Treatments and gadgets. 9 minute read.
PRP and stem cell injections for knee arthritis: what the guidelines say in 2026
By the PocketPhysio editorial team. Published 2026-10-01.
What the September 2026 guideline says
The American College of Rheumatology announced an updated osteoarthritis guideline on 14 September 2026, replacing the 2019 guideline it wrote with the Arthritis Foundation. Its summary document, approved by the ACR Board of Directors on 8 September 2026, settles both treatments on this page in one line each. "In people with knee and hip OA, we strongly recommend against PRP treatment", graded strong, with low certainty of evidence. "In people with knee and hip OA, we strongly recommend against stem cell injections", graded strong, with moderate certainty. They sit in the same block as the strong recommendations against glucosamine and chondroitin.
What the document recommends is led by exercise: "In people with knee and hip OA, we strongly recommend exercise", strong, moderate certainty, with weight loss also strong, also at moderate certainty, for people meeting criteria for overweight or obesity. Not every strong recommendation there is a behavioral one, and it would be wrong to tell you otherwise: the list also strongly recommends intraarticular glucocorticoid injections, meaning a steroid injection into the knee or hip, graded strong with low certainty. That is a different treatment from the two on this page, and reporting it is not the same as offering it to you as the alternative; whether any injection has a place is for you and the doctor who would give it. PRP and stem cells are nowhere on that recommended list. The only recommendation the summary document marks as new is a conditional one about GLP-1 receptor agonists in people who have knee osteoarthritis and obesity.
A word on those two gradings, because they are easy to misread. Strength describes the recommendation and certainty describes the evidence behind it. A strong recommendation against an intervention means the panel judged that almost everyone, given the same evidence, would decline it. It rests on net benefit, which takes in cost and burden as well as harm, so it is not the same thing as a safety alert about one particular risk. The PRP line is graded strong with low certainty, which tells you the panel's confidence is in the recommendation rather than in the trials behind it.
A conditional recommendation says the opposite: that reasonable people given the same facts would choose differently, so it belongs in a conversation.
Do PRP injections for the knee work?
Platelet-rich plasma is made from you. A tube of your blood is spun in a centrifuge so the platelets concentrate into a small volume of plasma, and that plasma is injected into the knee. The theory is that platelets carry growth factors, and growth factors do things to tissue. It is a reasonable theory, which is not the same as a result.
The result that matters most came from Australia. The RESTORE trial recruited 288 community-based adults in Sydney and Melbourne, aged 50 or over with painful knee osteoarthritis graded 2 or 3 on X-ray, and gave half of them three weekly injections of leukocyte-poor PRP and half three weekly injections of saline. Nobody involved knew which injection was which, including the people assessing the results; 93% finished the trial, and 97% of each group received all three injections. After 12 months knee pain had improved by 2.1 points in the PRP group and 1.8 in the saline group on an 11-point scale, a difference of 0.4 points (95% confidence interval minus 0.9 to 0.2) against the 1.8 points the trial had set in advance as the smallest change a person would notice. Medial tibial cartilage volume on MRI fell by 1.4% with PRP and 1.2% with saline (Bennell and colleagues, 2021).
Both groups got better, and that is the part people skip. The saline group improved by 1.8 points over the year, which shows how much an arthritic knee can improve with nothing active in the syringe at all. Symptoms swing up and down, people book an appointment during a bad patch, and being examined and listened to has its own effect. A study that only compares before with after cannot separate any of that from the treatment, which is why the comparison group decides what a trial is worth, and why a clinic's audit of its own patients tells you very little.
The honest counterweight: pooled analyses that also include trials with other comparison treatments do find a small average benefit. One from 2025 brought together six randomized trials in 1,162 people with mild to moderate knee osteoarthritis, in which PRP was compared with hyaluronic acid, a steroid injection or placebo. It reported a standardized mean difference of minus 0.32 for pain (95% confidence interval minus 0.48 to minus 0.15) and minus 0.28 for function (minus 0.44 to minus 0.12) at 6 and 12 months, with moderate variation between the studies. The same review stated that long-term structural improvement was not demonstrated and that larger trials are still needed (Nawaz and colleagues, 2025). An effect of about 0.3 is at the small end by the usual thresholds, and that is a fair summary of the literature: small, inconsistent, with nothing structural behind it.
Stem cell injections, and what is actually in the syringe
The name oversells the contents. A 2023 systematic review of mesenchymal stem cell injection for knee osteoarthritis graded 1 to 3 found 12 studies covering 539 patients and 576 knees. Seven of them used stromal vascular fraction taken from fat, four used bone marrow aspirate concentrate, and one used both. All were taken from the patient and injected with minimal processing, without being cultured or expanded in a laboratory, so what goes in is a mixed cell preparation in which stem cells are a minority ingredient nobody counts (Kyriakidis and colleagues, 2023).
That review reported pain improving across every included study, and it concluded that a single injection is a safe, reliable and effective option. Read its methods before you read its conclusion. Five of the 12 studies were randomized trials and seven were not, and only one of the 12 compared the injection against a placebo injection, with 13 knees in each arm. The authors wrote that pooling the results was inappropriate "due to the heterogeneity of the included studies and the significant risk of bias", and that the variation in cell sources, in how the preparation is made and in how it is given makes the treatment "not reproducible".
Four of the 12 had no comparison group at all. Studies showing that pain improves after an injection are exactly the evidence that RESTORE existed to test, and no equivalent of RESTORE appears in this literature for stem cells in a knee.
Regulation is the other half of the picture. The US Food and Drug Administration says the only approved stem cell products are blood-forming cells derived from umbilical cord blood, for disorders of blood production, and states that "none of these products have been approved for the treatment of any orthopedic condition, such as osteoarthritis, tendonitis, disc disease, tennis elbow, back pain, hip pain, knee pain, neck pain, or shoulder pain". On its patient information page about regenerative medicine therapies, the agency says it "has received reports of blindness, tumor formation, infections, and more" from the use of these unapproved products. Those harms came from clinics, not from trials, and they are the reason this page is blunt.
Why two clinicians give you two different answers
Because their professional bodies read the same evidence differently, and both are being honest. The American Academy of Orthopaedic Surgeons, in its 2021 guideline on managing knee osteoarthritis without surgery, says that "platelet-rich plasma (PRP) may reduce pain and improve function in patients with symptomatic osteoarthritis of the knee". It grades that Limited, which in its system means one or more low quality studies with consistent findings, or a single moderate quality study. Limited is the lowest of its three strengths that rest on evidence, with only a Consensus statement, meaning no supporting evidence, below it.
The same document says the panel downgraded this recommendation two levels, from Strong, because the evidence was inconsistent. Five years later the American College of Rheumatology read a similar literature and strongly recommended against the same treatment. Both panels were looking at small trials using different preparations, and they weighed that inconsistency differently.
The surgical side goes further in its newer documents. A 2025 review of orthobiologic guidance describes the AAOS technology overview on PRP and the European ESSKA consensus as both supporting its use mainly in mild to moderate knee osteoarthritis, while differing from each other on how much is injected, on how many injections make a course and on which preparation to use, and agreeing that evidence for structural change such as cartilage regeneration remains limited (Winkler and colleagues, 2025). For stem cell injections there is no such split to report: the AAOS guideline makes no recommendation on them at all, so between those two bodies the rheumatology recommendation against is the only position on the table.
So the disagreement is about symptoms, not structure. Nobody in that argument claims an injection rebuilds a joint. If a clinic tells you otherwise, that claim is outside what any of these bodies will support.
What to ask before you agree to an injection
Who pays is worth settling before anything else. Aetna's clinical policy bulletin 0784 lists platelet-poor and platelet-rich plasma injection for osteoarthritis of the hip, knee and temporomandibular joint, and bone marrow derived mesenchymal stromal cells for osteoarthritis, among procedures it considers experimental, investigational or unproven, a category the bulletin defines as covering approaches whose effectiveness has not been established. Policies differ between insurers and between countries, so check your own rather than assuming.
Three questions change the conversation in a clinic. What exactly is being injected, and was it grown in a laboratory or taken and injected on the same day? In the evidence you are being shown, what did the other group get, a placebo injection or nothing? And what would count as this not having worked, and what happens then?
Whether to have any injection at all is a decision for you and the doctor who would be giving it, and the guideline positions above belong in that conversation. Never stop or change a prescribed medicine on your own.
What the same guidelines put first
Exercise and weight management are the recommendations that survived every revision, and they are the two that nobody sells to you. Our knee osteoarthritis program is the staged version of that work. It begins with lying and sitting exercises for a painful knee and builds to stairs and single-leg work, and each exercise page carries its own dose, which your physio adjusts to your knee. If the arthritis is in your hip instead, the hip osteoarthritis program does the same job, and a knee that twisted and now catches or locks may be a meniscus tear rather than arthritis, which is a different problem with a different plan.
None of that is glamorous and it takes months. The trade is worth making: the benefit from strength work stays with you between appointments, it needs no equipment to keep going, and exercise is one of the strong recommendations in the 2026 guideline summary rather than one of the arguments.
When to see a physio (physical therapist)
This is the appointment to make first. A single assessment tells you whether the pain is coming from the joint surface, the muscle around it or somewhere else entirely, and that is the thing an injection cannot sort out for you. Bring what you have already tried, how long you managed it for, and what you want to get back to doing.
Come sooner if the knee gives way, if pain is keeping you awake, or if six to eight weeks of regular exercise has changed nothing. If you have already had an injection somewhere and the knee has become hot and swollen, that is not a question for a physio: the warning list below sets out how fast to act.
See a doctor promptly if
- Same day: the knee is hot, red and swollen, or you have knee pain with a high temperature or feel hot, cold or shivery. Get medical advice the same day, and go to an emergency department if you feel very unwell. This can be a sign of infection in the joint, which needs treatment quickly. If you have a very high or very low temperature, shivering you cannot control, you are breathing very fast or finding it hard to breathe, you are confused or your speech is slurred, your skin, lips or tongue look blue, gray, pale or blotchy (on brown or black skin this can be easier to see on the palms of the hands or soles of the feet), or you have a rash that does not fade when you press it, call emergency services, as this can be sepsis.
- Same day: the knee swells up quickly and a lot, with no clear reason. Get medical advice the same day.
- Same day: the knee locks and you cannot straighten it. Get medical advice the same day.
- Same day: a calf or thigh that is newly swollen, warm, tender, red or darker than usual, has swollen veins that are sore to touch, or has a throbbing or cramping pain that does not fit your injury or condition, or that feels different from normal muscle soreness after exercise. This can be a sign of a blood clot (deep vein thrombosis), especially after surgery, an injury, time in a cast or boot, a long trip or a spell of being much less mobile than usual. Get medical advice the same day. If you are also short of breath or have chest pain, call emergency services.
- Within a few days: the pain is there at night or at rest and keeps getting worse. See your doctor within a few days. If you have had cancer, now or in the past, or are losing weight without trying, see your doctor sooner, within a day or two, and mention it. This applies even if the pain is not getting worse. If you are being treated for cancer now, contact your cancer team the same day.
Common questions
Do PRP injections for knee osteoarthritis work?
A large, properly blinded, placebo-controlled trial says no. The RESTORE trial randomized 288 adults aged 50 and over with mild to moderate medial knee osteoarthritis in Sydney and Melbourne to three weekly injections of leukocyte-poor PRP or three weekly injections of saline, and nobody involved knew which was which. At 12 months the PRP group's knee pain score had fallen by 2.1 points on an 11-point scale and the saline group's by 1.8, a difference of 0.4 points when the trial had set 1.8 points as the smallest difference a person would notice. Of 31 prespecified secondary outcomes, 29 showed no significant difference between the groups (Bennell and colleagues, 2021). Reviews that also pool trials with other comparison treatments do find a small average benefit, which is why the subject is still argued about, and the 2026 American College of Rheumatology guideline summary strongly recommends against PRP for knee and hip osteoarthritis.
How many PRP injections are needed for a knee?
There is no settled number, because the protocols differ between clinics and between trials. The RESTORE trial gave three injections at weekly intervals. A 2025 review of orthobiologic guidance found that the two surgical society documents it compared suggest different courses from each other, and differ again on how much is injected and on which preparation to use (Winkler and colleagues, 2025). Nobody has shown which number works best, and the two documents do not agree with each other on it. We will not suggest a number for you: how many, and whether any at all, is a question for the doctor who would be giving them, and the 2026 American College of Rheumatology guideline summary strongly recommends against PRP for knee and hip osteoarthritis.
Do stem cell injections regrow knee cartilage?
No study has shown that, and the injections sold under that name usually contain no cultured stem cells at all. A 2023 systematic review of 12 studies covering 539 patients and 576 knees found that every one of them used fat-derived stromal vascular fraction, bone marrow aspirate concentrate, or both, which are mixed cell preparations taken from you and injected the same day without being grown in a laboratory (Kyriakidis and colleagues, 2023). The US Food and Drug Administration puts it bluntly: none of these products is approved for any orthopedic condition, including osteoarthritis. Cartilage repair is the claim to be most careful about, because it is the one no injection trial has delivered.
How much do PRP injections cost for a knee?
We do not publish prices, because they differ by country, and then by city and clinic. What is more useful is how payers class the treatment. Aetna's clinical policy bulletin 0784 lists platelet-poor and platelet-rich plasma injection for osteoarthritis of the hip, knee and temporomandibular joint, and bone marrow derived mesenchymal stromal cells for osteoarthritis, among procedures it considers experimental, investigational or unproven, a category it defines as covering approaches whose effectiveness has not been established. Policies differ between insurers and between countries, so check your own. Before cost comes into it, the 2026 American College of Rheumatology guideline summary strongly recommends against both treatments for knee and hip osteoarthritis, and that is the part to take to your doctor.
Are PRP and stem cell knee injections safe?
The short-term safety record in the published studies is reasonable, and the harms sit elsewhere: a 2025 pooled analysis of six randomized trials in 1,162 people with mild to moderate knee osteoarthritis reported that the side effects recorded were minor and self-limiting (Nawaz and colleagues, 2025). Of the 12 stem cell studies in the 2023 systematic review, three reported any complication at all: one listed temporary weakness in 2 knees, injection site pain in 1 knee, minor bleeding from the aspiration site in 1 knee and redness or swelling in 3 knees, and the other two each reported swelling in 1 knee after the injection (Kyriakidis and colleagues, 2023). The serious reports the US Food and Drug Administration has collected came from unapproved products sold outside trials, and the agency says they include blindness, tumor formation and infections; any injection into a joint can introduce infection. If a knee becomes hot, red and swollen, or you have knee pain with a high temperature or feel hot, cold or shivery, get medical advice the same day, and go to an emergency department if you feel very unwell. Safe is not the same as useful: the 2026 American College of Rheumatology guideline summary strongly recommends against both of these treatments for knee and hip osteoarthritis.
References
- American College of Rheumatology. 2026 update of the American College of Rheumatology recommendations for the management of osteoarthritis of the knee, hip, and hand. Guideline summary, approved by the ACR Board of Directors on 8 September 2026. Read 1 October 2026. https://assets.contentstack.io/v3/assets/bltee37abb6b278ab2c/bltb3d12c34020da842/oa-guideline-summary-2026.pdf
- American College of Rheumatology. American College of Rheumatology updates osteoarthritis guidance, reinforcing personalized, comprehensive care. Press release, 14 September 2026. https://rheumatology.org/press-releases/american-college-of-rheumatology-updates-osteoarthritis-guidance-reinforcing-personalized-comprehensive-care
- American Academy of Orthopaedic Surgeons. Management of osteoarthritis of the knee (non-arthroplasty) evidence-based clinical practice guideline. Published 31 August 2021. https://www.aaos.org/oak3cpg
- Bennell KL, Paterson KL, Metcalf BR, et al. Effect of intra-articular platelet-rich plasma vs placebo injection on pain and medial tibial cartilage volume in patients with knee osteoarthritis: the RESTORE randomized clinical trial. JAMA. 2021;326(20):2021-2030. doi:10.1001/jama.2021.19415
- Nawaz HMT, Jawwad MA, Khan MU, et al. Efficacy of platelet-rich plasma injections in knee osteoarthritis: a systematic review and meta-analysis. Cureus. 2025;17(10):e94288. doi:10.7759/cureus.94288
- Kyriakidis T, Pitsilos C, Iosifidou M, et al. Stem cells for the treatment of early to moderate osteoarthritis of the knee: a systematic review. Journal of Experimental Orthopaedics. 2023;10:102. doi:10.1186/s40634-023-00665-1
- Winkler T, Oehme S, Hildebrandt A, et al. Evidence-based guidelines on orthobiologics. EFORT Open Reviews. 2025;10(6):345-351. doi:10.1530/EOR-2025-0069
- US Food and Drug Administration. Consumer alert on regenerative medicine products including stem cells and exosomes. Content current as of 9 April 2024. https://www.fda.gov/vaccines-blood-biologics/consumers-biologics/consumer-alert-regenerative-medicine-products-including-stem-cells-and-exosomes
- US Food and Drug Administration. Important patient and consumer information about regenerative medicine therapies. Content current as of 8 April 2024. https://www.fda.gov/vaccines-blood-biologics/consumers-biologics/important-patient-and-consumer-information-about-regenerative-medicine-therapies
- Aetna. Blood and adipose tissue derived products for selected indications. Clinical Policy Bulletin 0784. Accessed 1 October 2026. https://www.aetna.com/cpb/medical/data/700_799/0784.html
Written and checked by the PocketPhysio editorial team. Last updated 2026-10-01.
This page is general education, not a diagnosis or a personal treatment plan. See a physiotherapist or doctor for advice about your own situation.