The 10-year gap that changes the decision

For an active patient facing a focal knee cartilage defect, the choice of repair technique carries consequences that only fully emerge at the ten-year mark. Randomised evidence conducted exclusively in athletes shows roughly double the rate of good or excellent outcomes — and a comparable gap in return-to-sport rates — with osteochondral autograft transfer (OATS/mosaicplasty) compared with microfracture over that horizon. Independent long-term follow-up has confirmed the same direction of difference, with mosaicplasty producing a statistically significant and clinically meaningful advantage at short, medium, and ten-year time points.

That advantage, however, is not uniform across all patients. Studies in general, mixed-activity populations have found no statistically significant difference between the two techniques at similar follow-up lengths — which identifies activity demand as the primary variable determining which procedure performs better. The divergence materialises most clearly under the repetitive high-load conditions of high-impact and pivoting sport.

The practical implication at assessment is that patient age, defect characteristics, and sporting demands belong in the same conversation as the procedure itself. For younger, highly active patients with a focal contained defect, the ten-year trajectory is a substantive part of the clinical picture.

Why microfracture breaks down under athletic load

The mechanism behind microfracture's midterm failure begins with what the procedure actually produces. When channels are drilled into the subchondral bone, marrow-derived stem cells migrate into the defect and form a repair tissue — but that tissue is fibrocartilage, not the native hyaline cartilage it replaces. The distinction matters because fibrocartilage is structurally inferior: it has lower stiffness, reduced wear resistance, and a limited capacity to distribute load across the joint surface. Under the repetitive high-force cycles of running, cutting, and pivoting, it degrades in ways that native cartilage does not.

DiBartola's 2016 meta-analysis, cited over 125 times, quantified this histologically: microfracture produces a significantly lower proportion of hyaline cartilage compared with osteochondral autograft techniques. This finding helps explain a recognised clinical pattern — good early symptom relief in the first twelve to eighteen months, followed by measurable deterioration by two to three years in active patients. A 2024 systematic review confirmed the same trajectory, describing midterm clinical decline as a predictable consequence of fibrocartilage's inferior biomechanical properties under load.

There is a further, practically important consequence. The drilling process disrupts the subchondral bone plate, and that disruption can compromise the biological and mechanical substrate that later cartilage restoration procedures depend on. For a patient who may want or need a more definitive repair in future — whether OATS, a cell-based technique, or another approach — prior microfracture can reduce the available options and complicate their delivery. The choice made at the first intervention therefore carries weight beyond the immediate result.

What OATS transplants and why hyaline cartilage holds

OATS works by relocating tissue rather than regenerating it. One or more cylindrical plugs — each comprising a column of subchondral bone capped with intact hyaline cartilage — are harvested from a low-load region of the knee, typically the medial trochlear ridge, and press-fitted into the prepared defect. Where a single plug is insufficient, mosaicplasty arrays several smaller cores across the defect, covering areas of up to approximately 4 cm².

The structural logic is direct. Because each graft carries its own native hyaline cartilage, the transplanted surface retains the collagen architecture, mechanical stiffness, and load-bearing capacity that fibrocartilage repair cannot replicate. Second-look arthroscopy and biopsy data from Radulescu's series of 32 patients confirmed that the transplanted plugs maintain viable hyaline cartilage; only the inter-graft spaces fill with fibrocartilaginous tissue, leaving the load-bearing cores structurally intact. By two years, graft margins were no longer visible at arthroscopy — the surface had integrated congruently with the surrounding native cartilage.

The procedure suits a specific candidate profile. A contained focal defect of 1–2 cm² is the primary target, though mosaicplasty can address larger lesions up to around 4 cm². Patients are typically under 45, with adequate subchondral bone, and engaged in high-impact or pivoting sport — the same conditions under which fibrocartilage repair is recognised to deteriorate. The 2026 constraint-based framework for chondral defect management underlines that lesion containment, subchondral bone status, and activity demands must all align before a structural transfer technique is appropriate; it is not the right choice in every scenario.

Donor-site morbidity, arising from the harvest zone, is a genuine consideration. Most published series report it as minor and well-tolerated, but the extent varies with the number of plugs required and warrants frank discussion at assessment.

What the 10-year trials actually show

The survival curves may be the starkest lens on the ten-year data. A 2018 long-term analysis found OAT graft survival remained above 80% through the first seven years and above 60% at fifteen years; the microfracture curve dropped substantially earlier. Those trajectories help contextualise the outcome figures introduced at the outset — the difference between techniques is not a snapshot at a single timepoint but a diverging pattern that widens as cumulative load accumulates on the repair tissue.

Gudas's 2012 randomised clinical trial, cited over 440 times, was designed exclusively around athletes — a deliberate choice, because the investigators understood that activity demand would determine whether repair tissue could endure. Its single-centre Lithuanian design is the principal limitation: replication across larger multi-centre athlete cohorts would add generalisability that the trial alone cannot provide. Solheim's 2017 independent long-term comparison addressed part of that concern: in a separate series, mosaicplasty's advantage over microfracture was statistically significant and clinically meaningful at every time point assessed — short-term, medium-term, and ten years.

The most instructive counter-evidence is Ulstein's 2014 Level II RCT, with a median 9.8-year follow-up. Among 25 patients drawn from a mixed-activity, non-exclusively-athletic population, no statistically significant difference emerged on Lysholm score (MF 69.7, OAT 62.6), KOOS, isokinetic strength, or radiographic OA — and both groups achieved modest absolute results. This is not a refutation of Gudas; it is its complement. Lower activity demand reduces the repetitive high-load cycling that exposes fibrocartilage's inferior biomechanical properties, so microfracture's biological disadvantage has less opportunity to translate into clinical failure. The Ulstein population was not the Gudas population. For a patient in high-impact or pivoting sport, the relevant evidence base is athlete-specific — and that evidence consistently points toward structural restoration.

Which patients are and are not strong candidates

Real-world practice reflects what the trial evidence suggests. In the German Cartilage Registry of 5,143 patients, marrow stimulation techniques dominated for smaller lesions and older or lower-demand patients, while structural restoration was preferred when activity demands were high — a distribution that maps closely onto the biological rationale covered in earlier sections.

Microfracture retains a place in that picture, but a defined one. For acute, contained defects under 2 cm² in patients with modest physical demands — or as a bridging procedure in younger patients ahead of a more definitive repair — it remains a reasonable choice. What the evidence does not support is its use as the default first step for high-impact sport: the fibrocartilage it produces will be tested hardest by exactly the patients most likely to be offered it at a younger age. Ulstein's 9.8-year follow-up, in a mixed-activity non-athlete cohort, recorded Lysholm scores of around 69 for microfracture and 63 for OAT mosaicplasty — modest absolute results for both groups, underlining that neither technique performs well when patient selection is imprecise.

For defects exceeding roughly 4 cm², or where available autograft volume is insufficient, the appropriate escalation is to fresh osteochondral allograft (OCA) for posttraumatic or extensive lesions, or to cell-based repair — MACI or ACI — when the defect size outstrips what a single-stage transfer can reliably cover.

One variable cuts across every technique: alignment. Uncorrected varus or valgus malalignment concentrates load on any repair surface, accelerating the same deterioration that inferior tissue quality produces through a different route. An osteotomy adjunct may be required and must be assessed before any repair is planned. Augmentation of OATS with bone marrow aspirate concentrate (BMAC) is an emerging consideration, with early evidence indicating meaningfully better structural healing scores at twelve months compared with standard technique alone.

Getting a cartilage assessment in London

Confirming which procedure fits requires a structured assessment rather than a technique preference applied in isolation. MRI characterises defect depth, size, and subchondral bone integrity; activity profile and sport demand set the biological threshold the repair tissue must meet; alignment and meniscal status determine whether any repair will be adequately unloaded. Together, those variables determine whether OATS, an alternative restoration strategy, or a different pathway altogether is appropriate for a given patient — the same constraint-based logic that runs through the evidence reviewed above.

For patients in London exploring osteochondral restoration, Professor Paul Y. F. Lee's specialist experience in knee cartilage procedures is the relevant clinical context for that assessment. A consultation at the London Cartilage Clinic, Harley Street, can be arranged via londoncartilage.com.

1. [1] Factors Affecting Choice of Surgical Treatment of Cartilage Lesions of the Knee: German Cartilage Registry (5143 patients, 2024). (2024). https://doi.org/10.1177/23259671241255672 https://doi.org/10.1177/23259671241255672
2. [2] Arthroscopic OAT in Patients with Focal Osteochondral/Chondral Lesions of the Knee — Mid-Term Clinical Outcome (2025). (2025). https://doi.org/10.5704/MOJ.2507.004 https://doi.org/10.5704/MOJ.2507.004
3. [3] Enhancing Bone–Cartilage Interface Healing in OAT: Effects of BMAC Augmentation and Rehabilitation Protocols (2025). (2025). https://doi.org/10.3390/life15071066 https://doi.org/10.3390/life15071066
4. [4] Contemporary Management of Knee Chondral Defects, Part I: A Framework for Treatment Selection (2026). (2026). https://doi.org/10.1055/a-2865-1757 https://doi.org/10.1055/a-2865-1757
5. [5] Microfracture and Microfracture Plus of the Knee Joint (2024, Clinical Sports Medicine). (2024). https://doi.org/10.1016/j.csm.2024.10.003 https://doi.org/10.1016/j.csm.2024.10.003