How Do Scientists Know a Stem Cell Therapy Is Safe Before It Reaches Human Trials? A Guide to Preclinical Testing

What This Article Covers

Before a stem cell therapy ever touches a human patient, it typically spends years moving through a structured pipeline of laboratory and animal testing designed to answer one question first: could this hurt someone? That pipeline — collectively called preclinical testing — includes characterizing the cells themselves in the lab, checking for the specific safety concern of tumor formation, running formal animal toxicology studies under federal quality standards, and ultimately packaging all of it into an Investigational New Drug (IND) application that the FDA reviews before any clinical trial can begin. This guide walks through each stage in plain language, and is honest about what preclinical testing can and cannot tell scientists about how a therapy will behave in people.

In Vitro Characterization: Getting to Know the Cells

Before any animal is involved, researchers spend enormous effort characterizing the cell product in the lab. This means confirming identity (are these the cells the manufacturer says they are), purity (is the population free of contaminating cell types or residual reagents), and potency — whether the cells actually do the biological job they're meant to do, such as differentiating into a target tissue. FDA's guidance on potency testing treats this as a foundational requirement, since a cell product's behavior can shift subtly between manufacturing batches. Scientists also run genetic stability tests to check that cells haven't accumulated worrisome mutations during manufacturing, which matters enormously for the next section.

Animal Safety and Efficacy Studies

Once a cell product is well-characterized, it moves into in vivo (animal) studies, which serve two purposes: gathering early evidence the therapy does something biologically useful, and — more importantly for safety — watching for adverse effects like unwanted immune reactions, abnormal cell migration, or organ toxicity. FDA guidance describes tracking where transplanted cells actually go in the body, into target tissue, off-target tissue, and even reproductive tissue, using labeled cells, genetic markers, or PCR-based detection of human-origin cells in animal tissue. Researchers monitor animals across multiple time points to catch not just immediate reactions but delayed-onset problems that might only appear weeks or months later — one reason preclinical programs for cell therapies often run longer than for a typical drug.

Tumorigenicity Testing: A Concern Specific to Stem Cells

This is where stem cell therapies face scrutiny most other drug classes don't. Pluripotent stem cells (like embryonic stem cells or induced pluripotent stem cells, iPSCs) can, by definition, form virtually any cell type — and that same flexibility creates a theoretical risk of forming teratomas (a type of tumor) if undifferentiated or incompletely differentiated cells persist after transplantation. FDA guidance is explicit that tumorigenicity testing must use the actual clinical product intended for people, not a generic stand-in, and that study design has to account for the cells' differentiation status, how heavily they were manipulated during manufacturing, and whether they carry transgenes that could influence growth. Practically, this means long-duration animal studies — long enough for transplanted cells to survive and, if they were going to form a tumor, to actually do so — with proper controls and enough animals for statistically meaningful conclusions. It's a slow, deliberately conservative process, because the failure mode being screened for is serious and largely specific to cells that retain some degree of "stemness."

GLP Toxicology Studies

The pivotal safety studies supporting an IND filing generally must be conducted under Good Laboratory Practice (GLP) regulations (21 CFR Part 58) — a federal quality framework covering data integrity, facility standards, and independent quality assurance, so regulators can trust results weren't cherry-picked or sloppily recorded. FDA guidance acknowledges a wrinkle specific to cell therapies: some specialized endpoints, like cell fate tracking or biodistribution in disease-specific animal models, aren't always available at fully GLP-certified facilities. When that happens, sponsors must document exactly where and why a study deviates from GLP, what that means for interpreting results, and bring in independent quality oversight anyway wherever possible. This isn't a loophole so much as an acknowledgment that stem cell science sometimes outpaces the standardized infrastructure built for conventional drugs.

The IND Application and FDA Review

All of this data — in vitro characterization, animal safety and efficacy results, tumorigenicity findings, and GLP toxicology reports — gets compiled into an Investigational New Drug (IND) application submitted to the FDA. For stem cell products, review sits with the Center for Biologics Evaluation and Research (CBER), specifically its Office of Therapeutic Products (OTP), a "super office" CBER established in February 2023 to handle the surge in cell and gene therapy submissions with more dedicated expertise. Once an IND is submitted, federal regulation requires a mandatory 30-calendar-day waiting period before a sponsor can start dosing human subjects — time the FDA uses to review the preclinical package and decide whether volunteers would face unreasonable risk. If reviewers have concerns, they can place the trial on clinical hold, pausing the study until the sponsor addresses the issue. No hold, and the 30 days run out, means the sponsor may proceed to human trials — the beginning of clinical testing, not a finish line.

The Limits of Preclinical Data

It's worth being direct about what this process does and doesn't guarantee. Animal models are chosen for being biologically informative, but no animal has an immune system, tissue architecture, or lifespan identical to a human's, and translational medicine researchers have long documented that animal toxicology results don't perfectly predict human outcomes. A therapy can clear every preclinical hurdle and still show unexpected side effects or insufficient efficacy once tested in larger, more diverse human populations — exactly why multi-phase clinical trials exist rather than being treated as a formality. Tumorigenicity testing, while serious and required, is a risk-reduction exercise, not a risk-elimination one: it lowers the odds of a catastrophic outcome without making that outcome impossible, especially for products with any residual pluripotent or highly proliferative characteristics.

Bottom Line

Preclinical testing is a genuinely substantial body of work — cell characterization, animal safety and efficacy studies, dedicated tumorigenicity testing, and GLP-grade toxicology, all reviewed by FDA scientists before a single human volunteer is ever dosed — and it exists because stem cell products carry real, distinctive risks that this system is specifically built to catch early. But it's a filter that reduces risk, not a guarantee of safety or effectiveness: passing preclinical testing earns a therapy the right to be tested carefully in humans, not a verdict that it works or that it's risk-free. Understanding that distinction is one of the most useful things a newcomer to this field can take away.