Which procedure holds up longer in an athletic knee?

Between two arthroscopic options for a focal knee cartilage defect, an athlete's most urgent question is not which procedure feels less invasive — it is which one still works a decade later.

The answer from the only randomised controlled trial conducted exclusively in young athletes to reach a ten-year endpoint is decisive: osteochondral autograft transfer (OAT/mosaicplasty) substantially outperforms microfracture. Gudas and colleagues recorded treatment failure in 14% of OAT patients versus 38% of microfracture patients at ten years (P < .05) — a gap that translates directly into athletic careers. Return-to-sport rates follow the same trajectory: 84–100% for mosaicplasty patients compared with 44–83% for those treated with microfracture, with the microfracture group more likely to be competing at a reduced level or to have stepped away from sport altogether.

One piece of context matters here: this advantage is specific to athletic populations under repetitive high-load conditions. In general mixed-activity patients, the superiority of OAT over microfracture is not reliably reproduced — which indicates that activity demand, rather than the procedures in isolation, is the primary variable determining which repair holds.

The sections that follow examine why the tissue biology drives this divergence, what fifteen years of survival data show, and how defect size and patient profile shape the choice between these two approaches.

What the Gudas trial found — and what the 15-year follow-up confirmed

Gudas and colleagues designed their trial specifically around the population most exposed to long-term cartilage failure: young athletes with focal osteochondral defects of the knee, randomised to mosaicplasty or microfracture, with a true ten-year primary endpoint. That design choice — an exclusively athletic cohort, no mixed-activity patients, a decade of follow-up — is precisely what makes it the definitive evidence anchor for this question. No other randomised controlled trial has replicated it.

What the subsequent cohort extension adds is more striking than the ten-year headline alone. When the same randomised patients were followed to 15–17 years (Gudas et al., Am J Sports Med 2018), the mosaicplasty functional advantage had not eroded. That matters because genuinely long-horizon RCT data are rare in cartilage surgery: patients move, surgeons retire, and funding lapses. Reaching 15–17 years of controlled follow-up on the same original cohort places the Gudas series in a very small group of trials. Functional outcome scores — IKDC and Tegner activity scales — tracked with the failure-rate difference at every interval, consistent with the principle that the structural gap between true hyaline cartilage and fibrocartilage repair tissue compounds progressively under athletic loading rather than remaining static.

One qualification deserves a brief note. A 2026 BMJ Open systematic review found only low certainty for patient-reported outcome differences between the two techniques across available RCTs — a finding that pertains to symptom scores at individual time points, and sits alongside rather than contradicting the durable failure-rate data the Gudas cohort provides.

Why microfracture repair tissue breaks down under athletic load

The divergence in survival curves has a straightforward biological explanation — one that makes sense of why procedure choice carries such weight for an athlete planning to return to pivoting sport.

Microfracture works by perforating the subchondral bone plate with small channels, triggering an influx of marrow-derived mesenchymal stem cells into the defect. Those cells form repair tissue, but it is fibrocartilage — a type I/II collagen scar matrix rather than the type II collagen network of native hyaline cartilage. Fibrocartilage is softer, less resistant to wear, and less effective at distributing compressive load across the joint surface. Under the repetitive high-impact and pivoting demands of athletic sport, those deficiencies are progressively exposed, which is why survival curves for the two procedures begin to separate almost immediately after surgery rather than converging over time.

OAT addresses this at source. Cylindrical plugs of intact hyaline cartilage and underlying subchondral bone are harvested from low-load donor zones of the patient's own knee and press-fitted into the defect, preserving native osteochondral architecture in the repair site.

Mosaicplasty — the multi-plug variant of OAT — carries its own structural caveat worth acknowledging honestly: the gaps between individual plugs tend to fill with fibrocartilage ingrowth rather than hyaline tissue, reducing the proportion of true hyaline coverage at the repair surface. A separate concern specific to microfracture is that repeated perforation of the subchondral bone plate risks structural damage that may limit the options available if the repair eventually fails and a revision procedure is needed.

Independent survival data: the Solheim cohort at fifteen years

Confirmation came from an independent source in 2018, when Solheim and colleagues published long-term Kaplan-Meier survival data on 203 patients who had undergone either OAT (n=84) or microfracture (n=119) — a cohort entirely separate from the Gudas randomised trial and drawn from a different operative and geographical context. The survival curves tell the essential story: OAT remained above 80% for the first seven years and stayed above 60% at fifteen years; microfracture dropped below 80% within twelve months of surgery and fell below 60% within three years.

Mean time to failure was 8.4 years for OAT versus 4.0 years for microfracture (P<0.001). For a 28-year-old footballer planning to compete into their forties, a repair most likely to fail inside four years is a bridge — and a short one. The OAT figure of 8.4 years is itself a conservative estimate, given that a meaningful proportion of the cohort had not yet reached failure by the end of the observation window.

The Solheim data carry weight precisely because they are independent: a different country, a different operative cohort, a different study design, and the same directional finding. Overall failure rates of 51% (OAT) versus 66% (microfracture) reached statistical significance (P=0.01), and the time-to-failure gap (P<0.001) reinforces that the OAT durability advantage is a robust, reproducible signal rather than a product of a single well-characterised trial population.

Selecting the right candidate: defect size, age, and activity level

Defect size, age, and activity level are the three variables that most reliably determine which procedure belongs in a given patient.

OAT is the preferred option for contained focal defects in the 1–4 cm² range in active patients. Mosaicplasty — the multi-plug configuration — can extend that upper boundary by tiling several smaller cores across a wider area, though each additional plug increases the volume of osteochondral tissue harvested from donor zones. Donor-site morbidity is a genuine trade-off in larger mosaic configurations: harvesting multiple cores from the non-weight-bearing margins of the knee can produce localised aching or stiffness at the harvest site that requires its own rehabilitation programme. Patients should expect this to form part of the recovery conversation before surgery.

Microfracture retains a narrower contemporary role: lesions below 2 cm² in patients with substantially lower activity demands — those who do not routinely subject the repair to repetitive high-impact or pivoting loads. That description does not fit the typical competitive or recreational athlete; in those patients, the durability argument set out in the preceding sections consistently favours OAT.

Among OAT candidates, the strongest positive prognostic factors for returning to sport are age under 25 years and a lesion below 2 cm². Recovery after mosaicplasty takes longer on average than after microfracture, which reflects the greater technical complexity of the reconstruction rather than any question about its durability.

Where defects exceed 4 cm² or disease is more diffuse, neither standalone procedure is adequate. Matrix-augmented microfracture (AMIC) and cell-based techniques such as MACI address that larger size range — brief mention here, as they warrant separate assessment in their own right.

Recovery timeline, return to sport, and what assessment involves

Recovery from mosaicplasty is a longer commitment than recovery from microfracture, and athletes entering the decision should understand that trade-off plainly. The osteochondral reconstruction — fitting, seating, and allowing biological integration of cylindrical bone-cartilage plugs — demands a rehabilitation programme proportionate to its complexity. The briefer microfracture timeline reflects a simpler intervention; the decade-level evidence reviewed in the preceding sections makes clear what that simplicity costs over time.

Across the wider evidence base, including a scoping review of pivoting-sport outcomes published in 2023, OAT return-to-sport rates range from 87% to 100%, with 67–93% of patients returning to their pre-injury competitive level — both figures substantially above the corresponding microfracture range. For an athlete whose priority is playing at full intensity years from now rather than weeks from now, the longer recovery trajectory is proportionate to the endpoint.

The most recent addition to the comparative data set is the SISMIC randomised controlled trial (2025), which tested microfracture augmented with a collagen scaffold — an attempt to address the biological vulnerability of unassisted marrow stimulation. Its two-year KOOS scores suggest incremental improvement over standard marrow stimulation, but two years of patient-reported symptom data do not resolve the decade-level question. That gap remains genuinely open.

The broader evidential limitation is the absence of large, multi-centre trials powered specifically around return-to-sport as a primary endpoint in well-characterised athletic cohorts. Most existing data use functional scoring as a proxy; the literature is not yet structured around the outcome that carries most weight for a competitive athlete.

Specialist assessment draws on high-resolution MRI characterisation of defect size, grade, and containment, combined with biomechanical evaluation and a realistic appraisal of sporting demand — the variables that together determine which procedure is appropriate for a given patient. Patients considering either approach can arrange a cartilage assessment via londoncartilage.com.

1. [1] Ten-year follow-up of a prospective, randomized clinical study of mosaic osteochondral autologous transplantation versus microfracture for the treatment of osteochondral defects in the knee joint of athletes (Gudas et al., 2012). (2012). https://pubmed.ncbi.nlm.nih.gov/23024150/ https://pubmed.ncbi.nlm.nih.gov/23024150/
2. [2] Return to Pivoting Sports after Cartilage Repair Surgery of the Knee: A Scoping Review (2023). (2023). https://doi.org/10.1177/19476035221141416 https://doi.org/10.1177/19476035221141416
3. [3] Long-Term Survival after Microfracture and Mosaicplasty for Knee Articular Cartilage Repair (Solheim et al., 2018). (2018). https://doi.org/10.1177/1947603518783482 https://doi.org/10.1177/1947603518783482
4. [4] Modern Surgical Methods of Articular Cartilage Repair: Review and Comparative Analysis (2025). (2025). https://doi.org/10.52889/1684-9280-2025-76-4-jto011 https://doi.org/10.52889/1684-9280-2025-76-4-jto011
5. [5] Randomized Study of Long-term (15-17 Years) Outcome After Microfracture Versus Mosaicplasty in Knee Articular Cartilage Defects (Gudas et al., 2018). (2018). https://doi.org/10.1177/0363546517745281 https://doi.org/10.1177/0363546517745281