Cartilage regrowth explained: what a mouse study found, and why it is not yet a treatment for arthritis

In old mice, a month of daily injections of an experimental drug thickened worn knee cartilage and restored proteins typical of young cartilage. Small pieces of human cartilage, kept alive in a dish for a week, moved in the same direction. No person with arthritis has been given the drug for their joints.

Those are the facts behind a headline that circulated this week asking whether joint replacements were on the way out. The research, from Stanford University, was published in the journal Science in November 2025, and an account of it was republished by ScienceDaily on October 6. It is a serious piece of laboratory work on why cartilage fails to repair itself. It is no basis for anyone to change how arthritis is treated, and the record of earlier attempts explains why.

Why cartilage does not regenerate on its own

Osteoarthritis is the gradual breakdown of the smooth cartilage that lets the ends of bones glide over each other. By one global estimate, 595 million people had osteoarthritis in 2020, and the knee was the joint most often affected. The same analysis states the problem flatly: there is no effective cure for osteoarthritis.

Care is aimed at symptoms. Education, exercise and weight loss are cornerstones of management, according to a 2021 review in the Journal of the American Medical Association, with pain relievers alongside. For advanced disease, total joint replacement effectively relieves pain.

What nobody has is a medicine that rebuilds the tissue. Aging or injury to the joints can lead to cartilage degeneration, the Stanford authors write, and for the disease that follows there are no FDA approved disease-modifying drugs. The FDA is the US Food and Drug Administration, and disease-modifying means acting on the damage itself and not just on pain.

Cartilage is a poor healer. Other tissues are restocked by stem cells, the body’s reserve of unspecialized cells. Efforts to find a comparable reserve inside joint cartilage have not succeeded.

What the mouse study of cartilage regrowth did

The Stanford group came to cartilage from muscle. In earlier work it had identified an enzyme called 15-PGDH as a hallmark of aged tissues, including skeletal muscle. The enzyme destroys prostaglandin E2, a signaling molecule that supports tissue repair. Blocking the enzyme in old mice increased aged muscle mass, strength, and exercise performance, as that paper put it. The group coined a word, gerozyme, for enzymes that rise with age and wear tissues down.

The new study asked whether the same thing happens in joints. It does, at least in mice. Abundance of 15-PGDH in cartilage of aged mice was twice that in young mouse joints, comparing animals aged 24 months with animals aged 4 months.

Three experiments followed.

Old mice. A cohort of aged male mice was treated daily for a month with a drug that blocks the enzyme, injected into the abdomen so that it reached the whole body. Untreated old mice had thin, frayed knee cartilage. In treated mice the cartilage was thicker and intact. Damage scores, assigned by assessors who did not know which animals had been treated, were lower. Tests for three characteristic proteins indicated the new tissue was hyaline cartilage, the smooth kind that joints need, and not the tougher, scar-like fibrocartilage.

Injured mice. In young mice, the researchers ruptured a knee ligament by mechanical compression. The ligament was the anterior cruciate ligament (ACL), the one often torn in sports. Left alone, these mice developed osteoarthritis within four weeks. The drug was injected straight into the knee, administered twice a week for 4 weeks starting 1 week after the injury. Six weeks after the injury, the treated knees showed less cartilage breakdown.

The team also looked for signs of pain, which in a mouse must be inferred. Treated animals placed the injured paw on the ground more like uninjured mice, reacted less to touch and tolerated more pressure on the knee.

Human tissue. Cartilage was collected from patients undergoing total knee replacement, tissue that would otherwise have been discarded. Discs four millimeters across were kept in culture and treated for one week. They accumulated more of the molecules that give cartilage its cushioning, and treated samples exhibited increased cartilage stiffness, a sign of firmer, more load-bearing tissue, not of a stiff joint. Inflammation was marginally reduced, with three of the signaling molecules measured falling and the others not.

How the cartilage regrew without stem cells

The surprise was in the mechanism. In muscle, bone and blood, the same pathway had previously been found to work by prompting stem cells to multiply. Cartilage turned out to behave differently. The researchers sorted the cartilage cells, called chondrocytes, into groups by which genes were switched on.

In old mice, treatment changed the mix. One group, which made the enzyme and carried a genetic signature of cartilage breakdown, dropped from 8% of cells to 3%. A second, tied to fibrocartilage, fell from 16% to 8%. A third, the cells that build smooth cartilage and its supporting matrix, the mesh of proteins around the cells, went the other way. Their share increased from 22% to 42%.

The change did not seem to come from new cells. The shift, the authors write, appears to result from changes in existing cells rather than proliferation and expansion of a stem or progenitor population, progenitor cells being the offspring of stem cells. Old cells were behaving like younger ones.

“We were looking for stem cells, but they are clearly not involved,” said Helen Blau, a Stanford professor of microbiology and immunology and one of the two senior authors.

The other, Nidhi Bhutani, an associate professor of orthopedic surgery, described the scale of the effect. “Cartilage regeneration to such an extent in aged mice took us by surprise,” Bhutani said.

Why mouse cartilage results often stall before reaching arthritis patients

Arthritis research has been here before. A 2013 review in Nature Reviews Rheumatology, written by two Australian specialists, set out why promising animal results in this field so often go nowhere.

Most animal experiments, they noted, produce arthritis through surgical or mechanical disruption of joint biomechanics in young individuals, meaning young animals, which is not how the disease arises in most older people. Arthritis triggered that way may work differently at the molecular level from the slow, age-related kind, which might explain the poor translation from preclinical to clinical trials. Preclinical means laboratory and animal work done before testing in people.

The Stanford study partly answers that objection. It used an injury model in young mice and also naturally aged mice, and saw a benefit in both: regrowth in the old mice and less breakdown in the injured ones.

A second lesson comes from a drug that got much further. Sprifermin, a growth factor injected into the knee, is under investigation as a disease-modifying osteoarthritis drug. In a trial of 549 people, it did what it was designed to do. After two years, cartilage at the highest dose was 0.05 millimeters thicker than with placebo.

Patients did not feel the difference. Scores for pain, stiffness and function were no better than with placebo at any dose. The trial’s authors called the gain in cartilage statistically real but of uncertain clinical importance. Durability of response also was uncertain.

Thicker cartilage, in other words, is not the same as a better knee. Any drug that follows will have to show both.

There is one place where a preventive drug would matter a great deal if it worked. A review of knee injuries found that 10 to 20 years after a torn ligament or meniscus, the knee’s shock-absorbing pad, about half of patients have osteoarthritis with associated pain and functional impairment: the young patient with an old knee. It also found a lack of evidence to support a protective role of repair or reconstructive surgery against that outcome.

What stands between the mouse study and a cartilage treatment

The animals. Only male mice were used for these experiments, the authors write of the injury work, and the aging experiment used males too.

The timescale. Treatment lasted about four weeks in both experiments. Whether regrown cartilage lasts, and what happens with longer dosing, was not tested.

The human evidence. It consists of cartilage discs in a dish for a week, taken from knees already damaged enough to be replaced. A dish carries no body weight and has no immune system.

Safety. Prostaglandin E2 is active throughout the body and has been implicated in inflammation and pain, as Blau acknowledges. Raising it for months in people with arthritis has not been studied.

The route to a trial. Blau said that “Phase 1 clinical trials of a 15-PGDH inhibitor for muscle weakness have shown that it is safe and active in healthy volunteers”. An inhibitor is a drug that blocks its target, and Phase 1 is the earliest stage of human testing. Those trials concerned muscle, and Blau described a cartilage trial as a hope for the future.

Who stands to gain. Blau, Bhutani and other coauthors are inventors on patent applications held by Stanford University covering the approach, which are licensed to a company called Epirium Bio. Blau is a cofounder of Myoforte/Epirium and holds equity and stock options in the company. The research itself was funded mainly by the US National Institutes of Health and several foundations.

Stanford’s account carries a caution that the headline over it on ScienceDaily does not. The results do not yet establish that the treatment can regrow cartilage or prevent osteoarthritis in people, it says, and clinical trials specifically testing cartilage regeneration will be needed to determine whether the approach is safe and effective for patients.

Whether and when a joint replacement is right for someone is assessed by their doctor, and nothing in this study bears on that today.

An experimental drug regrew knee cartilage in old male mice and nudged human cartilage samples in the same direction, and it has yet to be tested in the joint of a single person with arthritis.