What separates single-plug OATS from mosaicplasty

Both procedures work on the same principle: a surgeon removes a small cylinder of healthy cartilage and bone from a low-load region of the patient's own knee and presses it into the damaged area. The cylinder carries a surface layer of genuine hyaline cartilage — the smooth, load-bearing tissue that lines healthy joints — and its bone base integrates directly with the underlying recipient site without requiring fixation hardware or a second operation.

The practical difference lies in the number and diameter of those cylinders. In single-plug OATS, one larger graft — typically 8–12 mm across — fills the defect in a single piece. Because the surface is continuous, the repaired zone is covered by an unbroken sheet of hyaline cartilage. In mosaicplasty, several narrower cylinders (6 mm or less) are placed side by side in a mosaic pattern to resurface a wider area that a single plug could not span.

The mosaic arrangement introduces an unavoidable compromise: the small gaps between each cylinder do not regenerate as hyaline cartilage. They fill instead with fibrocartilage — a stiffer, less resilient tissue that may be mechanically inferior over time. The repair surface therefore becomes a composite of true hyaline islands and fibrocartilage filler, rather than the seamless coverage that a single well-fitted plug can provide for a smaller defect.

Why defect size is the primary decision factor

The 2 cm² threshold separating single-plug OATS from mosaicplasty is endorsed by NICE guidance (HTG463) and cited consistently across cartilage surgery clinical practice summaries. It translates in practice to a roughly 10–12 mm circular defect — narrow enough for one cylindrical plug to span the full diameter and leave the margins cleanly covered. Once the defect widens beyond that point, a single plug cannot reach all edges without leaving exposed bone, and tiling several smaller cylinders becomes necessary.

This geometry produces the three-band framework that guides most surgical decision-making:

• Under 2 cm²: a single OATS plug (8–12 mm diameter) covers the defect with a continuous hyaline surface and no interstitial gaps.
• 2–4 cm²: mosaicplasty tiles two or more smaller cylinders (2.7–8 mm) side by side; the repair surface is a composite of hyaline islands and the fibrocartilage filler that occupies the inter-plug spaces.
• Above 4 cm²: the usable autograft from the same knee is exhausted; cell-based therapies such as MACI or ACI, or fresh osteochondral allograft (OCA), become the appropriate options.

The 4 cm² upper boundary for mosaicplasty is not arbitrary — it reflects the finite supply of graft cylinders that can safely be taken from the non-weight-bearing donor zone of the ipsilateral knee, a point addressed in the harvest section below.

The 2–4 cm² band is where clinical judgement carries the most weight. No randomised trial has directly compared single-plug versus multi-plug constructs within this range; most comparative studies pool OATS and mosaicplasty as a single category rather than setting them against each other.

How defect location changes the surgical approach

Site matters as much as size. Even when a defect falls cleanly within the single-plug or mosaicplasty range, its precise location determines how the surgeon reaches it, how the graft must be angled, and how reliably the technique is likely to perform.

The medial femoral condyle — the inner surface of the thigh bone at the knee — accounts for the majority of focal osteochondral defects seen in clinical practice, and it is where OATS evidence is strongest. The anatomy offers straightforward access via either a standard arthroscopic approach or a small medial incision, and good-to-excellent outcomes are consistently reported across published series.

The lateral femoral condyle demands greater precision. The plug must sit exactly flush with the surrounding cartilage; even a fraction of a millimetre proud of the surface creates an edge that catches against the lateral meniscus during movement, raising the risk of meniscal damage and pain. Access typically requires a small lateral incision to get the instruments at the correct angle.

That angle is non-negotiable at every site. The harvesting and implantation tools must enter the recipient hole perfectly perpendicular to the joint surface — straight down, not tilted. If the plug is driven in at an angle, one edge sits higher than the other, creating uneven loading on the cartilage rim rather than distributing weight across the full surface. For defects on the posterior, weight-bearing part of the condyle, the knee may need to be held in a specific degree of flexion to achieve this geometry.

The trochlea and patellofemoral joint present the greatest difficulty. The trochlear groove follows a compound curve — it bends in more than one direction simultaneously — so a straight cylindrical plug cannot conform to its shape. The result is plug-proudness: part of the graft stands above the surrounding surface, concentrating load on that edge and causing pain with flexion. Small, well-contained trochlear lesions may still be amenable to OATS, but for larger or geometrically complex defects at this site the technique is unreliable, and osteochondral allograft or an alternative approach is usually preferred. It is worth noting that long-term outcome data for trochlear OATS are considerably less complete than for condylar sites.

Where the plugs come from and what limits the harvest

During the procedure, while the surgeon is repairing the damaged area, a second operative site is created elsewhere in the same knee. The harvest zone is the low-load periphery of the ipsilateral femoral condyle — specifically the trochlear ridge and the walls of the intercondylar notch — regions that carry minimal force during normal walking and are selected precisely because removing small cylinders here disrupts function as little as possible.

Donor-site morbidity is a genuine trade-off rather than a theoretical risk. Patients can experience localised pain, aching, and stiffness at the harvest site during recovery, and some series report subchondral cyst formation or cartilage thinning at plug extraction points when multiple cylinders are taken. That risk scales with the number of plugs: a single-plug OATS procedure places a much lower burden on the donor zone than mosaicplasty, and this is a practical advantage for smaller defects that a single cylinder can cover.

As noted earlier, once a defect exceeds 4 cm² the safe harvest capacity of this low-load zone is reached. At that point, rather than extending into load-bearing cartilage, the clinical pathway shifts to fresh osteochondral allograft (OCA) — which uses donor tissue and carries no harvest burden on the patient's knee — or to cell-based techniques such as MACI or ACI that grow cartilage cells from a small biopsy rather than requiring full-thickness bone-and-cartilage cylinders.

What the outcomes evidence shows at 2 and 10 years

Ten years of published follow-up data provide the strongest durability signal available for these procedures. A 2024 registry study of 63 consecutive patients established minimum clinically important difference (MCID) thresholds — the smallest score change a patient would personally notice in daily life. On the IKDC scale, which runs from 0 to 100 (where 100 represents full, pain-free knee function), the MCID sits at 9.3 points; on the KOS-ADLS disability scale it is 7.4 points. Approximately 78% of patients crossed the IKDC threshold at two years, and that proportion held steady through ten-year follow-up — suggesting that the functional gains achieved early in recovery are not systematically eroded over time.

A 2025 prospective series of 36 patients undergoing single-plug arthroscopic OAT for defects of 7–10 mm confirmed significant Lysholm score improvement over time (p < 0.0001), adding contemporary mid-term evidence to this durability picture.

Against microfracture, the most widely cited long-term benchmark is a ten-year randomised trial in athletic patients, which found roughly double the rate of good or excellent outcomes with autograft techniques. The advantage is partly biological: microfracture generates fibrocartilage, a structurally inferior tissue prone to deterioration, whereas an autograft plug restores genuine hyaline cartilage to the defect. In mixed-activity general populations the picture shifts. A 2024 network meta-analysis of 19 randomised trials found no statistically significant difference between cartilage repair techniques on patient-reported outcome measures. This does not mean technique choice is irrelevant — it identifies athletic demand as the key moderating variable. For patients who load their knees repeatedly at high force, autograft's hyaline surface appears to make a measurable difference that general-population pooling obscures.

One honest limitation is that direct head-to-head trial data comparing single-plug and multi-plug constructs within the mid-range 2–4 cm² band are absent; most studies group both approaches as one category. Technique selection within that band therefore rests on defect geometry and intraoperative judgement rather than on superiority data for either approach.

Who is a suitable candidate and how assessment works

The strongest candidates for either technique share a common profile: active adults, typically under 45, with a symptomatic full-thickness focal lesion (ICRS grade III–IV) measuring between 1 and 4 cm², on an otherwise stable and correctly aligned knee. Ligament laxity or significant malalignment should be addressed — through reconstruction or osteotomy as appropriate — before or alongside cartilage repair, because persistent mechanical instability compromises graft survival regardless of plug quality.

Several factors place a patient outside the autograft pathway. Defects larger than 4 cm² exhaust the safe donor supply from the same knee. Global or diffuse chondral change across a compartment points towards joint preservation by osteotomy or, in advanced cases, joint replacement rather than focal repair. These patients are generally better served by osteochondral allograft (OCA) or cell-based options such as MACI.

Accurate pre-operative sizing is essential: technique selection — single plug or mosaic — is made before the patient enters the theatre, and that decision rests on how precisely the defect dimensions are known. MRI with cartilage-sensitive sequences, including T2 mapping where available, gives the operating surgeon the detail needed. At Lincolnshire Knee, part of the MSK Doctors group, consultant-led assessment at Sleaford NG34 or Grantham NG31 is available without referral and can include onMRI™ cartilage analysis where clinically indicated. Further information and appointment booking are at lincolnshireknee.co.uk.

1. [1] Arthroscopic Osteochondral Autograft Transplantation (OAT) in Patients with Focal Osteochondral/Chondral Lesions of the Knee Mid-Term Clinical Outcome. (2025). https://doi.org/10.5704/MOJ.2507.004 https://doi.org/10.5704/MOJ.2507.004
2. [2] Defining clinically relevant outcome thresholds for pain and function after osteochondral autograft transplantation of the knee. (2024). https://doi.org/10.1002/ksa.12422 https://doi.org/10.1002/ksa.12422
3. [3] Autologous chondrocyte implantation, MACI, osteochondral autograft transplantation and osteochondral allograft: systematic review and meta-analysis. (2024). https://doi.org/10.1002/ksa.12525 https://doi.org/10.1002/ksa.12525
4. [4] Microfractures, AMIC, OCT and ACI for knee chondral defects: NMA of RCTs. (2024). https://doi.org/10.1530/EOR-23-0089 https://doi.org/10.1530/EOR-23-0089