PRACTICAL GUIDE | 2026
No checklist. One standard: convince the committee the product will be safe in humans.
By Julio G. Martinez-Clark
CEO, bioaccess®
Last verified: September 2026 | General information only—not legal or regulatory advice. Rules change frequently; confirm the strategy with qualified regulatory counsel.
How much preclinical data do you need before a first-in-human in Latin America? Sponsors usually arrive expecting a checklist — a fixed list of studies, species, and durations that unlocks the clinic. Latin American regulators do not work that way. There is no region-wide preclinical checklist and no codified rule that studies must be GLP-compliant as a condition of first-in-human authorization. There is one standard instead: prove the device or drug will be safe in humans. The burden of proof is on you.
Definitions first. Preclinical (nonclinical) data is the bench and animal testing done before a product enters humans — biocompatibility, toxicology, safety pharmacology, animal performance studies. GLP means good laboratory practice (21 CFR Part 58 in the US): the quality system for nonclinical lab studies. Non-GLP or “R&D-grade” data was generated without that formal quality system. The distinction matters less in Latin America than sponsors expect — and more, later, at FDA.
Is there a fixed preclinical checklist for a LATAM first-in-human?
No. No Latin American regulator publishes a 21 CFR Part 58-style list of required preclinical studies for first-in-human authorization. Colombia’s INVIMA publishes an inventory of approved and not-approved device studies — protocol, product, sponsor, site, investigator — but no required species, sample sizes, durations, or GLP stamp. Chile’s guidance requires preclinical testing and risk evaluation to be exhaustive and sufficient to support the investigation, under ISO 14971, while expressly disclaiming any device-specific test menu. The file that is actually read is a risk-management file plus an investigator’s brochure — not a checklist with boxes to tick.
Is non-GLP preclinical data enough?
For medical devices, yes — with rationale. “If you have the rationale to support why that drug doesn’t need that specific test, that should be sufficient.” R&D-grade data has supported real first-in-human device programs in the region: biocompatibility files, animal performance studies, and risk assessments accepted on the strength of their science rather than their quality-system stamp. A rationale can substitute for a specific test when it explains why the test is unnecessary — not merely that it was skipped.
For drugs and biologics, narrow the claim. Colombia’s medicines framework expects an ICH M3(R2)-type nonclinical safety package, and a non-GLP pivotal toxicology study for a new chemical entity or biologic will draw questions from INVIMA and the ethics committee in the same way it would from FDA. There is no Colombian rule that says “tox studies must be GLP” — but say “no explicit GLP mandate” rather than “non-GLP is accepted.” And note Brazil: ANVISA’s nonclinical guidance expects GLP-conducted safety studies for medicines — the closest thing to an explicit GLP expectation in the region, though it sits at guidance level and applies to drugs, not the device pathway.
Can we reuse the data we already generated abroad?
Yes — origin is not the criterion. Sponsors ask, “Can we just use everything we already used?” — testing done in China, non-GLP work from years ago. The answer: “They don’t care… GLP, non-GLP… regardless of where.” No LATAM regulator imposes a country-of-origin restriction on preclinical data. Three practical caveats make reuse succeed or fail:
- Language: reports must be submitted in Spanish, or with translation. Budget for it.
- Completeness: for novel products, reviewers expect full study reports — not summaries or slide decks.
- Traceability: show the standard used, the laboratory’s quality system (reviewers frequently ask about accreditation, such as ISO/IEC 17025 or a national GLP program, even where GLP is not mandated), and that the tested article matches the clinical product.
What does a convincing preclinical package look like?
Think of the package as an argument, not an inventory. Every study should answer a human-safety question, and the investigator’s brochure should tie the answers together:
- Map each test to a risk: what human harm does this study rule out, and for which patient population?
- Write the rationale for every test you did not run — the missing study with the best explanation beats the missing study with none.
- In Colombia, include the procedure risk matrix: regulators there want the risk analysis of the clinical procedure, not just the product. Confirm exactly what the committee expects before filing.
- Tie it to ISO 14971 risk management for devices: hazards, harms, controls, and residual risk, all traceable.
- Keep the test article consistent: the device or formulation tested preclinically must be the one going into humans, or the bridge must be explicit.
Does a LATAM-sufficient package satisfy the FDA later?
Not automatically — and this is the strategic point. FDA generally expects IND-enabling safety pharmacology and toxicology to follow GLP (21 CFR Part 58). A package that clears a Latin American first-in-human and a package that supports a US IND are not always the same thing. Design the preclinical program for both endpoints from the start: run the LATAM FIH on the rationale the region accepts, but build the tox package so it also serves the later FDA path. The cheapest preclinical program is the one you run once.
Frequently asked questions
Q: Is GLP required for a first-in-human trial in Colombia?
A: No codified GLP requirement exists for FIH authorization in Colombia — for devices, R&D-grade data with strong rationale has been accepted. For drugs and biologics, expect INVIMA and the ethics committee to probe non-GLP pivotal toxicology closely.
Q: Can we use preclinical data generated in China?
A: Yes. No LATAM regulator restricts preclinical data by country of origin. Submit full study reports in Spanish (or translated), with lab accreditation and test-article traceability documented.
Q: What if we skipped a standard preclinical test?
A: Provide the scientific rationale for why that test was unnecessary for your product. A reasoned omission is defensible; an unexplained gap is not.
Q: Who decides whether our preclinical package is sufficient?
A: The national regulator (e.g., INVIMA in Colombia) together with the ethics committee — which in Colombia does primary review. Their standard is sufficiency of evidence for human safety, not checklist compliance.
Q: How much preclinical data do you need before a first-in-human in Latin America — really?
A: Enough to convince a regulator and an ethics committee that the product will be safe in humans, documented so each study maps to a risk. That is the entire rule. Everything else is execution.
Q: Does non-GLP preclinical data hurt us with FDA later?
A: It can, for drugs: FDA generally expects IND-enabling safety studies under GLP. Plan the program so the LATAM FIH package and the future IND package are built together, not twice.
How much preclinical data do you need before a first-in-human in Latin America? As much as it takes to meet the burden of proof — no more, no less, and no checklist will tell you the number. Bring the rationale, not just the reports.
Talk with bioaccess® about your Latin America FIH strategy
Bring us your existing preclinical package — GLP, non-GLP, generated anywhere. We will assess it against the burden-of-proof standard for your target countries and tell you what is sufficient, what needs a rationale, and what needs to be run.
Talk with bioaccess® about your Latin America FIH strategy
References
- Colombia: Decreto 4725 de 2005 (medical-device regime, including investigational use of unregistered devices); Resolución 8430 de 1993 (ethical norms for health research); INVIMA device clinical-investigation guidance. Resolución 2378 de 2008 adopts GCP for medicines trials.
- Chile: ANDIM/ISP guidance — preclinical testing and risk evaluation must be exhaustive and sufficient under ISO 14971; no device-specific test menu.
- Brazil (medicines): ANVISA nonclinical guide expects GLP-conducted safety studies — guidance-level expectation; confirm current text before relying on it.
- US reference: 21 CFR Part 58 (GLP for nonclinical laboratory studies); ICH M3(R2) (nonclinical safety studies for human trials).
- bioaccess® preclinical sufficiency experience across LATAM FIH programs, 2021–2026. Last verified: September 2026.
Leave a Reply