PRP for Knee Osteoarthritis: What the Evidence Actually Shows, and What It Doesn’t

By Nikhil Verma, MD, fellowship-trained interventional spine and sports medicine physician, Essential Sports and Spine Solutions, Columbus, OH

Earlier this month a patient sat in my office with a knee that was bone on bone. He wasn’t asking me for a miracle. He was asking me a fair question that almost nobody had answered straight for him: does PRP actually work for this, or is it just an expensive shot?

So let me answer it the way I’d want it answered for my own family. Platelet-rich plasma can meaningfully improve pain and function for a lot of people with knee osteoarthritis, and for some it buys real quality of life for months at a time. It also does not regrow cartilage, and anyone promising you a rebuilt knee from an injection is telling you something the science does not support. Both of those things are true at the same time. That tension is the honest version, and it’s the version I think you deserve.

I’m an interventional spine and sports medicine physician here in Central Ohio, and I treat knee OA with orthobiologics every week. Here’s what I actually know.

What is PRP, really?

PRP is your own blood, concentrated. We draw a sample, spin it down to isolate the platelets, which carry the growth factors your body already uses to repair tissue, and then place that concentrate precisely into the joint. You’re not introducing anything foreign. You’re amplifying what’s already there.

That’s it. No donor cells, no lab-grown anything, nothing marketed as “millions of stem cells.” Just a concentrated version of your own biology, delivered where the problem is.

Does PRP work for knee osteoarthritis?

For symptoms, the honest answer is: often, yes, but not for everyone, and the quality of the PRP matters enormously.

Multiple meta-analyses have found PRP superior to hyaluronic acid for knee OA pain over 6 to 12 months, and a 2025 meta-analysis in The American Journal of Sports Medicine found the improvement clinically significant and influenced by how concentrated the platelets were. That last part is the whole ballgame, and I’ll come back to it.

Now, the part a lot of clinics conveniently skip. A large placebo-controlled trial published in JAMA in 2021, known as the RESTORE trial, compared PRP to a saline injection over 12 months and found no meaningful difference in pain or in cartilage volume on MRI. Both groups improved. That’s a real result from a rigorous study, and I’m not going to pretend it doesn’t exist.

Here’s how I hold both of those things. The evidence is genuinely mixed, and a big reason for the mixed signal is that “PRP” is not one thing. The preparation in a negative trial can be a fraction of the dose used in a positive one. When the studies with better outcomes are separated out, they tend to share something in common: a higher delivered platelet dose. So the question isn’t really “does PRP work.” The better question is “what kind of PRP, at what dose, for which patient.”

Why dose and preparation are not small details

Work published in Scientific Reports pointed to roughly 10 billion total platelets as an important threshold for sustained benefit in moderate knee OA out to a year, and a 2025 systematic review in Arthroscopy found that greater platelet doses tended to produce better clinical outcomes. Leukocyte-poor preparations, meaning fewer inflammatory white cells, also appear better tolerated inside a joint.

Translate that out of research language: if your PRP doesn’t reliably deliver a high enough dose of the right composition, you may be getting a shot that was never going to do much in the first place. This is why I don’t think it’s fair to say “PRP failed for me” until you know what was actually in the syringe.

At my practice this is not left to chance. We use systems built to deliver a high, consistent platelet yield in the range the dosing literature and leading orthobiologics groups like the Interventional Orthobiologics Foundation and The Orthobiologic Institute have been pointing to, and we make preparation decisions based on your specific pathology rather than a one-size-fits-all protocol. There’s more to that story, including what we do with the plasma most practices throw away, and I’ll get to it below.

We don’t do blind injections

Here’s something that follows directly from everything above. A perfect dose of the perfect preparation does nothing if it doesn’t land where it’s supposed to.

That sounds obvious, but a lot of injections are still done “blind,” by feel and surface landmarks alone. The data on that is not reassuring. Landmark-guided knee injections land in the joint somewhere in the range of 77 to 82 percent of the time across systematic reviews, which means roughly one in five misses the target. Add image guidance and accuracy climbs to around 95 percent, and studies consistently show better outcomes when the injection actually reaches where it’s aimed.

So every procedure I do is guided. For joints and most soft-tissue work, that’s ultrasound, which lets me watch the needle in real time and confirm the biologic is going exactly where the pathology is. For deeper spinal and certain bony targets, that’s fluoroscopy, live X-ray, often with contrast to verify placement before anything therapeutic goes in. Same principle in both cases: I want to see it, not guess at it.

When you’ve spent the effort to prepare a high-quality, correctly dosed biologic, delivering it precisely is not an upgrade. It’s the whole point. Precision is how the dose actually becomes a result.

What PRP does not do

It does not regrow cartilage. Once arthritis has thinned or worn the joint surface, an injection does not turn back that clock, and the placebo-controlled data on cartilage volume is pretty clear on that point. When you hear “regenerative,” read it as biological signaling that changes the environment inside the joint for a period of months, not as new cartilage.

I’m telling you this because it’s the truth, and because the clinics overclaiming on this are setting patients up to feel cheated. What PRP can realistically do is calm the joint, reduce pain, and improve how you function and move. For a lot of people, that’s the difference between sitting out their life and getting back into it. I just won’t sell it as something it isn’t.

The injection is one piece. The protocol is the point.

This is the part almost nobody talks about, and it’s the part I care about most.

The patients who do best are almost never the ones who got the “best” injection in isolation. They’re the ones who also addressed the terrain the injection lands in. Before we treat, we build a plan around diet, targeted supplements, sleep, stress, and how you load and move the joint. An inflamed, poorly-fueled, sleep-deprived system is a hard place for any biologic to work. A well-prepared one is a much better bet.

I’m not just doing an injection. I’m building a protocol. That framing isn’t marketing to me, it’s the actual mechanism of getting a good result.

PRP for knee OA in Columbus and Central Ohio

If you’re in Columbus, Dublin, Westerville, Gahanna, or anywhere in Central Ohio and you’ve been bounced between specialists or handed the same cortisone-then-surgery script, that’s exactly the gap this practice was built to fill. I look at your imaging myself. We talk through whether you’re actually a good candidate, because not everyone is, and I’ll tell you if you’re not.

If you want to go deeper on this, the specific protocols I use, how I think about system selection, and what I tell every patient after an injection, I cover all of that in my paid newsletter, Essentials of Healing. You can find it at nikhilvermamd.substack.com.

Frequently asked questions

Does PRP cure knee arthritis? No. Nothing injected cures arthritis or regrows worn cartilage. What PRP can do is improve pain and function for many people, often for several months to a year. That’s a symptom and function benefit, not structural repair, and any practice telling you otherwise is overselling.

How many PRP injections do I need for my knee? It depends on the severity and how you respond, but many patients are treated with one to three injections. I’d rather build the right plan for your knee than quote you a fixed package before I’ve looked at your imaging.

How long does PRP last for knee osteoarthritis? For most people who respond, relief tends to last somewhere in the range of 6 to 12 months. Higher-dose, higher-quality preparations tend to hold up longer, which is a big reason preparation matters.

Is PRP better than a cortisone shot? They do different jobs. Cortisone can be a useful tool for short-term flares, used deliberately and with informed consent, but its benefit is short and it isn’t something I reach for reflexively. PRP is aimed at a longer, more durable functional benefit. Which one fits depends on your situation.

Is PRP covered by insurance? Almost always no. PRP for OA is typically cash-pay. Part of the reason I’ve stepped away from routing this through insurance is that insurance economics tend to push the dose and quality down to a point where it stops being worth doing.

Am I too far gone for PRP if I’m bone on bone? Maybe, maybe not. Advanced arthritis is a harder case and I’ll be honest with you about expectations, but plenty of people with significant OA still get meaningful pain and function improvement. The only way to know is a real evaluation.

Do you use ultrasound or X-ray guidance for the injection? Yes, always. I don’t do blind injections. Joint and soft-tissue procedures are done under ultrasound so I can watch the needle in real time, and deeper spinal targets are done under fluoroscopy. Landmark-guided injections miss the knee joint roughly one in five times. Guidance takes accuracy to around 95 percent, and precise delivery is a big part of getting a real result.

Ready to talk through whether this is right for your knee? You can request an appointment at https://www.essentialsportsspine.com/request-appointment/.

This is for educational purposes only and does not constitute medical advice.