The short answer for most patients

In most patients asking about long-term knee preservation, the key decision is not simply which operation sounds best, but what is actually making the joint fail: a focal cartilage defect, abnormal loading, or both together. For a symptomatic focal knee cartilage defect, OATS or mosaicplasty generally has a stronger case than microfracture when the aim is lasting preservation, because it restores osteochondral plugs with native hyaline cartilage rather than relying on fibrocartilage repair tissue, which may be less durable over time.[^1][^2]

Alignment matters just as much. When varus or valgus malalignment is driving overload into one compartment, an otherwise well-chosen cartilage procedure may struggle unless that load is corrected; in those cases, high tibial osteotomy or distal femoral osteotomy is often the central joint-preservation step, either on its own or combined with cartilage restoration.[^3][^4] This discussion applies to focal defects and overload-driven preservation, not diffuse end-stage osteoarthritis where replacement may be the more appropriate pathway.

Why OATS and microfracture drift apart over time

The direct comparison that matters most here is not perfectly one-way at every timepoint. In a randomised trial of 25 patients with full-thickness femoral condylar lesions, 11 had microfracture and 14 had OAT or mosaicplasty, with a median follow-up of 9.8 years. At that stage, there were no significant between-group differences in Lysholm score, KOOS, muscle strength or radiographic osteoarthritis, so the 10-year picture should not be overstated as a clean win on every measure.

Where the gap becomes harder to ignore is later follow-up. A 15 to 17 year randomised follow-up reported more poor outcomes in the microfracture group and more excellent or good results in the mosaicplasty group, and a 2020 level-I network meta-analysis found OAT had more excellent or good results beyond 3 years, with higher return-to-activity rates, while microfracture ranked last overall. The clearest explanation is the tissue left behind in the knee: OATS transfers small osteochondral plugs, so the defect is resurfaced with native hyaline cartilage on bone, whereas microfracture creates channels through the subchondral plate and depends on fibrocartilaginous repair tissue filling in. That repair may help symptoms initially, but it is less like the original joint surface, which is a plausible reason durability looks weaker over 15 years than it does at about 10.

Who OATS tends to suit better

In practice, OATS tends to be considered for a symptomatic, contained focal cartilage defect in the knee — often a femoral condylar lesion — rather than established, diffuse osteoarthritis involving the whole joint. In everyday surgical teaching, it is usually discussed for relatively small defects, often around 1 to 2 cm², with mosaicplasty sometimes extending that towards about 4 cm², although the accessible evidence here does not fix one universal cut-off. Part of its appeal is that it is a single-stage procedure using osteochondral plugs with native hyaline cartilage and underlying bone, which can be attractive in younger or active patients where surface durability matters.

That still does not mean every small lesion should default to microfracture. A 2020 synthesis of level I randomised trials suggested better medium- to longer-term outcome patterns with OAT than microfracture, so current decision-making is usually broader than lesion size alone. Surgeons also weigh lesion containment, location, access, and the trade-off of donor-site morbidity, because the graft is harvested from a lower-load part of the same knee.

When alignment matters more than the cartilage defect

Some knees fail less because of the cartilage spot seen on MRI and more because the joint is being loaded unevenly every time the leg takes weight. In that setting, treating the surface alone may leave the same wear pattern in place. The central idea in high tibial osteotomy (HTO) and distal femoral osteotomy (DFO) is correction of the mechanical axis so the overloaded medial or lateral compartment is no longer taking the same chronic stress. A more precise analogy is a doorway with a dropped hinge: replacing the scuffed threshold does not help for long if the frame still hangs out of line.

That is why osteotomy is now used for more than isolated osteoarthritis. Reviews and technical papers in 2017–2024 describe HTO as a joint-preservation procedure for varus knees with medial overload, and also as an adjunct when malalignment would otherwise undermine cartilage restoration, meniscal transplantation or ligament surgery. In selected patients, this load-shifting strategy may delay arthroplasty rather than rush towards it. The long-term case for HTO is meaningful rather than absolute: a 2024 study reported 44% overall survivorship at 20 years, rising to 62% in more favourable patients, while a 2024 systematic review found average survivorship of 74.6% at 10 years even in selected advanced medial osteoarthritis. In valgus knees, where the deformity often comes from the distal femur, DFO is commonly the more logical correction.

HTO or DFO depends on where the deformity sits

A clearer rule of thumb is possible here. In 2023–2024 deformity-based reviews, HTO is usually matched to varus alignment with medial compartment overload when the correction belongs on the tibial side, whereas DFO is more often chosen for valgus overload driven by the distal femur. The decision is therefore about the source of the malalignment, not just whether the leg looks bow-legged or knock-kneed on first glance.

That distinction still has wrinkles. A 2024 cadaveric study in valgus knees found that both medial closing-wedge HTO and lateral opening-wedge DFO reduced lateral compartment loading, with HTO unloading more as flexion increased. That does not make HTO the default in every valgus knee; it shows why surgeons map deformity location, correction plane and functional loading before deciding which bone to realign.

The literature is also lopsided. In the sources retrieved here, long-term clinical outcome data are much fuller for HTO than for DFO. So DFO is presented mainly on deformity-analysis and mechanical-axis grounds, rather than on equally mature survivorship data. That helps explain why HTO is more familiar in published knee-preservation papers, even though DFO remains the usual femoral-side answer for many valgus patterns.

What a specialist assessment is trying to answer

By the end of a knee-preservation consultation, the practical questions are usually threefold: is there a focal, repairable cartilage defect; is that area being overloaded by malalignment; and is the rest of the joint still in a state worth preserving. That framing matters because the durable answer is not a procedure name in isolation. In published knee literature from 2012 to 2024, the final plan may be OATS alone, an osteotomy alone, or a combined strategy, depending on lesion pattern, tissue quality, limb alignment, age, activity, symptoms and the background level of arthritis.[^1][^2][^3][^4]

Equally, there are limits. When cartilage loss is diffuse rather than focal, and the knee is no longer realistically preservable, joint replacement may become the more appropriate pathway. In Harley Street, London Cartilage Clinic acts as one access point for this kind of advanced cartilage and alignment assessment, with the emphasis on clarifying the problem rather than promoting one operation. If an in-person assessment would help define those options, consultations can be arranged via londoncartilage.com.

1. [1] High survivorship rate and good clinical outcomes after high tibial osteotomy in patients with radiological advanced medial knee osteoarthritis: a systematic review. (2024). https://doi.org/10.1007/s00402-024-05254-0 https://doi.org/10.1007/s00402-024-05254-0