instruments that are not formally binding but exert regulatory influence because institutions treat them as authoritative. In the world of drug and chemical safety testing, three bodies produce the soft law that matters most: the International Council for Harmonisation (ICH), the Organisation for Economic Co-operation and Development (OECD), and the U.S. Environmental Protection Agency (EPA).
None of these bodies can pass laws. Their guidelines are technically “recommendations.” But when every major regulatory agency in the world — FDA, EMA, PMDA, Health Canada — adopts the same guideline as its operating expectation, the distinction between recommendation and requirement collapses in practice. A sponsor that deviates from an ICH guideline does not violate a statute. It simply cannot get its drug approved.
ICH Guidelines
The International Council for Harmonisation brings together regulatory authorities and pharmaceutical industry representatives from the U.S., EU, Japan, Canada, Switzerland, Brazil, China, South Korea, and others. Its guidelines are developed through a multi-step consensus process and, once finalized, are adopted — often verbatim — by member agencies. Three ICH guidelines are central to beagle use.
ICH M3(R2) — Nonclinical Safety Studies for Human Pharmaceuticals
The foundational guideline. M3(R2) establishes the expectation that nonclinical safety data supporting human clinical trials will include studies in one rodent and one non-rodent species. It does not say “beagle.” It says “non-rodent.” But over six decades of practice, “non-rodent” has become synonymous with “beagle” in the vast majority of pharmaceutical programs. M3(R2) covers the timing and scope of studies needed before first-in-human dosing, dose escalation, and long-term clinical trials — making it the gateway document for virtually every new drug.
ICH S7A/S7B — Safety Pharmacology (Cardiovascular)
S7A requires core safety pharmacology studies assessing effects on cardiovascular, central nervous, and respiratory systems. S7B specifically addresses the evaluation of human cardiac repolarization risk (QT prolongation) — one of the most common reasons drugs fail in development. Dogs, especially beagles, are the standard species for these studies because decades of cardiac electrophysiology data have been collected in beagles, creating a reference baseline that regulators know and trust. The hERG assay plus an in vivo QT study in dogs has become the default S7B package. Moving to a different species would mean abandoning that entire comparative dataset.
ICH S5(R3) — Reproductive and Developmental Toxicology
S5 governs studies evaluating a drug’s potential to harm fertility, embryo-fetal development, and pre/postnatal development. While rabbits are the standard non-rodent species for embryo-fetal studies, dogs are used in certain fertility and general toxicity programs that feed into the reproductive safety assessment. The guideline’s structure ensures that any drug intended for women of childbearing potential requires non-rodent data, which again defaults to species with extensive historical records.
OECD Test Guidelines
The OECD Test Guidelines are standardized protocols for safety testing of chemicals, pesticides, and industrial substances. Under the OECD Mutual Acceptance of Data (MAD) system, a study conducted according to OECD guidelines in one member country must be accepted by all other members — creating a powerful incentive to follow the exact protocol rather than innovate.
TG 409 — Repeated Dose 90-Day Oral Toxicity (Non-Rodent)
TG 409 is the workhorse guideline for subchronic toxicity testing in a non-rodent species. The guideline text states that the “dog is commonly used” and that “beagles are frequently used.” It specifies group sizes (typically 4 animals per sex per dose level), dosing protocols, clinical observation schedules, and necropsy requirements. The 90-day timeframe means each beagle endures three months of daily dosing followed by euthanasia and comprehensive organ examination.
TG 452 — Chronic Toxicity Studies
TG 452 extends the exposure period to 12 months or longer for non-rodent species. These chronic studies are required when humans will be exposed to a substance long-term. Dogs in chronic studies live in laboratory conditions for a full year of daily dosing. The guideline again identifies the dog as the standard non-rodent choice and relies on the same structural assumption: that historical control data exists for the selected species. Since the overwhelming majority of that data is in beagles, the guideline quietly perpetuates their use without ever mandating it by name.
EPA Guidelines (OPPTS 870 Series)
The EPA’s Office of Prevention, Pesticides, and Toxic Substances (OPPTS) publishes the 870 series of health effects test guidelines, harmonized with OECD protocols. These govern the safety testing of pesticides, industrial chemicals under TSCA, and other regulated substances.
| Guideline | Study Type | Dog Reference |
|---|---|---|
| OPPTS 870.3150 | 90-day oral toxicity (non-rodent) | Dogs, "preferably a defined breed" — beagles are the universal choice |
| OPPTS 870.4100 | Chronic toxicity | Non-rodent species, typically dogs for 1-year studies |
| OPPTS 870.3200 | 21/28-day dermal toxicity | Dogs used when oral route is not appropriate |
EPA’s 870 series sits atop the agency’s Good Laboratory Practice regulations (40 CFR Parts 160 and 792), which require standard operating procedures for animal housing, feeding, handling, and health evaluation. The GLP layer does not mandate a particular species, but it does require that studies be conducted according to “accepted” protocols — which in practice means OECD/EPA harmonized guidelines that default to dogs.
How Guidelines Become De Facto Requirements
The mechanism is straightforward. A sponsor submits a new drug application or chemical registration. A regulatory reviewer pulls the file and compares the nonclinical package against the relevant ICH, OECD, or EPA guideline. If the package matches the guideline — two species, appropriate duration, standard endpoints — the review proceeds smoothly. If it deviates, questions follow. Questions become delays. Delays cost millions.
No regulator formally rejects an application for using minipigs instead of dogs. What happens is subtler: an FDA reviewer issues an “information request” asking why the standard species was not used. The sponsor must justify the deviation, often with bridging studies or additional data. Months pass. Competitors with beagle data advance. The lesson is learned: follow the guideline, use the expected species, avoid friction.
The Lock-In Cycle
This cycle operates identically in every jurisdiction that follows ICH guidelines — which is every major pharmaceutical market on Earth.
Why Changing Guidelines Is So Slow
ICH guideline development follows a five-step process, from initial concept paper through expert working group drafting, regulatory consultation, adoption by member agencies, and implementation. A single guideline revision can take 5 to 10 years. OECD Test Guideline updates follow a similarly deliberate process involving member-country review cycles and consensus requirements.
Consensus requirement
ICH requires agreement among regulatory authorities and industry representatives across multiple regions. A single agency cannot unilaterally change a guideline — even if FDA wanted to remove the dog requirement tomorrow, it would need ICH-wide consensus.
Industry resistance
Pharmaceutical companies have invested billions in beagle-based testing infrastructure, historical databases, and regulatory relationships built on dog data. Any change that invalidates that investment faces organized opposition from within the guideline-development process itself — because industry sits at the ICH table as a full participant.
Validation burden
Replacing a standard species requires demonstrating that the alternative produces equivalent or superior predictive data. This requires years of parallel testing, peer-reviewed publication, and regulatory acceptance — all before a single guideline word changes.
Regulatory conservatism
Approving a drug that later causes harm is a career-ending event for a reviewer. Using the standard species and standard guideline is professionally safe. Innovation in species selection carries asymmetric risk: if the new approach works, nobody notices; if it misses a toxicity signal, the reviewer is blamed.
The FDA Modernization Act 2.0 as Potential Disruption
In December 2022, Congress passed the FDA Modernization Act 2.0, which amended the Federal Food, Drug, and Cosmetic Act to remove the mandate that drugs must be tested on animals before human clinical trials. The law replaced “animal testing” with “nonclinical testing” throughout the relevant statutory provisions, opening the door for New Approach Methodologies (NAMs) — organ-on-chip, computational models, organoids, and other human-relevant technologies.
The follow-on legislative effort, FDA Modernization Act 3.0 (S.355), advanced in late 2025, requiring an interim final rule to align FDA regulations with the statutory language change. Meanwhile, FDA announced in April 2025 a plan to reduce, refine, and replace animal testing requirements, encouraging NAMs immediately in investigational new drug contexts. In March 2026, FDA released draft guidance on general considerations for NAMs in drug development, including a validation framework.
But statutory permission is not the same as practical change. The lock-in cycle persists because historical control databases remain in beagles, regulatory reviewers still expect beagle data, and no single NAM has yet been validated to replace a full chronic-toxicity study in a non-rodent species. The statute opens a door. Walking through it requires dismantling the guideline infrastructure that has been built around dogs for more than half a century.