The headline figure: what 3-year data actually shows

Three years after a ChondroFiller knee injection, patients in the largest dedicated prospective study maintained a mean IKDC functional score of 80 — and that figure held without meaningful decline from the 12-month result.

The study in question is the Jerosch post-market clinical follow-up (PMCF, registration DRKS00009703), which followed 64 patients across multiple European centres for 36 months — the longest continuous ChondroFiller knee dataset currently published. Participants entered treatment with a mean IKDC score of approximately 48, a level at which everyday activities such as climbing stairs, kneeling, or walking on uneven ground are noticeably restricted. By 36 months, the group mean had reached 80, a gain of 32.4 points.

To put that number in context: the Minimal Clinically Important Difference (MCID) for the IKDC — the smallest improvement a patient actually notices in daily life — is 16.7 points. A 32.4-point gain is nearly double that threshold. Crucially, the Jerosch PMCF recorded no late deterioration: scores were sustained and marginally increased at the three-year mark rather than eroding from their one-year peak.

Pooled across published series at three to five years, roughly 70–85% of suitably selected patients maintain meaningful symptom relief — a figure worth holding in mind as a headline benchmark, since it applies specifically to patients with focal cartilage defects who meet the treatment criteria, not to every knee presenting with pain.

What the first year looks like: function and structural fill

Reaching that three-year result requires a strong first year — and the evidence for year one is the most consistent in the ChondroFiller evidence base.

Functional scores improve quickly. Across multiple cohorts, IKDC and Lysholm scores rise significantly within three months of the injection (p<0.05 versus baseline), and continue building through six months. The Simeonov 2024 cohort (17 patients, Journal of IMAB, 12-month follow-up) found that by six months the improvement was statistically significant at both scoring scales — but the difference between the six-month and twelve-month values was not significant (p>0.05). The functional plateau arrives by six months and is held rather than extended. The Jerosch 2016 controlled multicenter randomised trial (23 patients) showed the same pattern: significant IKDC improvement at three, six, and twelve months, with all treated defects showing defect fill on MRI and no adverse events recorded across the cohort.

Structural MRI data run in the same direction as the functional scores, which offers patients more than subjective reassurance. MOCART — a 0-to-100 MRI measure of how completely a focal cartilage defect has been filled and how well the new tissue integrates with its borders — progressed from a mean of 65.3 at four weeks to 81.6–84.3 at twelve months across European multi-site studies. A score above 80 reflects greater than 80% volumetric fill with good border integration. The convergence of functional improvement and structural fill at twelve months is the foundation on which longer-term durability rests.

Why durability comes from biology, not the scaffold

The scaffold that drives those results does not stay in the knee — and that is precisely why the benefits persist.

ChondroFiller is a cell-free Type I collagen hydrogel. Placed into the focal defect under ultrasound guidance as an outpatient injection, it polymerises within approximately three to five minutes, forming a dimensionally stable three-dimensional template that anchors to the defect walls and base. What it does not contain is cells. Instead, the collagen matrix acts as a chemotactic scaffold — it draws the patient's own progenitor cells, from the surrounding synovium and subchondral bone, into the defect. This mechanism is termed acellular matrix-induced chondrogenesis.

A 2025 ex vivo osteochondral study confirmed cell recruitment in human tissue directly: defect sites treated with ChondroFiller showed a 2.4-fold increase in DNA content by day 14, validating that the scaffold attracts and retains the body's own repair cells rather than relying on any externally delivered source.

Over the following 12–18 months, the collagen framework gradually biodegrades as those incoming cells lay down repair tissue in its place. By the time the scaffold has dissolved, the structure occupying the defect is the patient's own — not a degrading foreign material. A patient asking whether the injection will 'wear off' is asking about the wrong thing: after 18 months, there is no scaffold left to wear off. What remains is endogenous repair tissue that the collagen template prompted the body to produce.

That distinction has direct clinical relevance. It explains why the functional gains documented at one year are maintained rather than eroded at three — the durability is biological, not material.

Years 2 and 3: does the improvement hold?

Year-one results, described in the previous section, are the most consistent in the evidence base. The more pointed question is what happens at two and three years — and here it is worth being explicit about what is directly measured and what is reasonably inferred.

There is no dedicated 24-month measurement point in the published ChondroFiller knee literature. Stability through year two is therefore inferred: the Jerosch PMCF study (detailed in the opening section) followed patients to 36 months and found that the 12-month IKDC improvement was not only maintained but marginally exceeded at three years. No late deterioration was documented at any interval across that span. The absence of decline between 12 and 36 months is the strongest available basis for year-two stability — but it is an inference from the trajectory, not a discrete data point.

At year three, the evidence becomes direct again. The Jerosch PMCF's 36-month result — summarised in the opening section — represents the current outer boundary of published prospective data for ChondroFiller in the knee. That plateau is mechanistically consistent with what the previous section described: repair tissue established in year one has no particular biological reason to regress once the scaffold has dissolved and the patient's own cells have consolidated.

The reoperation record provides a practical durability check. In published series, the cumulative reoperation or failure rate for ChondroFiller in focal knee defects sits at approximately 3–8%. Microfracture — the most familiar surgical comparator for focal cartilage lesions — carries figures of up to 41% in comparable cohorts. That gap speaks to sustained tissue stability rather than a treatment that tends to break down and require revision.

Who gets lasting results and who may not

Not every knee pain patient is the right candidate — and the distinction matters as much as the data.

ChondroFiller is indicated for focal cartilage defects up to 6 cm² in area, typically where the damage is full or near-full thickness (what clinicians call Grade III or IV) but contained to a specific zone of the joint surface. Younger and physically active patients with one or two such lesions, and otherwise reasonable surrounding cartilage, tend to do best. What substantially reduces the probability of a good outcome is diffuse, end-stage osteoarthritis — where wear is widespread across the joint rather than concentrated in a repairable focal area.

The practical implication: the durability data described in the preceding sections reflect outcomes in that suitable subgroup, not the general knee pain population.

An MRI including cartilage mapping is the necessary starting point for any honest conversation about suitability. At Lincolnshire Knee, onMRI™ AI-assisted cartilage analysis helps characterise defect dimensions, border quality, and surrounding tissue condition — the factors that actually determine whether a focal scaffold approach is appropriate. A contained defect measuring under 6 cm², with a well-defined border and intact subchondral bone beneath it, represents a favourable pattern. Diffuse thinning across more than one compartment does not — regardless of how the symptoms present.

What the evidence can and cannot tell us yet

ChondroFiller's longest published dataset stops at 36 months — and that boundary is worth naming plainly. Compared with MACI, which has accumulating ten-year follow-up data, ChondroFiller's evidence base is genuinely younger and smaller. The Jerosch PMCF study represents the current outer edge of direct prospective evidence for the knee, and the 70–85% success rate quoted at three to five years reflects pooled series rather than a single controlled trial.

Two specific gaps deserve acknowledgement. There is no published head-to-head randomised trial comparing ChondroFiller with hyaluronic acid at matched follow-up intervals — the comparative reoperation data against microfracture exists, but the studies differ in design. And for the period beyond five years, peer-reviewed surveillance data in the knee has not yet been published.

None of that diminishes the 36-month signal, which is real and clinically significant — a 32-point IKDC gain held without late deterioration is not a small finding. The honest framing is that meaningful durability through three years is well-supported, the biological mechanism provides a plausible rationale for benefit beyond that window, and the five-plus-year picture awaits further study. Uncertainty about long-term data is a property of a relatively new device, which is a different statement from saying the treatment does not work.

Patients wanting to understand their individual outlook should discuss defect size, activity level, age, and MRI findings at a dedicated knee assessment — those specifics determine far more than the population-level statistics.

1. [1] Controlled, Randomized Multicenter Study: ChondroFiller Liquid vs Microfracturing in Focal Knee Cartilage Defects (Jerosch 2016). (2016). https://doi.org/10.5348/VNP05-2016-1-OA-1 https://doi.org/10.5348/VNP05-2016-1-OA-1
2. [2] Implantation of ChondroFiller Liquid® as a Scaffold Material for the Treatment of Chondral Lesions of the Knee Joint (Simeonov 2024). (2024). https://doi.org/10.5272/jimab.2024304.5936 https://doi.org/10.5272/jimab.2024304.5936
3. [3] Development of an Ex Vivo Osteochondral Biomimetic Platform for Cartilage Regeneration (2025). (2025). https://doi.org/10.3390/ijms262311759 https://doi.org/10.3390/ijms262311759