Who is a realistic candidate for this procedure

OATS and mosaicplasty suit a specific patient profile, and understanding that profile early saves time and avoids unnecessary disappointment.

The typical candidate is a younger, active adult — published registry data place the mean age at around 27 to 35 years — with a focal, full-thickness cartilage defect on the femoral condyle. The defect is usually caused by osteochondritis dissecans (OCD), where a fragment of cartilage and underlying bone becomes unstable or detached, or by a traumatic knock or twist that strips cartilage from the joint surface. Lesions graded ICRS III or IV — meaning cartilage loss extends to, or through, the subchondral bone — represent the typical surgical entry point.

Location matters as well as depth. The medial femoral condyle accounts for roughly two-thirds of defects treated with osteochondral procedures in the published literature. Applications on the patella carry a poorer mechanical fit due to contact-pressure geometry, and tibial plateau defects are procedurally difficult because perpendicular graft access is harder to achieve; both sites have a more limited evidence base.

Diffuse or multi-compartmental osteoarthritis sits outside the scope of these techniques entirely. They treat focal damage in an otherwise serviceable joint — not generalised wear. Patients who have not completed a structured course of physiotherapy, typically at least three months, are generally asked to do so before surgical options are confirmed.

Confirming suitability requires a consultant assessment with appropriate imaging, including MRI to map defect depth, size, and subchondral involvement.

How defect size shapes the surgical plan

Defect area is the single most decisive variable in planning an osteochondral autograft procedure. Before surgeon preference, patient age, or lesion location, the measured size of the cartilage defect determines which approach — if any — is viable.

Three size tiers guide the decision:

• Under ~2 cm²: A single OATS plug, typically 8–12 mm in diameter, is usually sufficient to cover the defect in one piece. Published prospective series in patients with defects averaging 1.0 cm² report mean Lysholm scores of around 92 at six months, with 84–93% achieving good-to-excellent functional results.
• 2–4 cm²: A single plug cannot tile the full area. Mosaicplasty — placing two or more smaller cylindrical plugs side by side — extends coverage within the same autograft framework and the same operation. In a registry cohort of 63 patients, a mean of 2.2 plugs per case was used to address a mean defect of 2.3 cm². This range sits squarely within Hangody and Fules' original recommended treatment window of 1–4 cm², which remains the most widely cited benchmark and is borne out by systematic review data covering defect sizes from 0.85 to 4.9 cm².
• Above 4 cm²: Autograft donor capacity is exhausted at this point. ACI, MACI, or fresh osteochondral allograft (OCA) become the standard alternatives for larger defects.

The boundary near 20 mm / ~4 cm² is not a clean clinical line. Whether a defect approaching that upper threshold is better served by a single large plug or a full mosaicplasty tiling pattern remains an active area of uncertainty, and individual anatomy will influence the final surgical plan.

What the procedure actually involves

The procedure is single-stage: harvest and implantation take place in the same operation, removing the cell-culture interval that separates the two stages of ACI or MACI.

Cylindrical plugs of bone and cartilage are taken from a quieter part of the same knee — the peripheral rim of the femoral trochlea, the intercondylar notch, or the medial femoral condyle margin — areas that bear little load during ordinary daily activity. Each plug travels with its full thickness of hyaline cartilage on top and the subchondral bone plate beneath it intact.

At the defect site, matching cylindrical sockets are prepared and the plugs are press-fitted in place, with the goal of a flush, congruent surface sitting level with the surrounding undamaged cartilage. For a single OATS plug, one 8–12 mm cylinder fills the lesion. For mosaicplasty, multiple smaller plugs (typically 2.7–8 mm) are arranged side by side across the defect, spaced roughly 3 mm apart to avoid merging the underlying bone tunnels.

What is transferred is genuine hyaline cartilage — the same type II collagen tissue the knee was built with. This is structurally different from the fibrocartilage produced by microfracture, which is formed from blood-clot repair tissue and tends to soften and thin from around two to three years onwards. Restoring the subchondral bone plate at the same time as the cartilage surface is a structural advantage that helps explain the durable improvements seen in long-term registry data.

Once the grafts are seated, empty donor sockets may be backfilled with allograft bone to reduce the risk of long-term irritation from unfilled cavities.

The donor site: what to expect from the harvest area

Taking cartilage from one part of the knee to repair another is the aspect of this procedure patients ask about most. The evidence on what that harvest actually costs is reassuring, though not without caveats.

A systematic review of 21 studies covering 1,726 patients found a mean donor-site morbidity rate of 5.9% for knee-to-knee procedures. The two most commonly reported complaints were crepitation — a clicking or grinding sensation at the harvest site — in 31% of patients who did experience a donor-site problem, and patellofemoral disturbances in 22%. These are sensations rather than structural failures in most cases, but they are real and worth naming clearly before any decision is made.

Critically, the review found no statistically significant correlation between total defect size and the morbidity rate. Whether the surgeon harvested grafts for a 1 cm² lesion or a 4 cm² mosaic tiling, the donor burden remained broadly comparable — a finding that supports the procedure's viability across its therapeutic range.

Harvest is confined to anatomically standardised non-weight-bearing zones: the peripheral trochlea rim and the intercondylar notch margins. These areas experience minimal load during ordinary daily activity, which is the anatomical rationale behind their selection.

What remains genuinely uncertain is the very long-term behaviour of the harvest site. Imaging follow-up beyond 10–15 years is sparse; for a patient in their late twenties today, what the donor area looks like at 50 or 60 is a question the current evidence cannot yet answer with confidence.

Recovery: what the rehabilitation timeline looks like

Recovery from OATS or mosaicplasty follows a predictable four-phase arc, though the precise pace depends on how many plugs were placed and what other structures were addressed at the time of surgery.

Phase 1 — Protected weight-bearing (weeks 1–6)

Crutches are essential for the first six weeks to shield the grafts while they begin integrating into their new bone bed. A hinged brace is typically worn for the first two to four weeks. Continuous passive motion (CPM) or guided passive range-of-motion exercises may be used in the early days to encourage cartilage nutrition without loading the repair site.

Phase 2 — Building strength (months 2–3)

Once crutches are phased out, the focus shifts to progressive neuromuscular strengthening and restoring full range of motion. Swelling should be settling; the knee should be tolerating low-load daily activity without significant pain.

Phase 3 — Low-impact exercise (months 4–6)

Stationary cycling and swimming are typically permitted from around month four, allowing cardiovascular fitness to be maintained without the impact forces that could stress maturing grafts. Running is generally not considered before four to six months at the earliest.

Phase 4 — Return to sport (months 6–12)

Return to sport is not permitted until at least six months post-operatively — this is where patient expectations most often diverge from clinical reality. Full graft maturation can take up to twelve months. Single-plug OATS repairs may trend toward the shorter end of this window; larger mosaicplasty cases involving multiple plugs typically trend longer.

The rehabilitation investment is worthwhile. Registry data show IKDC scores rising from 46.4 pre-operatively to 70.4 at ten years, with approximately 78% of patients sustaining a clinically meaningful improvement at two years and beyond. A 28.6% reoperation rate has been recorded over a mean four-year follow-up, but the majority of these procedures were minor secondary interventions — chondroplasty or loose body removal — rather than graft failures; only two of 63 patients in one registry ultimately required conversion to arthroplasty.

How OATS and mosaicplasty fit into the wider cartilage repair landscape

Selecting a cartilage repair technique ultimately comes down to defect size, procedural stage, and the patient's longer-term needs. OATS and mosaicplasty perform a specific role within that landscape.

Versus microfracture: A 2024 network meta-analysis of 19 RCTs found no statistically significant PROM difference between OATS, microfracture, AMIC, and ACI at any follow-up interval. Score equivalence at two years does not, however, capture longer-term trajectory: 10-year registry data for OATS show sustained functional gain, while microfracture's repair tissue carries a well-documented structural limitation for defects above roughly 2 cm² that becomes more evident over time. For a patient in their late twenties or thirties, that trajectory matters.

Versus ACI and MACI: Head-to-head trials show no clinically significant PROM difference here either. The operative distinction is procedural: ACI and MACI require two stages — a biopsy and cell-culture interval before implantation — and are better matched to larger defects in the 2–10 cm² range. OATS and mosaicplasty, completed in a single sitting, are the preferred option for focal lesions under 4 cm² in active younger patients.

Versus osteochondral allograft (OCA): When a defect exceeds approximately 4 cm² or involves extensive subchondral bone damage, autograft donor capacity runs short; fresh OCA is the appropriate next step.

Alignment adjuncts: In patients with varus or valgus malalignment, high tibial (HTO) or distal femoral (DFO) osteotomy may be combined with cartilage repair to offload the repaired compartment.

Within the four stages of a typical knee preservation pathway — symptom management, biologic and injection support, cartilage restoration, and joint replacement — OATS and mosaicplasty sit at stage three, suited to focal defects where earlier conservative measures have not provided sufficient relief. Determining which pathway fits a specific knee requires an individualised clinical assessment.

1. [1] Mega-OATS of the Knee without Specialised Instrumentation: A Low-Cost Option for Large Cartilage Defects. (2022). https://doi.org/10.17159/2309-8309/2022/v21n2a7 https://doi.org/10.17159/2309-8309/2022/v21n2a7
2. [2] Modern Surgical Methods of Articular Cartilage Repair: Review and Comparative Analysis. (2025). https://doi.org/10.52889/1684-9280-2025-76-4-jto011 https://doi.org/10.52889/1684-9280-2025-76-4-jto011
3. [3] Mosaicplasty/Osteochondral Autograft Transfer Remains a Durable Solution for Symptomatic Chondral Defects of the Knee: Two to Ten-Year Follow-up Analysis. (2024). https://doi.org/10.1177/2325967124s00003 https://doi.org/10.1177/2325967124s00003
4. [4] Functional Outcome after Arthroscopic Mosaicplasty for Localised Osteochondral Defects in the Knee. (2024). https://doi.org/10.1016/j.mjafi.2024.09.001 https://doi.org/10.1016/j.mjafi.2024.09.001
5. [5] Functional Outcome after Arthroscopic Mosaicplasty for Localized Osteochondral Defects in the Knee. (2024). https://doi.org/10.18203/issn.2455-4510.intjresorthop20241114 https://doi.org/10.18203/issn.2455-4510.intjresorthop20241114
6. [6] Long-Term Outcomes of Mosaicplasty for Spontaneous Osteonecrosis of the Knee with Varus Alignment. (2025). https://doi.org/10.18103/mra.v13i7.6647 https://doi.org/10.18103/mra.v13i7.6647
7. [7] Isolated In-Situ Mosaicplasty Fixation of Unstable Knee Osteochondritis Dissecans in Skeletally Mature Patients. (2024). https://doi.org/10.1016/j.jisako.2024.100322 https://doi.org/10.1016/j.jisako.2024.100322