OATS / mosaicplasty

Osteochondral autograft transfer moves cylindrical plugs of bone topped with living hyaline cartilage from low-weight-bearing areas of your own knee — the trochlear margins, the intercondylar notch — press-fit into the defect. One plug is OATS; several arranged together is mosaicplasty. Unlike marrow stimulation or cell therapy, the defect is resurfaced with mature hyaline cartilage on day one: the biology to get right is bony integration and surface congruity, not tissue regeneration.

The randomized evidence

A trial of 57 young athletes with femoral condyle defects randomized to OAT or microfracture is the flagship. At mean 37 months: 96% versus 52% excellent-or-good results, and 93% versus 52% returned to pre-injury sport. The same cohort at mean 10.4 years: OAT still superior, failure 14% versus 38%, and 75% versus 37% maintaining their pre-injury activity level. Both arms declined after year 3 — even the winner is not a cure.

Caveats: this is a single-center trial of 57 young athletes. Generalization to older or lower-demand patients is unsupported, and the effect size has not been independently replicated.

Long-term durability

Two systematic reviews frame the long game. Across 610 patients at mean 10.2 years: large score gains (IKDC +42.4, Lysholm +21.1) but 28% failure and 19% reoperation — older age, prior surgery, and larger defects predicted failure. Across 495 patients at mean 15.1 years: return to sport 86–100%, conversion to knee replacement 0–16%, with sustained score improvements. Both pool observational studies, with the survivor and publication bias that implies. Durability is better than microfracture, not perfect.

The donor site: the honest number

The originator's series of 831 patients reported good-to-excellent results in 92% of femoral condyle cases and claimed 3% donor-site morbidity. An independent systematic review of 1,726 patients tells it straighter: pooled donor-site morbidity was 5.9% when treating knee defects and 19.6% when harvesting the knee to treat the ankle — mainly patellofemoral pain and crepitus. The honest number is 6–20%, not 3%. You are trading a problem in one part of the joint for a smaller, usually tolerable one in another.

Who it suits

Single-stage, no cell culture, cheap relative to MACI — with a hard ceiling: donor tissue is finite, so the practical limit is about 2–4 cm². Beyond that, donor-site cost outweighs benefit, and large defects point to allograft or MACI instead. For athletes, this procedure has the fastest documented return to play among cartilage repairs — typically 6–9 months — with the strongest randomized return-to-sport data.

It is also technically demanding. Plug height mismatch or surface incongruity degrades results; surgeon volume matters.

Questions for your surgeon

• Is my defect within the 2–4 cm² window where autograft makes sense?
• How many of these do you do a year — and how do you handle plug congruity?
• What donor-site symptom rate do your patients actually report?
• If I am also a MACI candidate, why this one? (No adequately powered modern RCT compares them in the overlap zone.)

Safety

Standard knee surgery risks plus the donor site: patellofemoral pain, crepitus, rarely donor-site fracture. No allograft disease risk, no cultured-cell regulatory issues. Open questions include whether backfilling the donor holes reduces morbidity, and the long-term fate of the interface between plugs, where dead space fills with fibrocartilage.