Why cartilage repair needs more than microfracture alone

Left untreated, a full-thickness hole in the knee's cartilage surface does not close. Articular cartilage has no blood supply of its own and almost no capacity to regenerate — a property that makes even small focal defects clinically significant and determines why the surgical approach chosen matters far more than it might first appear.

Microfracture addresses this by puncturing the underlying bone so that marrow cells flood into the defect and solidify into a repair tissue. That tissue — fibrocartilage — is a type I collagen patch: serviceable in the short term but structurally weaker than the native surface it replaces. Published evidence and clinical experience consistently show that fibrocartilage repairs tend to soften and thin from around two to three years onwards, with the deterioration becoming more marked when the original defect exceeds roughly 2 cm².

OATS (Osteochondral Autograft Transfer System) and mosaicplasty take a fundamentally different approach. Rather than encouraging the body to form a substitute tissue, they transplant cylinders of the patient's own hyaline cartilage — the smooth, type II collagen–rich tissue that normally lines the joint — together with the subchondral bone beneath it, directly into the defect. The graft arrives as mature, load-bearing tissue and integrates into the surrounding joint surface rather than being grown from scratch.

That distinction — restoring the real tissue versus filling the space with a scar-like equivalent — is why the two approaches tend to diverge in outcome data at longer follow-up.

Which patients are suitable for OATS or mosaicplasty

Both OATS and mosaicplasty are second-line procedures. Before either is considered, candidates should have completed a course of conservative management — typically structured physiotherapy, activity modification, and where appropriate, injection therapy — without achieving adequate relief. Proceeding to surgery before that period has been given a genuine chance is itself a relative contraindication.

Age and general joint health are the next filters. Patients aged 50 or under, with a BMI no greater than 40, are generally the most appropriate candidates. The BMI threshold is not arbitrary: excess joint loading impairs graft integration and places greater mechanical stress on the repair site during the critical early healing period. On the joint health side, a Kellgren-Lawrence osteoarthritis grade of 2 or lower is required — meaning the surrounding cartilage is largely intact and the problem is a focal defect rather than widespread degeneration. Absolute exclusions include a history of knee infection, inflammatory arthritis such as rheumatoid disease, or any prior knee tumour.

The nature of the defect itself matters considerably. Focal osteochondral lesions — including osteochondritis dissecans (OCD) and post-traumatic injuries — are the target pathology. Such defects are genuinely common, appearing in around 19% of knee arthroscopies; however, only a subset of those patients meet all the filters above and are therefore realistic autograft candidates. Within that group, a single-plug OATS construct is generally suited to lesions in the 1–2 cm² range, while mosaicplasty can extend coverage to approximately 4 cm² by arranging multiple smaller plugs across a wider area.

Beyond that size limit, or in patients who fall outside the age and joint-health criteria, autograft is unlikely to be sufficient. ACI, MACI, or fresh osteochondral allograft become the more appropriate options — each serving a distinct part of the larger-defect spectrum and covered separately.

Single-plug OATS vs mosaicplasty: what the technique difference means clinically

The physical difference between the two techniques comes down to plug count. In a single-plug OATS procedure, the surgeon harvests one cylindrical core of cartilage and bone — typically from a low-load-bearing edge of the patient's own knee, such as the rim of the kneecap groove or the notch between the rounded femoral condyles — and press-fits it into the prepared defect. The result is a continuous disc of intact hyaline cartilage sitting flush with the surrounding joint surface.

Mosaicplasty, popularised by László Hangody, applies the same harvest-and-transfer principle but with multiple smaller plugs, typically 3.5–4.5 mm in diameter, arranged side by side in a mosaic pattern. This allows wider area coverage than a single core can provide. The trade-off is structural: small gaps between plugs cannot be filled with genuine hyaline tissue and instead heal with fibrocartilage — the same repair material that microfracture produces. The more plugs required, the smaller the proportion of true hyaline coverage at the treated site.

This is a recognised, manageable trade-off rather than a disqualifying flaw. A mosaic of mostly-hyaline plugs still tends to outperform a purely fibrocartilage repair in long-term follow-up, and for many patients it remains the most practical way to address larger focal defects using their own tissue.

Both approaches are single-stage procedures, usually performed arthroscopically or through a small incision.

Harvesting plugs from the patient's own knee introduces a secondary consideration: the donor site — that low-load-bearing rim described above — may cause some discomfort or stiffness during recovery. This is generally minor, but it is a real limitation worth discussing at consultation. Crucially, the amount of tissue that can safely be taken from one knee places a practical ceiling on total defect coverage, which is one reason very large defects ultimately require a different strategy altogether.

How defect size guides the choice between microfracture, OATS, and alternatives

Defect size is the most practical starting point for treatment selection, though it works as a guide rather than a fixed rule.

For lesions under 2 cm², both microfracture and OATS are clinically acceptable options. Microfracture is often chosen first for genuinely small defects because it is technically simpler and carries a lower procedural burden — the patient avoids a donor-site harvest altogether. For many in this size band, the short-term results are comparable.

Between 2 and 4 cm², the balance tips meaningfully towards OATS or mosaicplasty. Microfracture can still be performed at this size, but evidence suggests its long-term results deteriorate — the fibrocartilage repair it produces lacks the mechanical durability of a true osteochondral graft, particularly over a larger surface area. Patients and referring clinicians should be aware that choosing microfracture in this range may mean lower scores at follow-up and a higher chance of needing further surgery.

Above 4 cm², autograft harvest from the patient's own knee becomes the limiting factor. The available donor tissue is finite, and attempting to fill a defect of this size risks meaningful secondary morbidity at the harvest site. ACI, MACI, or fresh osteochondral allograft (OCA) are the more appropriate choices at this scale — each drawing on a different tissue source and covered separately within this series.

For defects of 3 cm² or greater specifically, the SUMMIT trial demonstrated that KOOS pain and function scores at both two and five years favoured MACI over microfracture, reinforcing the case against marrow stimulation for larger lesions.

Defect size is nonetheless only one variable. OA grade, patient age, activity demands, and MRI findings regarding subchondral bone involvement all shape the final decision — which is why a consultant assessment with dedicated cartilage imaging remains essential before any pathway is agreed.

Long-term outcomes vs microfracture: what the evidence actually shows

The clearest signal in the published data comes from Pareek et al.'s 2016 systematic review, which followed patients to 10 years after OAT and found significant improvements in both IKDC and Lysholm scores compared with microfracture. These are the two most widely used functional scoring tools in cartilage outcome research, and gains on both carry real clinical weight — they correspond to meaningful reductions in pain and improvements in knee function in daily life.

Tegner activity scores tell a more nuanced story. In the same body of evidence, Tegner scores — which measure return to a specific level of sport or physical activity — did not reach statistical significance. That gap matters: it suggests patients recover meaningful knee function after OATS, but not necessarily a full return to their pre-injury sport level across the board. Return-to-sport expectations should therefore be discussed carefully at the outset rather than assumed.

On that specific question, Campbell et al.'s systematic review of 1,117 patients (mean follow-up 3.6 years) found significantly higher return-to-sport rates after OAT than after microfracture. Across a large, pooled dataset, that advantage is harder to dismiss.

The evidence is not uniform, however. Lim et al., in a Level 2 study directly comparing microfracture, ACI, and OAT, found no statistically significant difference in either Lysholm or Tegner scores between the three groups. One plausible explanation is that Lim et al.'s cohort included smaller defects — a size range where microfracture performs adequately in the short term — while the Pareek and Campbell advantage for OAT is most pronounced in larger lesions and at longer follow-up. The discrepancy reflects a genuine feature of the evidence base rather than a contradiction: defect size and follow-up duration both influence which procedure looks better.

Few high-quality randomised controlled trials compare OATS directly to microfracture beyond five years, and robust data beyond the 10-year mark remains sparse. The available evidence nonetheless supports preferring OATS over microfracture for focal defects above 2 cm² in active patients — while being honest that the evidence hierarchy sits mostly at Level II to IV.

Recovery timeline and realistic return-to-sport expectations

Recovery from OATS or mosaicplasty proceeds in stages. The first four to six weeks typically involve protected weight-bearing on crutches while the plug integrates with the surrounding bone and cartilage. Progressive loading and physiotherapy follow, building through the third to sixth month. Sport-specific rehabilitation generally begins in the second half of the first year.

For most patients, six to twelve months is the realistic window for return to sport — towards the shorter end for single-plug repairs and lower-impact activities, and towards the longer end for larger mosaic repairs, high-impact sport, or cases where alignment work was carried out at the same time. Some patients take longer, and imposing a fixed deadline risks premature loading of an incompletely matured graft.

Meaningful improvements in pain and knee function are the most consistent finding across the outcome evidence. Full restoration of pre-injury sport level is not guaranteed for all patients; as the Tegner score data described above reflects, some gap commonly remains between functional recovery and return to peak sport demand, and this distinction is worth addressing directly in pre-operative discussions.

Donor-site discomfort at the harvest area is usually mild and transient, but it is a secondary morbidity that deserves honest discussion before any decision is made.

Objective monitoring supports confident, evidence-based return-to-sport clearance. Repeat MRI at defined intervals allows graft fill and tissue quality to be evaluated; gait analysis can detect compensatory movement patterns that clinical review alone may miss. Together, these tools allow the rehabilitation team to ground clearance decisions in measurable data rather than symptom reporting alone.

For patients considering whether OATS or mosaicplasty is the right pathway, a consultant-led cartilage assessment — including dedicated imaging where indicated — is the appropriate starting point. Lincolnshire Knee, part of the MSK Doctors group, accepts patients without referral, with consultations available at Sleaford and Grantham. Further information is at lincolnshireknee.co.uk.

1. [1] Osteochondritis dissecans – Wikipedia. https://en.wikipedia.org/?curid=3762029 https://en.wikipedia.org/?curid=3762029
2. [2] Articular cartilage repair – Wikipedia. https://en.wikipedia.org/?curid=19042351 https://en.wikipedia.org/?curid=19042351