Blog

Mesenchymal Stem Cells for Knee Osteoarthritis: Where the Trials Stand

Asare B. Christian, MD, MPH · 2026-11-19

Direct Answer

In small randomized trials, mesenchymal stem cell injections for knee osteoarthritis produce modest short-term improvements in pain and function scores, with effects generally comparable to less expensive injections. Placebo-controlled trials show smaller differences than open-label ones. No trial has convincingly demonstrated cartilage regrowth or a change in the disease course.

Why This Question Is Harder Than It Looks

Knee osteoarthritis research has three chronic problems that make the stem cell literature difficult to interpret. Trials use different cell sources, bone marrow, adipose, umbilical cord, and synovial. They use different doses, different preparation methods, and different comparator groups, including exercise, hyaluronic acid, corticosteroid, and, less often, placebo.

Pain and function scores improve in nearly every osteoarthritis study, including trials of placebo injection. That is not cynicism; it is what the data show, and it is why open-label improvement means very little.

What the Better Trials Report

The most informative trials share a design: randomization, a comparator a patient could plausibly accept, and follow-up of at least twelve months.

Across those, the consistent pattern is improvement in patient-reported outcomes at three to six months, with gradual attenuation. When compared against hyaluronic acid or corticosteroid, differences are small and often not statistically significant. When compared against saline, the effect is present but modest.

Imaging outcomes are the weak point. Several trials report cartilage thickness or MRI appearance changes, and almost none report differences that would be clinically recognizable to a patient. Regeneration, in the sense of new functional cartilage, remains unproven in human trials.

Why Comparisons to Other Injections Are Awkward

A physician asked to compare stem cells to PRP or to corticosteroid is being asked a question the trials answer badly, because the effect size of all three overlaps with the natural variation in this condition.

What is defensible: corticosteroids have the strongest short-term evidence and a documented tradeoff with repeated use. PRP has reasonable controlled evidence of modest benefit. Cell injections have weaker, more heterogeneous evidence at substantially higher cost.

What is defensible clinically: choose the least invasive, least expensive intervention with reasonable evidence for the stage of disease, and escalate only with a defined measure of success.

The Case Where It Might Matter

There is a legitimate research question about whether cell therapy changes the trajectory of early-stage disease differently than symptom control. That is the question worth funding, and it is being studied, including with rigorous placebo designs.

A patient should understand the difference between participating in that question through a well-designed trial and paying out of pocket for an unproven answer delivered as if it were settled.

What I Look For in a Candidate

Reasonable patients for any injection-based approach have mild to moderate radiographic disease, preserved joint space, no mechanical symptoms such as locking or gross instability, and a willingness to do the loading work that follows.

Poor candidates have advanced multicompartment disease, are aiming to avoid surgery indefinitely rather than delay it, or have expectations of cartilage regrowth.

Cost and Framing

Patients frequently pay several thousand dollars for a series of injections with no guaranteed outcome. A clinician who cannot tell you the cell dose, the source, the viability assay, and their own outcome data is not offering a measured intervention.

There is a question of honesty here that matters more than the biology: is the patient buying a plausible treatment with uncertain benefit at a price that reflects that uncertainty, or are they buying confidence?

Limitations

Trials are small, heterogeneous, and frequently industry-affiliated. Publication bias toward positive results is likely. Outcome measures are patient-reported in most, with objective imaging outcomes inconsistently captured. Long-term safety beyond two years is largely uncharacterized for intra-articular cell products.

Frequently Asked Questions

Will it regrow cartilage? No human trial has convincingly demonstrated that.

How does it compare to a cortisone shot? Corticosteroids have stronger short-term evidence and a cost of repeated use. Cell injections have weaker evidence at higher cost.

Is one injection enough? Most protocols use one or more, and repeat injections are commonly recommended. That should be factored into cost and expectation.

Can it delay knee replacement? Some patients report enough improvement to delay surgery. Whether that is attributable to the injection or to the rest, rehabilitation, and time that accompany it is difficult to separate.

References

  1. Meta-analyses of mesenchymal stem cell therapy for knee osteoarthritis, 2019 to 2024.
  2. OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis.
  3. American Academy of Orthopaedic Surgeons clinical practice guideline on management of osteoarthritis of the knee.
  4. FDA guidance on cell and tissue-based products: minimal manipulation and homologous use.

Educational information only. This is a draft for physician review and is not individualized medical advice.