Do I still have joint‑preserving options before replacement?

Symptoms that mainly come from one side of the knee (the inner/medial side or the outer/lateral side) do not automatically mean a knee replacement is the next step. When wear is still largely confined to one compartment, and the rest of the joint structure is reasonably preserved, joint-preserving strategies may reduce pain and, in some cases, delay arthroplasty.

A practical way to frame the options is as a staged pathway, moving from low-risk measures to more involved surgery:

• Symptom management and load modification: physiotherapy and activity modification, with unloader bracing used to shift load away from the painful compartment. Biomechanical studies consistently show valgus or varus unloader braces can reduce knee adduction moments and medial compartment forces during walking tasks, while increasing load on the opposite side. (This helps explain why bracing may improve symptoms but does not suit every pattern of disease.) [1]
• Joint-preserving surgery to change mechanics: osteotomy is used to realign the leg and offload the worn side (high tibial osteotomy/HTO is typically used when varus alignment drives medial overload; distal femoral osteotomy/DFO is used when valgus alignment drives lateral overload). In a 9.4-year follow-up series of medial closing-wedge DFO, alignment correction and patient-reported outcomes improved while osteoarthritis grade was generally stable when joint-line obliquity stayed within about 4°. [2]
• Focal cartilage restoration (when damage is localised): procedures such as AMIC, ACI/MACI, OATS and OCA aim to repair a discrete area of cartilage loss, often alongside correction of alignment when malalignment is part of the problem.
• Knee replacement: usually becomes the more realistic option once wear is more widespread (multiple compartments, diffuse cartilage loss) or when previous joint-preserving strategies have not controlled symptoms.

The key distinction is between a focal cartilage defect and more diffuse osteoarthritis. A focal defect is a localised “patch” of damage—often described as Outerbridge/ICRS grade III–IV when it is deep or full thickness—but still limited in surface area. Diffuse osteoarthritis is broader wear across much of a compartment (or more than one compartment), often with progressive changes elsewhere in the joint. Many cartilage-repair techniques are designed for the former pattern, not for established, widespread arthritis.

This article mainly fits situations where symptoms are still dominated by one compartment—medial or lateral—often with varus or valgus malalignment, and either a contained focal defect or early compartmental wear. In that setting, combining mechanical offloading with focal repair is one of the recurring themes in the evidence: for example, a 2009–2016 series of 66 varus knees treated with medial open-wedge HTO plus OATS reported substantial Knee Society Score improvements and 96.7% survivorship at a mean 9.49 years. [3]

The sections that follow map where each tool tends to sit by defect size and complexity: AMIC has published mid-term improvements for roughly 2–8 cm² lesions; MACI/ACI is often used for larger or more complex defects in a two-stage pathway; OATS is a single-stage option for small-to-moderate focal lesions; and OCA is generally reserved for larger, osteochondral, multisurface, or salvage scenarios, with series reporting meaningful reoperation and failure risks over 5–10 years. This guide stays focused on how these options fit clinically, rather than promoting any specific provider. [4–8]

Is microfracture still used for medium knee defects?

Microfracture was widely adopted as a keyhole (arthroscopic) option because it is a single operation: the surgeon makes multiple small holes in the bone just under the cartilage defect, aiming to release marrow cells and form a clot that “caps” the damaged area. The repair tissue produced is often described as more fibrous and less like native joint cartilage, which is one reason durability has been questioned for larger, higher-demand defects.

Historically, its appeal was practical rather than “high-tech”: one stage, no cultured cells, and no additional implant. That made it easier to offer at scale during the early growth of arthroscopic cartilage surgery.

For medium-sized focal defects (commonly in the ~2–5 cm² range), the direction of the evidence has shifted. A systematic review of randomised trials comparing third-generation ACI/MACI-type techniques with microfracture (follow-up 2–6 years) reported better patient outcomes after ACI in most studies, and lower reported failure in ACI groups (0–1.8%) than in microfracture groups (2.5–8.3%). The same body of evidence suggests the advantage is not just “scores on paper”: domains linked to pain and day-to-day function tended to improve more consistently after cell-based repair than after microfracture in these trials and trial syntheses. [9]

A separate meta-analysis of 12 RCTs (lesions 2.3–10 cm², follow-up up to 5 years) found that headline knee-function improvements could look similar between approaches, but pain, quality-of-life and activities-of-daily-living domains favoured ACI. [10] At ≥5 years, another review found collagen-membrane ACI and MACI produced significantly better knee scores and activity measures than microfracture, reinforcing a pattern that microfracture can be less durable as follow-up lengthens. [11] Longer-term concerns often discussed in the broader cartilage literature include the mechanical limits of fibrocartilage repair and potential changes in the underlying bone after marrow-stimulation; these issues are not fully pinned to any single trial, but they help explain why many teams now avoid microfracture as the default for medium defects.

In practice, surgeons’ indications have narrowed. In an ICRS survey of 385 cartilage surgeons, 56% reported limiting microfracture to lesions ≤2 cm², and 15% reported not using microfracture at all. [12]

• A typical “decision-changing” example is a relatively active person with a contained ~3 cm² full-thickness defect: published trial evidence suggests many surgeons now lean towards matrix- or cell-based repair rather than microfracture alone.
• By contrast, a smaller ≤2 cm² defect, or a lower-demand situation where simplicity is prioritised, remains one of the scenarios where microfracture is still used by some surgeons.

Where does AMIC fit compared with microfracture and MACI?

AMIC as a practical “middle ground”

AMIC (autologous matrix‑induced chondrogenesis) is usually described as microfracture plus a collagen membrane scaffold done in the same operation. The marrow stimulation creates a blood clot rich in repair cells, and the membrane is intended to help stabilise that clot and provide a structure for new repair tissue to form, without moving to a two‑stage, lab‑grown cell process. [4]

What published outcomes look like for medium‑sized defects

For the defect sizes where surgeons often feel microfracture alone becomes less reliable (roughly 2–8 cm²), AMIC has consistent short‑ to mid‑term results in clinical series. A multicentre retrospective study of 101 patients (mean defect 3.44 cm²) reported significant improvements in patient‑reported scores including KOOS and IKDC at a mean follow‑up of around 30 months. MRI assessment also suggested good defect fill, with a mean MOCART score of 75% at 2 years. [4]

A larger 2024 systematic review pooling 18 studies (490 patients) reported clinically important improvements across commonly used knee scores (including Lysholm, IKDC and KOOS) alongside substantial reductions in VAS pain. Taken together, those data support AMIC as a credible single‑stage option when the target is meaningful symptom and day‑to‑day function improvement in a focal defect of about 3–4 cm² (and, in some series, larger). [5]

The “activity ceiling” seen in systematic review data

One recurring pattern in the 2024 review was that activity level, measured using the Tegner score, improved only modestly and not significantly overall. That gap matters in practice: published outcomes often support better pain and function for walking, stairs and general fitness, while return to high‑demand pivoting sport may remain less predictable even when the cartilage fill on MRI appears good. [5]

AMIC versus microfracture and MACI: what tends to drive choice

Compared with traditional microfracture, AMIC keeps the one‑stage appeal but adds a stabilising matrix, and the AMIC series above report both clinical improvement and MRI fill that is often used as a proxy for repair quality. In parallel, randomised trial syntheses in the ~2–5 cm² range have tended to favour cell‑based ACI/MACI‑type techniques over microfracture on patient‑reported outcomes at 2–6 years, reinforcing the broader trend away from “bare” microfracture for medium lesions where durability matters. [9, 11]

Head‑to‑head comparisons between AMIC and matrix‑associated ACI (mACI/MACI) remain limited. A systematic review comparing the two found that both approaches produce substantial clinical improvement, and some mid‑term series suggest AMIC can be similar or slightly better on certain outcomes; however, the included studies were largely non‑randomised and heterogeneous, so a definitive “best procedure” conclusion is not supported by the current evidence base. In day‑to‑day decision‑making, AMIC is therefore often positioned for contained, medium‑sized focal defects where a single‑stage operation is preferred, while MACI/ACI is more often considered when defects are larger, multiple, or otherwise complex and a two‑stage pathway is acceptable. [18]

When is cell‑based ACI or MACI still worth considering?

Two‑stage cell‑based repair is usually considered when the shape and scale of the cartilage problem make a single‑stage option less suitable, even if that means accepting two procedures. Put simply, the extra stage is “buying” the ability to cover a broader surface area with implanted cartilage cells, which can matter when a defect is too large, too shallow‑wide, or too multi‑focal for plug techniques.

What ACI and MACI are trying to achieve (and why they take two stages)

ACI and MACI sit in the “cell‑based restoration” category: cartilage cells are harvested first, expanded in a lab, then re‑implanted at a second procedure. In third‑generation ACI techniques, those cells are delivered within a collagen membrane (often grouped under MACI‑type procedures), aiming for a more durable repair than marrow‑stimulation alone when defects are in the ~2–5 cm² range. [9]

Where the evidence still supports choosing ACI/MACI over microfracture

The practical niche for ACI/MACI is strongest where surgeons worry about durability: medium‑sized full‑thickness defects (for example around 3–5 cm²) in people who need reliable symptom control over several years. In randomised trial syntheses with 2–6 years follow‑up, third‑generation ACI/MACI‑type approaches have generally shown better patient‑reported outcomes than microfracture in this size band, and reported lower failure ranges in ACI groups than microfracture groups. At ≥5 years, reviews have also reported better knee scores and activity measures after collagen‑membrane ACI and MACI than after microfracture, supporting the idea that the difference becomes clearer as follow‑up lengthens. [9, 11]

Mid‑ to long‑term case series of next‑generation MACI variants report sustained clinical and radiological improvements, and technical reports describe MACI being used to address multiple and large lesions within a compartment, often with high satisfaction and quality‑of‑life gains in published follow‑up. These studies are not the same as large head‑to‑head randomised comparisons across every technique, but they help explain why some teams still accept the logistics of a two‑stage pathway when the defect pattern is complex. [6, 13]

How this compares with single‑stage alternatives (OATS and OCA)

A useful way to hold the options in mind is to match the procedure to the footprint (surface area) and depth (pure cartilage vs bone‑and‑cartilage) of the injury.

• OATS / mosaicplasty tends to fit small to moderate, contained focal defects (often around 1–3 cm², and in some settings mosaicked towards ~4 cm²). It is single‑stage and transfers the patient’s own osteochondral plugs. In a 2026 systematic review/meta‑analysis of single‑stage treatments (mean lesion about 3.3 cm²), OATS showed large improvements in knee scores (for example, an IKDC gain of roughly 40 points) and substantial pain reduction (VAS about −4.27) at a mean follow‑up of about 4.5 years. The main trade‑off often discussed clinically is the “borrowed” graft from elsewhere in the knee, so donor‑site symptoms can be a relevant consideration in higher‑demand patients. [14]
• OCA (osteochondral allograft) is a different category again: larger grafts from a tissue bank, commonly used for bigger defects (often >4–5 cm²), deeper osteochondral problems, bipolar lesions, or salvage after a previous cartilage procedure has failed. In a systematic review of secondary (salvage) OCA, reported graft survival was roughly 79–88% at 5 years and 61–82% at 10 years, but with a 42.8% reoperation rate and 16.6% overall failure across included studies; outcomes were notably worse in very large defects (for example 9–10 cm² lesions were associated with much higher failure). A prospective cohort of 186 primary OCAT cases reported 23.1% failures at about 6.5 years, with higher risk in multisurface/bipolar grafts and when combined with other major procedures. [7, 8]

A simple “pattern of damage” way to think about selection

Rather than treating the choice as a yes/no question about one operation, many services use a broad pattern‑matching approach:

• Small, contained defects: OATS is often considered ahead of microfracture in higher‑demand settings; paediatric/young‑cohort evidence also reflects this size‑led logic. [15]
• Small–moderate focal defects (~2–3.5 cm²): OATS or other single‑stage scaffold approaches may be considered, depending on containment and goals. [14]
• Medium–large single or multiple defects (often ~3–8 cm²) in a reasonably healthy joint: two‑stage MACI/ACI remains in play when the surface area is too broad for plugs or there are multiple lesions to address in one compartment. [6, 13]
• Very large, deep osteochondral, multisurface or bipolar problems, or failed prior cartilage surgery: OCA is often positioned as a joint‑salvage option, with meaningful 5–10‑year survival reported in series but higher reoperation and failure risk as complexity rises. [7, 8]

Across all of these choices, the main uncertainty is not whether each technique “can work” in published studies, but the lack of perfect, long‑term, like‑for‑like comparisons across AMIC, OATS, MACI/ACI and OCA for every defect pattern—so decisions often hinge on defect mapping, joint health, and how much complexity is reasonable for the expected gain. [9]

How do osteotomy and focal cartilage repair work together?

The key message up front (rather than treating it as a yes/no choice) is that load and alignment often decide whether a focal cartilage repair will cope over time. In a bow‑legged (varus) knee, the inner (medial) side is asked to carry more force with every step; in a knock‑kneed (valgus) knee, the outer (lateral) side takes the hit. When that overload is the main driver, “fixing the pothole” in the cartilage without also shifting the load may leave the repaired area doing the same job that wore it out in the first place.

HTO (high tibial osteotomy) and DFO (distal femoral osteotomy) are the main joint‑preserving ways to address that problem. They use a controlled bone cut to subtly change the leg’s overall line so the knee is loaded more evenly, aiming to offload the worn compartment while keeping the person’s own joint surfaces. In practice, osteotomy is typically considered in younger or middle‑aged people with one clearly overloaded compartment and good movement, because it is a major operation with its own risks and rehabilitation demands.

When there is both (1) malalignment and (2) a discrete, repairable defect, surgeons may combine realignment with a focal cartilage procedure in the same plan. A concrete example is a 2024 report of 66 varus knees with medial compartment osteoarthritis treated with medial open‑wedge HTO plus OATS (operations performed between 2009 and 2016): knee and function scores improved substantially, and survivorship was 96.7% at a mean 9.49 years, with only 2 conversions to total knee replacement. Where second‑look arthroscopy was performed (57 knees), the authors reported complete cartilage regeneration, described as hyaline‑like in 49/57 cases—supporting the idea that correcting mechanics can create a more favourable environment for a focal repair to hold up. [3]

Choosing between HTO and DFO is mainly about where the deformity comes from and which side of the knee is failing. Long‑term clinical data for valgus (lateral‑side) arthritic knees show medial closing‑wedge DFO can give durable correction and sustained improvements in pain and patient‑reported scores at around 9.4 years follow‑up, particularly when surgeons avoid creating an excessive “tilt” at the joint line (kept within about 4° in that series). The practical point of that threshold is not the number itself, but the principle: realignment needs to shift load without creating a new, awkward joint orientation that could cause fresh symptoms elsewhere. [2]

Biomechanical work helps explain why the details of correction matter. In a cadaveric comparison using an 8° varus correction for valgus knees, a varus‑producing HTO reduced lateral compartment contact pressures more consistently through 30°–90° of flexion than a lateral opening‑wedge DFO, whereas after DFO the lateral pressures in deeper flexion approached the native state. This type of data does not “pick a winner” for every person, but it supports the broader planning logic: osteotomy choice is about the pattern of loading across the arc of motion, and cartilage repair (AMIC, MACI, OATS) is usually layered on only when there is a focal target—rather than diffuse, end‑stage wear throughout the compartment. [16]

Where do unloader braces fit in this pathway?

A well‑fitted unloader brace is the most reversible way to change how force travels through a knee with one clearly painful compartment (medial with varus “bow‑leg”, or lateral with valgus “knock‑knee”). In plain terms, it applies a gentle corrective push so the knee spends more of each step on the healthier side, which is why many people describe the main benefit as pain easing while the brace is on.

Biomechanical studies of valgus/varus unloader braces have consistently reported a reduction in medial “adduction” loading and medial compartment forces during tasks such as level walking, stairs and slope walking, with a corresponding increase in load on the opposite (less worn) side. That trade‑off matches the intended job of the brace: symptom relief by shifting pressure away from the sore compartment rather than “fixing” the cartilage itself. [1]

When the question becomes “brace versus a structural correction”, the clearest head‑to‑head evidence comes from a multicentre randomised trial in adults aged 18–65 with symptomatic medial compartment osteoarthritis. At 12 months, valgus high tibial osteotomy (HTO) produced a much larger improvement in KOOS pain than a valgus unloader brace (about a 28‑point better group difference), supporting osteotomy as the stronger option for people who are suitable for surgery and want a longer‑acting change in joint mechanics. [17]

Expectation‑setting matters because bracing sits mainly in the symptom‑modifying part of the joint‑preservation ladder. Evidence is limited on whether bracing meaningfully slows structural progression or reliably delays arthroplasty over the long term; the more predictable effect is improved comfort and function during use. [1]

In practice, unloader bracing is often used in three ways in unicompartmental disease:

• alongside physiotherapy and weight management as an early, non‑operative step in medial or lateral compartment overload
• as a “test‑drive” of unloading: if symptoms improve when load is shifted, that can support the logic of considering an osteotomy or a cartilage procedure in a corrected mechanical environment
• as an adjunct for selected higher‑load days even after joint‑preserving surgery, recognising it helps manage forces rather than regenerate tissue

Seen this way, a brace is a reversible trial of unloading; an osteotomy is the structural, longer‑term version. Search MSK can be used to find UK specialists who assess unicompartmental knee pain and offer bracing, osteotomy and cartilage‑preserving options—filtered by region and treatment.