What the evidence says about avoiding knee replacement

For many patients facing that question, the honest answer — backed by a decade of follow-up data — is yes, provided the damage is the right kind and the knee is otherwise in good shape.

The strongest evidence comes from two independent long-term cohort studies tracking patients who had undergone matrix-associated autologous chondrocyte implantation (MACI), a technique that uses the patient's own cartilage cells grown on a collagen scaffold. One cohort followed 168 patients for ten years; a separate group tracked 204. Both arrived at strikingly similar findings: total knee arthroplasty (TKA) conversion rates of just 7.4% and 9.3% respectively. In plain terms, roughly nine in ten appropriately selected patients avoided knee replacement over an entire decade. Of those who retained their own joint, 92% reported lasting satisfaction with their pain relief — and outcomes held steady throughout the follow-up period, ruling out any short-lived honeymoon effect in the early years after surgery.

These figures represent the most robust TKA-avoidance data currently available in the cartilage repair field.

One important distinction is worth stating early: the procedure in these studies is biological restoration — rebuilding the cartilage surface at a cellular level. This is not the same as a simple arthroscopic washout or debridement, a historical procedure that involves flushing or tidying the joint. Washout has no evidence of delaying or avoiding knee replacement, and the two should not be conflated.

The 'nine in ten' result does not apply universally. It reflects a specific patient profile — explored in the next section — and that distinction matters enormously.

Which patients are realistic candidates

Repair works best when the damage is a single, well-defined area — think of it as a pothole in an otherwise sound road, with healthy cartilage around the edges to anchor new tissue against. That stands in contrast to the crumbling road surface of diffuse osteoarthritis, where there are no intact borders left to build against. Once damage has spread across the whole joint surface — the classic 'bone-on-bone' picture — no current repair technique can succeed, and the appropriate pathway shifts toward joint replacement. Knowing this early is genuinely useful: it focuses patients toward the right decision rather than delaying it.

Several features make repair realistic:

• Defect type and size. A focal, isolated defect — most often from a sports injury or osteochondral injury — is the target. Defect size shapes which technique applies: marrow-stimulation approaches suit smaller areas (historically under 2 cm²), while cell-based options such as MACI and allograft transfer extend to larger lesions.
• Joint alignment. A knee that loads unevenly will stress a repair site from day one. Malalignment is not a disqualifier, but it must be corrected — often with an osteotomy procedure — before or alongside repair.
• Meniscal integrity. A deficient or absent meniscus alters load distribution in ways that can undermine even technically sound repairs; meniscal preservation or transplant may need to form part of the plan.
• Age and activity. The literature consistently refers to 'younger, active patients', though no firm numerical cut-off is established. Age is one factor among several, not a stand-alone criterion.

A thorough assessment — covering imaging, alignment, and activity goals — is needed to determine where any individual sits on this spectrum.

The repair and restoration techniques available

Several techniques now sit within the joint-preservation menu, and the choice between them turns on defect size, depth, and what the joint can sustain. Worth foregrounding: a 2025 matched comparison of MACI, AMIC, and minced cartilage implantation across 48 patients found all three produced equivalent improvements in pain and function at two years — a result that cautions against treating technique hierarchy as fixed or settled.

ChondroFiller injection sits at the least invasive end of the restorative pathway. Delivered via ultrasound-guided outpatient injection, this acellular collagen scaffold works through matrix-induced chondrogenesis — providing a structure into which the body's own progenitor cells migrate and begin organising repair tissue. No theatre admission is required.

Microfracture has historically been the starting point for small defects under 2 cm², using fine drilling to draw marrow cells to the surface. The repair tissue it generates is fibrocartilage rather than native hyaline cartilage — more scar-like and less durable — with breakdown reported at two to three years and a risk of subchondral bone damage that can limit future options. Its use as a routine first choice is now declining.

AMIC (autologous matrix-induced chondrogenesis) augments microfracture with a collagen scaffold, giving progenitor cells a structure to form around. Single-stage and less resource-intensive than cell-based approaches, it offers a practical middle ground for intermediate defects — and the 2025 comparison places it on equal footing with MACI at two years.

MACI and ACI are two-stage, cell-based procedures for larger defects — typically 2–10 cm². Cartilage cells are harvested, cultured in a laboratory, and implanted on a membrane at a second operation. The tissue produced sits closer to native hyaline cartilage, which likely underpins the durable long-term outcomes seen in cohort data.

OATS and mosaicplasty transfer bone-cartilage plugs from a low-load zone of the same knee to fill defects in the 1–4 cm² range. Single-stage and reliable for that size, though donor-site discomfort is a meaningful trade-off to weigh.

Osteochondral allograft (OCA) uses donor tissue for larger or deeper posttraumatic lesions where autograft is insufficient — extending the restorative option to defects beyond the scope of the techniques above.

When alignment or meniscal problems are part of the picture

Cartilage repair does not work in isolation. Even technically sound grafts can fail prematurely if the mechanical environment works against them — and two factors account for most of those failures: abnormal limb alignment and meniscal insufficiency.

Osteotomy (high tibial or distal femoral, depending on which compartment is overloaded) reshapes the bone to shift weight-bearing force away from the damaged area. It does not repair cartilage directly, but it creates the mechanical conditions under which a repair can survive. In younger patients with unicompartmental loading and early cartilage change, osteotomy alone can serve as a meaningful joint-preservation strategy, deferring replacement while the biology remains favourable.

Meniscal preservation or transplant addresses the other common risk: a torn or absent meniscus removes the joint's natural load-sharing cushion, concentrating force on the repaired surface regardless of which technique was used. Restoring meniscal function — through repair where tissue remains viable, or transplant where it has been lost — protects the cartilage repair from the excessive stress that would otherwise undermine it.

Both procedures are typically combined with cartilage restoration rather than offered instead of it. Identifying which mechanical factors are present — and how significant they are — is best done before any repair is planned. Objective biomechanical assessment, including AI-assisted markerless motion capture such as MAI Motion®, can quantify abnormal loading patterns that static imaging alone may not reveal, giving the clinical team a clearer picture before planning begins.

For patients not yet at the point of surgery, unloader bracing offers a non-operative way to reduce compartment loading — a more modest effect than osteotomy, but a relevant option at earlier decision stages.

Realistic expectations for recovery and long-term durability

Knowing the headline outcomes is different from understanding what they mean in practice — especially for a patient deciding whether to proceed with surgery.

The approximately 8–10% TKA conversion rate at ten years in MACI cohorts deserves honest framing in both directions: the large majority of appropriately selected patients avoid replacement within a decade, but a meaningful minority do not. Patient selection is carrying much of the weight in those figures; they cannot be applied to knees with diffuse damage, significant malalignment, or absent meniscal support.

Emerging single-stage techniques add to the picture over shorter horizons. A five-year study of 28 patients treated with single-step autologous minced cartilage — mean age 29.5 years, mean defect size 3.5 cm² — recorded pain scores (NAS) falling from a median of 7 before surgery to 2 at five years, with 75% reaching the patient-acceptable symptom state on IKDC scoring and a revision surgery rate of 3.5%. These figures are encouraging, though five-year follow-up does not yet match the decade-long MACI data.

Mesenchymal stem cell (MSC) injections for knee osteoarthritis represent a distinct and earlier-stage category. A systematic review of 17 studies found improved pain and function in most cases at under 28 months, with the majority also reporting MRI evidence of cartilage change. What that evidence does not yet provide is a long-term TKA-avoidance rate comparable to established cell-based repair data. MSC therapy for OA remains an area of active investigation rather than settled proof.

Recovery timelines vary considerably by procedure and individual circumstances. Two-stage cell-based techniques typically involve a longer rehabilitation arc than single-stage approaches; a consultant assessment is needed to give realistic, personalised guidance rather than fixed numbers. One reassurance worth stating plainly: a repair that does not hold its result does not foreclose joint replacement. For most appropriately selected patients it averts that step entirely — and for the minority for whom it does not, the pathway forward remains open.

Starting the joint-preservation conversation

The article's central answer is conditional but evidence-supported: cartilage repair can delay or avoid knee replacement, provided the damage is focal, the joint is otherwise mechanically sound, and any alignment or meniscal issues are addressed alongside the cartilage itself. Whether those conditions apply to a particular patient is precisely what a structured specialist assessment is designed to establish — looking at defect grade, limb alignment, meniscal integrity, and activity goals as a whole rather than in isolation.

The NHS treatment pathway supports this sequence: conservative management first, surgical options where indicated. A specialist consultation through MSK Doctors works within that clinical logic but without NHS waiting times or a GP referral requirement. Sleaford and Grantham both offer on-site Open MRI and biomechanical evaluation to support that assessment; London-based patients are seen through the London Cartilage Clinic. Consultations can be booked directly at mskdoctors.com.

A decade of follow-up data is encouraging. What it cannot do is answer the question for any individual knee — that is what the assessment is for.

1. [1] Comparison of MACI vs AMIC vs Minced Cartilage: 2-Year Patient-Reported Pain and Function. (2025). https://doi.org/10.3390/jcm14072194 https://doi.org/10.3390/jcm14072194
2. [2] Intra-articular Mesenchymal Stem Cells in Osteoarthritis of the Knee: A Systematic Review. (2019). https://doi.org/10.1016/j.arthro.2018.07.028 https://doi.org/10.1016/j.arthro.2018.07.028
3. [3] Autologous minced cartilage repair for chondral and osteochondral lesions of the knee: good outcomes and low reoperation rates at minimum 5-year follow-up. (2023). https://doi.org/10.1007/s00167-023-07546-1 https://doi.org/10.1007/s00167-023-07546-1
4. [4] CO2 Insufflation in Cartilage Repair Using Minced Autologous Cartilage, PRP, and Autologous Thrombin. (2025). https://doi.org/10.1016/j.eats.2025.103835 https://doi.org/10.1016/j.eats.2025.103835