Adipose-derived cells (MFAT / SVF)

Two distinct products get sold under the same "fat-derived cells" banner. Microfragmented adipose tissue (MFAT) is lipoaspirate mechanically washed and resized — Lipogems is the best-known device — retaining intact stromal niches. Stromal vascular fraction (SVF) is a cell pellet isolated by enzymatic or mechanical digestion. The proposed action for both is trophic, anti-inflammatory signaling from perivascular cells, not engraftment or cartilage regrowth: the best comparative RCT found no imaging change after injection.

The closest thing to a placebo-controlled win

Richter 2025 randomized 75 patients to MFAT, corticosteroid, or saline. MFAT beat saline on all primary pain and function outcomes across one year, while the steroid's advantage lasted only 2–6 weeks. That makes it the strongest single result in the fat-derived literature — with caveats that matter: the trial was partially blinded, arms were modest, and the lead author reports equipment and supplies from Lipogems. Independent replication does not yet exist.

Equivalence to PRP is the main finding

Zaffagnini 2022 randomized 118 patients to a single MFAT injection or PRP and followed them 24 months: both improved significantly, no between-group differences, and no structural change on imaging. A meta-analysis of 6 RCTs reaches the same conclusion — comparable pain and function benefit through 12–24 months, both usually reaching the clinically important threshold, with a small MFAT edge at 6 months only. Since PRP itself shows only small, short-lived benefit over placebo, equivalence to PRP is a modest claim.

The large trial cuts the other way. In Mautner 2023 (480 patients, phase 3), the SVF arm — like BMAC and umbilical-cord cells — was not superior to a single corticosteroid injection at 12 months.

The one SVF trial with an imaging signal, Hong 2019, reported cartilage improvement on MRI scoring — in 16 patients, using a self-controlled bilateral design, observer-dependent scoring, and an enzymatic preparation that is not lawful point-of-care practice in the US.

The regulatory gray zone

FDA regards enzymatically digested SVF as more-than-minimally-manipulated — a drug requiring a biologics license; clinics selling it have received warning letters, and federal courts upheld FDA's position in 2021. MFAT devices are 510(k)-cleared for processing adipose tissue, but marketing MFAT injections as a "stem cell treatment" for OA exceeds that clearance and has drawn FDA untitled letters. Nothing fat-derived is FDA-approved as a knee OA therapeutic — patients are paying cash inside an enforcement-discretion gap.

Practical notes

All cited RCTs used a single injection; MFAT volume is typically 5–10 mL from a ~100–150 mL lipoaspirate, and SVF cell doses are rarely standardized. The procedure adds liposuction — donor-site bruising, rare hematoma or infection — and typical US cash prices run in the thousands. Nothing in the trials justifies choosing MFAT over PRP on efficacy; the argument would be single-visit convenience for PRP non-responders, which is untested.

Safety

No serious adverse events in the cited RCTs, including the 480-patient trial; expect donor-site soreness and a transient knee flare. The larger hazards are financial and the unregulated-clinic setting — the SVF gray market outside these trials has documented infections and vision loss.

What would change the tier

Independent, non-device-funded replication of the MFAT-versus-saline result; any structural endpoint success in an adequately powered trial; and a head-to-head MFAT-versus-BMAC readout.