Which patients are OATS candidates?

Most patients who reach a conversation about OATS have already tried, or been assessed for, a simpler approach — and found it insufficient. OATS is a surgical escalation, not a first-line treatment, and it is most clearly indicated when the size or complexity of a talar osteochondral lesion (OLT) places it beyond the reliable reach of bone marrow stimulation (BMS).

The clearest evidence for where that boundary lies comes from two sizeable clinical series. Chuckpaiwong et al. (2008), studying 105 ankle osteochondral lesions, found that BMS produced only a 3% success rate once average lesion diameter reached 15 mm or more. Choi et al. (2009) corroborated this in 168 lesions, reporting that a defect area exceeding 150 mm² on MRI — roughly 1.5 cm² — reliably predicted BMS failure. Below these thresholds, BMS tends to work well; at or above them, it almost invariably does not.

OATS is therefore considered in three broad situations: as a primary procedure for lesions meeting or exceeding the 15 mm / 150 mm² threshold; as a salvage option when BMS has already been attempted on a smaller lesion and failed; and for cystic OLTs where subchondral bone has collapsed and marrow stimulation alone cannot restore structural support.

In terms of patient profile, those most likely to benefit are typically younger, active individuals whose cartilage retains regenerative potential and who can tolerate the procedural complexity involved — including the implications of a donor harvest from the knee.

The 15 mm size threshold and what the evidence shows

Those two size boundaries — 15 mm diameter and 150 mm² area — mark where BMS stops working. What they do not tell you is what to do next, and that is where a second layer of evidence becomes relevant.

For lesions in the 2–4 cm² range, the 2021 Cartilage Book notes that both microfracture and osteochondral autografting (mosaicplasty/OATS) are recognised options — neither has an automatic claim on this size band. Osteochondral autografting is described as a legitimate first-choice or co-equal treatment at this scale, not simply a fallback. To put that in perspective, 2 cm² is roughly the surface area of a small fingernail; 4 cm² is closer to that of a thumbnail. A lesion that might seem modest on an X-ray can sit squarely within the range where the two techniques compete on evidence.

The argument for favouring OATS in this zone strengthens when follow-up extends beyond the first year or two. Shorter-term studies can give microfracture a misleading advantage, because the fibrocartilage it generates tends to perform adequately early on but degrades under repeated joint loading. In long-term comparative follow-up, mosaicplasty has been shown to produce higher clinical scores than microfracture — a difference that may not be visible at twelve months but becomes clinically meaningful over several years.

For borderline lesions — those sitting at, say, 148 mm² on MRI — the decision often hinges on factors beyond area alone: subchondral bone quality, cyst depth, prior procedures, and patient age. Size is the starting point, not the whole equation.

How OATS works: restoring native cartilage from the knee

The operation itself takes place in a single theatre session — an important practical distinction from cell-based options such as ACI or MACI, which require a harvest procedure followed weeks later by a separate implantation.

The surgeon begins at the knee, taking one or more small cylindrical cores of bone with their overlying cartilage from areas that bear little load in normal walking — typically the lateral margin of the trochlea or the intercondylar notch. Each plug is roughly the diameter of a pencil. The talar defect is then prepared to receive them, and the plugs are press-fitted into place, restoring the joint surface from the inside out.

What matters biologically is what those plugs contain. Healthy articular cartilage — the smooth, glassy tissue that lines every joint surface — is made predominantly of type II collagen, which gives it the stiffness and resilience to absorb load over decades. Microfracture does not restore this tissue. The blood clot it stimulates matures into fibrocartilage: a structurally simpler, scar-like repair tissue built mainly from type I collagen that tends to break down under sustained joint loading within two to three years. OATS transplants the real thing.

When the defect is larger or irregularly shaped, surgeons may use several smaller plugs placed side by side — a configuration known as mosaicplasty. Smaller, more discrete lesions are typically filled with a single plug. Where harvesting from the knee is not appropriate, autologous chondrocyte implantation (ACI) offers an alternative that avoids knee donor-site surgery entirely — a distinction covered later in this article.

Surgical access and the curvature challenge

Around six in ten talar osteochondral lesions sit on the medial dome — the inner side of the ankle joint — and that location creates an immediate access problem. The medial malleolus (the bony prominence on the inner ankle) partially shields this area, leaving insufficient room for a surgeon to reach the defect at the perpendicular angle needed to seat a plug correctly.

To solve this, surgeons often perform a chevron-type medial malleolar osteotomy: a controlled, planned cut through the medial malleolus that opens a window into the joint. Once the plug is placed, the bone is repositioned and fixed back with screws. The osteotomy itself then needs to heal alongside the transplanted cartilage — a secondary burden that extends overall recovery. Lamb et al. (KSSTA 2012) documented functional, radiographic, and MRI outcomes for this access approach. Patients considering OATS for a medial lesion should be aware it may form part of their procedure before giving consent.

Lateral talar lesions occupy a more forgiving position: they tend to sit further forward on the dome and are generally more accessible without a bone cut, often manageable arthroscopically.

A separate technical challenge applies regardless of lesion location. The talar articular surface curves more tightly — approximately 3 cm radius — than the flatter knee surface from which plugs are harvested. Standard cylindrical plugs can sit slightly proud or create edge-loading at their margins after transfer. This congruency mismatch is a known limiting factor specific to ankle OATS and one reason why consultant-level familiarity with the ankle variant of this procedure — rather than its knee counterpart — matters for graft integration and durability.

What outcomes to expect — and the donor-site question

Hangody's foundational mosaicplasty series established the clinical benchmark: 80–94% of patients achieved good-to-excellent results at two to seven years' follow-up. For a joint-preserving procedure tackling lesions that had already defeated simpler approaches, those figures represent a meaningful floor — and they come from a technique with several decades of use in specialist centres.

The more candid part of the picture is the donor-site question. Rates of knee symptoms after harvest — persistent pain, swelling, or stiffness at the collection site — span roughly 2% to over 36% across published series, a range that reflects genuine variability in how and when centres measure it rather than an error in reporting. For a patient who relies heavily on knee function, whether as an athlete, a manual worker, or simply an active adult, the donor-site risk sits at the centre of the candidacy conversation rather than the margins.

On durability, the honest reading of two-to-seven-year data is cautiously encouraging: hyaline cartilage transplanted under the right conditions tends to behave more like native tissue than fibrocartilage does, and the clinical scores held across the follow-up periods studied. Whether that trajectory extends reliably past ten years in the ankle specifically — rather than the knee, where longer series exist — remains less well documented, as do outcomes stratified by medial versus lateral lesion position. These are genuine evidence gaps; the technique's track record to this point gives reason for measured confidence, but for any individual patient, expected recovery and longer-term durability are properly discussed at consultation.

Where OATS sits in the treatment pathway

For smaller primary talar OLTs — those below the 15 mm / 150 mm² boundaries covered earlier — arthroscopic LDFF (lift, drill, fill, fix) is the validated first step. Lambers et al. (KSSTA 2019) reported consistent success with this approach in primary defects; it avoids donor-site exposure and the access challenges that come with the full OATS procedure.

OATS enters the picture in two scenarios: when a lesion has already failed BMS or LDFF, and when the defect exceeds the size threshold from the outset — particularly where substantial cystic components are present. In those cases, waiting to exhaust marrow stimulation is not standard practice; the evidence supports moving directly to a restorative approach.

For larger defects where knee harvest is a concern — an existing knee problem, bilateral sport demands, or high physical load — autologous chondrocyte implantation (ACI) is the main alternative. Giannini et al. (OAC 2005) demonstrated cell harvest from the detached osteochondral fragment itself as a donor-site-sparing approach. ACI is two-stage and resource-intensive, but sidesteps knee morbidity entirely; the choice between OATS and ACI rests on anatomy, prior surgical history, and defect geometry rather than a fixed universal hierarchy.

A separate branch applies to contained cystic OLTs whose overlying cartilage remains structurally intact. Artioli et al.'s 2023 systematic review confirmed retrograde drilling — under fluoroscopic or CT guidance — as effective in this subtype, allowing deferral or avoidance of more invasive grafting in suitable patients.

The appropriate path depends on lesion size, depth, location, and each patient's overall circumstances — factors that require careful MRI-based characterisation before any recommendation can be made.