Two intravenous injections, 14 days apart, and a 90-day scorecard: that is the shape of the first randomised, placebo-controlled trial of stem cells for feline arthritis published in full. Of 35 cats with X-ray-confirmed osteoarthritis, 7 of 9 on the low dose and 10 of 12 on the high dose met the owner-rated success definition at day 90. So did 4 of 11 cats that received saline.

What the six clinics actually measured

The study ran at six US practices and screened 47 cats; 37 were randomised and 35 treated — 10 at 5 million cells, 13 at 20 million, 12 on saline. Mean age was 12.7 years (range 4–19) and mean weight 5.6 kg. Seventy-one percent of the cats had at least three joints involved and 43% had four; the stifle was affected in 68.6%, the hip in 65.7%, the elbow in 60%.

Owners filled in a Client-Specific Outcome Measure, a list of that individual cat's daily activities scored for difficulty. The most flagged activities at baseline were difficulty jumping up (37%), reduced play (18%) and difficulty climbing (16%); visible lameness accounted for only 4.6%. Treatment success was pre-defined as a fall of at least 2 points with no single activity worse than baseline.

The day-90 numbers

Outcome at day 905 million cells20 million cellsSaline
Owner activity score (CSOM)7/9 (78%)10/12 (83%)4/11 (36%)
Owner overall assessment7/9 (78%)10/12 (83%)4/11 (36%)
Veterinarian pain score6/8 (75%)10/12 (83%)5/11 (45%)
Veterinarian overall assessment4/8 (50%)8/12 (67%)2/10 (20%)

The confidence intervals are wide — 45% to 96% for the low-dose activity score — because each group held roughly ten cats. Success accumulated rather than appearing after the first injection: the low-dose group moved from 50% at day 30 to 80% at day 60, the high-dose group from 64% to 83%.

The placebo arm is the part to read twice

A 36% success rate in cats injected with salt water is not a footnote. It is the baseline every arthritis treatment has to beat, and caregiver placebo is well documented in feline pain trials, which the authors concede. The same paper names the licensed comparator: in the frunevetmab field study, 76% of treated cats were activity-score successes at day 84, and 68% of placebo cats were too. A 42-point gap between cell-treated and saline cats, as here, is a different order of signal from an 8-point gap. It is also one pilot of 35 cats, funded by the company that makes the cells.

What was in the syringe

The cells were allogeneic, grown from uterine tissue collected during a routine spay from a single screened donor, expanded to passage six, frozen at 20 million viable cells per 2 mL vial in DMSO and shipped overnight on dry ice. They were negative for CD34 and CD45, positive for CD44 and CD105, and negative for MHC class II. Nothing was harvested from the treated cat, no anaesthetic was needed, and the dose went into a vein rather than into a joint.

That is a different product from the autologous fat course sold in clinics, where fat is collected under anaesthesia, cultured for 2–4 weeks and reinjected. An off-the-shelf vial can be described by cell count and passage; a fat sample cannot.

Side effects: 10%, 38%, 17%

Events judged possibly or probably treatment-related over 90 days occurred in 1 of 10 cats at 5 million cells (10%), 5 of 13 at 20 million (38%) and 2 of 12 on saline (16.7%). Within 24 hours the usual signs were transient nausea or vomiting, weakness, lethargy, hypersalivation, fast breathing and an odd smell to the breath; nystagmus appeared in one cat in each cell group.

One serious event in the high-dose group was judged treatment-related: lethargy, nausea and drooling after dose one, then weakness, rapid breathing, vocalising and nystagmus during dose two. That cat was treated with dexamethasone, maropitant, ondansetron and oxygen and recovered fully. Blood counts and chemistry stayed stable or improved in all three groups. The authors suggest the DMSO used to freeze the cells, not the cells themselves, may drive some reactions — one reason the 5 million dose is the one being taken forward.

“Eligible for expanded conditional approval” is not approval

In April 2026 the manufacturer announced that the FDA's Center for Veterinary Medicine had granted eligibility for the expanded conditional approval pathway. That pathway, created by Congress in 2018 for drugs outside the minor-use route, requires safety plus a reasonable expectation of effectiveness, is renewed annually and runs at most five years before the company must show full effectiveness or withdraw. Eligibility means an application may be filed. It does not mean a product exists: no stem cell therapy is licensed for cats in the United States today, and a one-year continuation study of the 5 million dose is still enrolling.

What it costs, and what the alternative costs

  • Autologous fat course at a private clinic: USD 2,500–6,000, GBP 2,000–5,000 or CNY 12,000–40,000, with no licensed product behind the price.
  • Frunevetmab injection, approved 13 January 2022: USD 75–115 per monthly dose, plus the visit.
  • Trial participation: owners usually pay screening X-rays and travel; the injections are free.
  • Off-the-shelf allogeneic price: not set. Any figure quoted now is a guess.

Who was excluded, and what to ask

The trial excluded end-stage joints (bridging bone, collapsed joint space), unstable joints, cruciate rupture within 6 months, cancer, neurological disease, FeLV or FIV positivity, and any cat that had already received stem cells. NSAIDs stopped 7 days before enrolment, monoclonal antibodies 30 days, steroids 14 days. In plain terms: a cat with moderate, multi-joint arthritis and no other major illness.

If cell therapy is offered to you, ask for the cell count, the passage number, the donor screening protocol, whether DMSO is in the formulation, and which scoring instrument will decide success. Then ask what the plan is at 6 and 12 months; this evidence stops at 90 days.

If your cat is stiff this month

Radiographs are still the entry test, but they under-read. In cats over 12 years old, X-rays show arthritis of the limbs in only 26% while degenerative joint disease of some type is present in 90%; in cats over 6 years, 61% have arthritis in at least one joint, 48% in more than one, and 82% of cats over 14. That is why a mobility score and a phone video of the jump onto the counter often carry as much weight as the film. Weight control, pain relief through the standard treatment ladder, ramps and low-sided trays come first, and the monthly antibody injection is the licensed drug option. Cells are, for now, the experimental one.

This article is for information only and does not replace examination, imaging and diagnosis by a veterinarian. Stem cell products for cats are not licensed in the US, the UK or the EU; check the local regulatory position before any money changes hands.

Terms mentioned in this article

FIV

Frequently asked

Can I get stem cell therapy for my cat's arthritis today?

Not as a licensed product. No feline stem cell therapy is approved in the US, UK or EU. Clinics do sell autologous fat-derived courses, which are unlicensed and priced by the clinic, and a few universities and practices recruit cats into trials. The 2026 trial described here used an off-the-shelf allogeneic product that has only been granted eligibility to apply for expanded conditional approval, not approval itself.

How much does it cost compared with the monthly injection?

An autologous fat course runs about USD 2,500-6,000 (GBP 2,000-5,000; CNY 12,000-40,000) and buys a process with no licensed product behind it. Frunevetmab, approved in January 2022, costs roughly USD 75-115 per monthly dose plus the visit, so a year of the licensed drug lands in the same range as one unlicensed cell course. Trial injections are usually free but you pay for screening X-rays.

How long does the effect last?

The published data stop at 90 days, and success rates were still climbing at day 60 with no sign of reversal by day 90. A one-year continuation study of the 5 million cell dose is running, so duration beyond three months is unknown. Owners should plan a mobility re-check at 3, 6 and 12 months rather than assume one course is permanent.

Related guides