What the long-term evidence actually shows

For active patients with a focal knee cartilage defect, the long-term survival data point clearly in one direction: mosaicplasty substantially outperforms microfracture over time, even if the gap is less obvious in the short term.

The most informative dataset comes from a comparative cohort study (203 patients; microfracture n=119, osteochondral autograft n=84) published in 2018. Using Kaplan-Meier analysis, the long-term failure rate was 66% for microfracture versus 51% for osteochondral autograft (P=0.01). More striking than the raw failure rates is the speed of deterioration: mean time to failure averaged 4.0 years after microfracture but 8.4 years after osteochondral autograft (P<0.001). The survival curves tell the same story — osteochondral autograft held above 80% survival for the first seven post-operative years and above 60% at 15 years, whereas microfracture fell below 80% within just 12 months. A matched subgroup of patients under 51 years with lesion area below 500 mm² — the population most representative of active, younger patients — showed the same advantage (log-rank P=0.001).

The Gudas randomised controlled trial, the longest in the field at 15–17 years of follow-up, adds further weight: mosaicplasty demonstrated sustained superiority, particularly in younger, more active cohorts.

A 2026 systematic review of RCTs complicates the picture somewhat. It found no statistically significant difference in patient-reported outcomes between OATS and microfracture — though the certainty of that evidence was rated as low. The likely explanation is that most RCT follow-up periods are too short to capture the divergence that becomes apparent beyond five to seven years; patient-reported scores and joint survival are not the same measurement.

The combined long-term failure rate across both techniques is approximately 62%. Mosaicplasty delays that failure substantially — roughly doubling mean functional survival — but neither procedure eliminates the underlying vulnerability of repaired cartilage.

Why fibrocartilage wears out faster than hyaline cartilage

The difference in durability comes down to biology. Microfracture works by puncturing the subchondral bone plate with an awl, releasing marrow cells that migrate into the defect and solidify into a repair tissue. That tissue is predominantly fibrocartilage — a denser, scar-like material that fills the gap but lacks the mechanical stiffness and load-distributing architecture of native hyaline cartilage. Think of it as a patch rather than a transplant: it covers the wound, but it is not the original material.

Under the repetitive loading of an active lifestyle, fibrocartilage tends to break down, commonly within two to three years — which is why the microfracture survival curve deteriorates so rapidly in high-demand knees. There is a further downstream consequence worth noting: the awl perforations damage the subchondral bone plate itself. A compromised bone plate can make subsequent cartilage repair procedures technically harder and less likely to succeed, should the microfracture eventually fail. For a younger, active patient who still has decades of knee use ahead, this is a meaningful risk.

OATS takes a different approach entirely. Rather than stimulating scar-like repair tissue, it transplants cylindrical plugs of intact bone and overlying hyaline-like cartilage from a lower-load area of the same knee into the defect. Because the plugs arrive with their full structural architecture — including the bone-to-cartilage interface — preserved, they are better equipped to handle cyclic loading over time.

That said, even transferred hyaline-like cartilage is not permanent. Mosaicplasty prolongs durable joint function rather than guaranteeing it indefinitely, and that distinction matters when setting realistic expectations.

Which patients are most likely to benefit from OATS

Deciding between OATS and microfracture hinges on three intersecting factors: the size of the defect, the patient's age and physical demands, and whether load is being distributed correctly across the joint.

Lesion size provides the clearest boundary. Microfracture has historically been applied to defects smaller than 2 cm², where its relative simplicity — single-stage, no donor harvest — carries practical weight for less active patients. OATS is typically considered for lesions between 1 and 4 cm², the range where the biological case for hyaline-like tissue quality is strongest and donor plugs from the patellofemoral sulcus remain available. Beyond 4 cm², the number of plugs required exhausts autograft supply; at that point, ACI, MACI, or a fresh osteochondral allograft generally becomes the more realistic option.

Age and activity level refine the picture further. The 2018 matched subgroup analysis — patients under 51 with lesion area below 500 mm², the demographic that best represents active, working-age individuals — confirmed the OATS survival advantage directly in that population (P=0.001). The important caveat is that randomised trials have not been powered specifically for athletic subgroups, so the precise age or activity threshold at which OATS definitively outperforms microfracture remains directional rather than precisely fixed by the current evidence base.

Mechanical alignment deserves assessment alongside the defect itself. In a varus-aligned knee, for example, the medial compartment continues to absorb disproportionate load even after repair — and there is evidence that combining mosaicplasty with concurrent high tibial osteotomy improves plug union and MRI repair scores compared with cartilage repair alone. Load correction and tissue restoration address different problems; treating only one while ignoring the other is likely to produce a suboptimal result.

For patients with widespread, diffuse arthritis involving multiple compartments, none of these focal repair procedures is appropriate — that is a different clinical conversation altogether.

Technical factors that shape mosaicplasty outcomes

Several procedural decisions — made before and during the operation — can shift the balance between a durable repair and an early failure.

Plug diameter and seating depth are among the most technically sensitive variables. A 2023 study specifically examined how plug diameter influences repair quality in mosaicplasty; the broader principle is that each graft must sit flush with the surrounding cartilage surface. A plug placed even marginally proud concentrates shear stress at its edges under load, while a recessed plug creates a step that disrupts smooth articulation. Perpendicular insertion matters for the same reason — a tilted plug produces a sloped surface rather than a continuous one.

The plugs are harvested from lower-load regions of the same knee — typically the patellofemoral sulcus or the intercondylar notch. At the standard two to four plugs, these donor sites tolerate harvest reasonably well. Beyond roughly 4 cm² of coverage, however, the number of plugs required begins to strain what those areas can comfortably provide. It is worth knowing that the harvest site carries its own recovery: some patients experience localised soreness or stiffness there, and this can overlap with the rehabilitation period. Evidence to date does not support drug treatments as a way to meaningfully accelerate that recovery in clinical terms.

Pre-operative alignment assessment operates at the planning stage rather than the operating table, but its effect on outcome is no less real. Where varus malalignment places disproportionate load on the medial compartment, a cartilage repair procedure alone leaves the mechanical cause unaddressed — increasing the risk of premature graft failure. A concurrent high tibial osteotomy, where indicated, addresses both problems together, and that planning conversation is worth raising with the surgeon well in advance.

Recovery and return to sport after mosaicplasty

Recovery from mosaicplasty is measured in months, not weeks — and the length of the rehabilitation programme itself has a measurable effect on the result.

An RCT of 37 patients with chondral defects larger than 3 cm² in the weight-bearing femoral condyle found that a structured 12-week two-phase programme produced significantly better WOMAC functional scores at intermediate and final follow-up than a shorter 6-week protocol. That finding is directly actionable: committing to the full rehabilitation course is not optional extra effort — it is part of the treatment.

Weight-bearing is reintroduced progressively. Loading the graft too early, before bone integration is established, risks plug subsidence and disruption of the repair surface — which is why the phased approach in that study produced better outcomes than the abbreviated version.

Return to higher-demand activity typically falls somewhere in the 6–12 month window, depending on defect size, the number of plugs placed, and whether concurrent procedures such as osteotomy were performed. Active patients who follow a structured programme tend to achieve better functional outcomes than sedentary individuals — controlled loading during rehabilitation appears to support, rather than threaten, graft integration, provided the timing is appropriate.

There is no single clearance date. Sport-specific load testing — assessing the knee under the movement patterns of the activity in question — is advisable before returning to competition or high-demand training. A consultant or specialist physiotherapist can guide that final progression.

Where OATS sits in the full cartilage repair pathway

Cartilage repair is a tiered field, and OATS/mosaicplasty sits roughly in the middle of that hierarchy — beyond the purely marrow-stimulating techniques but short of the options designed for much larger or more complex damage.

The broad pathway runs from conservative management and load reduction, through biologic or injection support, to surgical restorative intervention for focal structural damage, and eventually to joint replacement when preservation is no longer viable. OATS belongs at the third stage only.

Within the restorative tier, microfracture and AMIC (matrix-augmented marrow stimulation) cover the smaller end of the spectrum. Mosaicplasty steps in where hyaline-quality tissue is a priority and donor supply allows — a Cochrane systematic review of surgical cartilage interventions in adults placed it in exactly this bounded role alongside drilling, microfracture, and allograft options. Where defect dimensions exceed what autograft harvest sites can comfortably provide, osteochondral allograft (OCA) removes that constraint by using cadaveric tissue. For defects where a two-stage approach is acceptable, MACI or ACI — in which harvested cells are cultured then reimplanted — sidesteps donor-site morbidity entirely and is well-suited to larger lesions.

OATS is not indicated for diffuse or end-stage osteoarthritis; those patients belong in a joint-preservation or replacement conversation, not a focal repair pathway.

The practical upshot is that the right technique shifts with lesion dimensions, surgical history, and patient goals. A younger active patient with a small-to-moderate focal defect is in the core territory where mosaicplasty evidence is strongest; the same patient with a substantially larger lesion sits in allograft or cell-therapy territory regardless of fitness level. Across every option in this hierarchy, restorative surgery delays the trajectory towards arthritis rather than reversing it — a realistic expectation that should be central to any pre-operative discussion, and the question most worth putting directly to a surgeon.

1. [1] Long-Term Survival after Microfracture and Mosaicplasty for Knee Articular Cartilage Repair: A Comparative Study Between Two Treatments Cohorts. (2018). https://doi.org/10.1177/1947603518783482 https://doi.org/10.1177/1947603518783482
2. [2] Third-generation autologous chondrocyte implantation versus mosaicplasty for knee cartilage injury: 2-year randomized trial. (2016). https://doi.org/10.1002/jor.23152 https://doi.org/10.1002/jor.23152
3. [3] Randomized Study of Long-term (15–17 Years) Outcome After Microfracture Versus Mosaicplasty in Knee Articular Cartilage Defects. (2018). https://doi.org/10.1177/0363546517745281 https://doi.org/10.1177/0363546517745281
4. [4] Surgical interventions (microfracture, drilling, mosaicplasty, and allograft transplantation) for treating isolated cartilage defects of the knee in adults. (2016). https://doi.org/10.1002/14651858.CD010675.pub2 https://doi.org/10.1002/14651858.CD010675.pub2
5. [5] Microfracture with or without collagen scaffold insertion for adults with chondral or osteochondral defects of the knee: the SISMIC RCT and its challenges during and after the COVID-19 pandemic. (2025). https://doi.org/10.3310/BRTS2415 https://doi.org/10.3310/BRTS2415
6. [6] Fixation of unstable osteochondritis dissecans lesions of the knee using arthroscopic autogenous osteochondral grafting (mosaicplasty). (2007). https://doi.org/10.1016/J.ARTHRO.2007.02.017 https://doi.org/10.1016/J.ARTHRO.2007.02.017
7. [7] A New Approach to Postoperative Rehabilitation following Mosaicplasty and Bone Marrow Aspiration Concentrate (BMAC) Augmentation. (2024). https://doi.org/10.3390/biomedicines12061164 https://doi.org/10.3390/biomedicines12061164
8. [8] The effects of alendronate on the suppression of bone resorption and the promotion of cartilage formation in the human mosaicplasty donor site: A randomized, double-blind, placebo-controlled prospective study. (2021). https://doi.org/10.1016/j.jos.2021.05.007 https://doi.org/10.1016/j.jos.2021.05.007