On this page
This is the deepest evidence base in the whole of regenerative medicine marketing, which makes it the fairest place to see what “good evidence” actually looks like in this field. The knee osteoarthritis studies include randomised controlled trials, a network meta-analysis and a 2025 systematic review — and they still do not settle the question.
What is being asked
Whether injecting mesenchymal stromal cells into an osteoarthritic knee reduces pain and improves function more than the alternatives, and whether it changes the joint itself.
Those are two separate questions and they have different answers. Most trials measure the first. The second is measured less often and answered less clearly.
A third question sits underneath both: whether any benefit lasts long enough to matter in a condition that runs for decades.
Worth stating at the outset: none of the trials below compared cells against knee replacement, and none followed patients long enough to say whether an operation was deferred or merely delayed. Those are the questions patients most often ask and the literature least often answers.
The pooled evidence
A 2025 systematic review and meta-analysis of randomised controlled trials is the most current synthesis — Efficacy and safety of mesenchymal stem cells in knee osteoarthritis: a systematic review and meta-analysis of randomized controlled trials.
A 2024 network meta-analysis went further and compared cell types against each other rather than each against control — Transplantation of three mesenchymal stem cells for knee osteoarthritis, which cell and type are more beneficial?.
That second design matters. Most patients are not choosing between cells and nothing; they are choosing between cell types, doses and clinics, and a network analysis is the only way to address that with pooled data.
Both reviews carry the same structural caveat: they pool trials that differ in cell source, dose, preparation and follow-up, and that heterogeneity limits how much weight a summary figure can bear.
The individual randomised trials
A phase III randomised trial examined intra-articular autologous adipose-derived MSCs, reporting on efficacy and safety — Clinical Efficacy and Safety of the Intra-articular Injection of Autologous Adipose-Derived Mesenchymal Stem Cells for Knee Osteoarthritis.
An earlier randomised trial used umbilical cord-derived cells and compared dosing schedules, finding repeated dosing outperformed a single dose — Umbilical Cord-Derived Mesenchymal Stromal Cells for Knee Osteoarthritis: Repeated MSC Dosing Is Superior to a Single MSC Dose.
A further trial injected autologous adipose tissue-derived cells — Intra-Articular Injection of Autologous Adipose Tissue-Derived Mesenchymal Stem Cells — and an older study used cultured bone marrow-derived cells in varus knees with cartilage defects — Injectable cultured bone marrow-derived mesenchymal stem cells in varus knees with cartilage defects.
What was measured
Almost all of these trials use patient-reported scores — pain and function questionnaires completed by the patient.
Those instruments are validated and widely used, and they are subjective by design. That is not a flaw, but it does mean blinding matters enormously, and blinding is hard when one arm involves a liposuction or a marrow aspiration.
A patient who has undergone a harvest knows they were in the treatment group. That knowledge is not dishonesty, it is unavoidable, and it inflates reported benefit in every trial where it happens.
A 2025 prospective randomised study looked specifically at whether cartilage changes in early disease — Cartilage Regeneration Potential in Early Osteoarthritis of the Knee: A Prospective, Randomized study — which is the structural question rather than the symptom one.
How long anyone was followed
Typically six to twenty-four months. A small number of studies extend further.
Against a condition that progresses over twenty or thirty years, that is a short window. It is long enough to see whether symptoms improve and far too short to see whether the joint’s trajectory changed.
Anyone quoting durability beyond the follow-up period of the trials is extrapolating, and should say so.
This cuts both ways. Short follow-up does not show that benefit stops — it shows that nobody looked. The honest statement is that durability is unknown, which is less satisfying than either optimism or dismissal.
The placebo problem
Knee injection trials have a well-described and substantial placebo response. The act of having a needle put into a painful joint by an attentive clinician produces measurable improvement on its own.
This is why a trial comparing cells against baseline is nearly uninformative, and a trial comparing cells against a saline sham is worth a great deal more.
When reading any claim about this treatment, the first question is what the comparison group received — see what the evidence levels mean.
Where this leaves a patient
The honest position is that there is real randomised evidence of symptom benefit in appropriately selected patients, with meaningful uncertainty about effect size, durability and structural change.
That is considerably better than most of what is sold in this field, and considerably weaker than the language used to sell it usually implies.
The practical translation is that a well-selected patient has a reasonable chance of feeling better for a year or two, an uncertain chance of that lasting, and little expectation of the joint itself being different. Whether that is worth the price is a judgement, not a medical fact, and it is yours to make.
It also sits alongside exercise therapy, which has evidence at least as good and costs almost nothing — see the knee osteoarthritis page.
Related reading
Cartilage regeneration research covers the structural question in more detail.
MSC dosage research covers why doses vary so widely between these trials, and autologous versus allogeneic covers the cell-source choice.
MSC therapy describes the cells; the research library indexes everything else.
Frequently asked questions
Is there randomised evidence?
Yes, several trials, plus a 2025 systematic review and meta-analysis and a 2024 network meta-analysis. It is the strongest evidence base in this field.
Does it change the joint or just the symptoms?
Mostly symptoms in the published work. Structural change is measured less often and demonstrated less consistently.
How long does benefit last?
Trial follow-up is usually six to twenty-four months. Beyond that nobody has good data, and claims of permanence are extrapolation.
Why does the placebo comparison matter so much?
Knee injections produce a large placebo response. A trial without a sham comparison cannot separate the treatment from the needle.
Which cell type performed best?
The 2024 network meta-analysis addresses exactly this. It found differences, which is why “stem cell therapy” as a single category is not a useful description.
Is this better than exercise therapy?
No trial has shown that. Exercise has evidence at least as good, and is the option most often skipped before people consider injections.
Requesting a medical evaluation
Nothing on this page establishes whether any treatment is appropriate for you. That needs your history, your imaging and your current medications read by a clinician. If you would like that review, send your records and we will tell you honestly whether there is anything worth discussing — including when the answer is no.
This page is general information, not medical advice, and does not create a doctor–patient relationship. Regenerative treatments discussed here are in most cases investigational. Discuss any treatment with a clinician who knows your history.