What the numbers actually show

Across published clinical evidence, roughly 70–85% of patients treated with ChondroFiller report significant, sustained symptom relief — a figure that holds across knee, hip, and small-joint applications at follow-up periods extending to five years. Joint by joint, the headline picture looks like this.

Knee carries the most developed evidence base. Four independent prospective cohorts — including the Jerosch et al. Post-Market Clinical Follow-up study and Simeonov's 2024 series (n=17, mean age 31) — each record a mean IKDC score improvement of approximately 30 points at 12 months, rising from a baseline of around 48 to approximately 78–80. That gain sits at roughly 1.8 times the validated minimum clinically important difference of 16.7 points, meaning it represents a real, patient-felt change rather than statistical movement. The improvement holds and marginally increases at three years, with the Jerosch cohort reaching a mean IKDC of 80.1.

Hip evidence is smaller but credible. Mazek et al. (2021) followed 26 patients with acetabular cartilage lesions for up to five years: 17 of 21 patients available at 3–5 year follow-up achieved good or excellent results, with modified Harris Hip Score gains averaging up to 33 points. Two patients eventually required total hip replacement.

Ankle data are directionally positive — reported pain-reduction success rates of 70–85% — but currently rest on case studies and registry data rather than dedicated prospective controlled trials.

All current evidence is prospective observational in design; no large randomised controlled trial exists. That is worth stating plainly, though it does not diminish the consistency of the findings across independent cohorts. The adverse event rate across clinical programmes is approximately 0.06%, and more than 80% of patients report high satisfaction.

Knee outcomes: IKDC and MOCART scores in detail

The consistency of the knee evidence is as notable as its magnitude. Jerosch et al.'s post-market follow-up recorded a mean gain of 32.4 IKDC points — a figure that held and slightly increased when the same patients were reassessed at three years. Two independent groups replicate this pattern: Simeonov's 2024 cohort (n=17, mean age 31) and a 2016 multi-centre prospective trial both recorded statistically significant IKDC and Lysholm improvements at three, six, and 12 months (p<0.05). The same result across four separate cohorts, different populations, and different time points narrows the space for any single study's methodology to account for the outcome.

MRI scoring provides a structural counterpart to these functional gains. MOCART (Magnetic Resonance Observation of Cartilage Repair Tissue) assesses repair quality on a 0–100 scale, measuring how completely the defect has filled and how well the new tissue integrates with surrounding native cartilage. In European knee studies, mean MOCART began at approximately 65.3 at four weeks — a partial-fill reading consistent with early scaffold occupancy — and consolidated to between 81.6 and 84.3 by 12 months. A score above 80 indicates that more than 80% of the defect is filled with well-integrated tissue. Across published one-year data the range sits between 70 and 87, indicating consistently strong structural repair rather than a single favourable outlier.

The maturation trajectory matters because it reflects the biology rather than a slow start. The roughly 16-point rise between four-week and 12-month MOCART readings corresponds to the scaffold being progressively colonised and replaced by the patient's own progenitor cells. Functional scores and structural scores therefore tell the same progressive story, with both consolidating the bulk of their gains within the first 12 months and modest further improvement recorded at three years.

Hip outcomes: Harris Hip Score and what patient selection means

The hip findings from Mazek et al. (2021) carry a message that is equal parts encouraging and conditional: for appropriate patients the results are durable and clinically meaningful, but patient selection is what separates good outcomes from poor ones.

The modified Harris Hip Score (mHHS) is a 100-point measure of hip function weighted towards pain, mobility, and daily activity capacity. In the Mazek cohort of 26 adults with femoroacetabular impingement and acetabular lesions larger than 2 cm², mean mHHS gains averaged up to 33 points across a follow-up window of 12 to 60 months. One case illustrates the range available to well-selected patients: a pre-treatment score of 36 — severe functional limitation — climbed to 89 at one year. Seventeen of the 21 patients available at three-to-five year review met the good-to-excellent threshold; two progressed to total hip replacement.

OA grade as the decisive filter

That 81% figure is not evenly distributed across the cohort. Patients who entered with Tönnis grade 2–3 osteoarthritis — indicating moderate-to-severe joint-space narrowing and established articular damage — did not achieve meaningful benefit. This is a clear clinical boundary rather than a minor statistical nuance. The acellular scaffold mechanism depends on viable surrounding tissue to recruit progenitor cells; where generalised OA has already compromised that biological environment, the repair process has less to work with.

Pre-treatment imaging review — MRI or weight-bearing X-ray sufficient to assess Tönnis grade — is therefore part of any responsible pre-treatment assessment. The delayed weight-bearing protocol typically observed after hip procedures (often three weeks non-weight-bearing) is mechanistically consistent with allowing stable scaffold occupancy before loading begins.

The 12–60 month follow-up range is a relative strength of this dataset: most injectable cartilage studies report to 12 months only, giving the Mazek cohort's five-year tail a durability signal that short-term registry data cannot match. The hip evidence base remains smaller than the knee literature, but its prospective design and extended follow-up give it moderate-to-good credibility.

Ankle and small-joint evidence: what is known so far

Dedicated prospective trial data for ChondroFiller in ankle cartilage defects have not yet been published at the scale seen for the knee or hip. Available evidence consists of case series and registry entries reporting MRI-confirmed tissue regeneration in focal defects and pain-reduction rates drawn from smaller series; where success rates in the 70–85% range appear in ankle-specific data, they carry less statistical weight than the multi-centre replicated knee cohorts underlying the same figure — a distinction worth keeping in mind when calibrating expectations.

Some ankle conclusions currently rest on extrapolation from knee and hip trajectories — reasonable given that the acellular scaffold mechanism (recruiting the patient's own progenitor cells into a collagen framework) is not inherently joint-specific, but extrapolation from larger-joint data is not equivalent to dedicated ankle evidence.

The most instructive published small-joint comparison comes from Demmer et al. (2025), a wrist cartilage study chosen partly because the wrist's confined articular geometry and focal traumatic defect patterns more closely resemble ankle conditions than the knee does. At follow-up arthroscopy, ChondroFiller-treated wrists showed significantly better cartilage quality than controls: median Outerbridge score 1.5 versus 3 (P=0.006) and ICRS grade 1 versus 3 (P=0.002), with fibrous tissue formation limited to overfilled defects. That structural difference supports the inference that the scaffold principle extends to smaller joint surfaces. Ankle-specific trials are ongoing, and dedicated cohort data are expected to emerge over the coming years as clinical programmes mature.

Why scores improve progressively: how the scaffold works

The scaffold placed during an ultrasound-guided outpatient appointment is cell-free — it contains no transplanted cells of its own. Rather than delivering repair tissue directly, it acts as a temporary collagen framework that the patient's own progenitor cells, drawn from the surrounding synovium and subchondral bone, gradually migrate into and colonise. This process — acellular matrix-induced chondrogenesis — explains why improvement follows a curve rather than arriving immediately: the tissue has to be built, not delivered.

As those cells establish themselves and begin producing repair matrix, the scaffold progressively resorbs, yielding native-like tissue in its place. The structural healing trajectory visible on MRI scans — discussed in the knee section above — directly reflects this biology: early readings show partial occupancy; by twelve months, consolidation is largely complete, and functional scores follow accordingly.

The same timetable makes post-injection loading protocols clinically meaningful rather than merely precautionary. In the early weeks after placement, the scaffold has not yet achieved stable integration, and Pieringer et al. (2024) demonstrated in vitro that loading before that integration point carries a real biomechanical risk to the opposing articular surface. Patients are typically asked to limit weight-bearing for the first few weeks; understanding the reason — that the forming repair tissue needs time to anchor before it can tolerate load — tends to support adherence.

One technique-dependent variable is fill volume. Overfilling the defect promotes fibrous rather than hyaline-like tissue formation, as Demmer et al. (2025) observed in overfilled wrist defects. Flush, accurate placement under image guidance is therefore both a quality-control step and a determinant of the tissue type that ultimately develops.

What shapes your individual outcome

Three variables — defect characteristics, joint background, and patient age — do the most to predict where an individual outcome will land.

Defect size and containment matter most. The published evidence reflects patients with focal, contained cartilage lesions; outcomes deteriorate when a defect is large, fragmented, or set against a background of degenerative change across the joint. For the hip, that relationship is direct and well-documented (see the Tönnis grade analysis above); knee evidence points in the same direction, favouring isolated traumatic or activity-related lesions over generalised wear. Age correlates closely with defect type: Simeonov's 2024 cohort, with a mean age of 31, sits at the top of the evidence hierarchy partly because younger patients tend to present with exactly this pattern — focal, activity-related, and contained.

The surgical revision rate across published programmes runs at approximately 3–8%. In straightforward cases this figure reflects the complexity of the underlying defect rather than a fundamental limitation of the treatment; it is not a meaningful signal in low-complexity presentations.

Pre-treatment MRI is the gatekeeping step — not a formality. Characterising defect depth, containment, and the condition of surrounding cartilage and subchondral bone is where individual suitability is established rather than assumed. AI-assisted MRI analysis, such as the onMRI™ platform available at MSK Doctors, can add objective defect characterisation to that assessment where clinically relevant.

ChondroFiller is not currently NHS-commissioned in the UK; access is through private providers offering ultrasound-guided outpatient placement. No formal GP referral is needed to be assessed for it — London-based patients can be seen through the London Cartilage Clinic, and appointments elsewhere in the country are available directly through MSK Doctors.