Which procedure holds up better for active knees over a decade

Both osteochondral autograft transfer (OATS) and microfracture are used to treat the same focal knee cartilage defects, yet their long-term records differ in ways that matter considerably for anyone planning to stay active. For athletes and high-demand patients, OATS is the more durable option — and the evidence at ten years makes that gap difficult to ignore.

The clearest data come from the Gudas et al. prospective randomised trial published in the American Journal of Sports Medicine in 2012 — the only RCT in young athletes to reach a genuine 10-year endpoint. Treatment failure occurred in 14% of OATS patients versus 38% of those who had microfracture, a statistically significant difference. Return-to-sport rates follow the same pattern: published series report 84–100% for OATS in athletic populations, against 44–83% for microfracture; athletes returning after microfracture are more likely to compete at a reduced level or to retire from sport altogether.

The detail that matters most for clinical decision-making is when the gap appears. At two years, both procedures can look reasonably similar on patient-reported scores. The divergence opens decisively between years five and ten, which means short-term follow-up data systematically underestimate how differently these two techniques perform over a sporting career.

Microfracture held an important historical role as the simplest, quickest option, but current evidence no longer supports it as a first-line choice for active patients. The sections that follow explain the biological reasons for this divergence and identify which patients and defect characteristics are best suited to OATS.

What the two procedures actually do to the knee

The distinction between these two procedures comes down to what the surgeon puts into the defect — and what the body generates in response.

OATS works by harvesting small cylinders of bone and cartilage — called osteochondral plugs — from a lower-load region of the same knee, typically the periphery of the femoral condyle. These plugs are press-fitted into the damaged site. Because they carry native hyaline cartilage, the repair tissue is structurally identical to what was lost: stiff, wear-resistant, and well-suited to distributing load across the joint surface. Defects of 1–2 cm² are typical candidates; the mosaic arrangement of multiple smaller plugs can address lesions up to around 4 cm². The trade-off is a secondary harvest wound — donor-site morbidity that most patients tolerate well but that requires careful planning in terms of lesion sizing and plug number.

Microfracture takes a different approach. Small holes are drilled through the exposed bone into the underlying marrow space, releasing stem cells and growth factors that gradually form a repair tissue called fibrocartilage. Fibrocartilage is not the same material as hyaline cartilage — it is softer, less wear-resistant, and less effective at distributing mechanical load. Under the repetitive high-impact and pivoting forces of athletic activity, it tends to break down progressively over time. Microfracture also disturbs the subchondral bone plate, which can complicate — or reduce the success of — any future cartilage procedure if the initial repair fails.

Both techniques are applied to focal osteochondral defects and osteochondritis dissecans. These conditions arise disproportionately in athletes and active adults — precisely the population whose joint demands fibrocartilage is least able to sustain over the long term, and where the biological difference between the two repair tissues carries its greatest clinical weight.

What the 10-year data actually show

Taken together, four studies published between 2012 and 2026 form the most direct evidence base at the ten-year mark — and they point in a consistent direction, with one important caveat.

Gudas et al. (AJSM, 2012) remains the only prospective randomised trial to follow young athletes all the way to ten years. The failure rates — 14% for OATS versus 38% for microfracture (P<0.05) — have already been noted; what matters here is that this single RCT carries the bulk of the randomised weight at this time horizon. The evidence base is thinner than clinicians would ideally want.

Solheim et al. (2018, n=203) adds survival-curve granularity from a large cohort. OAT survival remained above 80% through the first seven years and above 60% at fifteen years; microfracture dropped below 80% within twelve months of surgery and below 60% within three years. Overall long-term failure was 51% for OAT versus 66% for microfracture (P=0.01). The mean time to failure tells a similar story: 8.4 years for OAT against 4.0 years for microfracture (P<0.001). These are not marginal differences.

Toyooka et al. (scoping review, 2023, 16 studies) translates the survival data into functional terms for athletes in pivoting sports. Return-to-sport rates after OAT ranged from 87–100%, with return to pre-injury level at 67–93%; after microfracture the figures were 44–83% and 25–75% respectively. The wide ranges reflect genuine heterogeneity in sport type, defect size, and follow-up length across the included studies.

Amoo-Achampong et al. (2026, n=63, mean age 27.4 years, mean lesion 2.3 cm²) provides the most granular OAT trajectory: IKDC scores rose from 46.4 preoperatively to 76.1 at two years, 78.0 at five years, and 70.4 at ten years (P<0.001). The modest attenuation between five and ten years is a finding worth noting — sustained benefit, but not entirely static.

The one study pointing in a different direction is Ulstein et al. (KSSTA, 2014, n=25, median 9.8-year follow-up), which found no statistically significant difference in Lysholm score (microfracture 69.7 vs OAT 62.6), KOOS, or radiographic osteoarthritis. With only 25 patients, the trial was almost certainly underpowered to detect a clinically meaningful difference, and the absolute scores in both groups were modest — neither group came close to a normal knee. The null result should not be read as equivalence.

What remains genuinely uncertain is how outcomes vary by specific sport, defect location, and concurrent ligament pathology. The ten-year IKDC attenuation seen in Amoo-Achampong also warrants further investigation.

Why the gap between the two procedures widens after year five

The timing of divergence is not accidental — it reflects how fibrocartilage behaves under load over months and years.

In the early postoperative period, fibrocartilage can fill a cartilage defect adequately, producing meaningful pain relief and acceptable outcome scores at two years. This is precisely why short follow-up studies mislead: they sample a window when the fibrocartilage is still structurally intact and the gap between procedures has not yet opened.

The problem emerges later. Fibrocartilage is softer than native hyaline cartilage and wears out faster under repeated loading. In pivoting and high-impact sports — football, rugby, basketball — the repetitive shear and compressive forces progressively degrade the repair tissue through years three to seven, in a process that tends to accelerate rather than stabilise. Native hyaline cartilage, restored by OATS, distributes those same loads without progressive breakdown. Steady-state endurance activity is less demanding, which is one reason outcome differences between the two procedures are less consistent in mixed-activity or lower-demand populations.

Solheim et al.'s 2018 survival curves capture this timing clearly. Microfracture survival fell below 80% within twelve months of surgery and below 60% within three years; OAT survival remained above 80% through year seven. The mean time-to-failure difference — 4.0 years for microfracture versus 8.4 years for OAT — reflects not just how often repairs fail but when, and that divergence falls almost entirely between the typical two-year endpoint and the ten-year mark.

There is also a downstream consequence worth acknowledging: a failed microfracture does not leave the knee as it was. Subchondral bone disturbance and cyst formation can limit the options — and reduce the success rates — of any revision cartilage procedure. For patients reviewing two-year RCT data when weighing their options, the message is straightforward: early equivalence in scores does not predict long-term equivalence in function or revision risk.

Who is suited to OATS and where its limits lie

Selecting the right patient matters as much as selecting the right technique. OATS sits within a defined clinical window, and outcomes outside that window are considerably less predictable.

The good-candidate profile centres on three intersecting factors: younger age (typically under 45–50), a focal defect in the range of 1–2 cm² (mosaic configurations can extend to approximately 4 cm²), and meaningful athletic or physical demand that makes durable hyaline cartilage — rather than fibrocartilage — the clinically relevant goal. Amoo-Achampong et al.'s 2026 series, which reported sustained IKDC benefit through ten years in a cohort with a mean age of 27.4 years and a mean lesion size of 2.3 cm², broadly reflects this profile.

Factors associated with worse outcomes in published series include increasing patient age, prior cartilage surgery — particularly previous microfracture, which can compromise the subchondral bone plate — larger lesions, and compartment malalignment. Pareek et al.'s 2016 systematic review identified age, prior surgery, and defect size as the primary correlates of poorer long-term results. Where compartment loading is a concern, alignment correction via osteotomy may be considered as an adjunct rather than treating the cartilage defect in isolation.

Defect size is a firm practical limit. Lesions beyond approximately 4 cm², post-traumatic multi-lesion patterns, or cases where donor-site morbidity from harvesting multiple plugs becomes prohibitive are generally outside the OATS indication. In those circumstances, cell-based approaches such as MACI or ACI, or fresh osteochondral allograft for larger posttraumatic defects, become the more appropriate considerations — each carrying its own evidence base and procedural profile.

The ten-year IKDC of 70.4 in Amoo-Achampong's series is a useful reference point for expectation-setting: meaningful and statistically significant improvement sustained over a decade, but with modest attenuation from the five-year peak of 78.0. OATS, within its indicated range, is well-evidenced — it is not, however, a guarantee of full functional restoration for every patient.

Getting a cartilage assessment and what to expect

Translating survival curves and return-to-sport rates into an individual decision requires something the published trials cannot supply: accurate characterisation of the specific defect — its size, ICRS grade, location on the condyle or trochlea, and the condition of the underlying subchondral bone. That information determines which technique is appropriate, whether alignment correction should accompany it, and what a realistic recovery trajectory looks like.

MRI with cartilage-specific sequencing is the standard pre-operative imaging tool for this planning stage; AI-driven cartilage segmentation and T2 mapping — as available through onMRI™ — can add objective lesion measurement where it would otherwise rely on visual estimate alone. For athletes, a structured loading assessment identifying gait asymmetries that contributed to the injury or will affect rehabilitation (MAI Motion® provides one framework for this) adds biomechanical context that imaging does not capture.

What the ten-year data collectively establish is that the gap between OATS and microfracture is not visible at two years but is substantial by year five, and continues to widen. For younger, active patients with focal defects, that durability asymmetry is the decision-critical finding — and getting the defect characterised accurately at the outset is what determines whether the repair holds.

Lincolnshire Knee accepts patients without GP referral at Sleaford NG34 and Grantham NG31. Book an assessment at lincolnshireknee.co.uk.