Orthopedic Regenerative Medicine for Knee Arthritis: What the Evidence Actually Shows

"Orthopedic regenerative medicine" is one of the most common things people in our area search for when their knee hurts, and there are a lot of clinics that will happily sell you an injection. Surgery for arthritis should still be a last resort, after conservative treatment has failed. By the time a patient gets to a knee replacement with me, I want them to have tried to manage the arthritis with many different treatments, so that we know they have tried everything before surgery. For this reason, I would be very interested in an injection that could delay the need for surgery. Unfortunately, I see many patients who have had stem cell injections and paid a significant amount of money, only to be left with the same knee pain and less money. In fact, every injection we have today, including stem cells, is pain control. None of them regrows cartilage or reverses arthritis.

This post walks through what the controlled trials show for stem cells and for PRP (platelet-rich plasma), why cartilage is so hard to grow back, and why PRP is the regenerative injection our group offers.

What "regenerative medicine" usually means

When a clinic advertises regenerative medicine for the knee, it usually means one of a few injections. PRP is made from your own blood, spun down to concentrate the platelets. Bone marrow aspirate concentrate (BMAC) is drawn from the pelvis and spun down the same way. Fat-derived injections use liposuction and either process the fat on the same day or grow the cells in a lab. Amniotic and umbilical cord products and exosomes come from donor tissue. The same-day bone marrow and fat injections are often called "stem cell" shots, but true stem cells make up only a small fraction of what is injected.

The FDA states that no regenerative medicine product (stem cells, exosomes or fat-derived cells) has been approved to treat any orthopedic condition, including knee arthritis¹.

What the stem cell trials show

The best summary is a Cochrane review that pooled 25 trials and 1,341 patients². Compared with a placebo injection, stem cells may improve pain by about 1 point on a 10-point scale at 6 months, and function by about 14 points out of 100. The authors rated the certainty as low, most trials were small, and none of them measured whether the arthritis progressed. Up to three larger trials were also stopped without publishing results.

The largest U.S. trial randomized 480 patients to bone marrow concentrate, same-day fat cells, donor umbilical cord cells or a cortisone shot³. At one year, none of the cell injections relieved pain better than cortisone, and no group showed a change in the arthritis on MRI.

The newest well-designed trial, ADIPOA2, was published in the Annals of the Rheumatic Diseases⁴. Patients with mild to moderate arthritis had liposuction, their fat cells were grown in a lab for about two weeks, and they received either a low dose or a high dose of their own cells or a saline injection. Patients in the saline group even had a sham liposuction so nobody knew which group they were in. At 6 months, 47% of the stem cell patients were strong responders, compared with 55% of the saline patients. There was no difference in pain or function at any point through one year, and no improvement in cartilage on MRI in the patients who had it measured.

That last trial says a lot about why patients feel better after these shots. About 70% of the people in both groups improved, including the people who got salt water⁴. A separate analysis of eight placebo-controlled trials estimated that half to two-thirds of the improvement after a stem cell injection comes from the placebo effect and the natural ups and downs of arthritis⁵. The trials that do show a benefit are usually funded by the company that makes the product, and even those have not shown cartilage growing back when compared with placebo on MRI^6,7.

Why an injection can not grow back cartilage

When cartilage starts to thin as arthritis progresses, many other things in the knee have already changed. The alignment has changed, and the forces through the joint surface have become abnormal. The meniscus is often torn or worn, and bone spurs and changes in the shape of the joint surfaces mean the two sides no longer fit together smoothly. The thought that an injection can make stem cells stick to that surface and regrow cartilage is just not the reality. When we do a cartilage replacement surgery like MACI, we have to be very careful to get down to the bone and remove a layer called the calcified cartilage, which links the cartilage to the bone. Without this special preparation the new cartilage will not adhere properly. It is one of the most important parts of that operation, and it obviously can not be done with an injection. The surface where we would be hoping the cartilage regrows does not have the ability to hold on to new cells and start the regrowth process. Cartilage also has no blood supply, and injected cells do not stay in the knee for long, which the authors of the ADIPOA2 trial point to as one reason the shots do not work⁴.

When I do restore cartilage, it is surgery. MACI (cells grown from your own cartilage) or an osteochondral allograft works on a contained defect, and they only last if the alignment, the meniscus and the kneecap tracking are corrected too. You can see an example in my kneecap dislocation case, where the cartilage repair was only part of the operation. The cartilage research in Japan and at Stanford is real and exciting, but it is still in the lab and in early trials.

Why we offer PRP

Of the biologic injections, PRP has the most and best controlled evidence. A meta-analysis of 18 placebo-controlled trials with nearly 2,000 patients found that PRP improved function for a full year, and high-platelet PRP (over 1 million platelets per microliter) relieved pain through 12 months⁸. The national orthopedic guideline lists PRP as an option for knee arthritis⁹. A trial comparing bone marrow concentrate with PRP found no difference between them at one year¹⁰, and bone marrow concentrate costs much more and requires a needle into the pelvis.

The evidence is not perfect. A large Australian trial found that PRP was no better than saline¹¹, and the rheumatology guideline recommends against both PRP and stem cell injections¹². I think the preparation explains a lot of the mixed results. Reviews that separated PRP by white blood cell content found the best pain and function results with leukocyte-poor PRP^13,14, although few trials have compared the types head to head. That is why our group uses a high-platelet, leukocyte-poor PRP. The injections are done by my partner Dr. David Lessman, a non-operative sports medicine physician, under ultrasound guidance. In addition, I am clear with patients that PRP is a treatment for pain and function. It does not regrow cartilage.

PRP works best in mild to moderate arthritis, before the knee is bone-on-bone. It can be given as a single shot or as a series of three. I am honest with patients that the results are still a little mixed. Some people respond well and others do not, and unfortunately we do not know until we try. Even so, there are lots of patients who have had significant pain relief for many months from a PRP injection. The visit takes about an hour, and you stay off anti-inflammatory medications for 5 days before and 2 weeks after. It is usually self-pay, since most insurance plans still consider it investigational. It also works best alongside the things that take load off the knee: strengthening, weight loss when it applies, and sometimes a brace.

If you are weighing an injection for knee arthritis, come see me or Dr. Lessman. We will look at your X-rays and MRI together, talk about where your knee stands, and decide whether PRP, another non-surgical option or something else makes sense.

This article is general education, not medical advice for your situation. Every knee and hip is different — talk with your own doctor about your care.