A reading guide
How to read stem cell evidence without getting lost.
A field guide for patients and families researching regenerative medicine — what each kind of evidence can tell you, what it cannot, and how to read past the marketing.
Most patients researching stem cells do not need a statistics lecture. They need to know what kind of evidence they are looking at — a lab idea, an animal study, a case report, a small human trial, a larger controlled trial, a registry, a guideline, or a regulator's decision. The trick is not to dismiss early science, and not to mistake it for a proven treatment.1
Regenerative medicine is a real and active field. Some of the early biology is genuinely interesting. Some international approvals matter. Some patients are right to be curious — about their own joint, their own diagnosis, their own options. The problem is not curiosity. The problem is that hope, biology, and a published abstract get blended together until a wellness IV in a strip mall sounds like a treatment your oncologist forgot to mention.
This page is not a verdict machine. It is a reading guide. It walks through the kinds of evidence you will see cited, what each one can and cannot tell you, and the questions that turn a confident-sounding claim into something you can actually weigh.
The ladder is not a moral ranking.
It helps to say this out loud: early evidence is not bad. It is early. Every standard-of-care treatment in any modern hospital began as a hunch, then a lab idea, then an animal study, then a small human trial, then something larger. The ladder is simply how a field learns whether a hunch survives contact with real bodies. Skipping rungs is not faster science; it is just less science.
The trouble starts when an early rung gets sold as a top rung. When a clinic charges twenty thousand dollars for an intervention that has been studied in twenty people, with no comparison group and no long-term follow-up, the issue is not that the biology is uninteresting. The issue is that the confidence has been priced in years before the evidence justifies it.
Early evidence can be worth watching. It should not be priced, marketed, or promised like settled evidence.
Four bands of evidence, in plain English.
Most of what you will read about a stem cell, PRP, exosome, BMAC, or MSC intervention falls into one of four rough bands. The bands are not labels we paste onto every condition page. They are reading sections — places to pause and ask what a study can and cannot tell you.2
Band 1
The biology makes sense.
The cells, the molecule, or the mechanism behave in a lab dish or an animal model the way a clinician would hope. The idea is plausible. That is genuinely useful — most useless ideas die here without ever reaching a person.
What it can tell you
That a hypothesis is worth testing in humans. That a particular cell type does something interesting under controlled lab conditions.
What it cannot tell you
Whether it works in a person. Whether it is safe in a person. Whether the dose, route, or preparation a clinic uses resembles the one in the paper at all.
What a clinic should not claim from this
That a treatment “works” or is “backed by science” because there is a cell-biology paper somewhere behind it.
What to ask
Is there any human evidence for this specific use, or only lab and animal work? If only lab and animal work — what would change your mind that you are paying for an idea, not a treatment?
Band 2
There is an early human signal.
People have actually received the intervention. There may be case reports, single-arm pilot studies, small uncontrolled series, or early-phase trials. Some patients did better. Some did not. There is no large comparison group, no blinding, no long follow-up — but it is real human data, not just biology.3
What it can tell you
That the intervention reaches people. That it is, at least in small numbers, tolerable. That an effect is plausible at roughly the dose and route being tried.
What it cannot tell you
How big the real effect is once you account for the natural ups and downs of the condition, the placebo response, and the tendency of people who pay out of pocket to report improvement. Most early signals shrink — sometimes to zero — once they are tested against a proper comparison group.
What a clinic should not claim from this
That the intervention is “proven”, “effective”, or “clinically validated” based on a handful of testimonials and one open-label pilot.
What to ask
How many patients in total? Was there a comparison group? Was the outcome something patients felt — pain, function, quality of life — or only a lab marker? Was the paper peer reviewed, and by whom?
Band 3
There is controlled human evidence.
Randomized trials with a real comparison group — sometimes a sham injection, sometimes standard care — have looked at the intervention in the population it is being marketed to. Systematic reviewers, such as Cochrane, have summarized the picture across studies. The effect, if any, has been measured against the noise.4
What it can tell you
How the intervention compares to doing something else — or nothing. Whether the effect is modest or large. Whether the certainty of the estimate is high, moderate, low, or very low.
What it cannot tell you
That every product on the market is the same as the one in the trial. Trial protocols are specific: a particular cell type, from a particular source, prepared a particular way, at a particular dose, for a particular condition, in a particular patient. A clinic that cites the trial may be doing something subtly — or completely — different.
What a clinic should not claim from this
That a trial of one product, in one indication, generalizes to their cash-pay version for a different indication. The evidence belongs to the exact thing that was tested.
What to ask
Is the trial they cite about this product, in this condition, with this route of delivery? Did the trial actually show a meaningful improvement in a patient-felt outcome — or a small change on a scan?
Band 4
A specific product is authorized — or carries guideline-level support — for a specific use.
A regulator has reviewed the data and authorized a named product for a defined indication, or a major clinical guideline (a national society, a Cochrane recommendation, a national health system) has endorsed its use for a defined population. Examples include FDA-approved hematopoietic stem-cell products for certain blood and immune disorders, and certain cartilage-repair products approved in Japan, Korea, or the European Union for narrowly defined surgical uses.5
What it can tell you
That, for this exact product and this exact indication, an external reviewer found the evidence strong enough to permit clinical use under defined conditions.
What it cannot tell you
That any clinic using cells from the same family, on a different condition, by a different route, is offering the same thing. The authorization is for the product as defined, the indication as defined, the patient group as defined. Outside that envelope, you are back in Band 2 or 3 — or, in some cases, off the ladder entirely.
What a clinic should not claim from this
That because some stem-cell product is approved somewhere, their stem-cell injection for your condition is too.
What to ask
Approved where? For what named product? For what condition? By what route — surgical implant, joint injection, IV infusion? For which patients? Under what restrictions?
“Approved” still needs a second question.
The word approved does a lot of work in clinic marketing, and not all of it is honest. An approval is always attached to a thing — a named product, a defined condition, a defined patient group, a defined route of delivery, under defined restrictions. Strip any of those out and the same word stops meaning the same thing.
When you see “approved” or “approved abroad” in a brochure, it is fair to ask:
- Approved where — by which regulator?
- Approved for which exact product, by which manufacturer?
- Approved for which condition — and is it your condition?
- Approved for which route — surgical implant, joint injection, topical application, IV infusion?
- Approved for which patient group — and do you fall inside it?
- Approved under what restrictions — and is the clinic offering it operating under those restrictions?
A cartilage-repair product approved in Japan, Korea, or the EU is a meaningful regulatory decision, made by reviewers reading the same kind of trial data the FDA reads. It is not a justification for a different clinic's cash-pay knee injection for a different condition. FDA non-approval, in parallel, matters for U.S. readers — but it does not automatically mean the underlying biology is useless, and it does not automatically mean a foreign approval is suspect. Permission abroad can be real. It just doesn't travel farther than the product, the indication, and the setting it was granted for.1
How to read what a clinic actually tells you.
Clinic copy tends to lean on a small set of phrases that sound stronger than they are. None of them are inherently dishonest. They just leave most of the work to the reader. The honest question underneath each one is usually some version of: which exact product, for which exact condition, with which exact evidence, reviewed by whom?
“Backed by science.”
Backed by which science — a cell-biology paper, an animal study, a clinical trial, a systematic review? In what indication? For this product or a related one?
“Published studies.”
Published where? In what kind of journal — peer-reviewed, indexed, with a real editorial process? How many patients? What did the studies actually measure, and what did they find?
“Clinical trial.”
A registered trial is a protocol, not a verdict. Is it registered on a public database? Is it actively recruiting, or finished? Was there a comparison group? Did the result support the marketing claim being made now?3
“Patient results.”
How many patients in total — including the ones who did not improve, the ones lost to follow-up, the ones who asked for a refund? Was the outcome something a patient could feel, or only a number on a scan or a blood test?
“Used internationally.”
Used by whom, where, in what setting? Is the international use inside a regulated, authorized indication — or is it the same cash-pay model in a different country?
“Not FDA-approved, but available abroad.”
Available abroad for this exact condition, under authorized use? Or available in the sense that the local regulator does not actively prohibit it? Both are technically true; only one is meaningful.
“Doctor-developed protocol.”
Developed by which physician, with what training, reviewed by which independent body? An in-house protocol is not the same as a peer-reviewed protocol, and a peer-reviewed protocol is not the same as an authorized treatment.
The bigger the promise, the bigger the burden.
How strong the evidence needs to be is not a fixed number. It shifts with the decision in front of you. A low-cost, low-burden, locally available intervention with a modest claim can sit comfortably on early evidence. A high-cost, high-burden, far-from-home intervention with a sweeping claim cannot.
Evidence requirements should rise when:
- the procedure is expensive,
- travel — especially international travel — is involved,
- cells are expanded in a laboratory before being given back,
- donor tissue, rather than your own, is used,
- delivery is intravenous rather than directly into the target tissue,
- repeated, ongoing treatments are recommended,
- or the claim is anything in the family of disease reversal, tissue regeneration, fertility restoration, immune reset, anti-aging, or neurological recovery.
The bigger the promise and the bigger the burden, the stronger the evidence should be.
This is especially worth holding onto for medical-tourism decisions. A guideline that sounds reasonable at home — a small trial, a positive case series — has to carry the additional weight of a flight, a hotel, a foreign clinic, an unfamiliar regulator, and the absence of your usual physician in the weeks after the procedure. The same evidence supports less, the farther you have to travel to receive it.6
How we talk about uncertainty here.
When you read other pages on this site, you will see careful phrasing — and we want to be honest about what that phrasing means. None of these are verdicts. They are descriptions of where the evidence currently sits.
- Early but plausible — the biology is interesting, the human evidence is small, and a careful person can reasonably wait for more.
- Interesting biology, limited human evidence — there is a real reason researchers are studying it, but treatment-level claims are getting ahead of the data.
- Available abroad, evidence still developing — another regulator has permitted some form of use; what that use is, and how closely it matches what a clinic is selling, is its own question.
- Approved for a narrow indication, not the broader claim — yes, something is approved; no, the approval does not cover the marketing copy beside it.
- Worth discussing with a qualified physician — the option exists, the trade-offs are real, and this is not a decision to make from a brochure or a sales call.
- Not enough evidence to rely on as a main treatment — there may be a role in a research setting or alongside standard care, but treating it as the primary plan would be outrunning what is known.
None of these phrases is a recommendation, and none is an attack. They are an attempt to describe a moving field honestly — to leave room for hope where hope is reasonable, and to leave room for caution where caution is reasonable too.2
Where to go from here.
If you want to keep reading, these pages go deeper on the method, the sources, and the practical questions to bring to a clinic conversation.
How CellDecide reads a paper and writes a page →
The full source list, by category →
Stem cells vs PRP vs exosomes vs BMAC, in plain English →
What this is not.
- Not medical advice. Nothing here is tailored to your case. The job of this page is to help you read evidence, not to prescribe.
- Not a clinic recommendation. We do not name clinics as places to go or to avoid. Where clinics appear on this site, it is to discuss claims and regulatory context, not to refer.
- Not anti-treatment. Some patients will find that the evidence, in their specific case, supports trying something. Some will not. Both are reasonable conclusions to reach.
- Not a substitute for a clinician. Use this page to ask sharper questions, then take those questions to a physician who can read your records and examine you.
For the long version of how we source, see our methodology. For the current state of clinician review on this site, see our medical-review policy.
Sources & footnotes
- U.S. Food & Drug Administration. “Important Patient and Consumer Information About Regenerative Medicine Therapies.” fda.gov · consumer-information page distinguishing FDA-approved stem-cell products from marketed claims, and outlining how the agency reviews investigational products. Verified 2026-05-14. ↩
- International Society for Stem Cell Research. “ISSCR Guidelines for Stem Cell Research and Clinical Translation.” isscr.org · the Clinical Translation chapter sets out how emerging cell-based interventions should be communicated to patients, including the limits of autologous-origin and early-signal reasoning. Verified 2026-05-14. ↩
- U.S. National Library of Medicine. “ClinicalTrials.gov public registry.” clinicaltrials.gov · registration and results database for clinical studies. A listing reflects registration, not endorsement, approval, or proof of efficacy. Verified 2026-05-14. ↩
- Cochrane Musculoskeletal Group. Systematic reviews of platelet-rich therapies for musculoskeletal soft-tissue injuries. cochranelibrary.com · meta-analyses of PRP for knee osteoarthritis and adjacent indications, reporting low-certainty evidence with substantial heterogeneity and meaningful sham-comparator effects. Verified 2026-05-14. ↩
- International Society for Stem Cell Research. “A Closer Look at Stem Cell Treatments — Patient Handbook.” closerlookatstemcells.org · ISSCR patient-facing handbook on what is approved, what is investigational, and how country-by-country authorization for specific cell-based products differs from a general “stem cell” claim. Verified 2026-05-14. ↩
- Kuriyan AE, Albini TA, Townsend JH, et al. “Vision Loss after Intravitreal Injection of Autologous ‘Stem Cells’ for AMD.” New England Journal of Medicine2017; 376:1047–1053 · documented case series of severe vision loss after a non-homologous stem-cell procedure at a clinic operating outside trial oversight; cited here as a reminder that the gap between sweeping claim and supporting evidence has real downside. Verified 2026-05-14. ↩