Does PRP Work for Bone-on-Bone Knee Arthritis?
- 15 hours ago
- 6 min read
By Dr. Jeffrey Peng, MD · Published July 4, 2026 · 7 min read
Watch the Full Video
If you have bone-on-bone arthritis in your knee, you have probably been told that platelet-rich plasma (PRP) won't help you — that you waited too long, and that once the cartilage is this far gone, a knee replacement is your only real option. New research is challenging that assumption.
I'm Dr. Jeffrey Peng, a sports medicine physician practicing in the San Francisco Bay Area, and I've helped thousands of patients navigate orthobiologic treatments like PRP for knee osteoarthritis. A recently published randomized controlled trial did something almost no other PRP study had done: it measured what was actually happening inside patients' bodies after treatment. What those blood markers showed is the real reason I wanted to write this article — because it runs counter to what current guidelines would predict, and it changes how we think about treating advanced knee arthritis.
What Does “Bone-on-Bone” Knee Arthritis Actually Mean?
A healthy knee works like a smooth, well-oiled machine. The ends of the bones are capped with articular cartilage, which lets the joint glide with almost no friction. Osteoarthritis is the slow wearing away of that cartilage, and physicians grade it on a scale from 0 to 4: grade 0 is none, grades 1 and 2 are mild, grade 3 is moderate, and grade 4 is severe — so much cartilage has worn away that the space between the bones has collapsed on X-ray. That collapsed joint space is what people mean by “bone-on-bone.”
Here is a detail most patients are never told: much of the pain from arthritis does not actually come from the cartilage itself. Cartilage has no nerves. As the surfaces grind together, the joint sets off a chronic, low-grade inflammatory process — the synovial lining becomes inflamed, the joint swells, and over time the knee becomes hypersensitized. That underlying inflammation, not the worn cartilage alone, is a major driver of the pain.
Why Do So Many People Assume PRP Can't Help Severe Arthritis?
The reasoning sounds logical. PRP is thought to work by delivering concentrated growth factors that help tissue heal. But if the cartilage is already gone, the argument goes, there is nothing left to heal and no surface to regrow — so a knee replacement must be the only option. Several major organizations echo this position. The American Academy of Orthopaedic Surgeons has concluded that PRP is not recommended for severe, grade 4 knee arthritis, and the American Academy of Physical Medicine and Rehabilitation has stated that the evidence for severe disease is insufficient to make a recommendation.
What Did the New PRP Study Find?
The recently published trial tested PRP in exactly the patients guidelines say it shouldn't help. Ninety patients with symptomatic grade 3–4 knee osteoarthritis — all already on the waiting list for knee replacement surgery — were randomized into three groups. One group received two PRP injections one week apart, a second received a single corticosteroid (cortisone) injection, and a third took an oral anti-inflammatory medication. Researchers then followed them for six months.
The PRP group improved significantly more than either comparison group in pain, knee function, and stiffness, and used fewer opioids. Just as telling was who stayed the course: by the end of the study, no patients dropped out of the PRP group, while several patients in the cortisone and medication groups moved their surgery up and had their knees replaced early. These patients started from the same place — bone-on-bone arthritis with surgery already scheduled — yet six months later, the ones who received PRP were the ones who stayed out of the operating room.
What the Blood Tests Revealed: PRP Changed the Biology
This is what set the study apart. Instead of tracking symptoms alone, the researchers drew blood at baseline, three months, and six months and measured a panel of serum biomarkers — molecules that reflect cartilage breakdown and joint inflammation, as well as protective markers that rise when the body is calming inflammation down.
In the PRP group, the destructive markers fell. Cartilage-breakdown proteins such as COMP and MMP-3 dropped — COMP appears in the blood when cartilage is actively being damaged, so lower levels point to less breakdown. Key inflammatory drivers including IL-6, IL-18, and TNF-α decreased, as did CGRP, a marker tied to how much pain the nerves around the joint report. At the same time, protective and anti-inflammatory markers — including the anabolic growth factor TGF-β1, along with sTREM2 and sRAGE — went up. None of these shifts appeared in the cortisone or the oral medication groups, and several of the changes persisted through the six-month mark.
That combination matters. It suggests the relief patients felt wasn't just expectation or placebo — there were real, measurable changes moving the joint's environment toward a calmer, less inflammatory state. Of the three treatments, only PRP was able to change the underlying biology.
Does This Mean PRP Regrows Cartilage?
No — and this distinction is important. The study did not use MRI to assess whether cartilage grew back. It found that markers of cartilage damage slowed down, which is very different from cartilage regeneration. While a handful of small studies have suggested PRP might stimulate cartilage growth, we do not yet have solid, reproducible evidence that it can. What this and a growing body of research point to is more modest but still meaningful: PRP appears to slow the breakdown of cartilage and neutralize the chronic, low-grade inflammation inside an arthritic knee. By doing so, it can slow the progression of arthritis while reducing pain, improving function, and easing stiffness.
How Platelet Dose Affects PRP Results
There is one place where I think these already-encouraging results could have been even stronger: the dose. In the study, researchers drew about 30 cc of blood per treatment, which contains an estimated 5 billion platelets. Recent dosing guidance suggests aiming for at least 10 billion platelets to reach a therapeutic response — so these outcomes were achieved with roughly half of what the evidence suggests you want. In other words, this is likely the floor for what PRP can do in severe arthritis, not the ceiling.
That is quite different from what I do in my clinical practice. I draw about 120 cc of blood at each treatment — roughly four times the volume used in the study, which puts me in the neighborhood of 20 billion platelets, about double the minimum that guidelines recommend. For severe arthritis, I also perform two high-dose injections rather than one. Among the patients I've treated who have experienced both a lower dose and this higher-dose approach, nearly all prefer the higher dose, reporting better pain relief, better function, and less stiffness.
Is PRP a Replacement for Knee Surgery?
I still have an honest conversation with every patient. When arthritis is this advanced, the most likely long-term solution for lasting relief is still a knee replacement. PRP does not rebuild the joint and does not regrow cartilage. What it can offer is meaningful pain relief and improved function — and, just as importantly, time. That time lets patients get stronger before surgery and prepare for the operation on their own terms and timeline. For many people, regular PRP injections provide enough relief that surgery moves further down the road, and for some it moves much further than they expected.
So can PRP do anything once you've lost most of the cartilage in your knee? Based on this research and what I see in clinic, yes — not by rebuilding the joint, but by controlling the inflammation that drives so much of the pain. If you'd like to discuss whether PRP is a reasonable option for your knee, you're welcome to schedule a consultation or explore more articles in my knee arthritis library.
References
1. Lacko M, Awad O, Matúška M, et al. Intra-articular platelet-rich plasma demonstrates superior clinical and serum biomarker outcomes compared with corticosteroids and NSAIDs in late-stage knee osteoarthritis: a randomised controlled trial. Journal of Orthopaedic Surgery and Research. 2026. doi:10.1186/s13018-026-07013-w. (via PubMed, PMID 42231445)
Medical Disclaimer: This content is for educational purposes only and does not substitute for the medical advice of a physician. Always consult your healthcare provider before beginning any new treatment program. The information presented reflects the opinion of Dr. Jeffrey Peng and does not represent the views of his employers or affiliated hospital systems.
Comments