Can Stem Cells Regrow Cartilage? What MRI and Arthroscopy Studies Show

"Will it grow my cartilage back?" is probably the single most common question we hear from people with knee arthritis. It is an understandable hope. Cartilage is the smooth tissue that lets the joint glide, and once it wears thin it does not repair itself well. Surgery replaces the joint rather than restoring it, so the idea of an injection that rebuilds what has been lost is very appealing.

This article looks only at that question: not whether stem cells relieve pain, but whether they measurably restore cartilage. The two are often blurred together in marketing, and they should not be. We go through what the imaging and arthroscopy studies actually found, how strong that evidence is, and what it means if you are deciding whether to travel for treatment.

The quick answer

Possibly, in some patients, to a limited degree — but it has not been shown reliably. A 2020 meta-analysis of six randomized trials (203 patients) of culture-expanded MSC injections without additional surgery found improvement in pain at 6 to 12 months, but no significant difference in cartilage repair measured by MRI. The authors concluded that evidence on function and cartilage repair remains limited.[8]

Individual trials are more mixed. Some reported a smaller cartilage defect, better cartilage quality on specialised MRI, or less joint-space narrowing on X-ray in the higher-dose groups.[1][3][4] Others found pain and function improved with no detectable change on MRI.[6] The most striking structural results came from a study that combined cells with knee-realignment surgery, which is a different treatment from an injection on its own.[2]

If you have advanced, bone-on-bone arthritis, nothing in the current evidence suggests an injection will rebuild a normal joint surface.

Why "it hurts less" and "the cartilage grew back" are different questions

MSCs are thought to work mainly by signalling: they release molecules that dampen inflammation and influence the cells already in the joint. That can reduce pain and swelling without any change in how much cartilage you have. It is also why a patient can feel substantially better while their MRI looks the same.

A good example is a 2018 triple-blind, placebo-controlled trial by Emadedin and colleagues in Cytotherapy. Forty-three patients with knee osteoarthritis received either 40 million of their own bone marrow-derived MSCs or a saline injection. Over six months, the MSC group had significantly greater improvement in pain and function scores, with no major adverse events.[7] That is a meaningful symptom result from a well-designed trial — but it was not a study of cartilage regrowth, and it should not be quoted as one.

For most patients, symptoms are what matter day to day. The reason structure still matters is long-term: if an injection only calms the joint, the underlying wear continues, and expectations about "avoiding surgery forever" need to be set accordingly. Our comparison of stem cells vs knee replacement covers that decision in more detail.

How cartilage change is measured

When you read a claim that stem cells "regenerated cartilage", it helps to know which test was used, because they measure very different things:

• X-ray (joint-space width). Cartilage does not show on X-ray, but the gap between the bones is an indirect proxy for how much is left. It is crude and changes slowly.
• Standard MRI scoring systems such as WORMS and MOAKS. Radiologists grade cartilage, bone marrow and other structures across the whole joint. Useful, but small changes can fall within measurement error.
• Quantitative MRI such as T2 mapping, which estimates cartilage composition and quality rather than just thickness.
• MOCART scoring, designed to assess repair tissue filling a specific cartilage defect, usually after surgery.
• Arthroscopy and biopsy. Looking directly inside the joint, and examining a tissue sample under the microscope, is the most direct evidence — and the least commonly done, because it is invasive.

The type of tissue matters too. Normal joint cartilage is hyaline cartilage. Repair tissue is often fibrocartilage, which is tougher and less smooth. A finding of "hyaline-like" tissue on biopsy is therefore more meaningful than a finding that a defect has simply filled in.

What the trials found, study by study

A few points stand out. First, every individual trial here is small, with roughly 18 to 56 patients. Second, the most encouraging structural results tended to come with high cell doses or with surgery, and several trials compared MSCs against hyaluronic acid rather than a true placebo. Third, when the trials of injections alone were combined, the structural signal did not hold up as statistically significant.[8]

The strongest-looking result

The 2014 proof-of-concept trial by Jo and colleagues in Seoul is the study most often cited for cartilage regeneration. Eighteen patients received one of three doses of their own fat-derived MSCs. In the high-dose group (100 million cells), pain and function scores improved at six months, the size of cartilage defects decreased and cartilage volume increased on MRI, arthroscopy showed smaller defects, and biopsies showed thick, hyaline-like cartilage.[1] Those are genuinely interesting findings. They come from a small, early-phase study without a comparison group, and they have not been consistently reproduced in randomized trials since.

Cells plus surgery is a different treatment

Wong and colleagues in Singapore randomized 56 patients aged 55 or under, with bow-legged knees and cartilage defects, who were all having a high tibial osteotomy (a realignment operation) and microfracture. Three weeks after surgery, half also received an injection of cultured MSCs. At one year, the MSC group had significantly better MOCART cartilage repair scores, and better clinical scores over two years.[2] This supports MSCs as an add-on to surgery in a carefully selected group. It does not show that an injection alone will regrow cartilage in an arthritic knee.

So what can a stem cell injection realistically do?

Putting the evidence together, a fair summary is:

• Pain relief over 6 to 12 months is the outcome with the most support, though even that varies between studies and some larger analyses are less positive. We compare the evidence against a common alternative in stem cells vs PRP for knee arthritis.
• Slowing cartilage loss has been reported in some small trials, particularly at higher doses. It is a plausible benefit, not an established one.
• Visible regrowth has been documented in a few patients in early studies, mainly in focal defects rather than widespread wear. It is the least reliable outcome and should be treated as a possibility, not an expectation.
• Restoring a bone-on-bone joint to normal is not supported by any controlled evidence we are aware of.

Our longer guide to stem cells for knee osteoarthritis goes through the full outcome data, including timelines for when patients typically notice change.

Who may be a reasonable candidate

If structural benefit is possible at all, the published trials suggest it is most plausible in people who:

• Have mild-to-moderate osteoarthritis (roughly Kellgren-Lawrence grade 2 to 3) or a focal cartilage defect, confirmed on a recent MRI
• Still have meaningful cartilage left in the affected part of the joint
• Have reasonable knee alignment, or are addressing alignment separately with a surgeon
• Have tried exercise therapy, weight management and standard pain relief without enough benefit
• Are in reasonable general health and understand that pain relief, not regrowth, is the more realistic goal

How we assess knees, shoulders, hips and other joints is set out on our page on stem cell therapy for orthopedic injuries and joint conditions.

When this may not be right for you

We would advise against treatment, or at least against travelling for it, if:

• Your main goal is to regrow cartilage. If an injection would only feel worthwhile with a visibly rebuilt joint on your next MRI, the evidence does not support that expectation.
• You have end-stage, bone-on-bone arthritis with significant deformity or loss of motion. Joint replacement is the evidence-based option there.
• The knee locks or gives way because of a mechanical problem such as a loose fragment or a displaced meniscal tear, which needs a surgical assessment first.
• You have an active infection, active cancer, are pregnant or breastfeeding, or have a bleeding disorder that makes a harvest or injection unsafe.
• You are being promised a cure or guaranteed regrowth anywhere. No honest clinic can promise either.

Questions to ask if a clinic claims cartilage regrowth

• Which study are you basing that on, and did it use the same cell type, dose and technique you are offering?
• Was the cartilage change measured on MRI, arthroscopy or biopsy — and against a control group?
• Do you have before-and-after MRIs from your own patients, read by an independent radiologist?
• What proportion of your patients show no structural change?

The bottom line

Stem cells can, in some trials and some patients, produce measurable signs of cartilage repair or slower cartilage loss. That is a real and active area of research, and it is why the question is worth asking. But the pooled randomized evidence does not yet show reliable cartilage regrowth from an injection alone, and the strongest structural results come from small studies, high doses or combined surgery.

The most realistic reason to consider treatment is the chance of less pain and better function for a period, in a joint that still has cartilage to work with. If a follow-up MRI shows improvement, that is a welcome bonus rather than the promise. If you would like an honest view of where your own knee sits on that spectrum, send us your recent imaging and we will tell you plainly what we think is achievable — including when the answer is surgery.