Which repair lasts longer — the direct answer

For patients with a focal knee cartilage defect of 3 cm² or larger, the best available evidence consistently points to MACI as the more durable option. The SUMMIT trial — a Level 1 prospective randomised controlled trial of 144 patients — is the only head-to-head study with five-year follow-up, and at both the two- and five-year marks MACI produced significantly greater improvements in pain, function, and quality of life than microfracture.

Microfracture is not without short-term merit: many patients notice meaningful symptom relief in the first year or two. The problem is longevity. Clinical gains from microfracture tend to plateau and then decline around the three-to-five-year mark, a trajectory that reflects the nature of the tissue it produces. MACI, by contrast, maintained its functional advantage at five years, with statistically significant superiority in KOOS pain (P=.022) and activities of daily living (P=.007) confirmed by Brittberg et al. in 2018.

This advantage applies most clearly to femoral condyle and trochlear defects ≥3 cm². For smaller lesions — roughly below 2–3 cm² — the comparison is less settled, and other options remain reasonable.

Beyond five years, no randomised head-to-head data exist. Longer-term figures come from observational studies and carry important caveats that the sections below address in detail.

Why the two techniques produce different repair tissue

The durability gap between the two procedures starts at the cellular level, and understanding it makes the clinical story much easier to follow.

Microfracture works by drilling or picking small holes through the hard subchondral bone beneath the defect. This releases bone marrow cells and growth factors, which form a blood clot that gradually solidifies into repair tissue. The result is fibrocartilage — a scar-like patch made predominantly of type I collagen. Type I collagen is the same structural protein found in tendons and skin: tough in tension, but poorly suited to the repetitive compressive loading of a knee joint. Under that loading, fibrocartilage degrades more quickly than the surface it replaced.

MACI takes a fundamentally different route. A small sample of the patient's own cartilage cells (chondrocytes) is harvested, expanded in a laboratory, then seeded onto a collagen scaffold and secured over the defect. Given the right environment, these cells aim to regenerate hyaline-like cartilage — the original surface — which is rich in type II collagen and far more resilient under load.

T2-mapping MRI, which measures the water and collagen arrangement within repair tissue, demonstrates this difference directly. In a patellar and trochlear cohort followed to 48 months, MACI repair tissue reached T2 values approaching native cartilage (36.54 ms), while microfracture tissue remained significantly further away (48.37 ms, P=.005) and never reached comparable native values.

There is a further practical concern with microfracture: penetrating the subchondral bone plate can cause cyst formation and structural changes in the underlying bone, which may narrow the options for any future repair procedure — a meaningful consideration for younger, active patients.

What the clinical trials actually show

Numbers from the SUMMIT trial put shape around the clinical picture. The 144 participants had a mean defect size of 4.8 cm² and a mean age of 33.8 years — a relatively young, active cohort with substantial lesions, which is worth bearing in mind when applying the findings more broadly.

At two years, MACI produced significantly greater improvements across all five KOOS domains: pain scores averaged 82.5 with MACI versus 70.9 with microfracture (P=.001), sport and function scores 60.9 versus 48.7 (P=.001), activities of daily living 87.2 versus 75.8 (P<.001), symptoms 83.7 versus 72.2 (P<.001), and quality of life 56.2 versus 47.3 (P=.029). These are clinically meaningful margins, not statistical artefacts.

At five years, those gains held. MACI remained statistically superior on KOOS pain (P=.022) and activities of daily living (P=.007) — the same domains patients most readily feel in daily life.

The most instructive finding, however, is what MRI did not show. At five years, defect filling on imaging was significant in both groups, and the scans showed no statistically meaningful structural difference between them. Yet patient-reported outcomes continued to diverge. In plain terms: the repairs looked broadly similar on the scan, but patients who received MACI reported meaningfully better pain and function. This dissociation matters because it suggests that tissue quality — how the repair cartilage is organised and how it behaves under load — is driving the clinical gap, and that difference does not show up on standard volumetric MRI. T2-mapping sequences, which assess collagen and water arrangement rather than mere fill volume, capture it more faithfully, as the 48-month patellar cohort data described above demonstrate.

What happens beyond five years

Beyond the five-year mark, the evidence thins considerably — and patients deserve to know that plainly.

No randomised controlled trial has followed MACI and microfracture head-to-head past five years. The most detailed long-term signal comes from a New Zealand prospective series of 15 patients who received MACI after an earlier microfracture procedure had already failed. By 10 years, MRI defect fill had fallen from 90% at two years to 49%, bone oedema and subchondral cysts were present in every patient, and only 27% of biopsies showed hyaline-like cartilage. Graft survival stood at 87% at a mean of eight years. These are sobering figures — but this was a salvage cohort whose subchondral bone had already been altered by a prior failed procedure; the results cannot be read as a direct projection onto patients receiving MACI as a first-line treatment.

For those undergoing MACI as an initial procedure, observational data suggest approximately 97% graft survivorship at 10 or more years. This figure comes from registry and cohort studies rather than randomised trials, so it carries less certainty than the five-year RCT evidence described above.

Microfracture's long-term trajectory is less ambiguous: the fibrocartilage breakdown documented at three to five years in multiple studies does not reverse, and functional scores generally continue to decline thereafter.

One further prognostic variable belongs in this picture. In cell-based cartilage repair, age under 30 has been identified in some cohorts as the strongest independent predictor of failure — a finding that is counterintuitive at first glance. The explanation lies not in biology alone but in time: a patient who is 25 years old at the time of repair needs that tissue to hold up for several more decades under active loading conditions than a patient who is 45. This does not counsel against MACI in younger patients — if anything, the evidence supports early, high-quality repair before secondary joint changes accumulate — but it does mean that long-term follow-up, activity counselling, and realistic expectations about durability carry particular weight for this group.

Who is each technique best suited for

The choice between procedures is less about a single threshold and more about the intersection of defect size, location, treatment history, and what the patient's life demands of their knee.

Defect size and location are the starting point. MACI carries its strongest evidence for focal femoral condyle and trochlear defects of 3 cm² or larger — the population enrolled in the SUMMIT trial. For smaller defects, roughly 2 cm² and below, other options carry a reasonable evidence base: OATS/mosaicplasty transfers osteochondral plugs in a single stage and suits defects of 1–2 cm², while AMIC (matrix-augmented microfracture) offers a single-stage bridge for patients who want to avoid a two-stage procedure but need more than bare microfracture can reliably provide.

Microfracture as a standalone intervention is increasingly difficult to justify as first-line care. The subchondral bone disruption it causes can narrow future repair options if it fails — which is why a prior microfracture is not a neutral item in a patient's history. Altered subchondral bone architecture may reduce graft integration in any subsequent cell-based repair. Sequencing decisions matter, and earlier specialist input — before that first procedure is chosen — is the cleanest way to preserve future options.

Age and activity level refine the picture further. Younger, more active patients stand to gain most from hyaline-like repair because their knee will face decades of loading. For older or lower-demand patients, the risk-benefit balance shifts, particularly when weighed against MACI's two-stage demands: a biopsy under anaesthetic, six to eight weeks of cell culture, and then a second anaesthetic for implantation — two separate recovery periods in all. Patients with anaesthetic risk factors or time constraints that make two operative episodes impractical may find a single-stage approach such as AMIC a more realistic fit.

No single algorithm determines the right choice. An individual assessment covering MRI morphology, activity goals, defect characteristics, and treatment history is what drives the decision in practice.

Getting an accurate assessment

Choosing between these procedures requires a precise picture of the defect — its size, depth, location, and the state of the underlying subchondral bone. That information, combined with any history of prior procedures, is what allows a specialist in cartilage repair to recommend a realistic pathway rather than a default one.

Three questions are worth raising at any assessment: Does the defect fall within the size range where the planned technique carries reliable evidence? Has prior marrow stimulation already altered the subchondral bone? And is a two-stage approach — two separate anaesthetics, two recovery periods — practical given the patient's circumstances and timeline?

Patients who have already undergone microfracture and are noticing declining function should not delay seeking advice: subchondral bone changes accumulate with time and can narrow the options available for any subsequent repair procedure.

MSK Doctors consultants see knee cartilage patients without a GP referral, with access to Open MRI at the Sleaford Regeneration Hub and research-linked pathways through the MFO Life Sciences Lab in Grantham. For patients based in London, the London Cartilage Clinic offers equivalent specialist access. Appointments can be booked directly at mskdoctors.com — no referral needed.

1. [1] A Prospective Outcome, MRI and Biopsy Study of MACI Cartilage Transplantation. (2017). https://doi.org/10.1177/2325967117S00186 https://doi.org/10.1177/2325967117S00186
2. [2] Arthroscopic Treatment of Patellar/Trochlear Lesions with MACI versus Microfracture: T2-Mapping and MOCART at 4-Year Follow-up. (2019). https://doi.org/10.1177/1947603519835909 https://doi.org/10.1177/1947603519835909
3. [3] Autologous Chondrocyte Implantation — Wikipedia. https://en.wikipedia.org/?curid=19074150 https://en.wikipedia.org/?curid=19074150
4. [4] ACI, MACI, OAT and OCA improve knee function and pain: Systematic review and meta-analysis (2024). (2024). https://doi.org/10.1002/ksa.12525 https://doi.org/10.1002/ksa.12525
5. [5] Microfracture Surgery — Wikipedia. https://en.wikipedia.org/?curid=8840994 https://en.wikipedia.org/?curid=8840994