These two treatments rarely compete — here's why

Most patients researching these options assume the decision is a straight contest — pick ChondroFiller or pick a stem cell injection. In practice, the two treatments rarely occupy the same clinical lane, and understanding why makes the choice far clearer.

ChondroFiller is a structural intervention: an injectable collagen scaffold designed for focal, full-thickness cartilage defects — typically up to around 6 cm² — in a joint that is mechanically well-aligned. It works by providing a physical matrix that recruits the patient's own repair cells, a process called acellular matrix-induced chondrogenesis. Think of it as filling a pothole: there is a defined gap, and the goal is to restore that specific area.

Stem cell injections — whether bone marrow aspirate concentrate (BMAC) or mesenchymal stem cells (MSCs) — work differently. Their primary action is immunomodulatory: reducing inflammation and modifying the joint environment through paracrine signalling rather than laying down a new structural matrix. They are generally best matched to patients with diffuse, mild-to-moderate osteoarthritis, where calming the whole joint surface is the aim.

No published head-to-head randomised trial has compared the two approaches directly, which itself reflects the reality: they are answering different clinical questions. The more useful question is not which treatment is better overall, but which indication fits the individual patient.

How ChondroFiller works as an injectable scaffold

The treatment itself takes place at an outpatient clinic appointment. ChondroFiller is a CE-marked Class III medical device — a purified Type I collagen solution that, once placed precisely into the cartilage defect under ultrasound guidance, gels in situ to fill the space. There are no incisions, no general anaesthetic, and no theatre admission.

Once the scaffold is seated, it creates a chemotactic environment: the patient's own progenitor cells migrate from the surrounding synovium and subchondral bone into the collagen matrix, where conditions favour differentiation into cartilage-like tissue. This is why the treatment is described as supporting the body's own repair processes rather than replacing tissue directly. It is designed for focal, full-thickness defects; it is not an appropriate tool for diffuse cartilage thinning caused by established osteoarthritis. In the hip, published cohort data show that patients with Tönnis grade 2–3 degenerative change achieve poor results — a finding that underlines why thorough joint assessment before any decision is made matters as much as the treatment itself.

Two practical points shape post-injection care. First, the scaffold needs time to stabilise: a 2024 in-vitro biomechanical study found that the immature gel does not protect the opposing joint surface under immediate cyclic loading, making structured, protected rehabilitation a clinical requirement during the maturation window. Second, volume precision matters — a 2025 wrist study found that overfilling the defect produced fibrous rather than cartilage-like tissue. The scaffold must be matched to the defect, not exceed it.

How stem cell injections work — and where they fit

BMAC and MSC injections are better understood as biologic support tools than as structural repair agents. When bone marrow aspirate concentrate is used, the process involves a same-day outpatient appointment: a small volume of marrow is drawn — typically from the iliac crest — processed through centrifugation to concentrate the biologically active fraction, then injected into the affected joint under image guidance. The procedure does not involve surgery or theatre admission, though the harvesting step means a brief additional preparation phase before the injection itself.

The mechanism is primarily paracrine: the concentrated cells and signalling molecules suppress local inflammation, modulate the immune environment, and may slow degenerative progression — but they do not anchor new tissue within a focal cartilage hole the way a scaffold does. This distinction matters clinically. For a patient with diffuse, mild-to-moderate osteoarthritis affecting much of a joint surface, modifying the inflammatory environment throughout the joint is exactly what is needed; symptomatic benefit is frequently reported in that setting. For a patient with an isolated full-thickness defect in an otherwise structurally sound joint, an injection that does not fill the physical gap is addressing the wrong problem.

The evidence for structural imaging improvement with BMAC and MSC injections in humans remains inconsistent — pain and function scores tend to move more reliably than cartilage appearance on MRI. Protocols also vary considerably between centres, which makes drawing firm comparisons difficult. Longer-term durability data, particularly for the injected-only (non-scaffold) approach, are still limited. Where scaffold-embedded BMAC has been studied — a distinct combined strategy — outcomes for larger defects have been more durable, a point returned to later in this article.

What the evidence actually shows for each

Published figures for ChondroFiller in the knee are consistent across four separate cohorts: IKDC scores improve by approximately 30 points, comfortably exceeding the established minimum clinically important difference of 16.7 points. The Jerosch et al. prospective post-market clinical follow-up study — the longest dataset currently available — recorded a mean improvement of 32.4 points sustained at three years, with the group reaching an absolute IKDC score of 80. Structural MRI tells a similar story: MOCART scores at 12 months range from 81.6 to 84.3 out of 100, with one study tracing progressive maturation from 65.3 at four weeks upward to that 12-month figure, indicating the scaffold continues to develop well after the initial healing window.

The hip cohort data (n=26, three-to-five year follow-up) add a useful indication bracket: 17 of 21 evaluable patients achieved good or excellent results, a finding already noted alongside the Tönnis 2–3 contraindication.

Two honest caveats apply to all of this. First, the evidence base is predominantly manufacturer-sponsored case series rather than independent large-scale trials. The only published randomised controlled trial — comparing ChondroFiller against microfracture in 23 patients — saw a 60% dropout in the comparator arm, making formal comparison impossible; the ChondroFiller arm showed significant IKDC improvement and no adverse events, but the trial cannot support superiority claims on its own. Second, biopsy data confirming whether the regenerated tissue is true hyaline cartilage or a fibrocartilage-like blend are largely absent — a question the current evidence cannot yet answer.

For BMAC and MSC injections, the 2026 systematic review of cell-based cartilage repair puts the picture plainly: short-term symptomatic improvements are frequently reported, but structural imaging outcomes remain inconsistent across studies, and longer-term durability with the injected-only approach is uncertain. Pain and function scores move more reliably than cartilage appearance on MRI — a pattern that reflects the primarily immunomodulatory rather than tissue-building mechanism described in the previous section.

The rehabilitation window that makes ChondroFiller work

Recovery from a ChondroFiller treatment is not passive rest — it follows a specific protected loading protocol, and the science behind that protocol is worth understanding before committing to the treatment.

A 2024 in-vitro biomechanical study applied cyclic loading of 33 N to porcine osteochondral samples and found that the immature ChondroFiller gel offered no measurable protection to the opposing cartilage surface — damage scores were statistically equivalent to untreated defects. The gel needs time to stabilise and recruit the patient's own progenitor cells before the joint can bear normal load safely. That maturation process is the mechanism of repair; rushing it does not simply slow healing, it undermines the scaffold's structural role.

For patients planning return to sport or physically demanding work, this has direct practical implications. The protected rehabilitation phase — typically structured weight-bearing restrictions followed by graduated loading — is clinically non-negotiable, not a precautionary afterthought.

Stem cell injections carry a meaningfully different expectation here. Because BMAC and MSC injections work through paracrine signalling and inflammation modulation rather than through physical scaffold maturation, they do not require the same guarded loading window. Patients typically resume normal activity much sooner, which is a genuine practical advantage for those who cannot manage a structured restricted-weight-bearing period.

Pre-treatment consultations at MSK Doctors routinely cover this distinction, since the rehabilitation timeline is a meaningful variable in treatment choice — not just a post-procedure detail.

Matching the right treatment to the right patient

Choosing between these two approaches is largely a matter of matching the clinical picture to what each treatment is designed to do.

For patients with a focal, full-thickness cartilage defect — typically in a mechanically aligned joint without significant background arthritis — ChondroFiller is the scaffold-based option worth exploring. The scaffold needs a contained defect to fill and a joint environment that can support repair; that is the indication the published cohort data reflect.

For patients with diffuse, mild-to-moderate osteoarthritis — particularly where systemic joint inflammation is part of the picture — BMAC or MSC injections are the more appropriate tool. They address the inflammatory environment driving symptoms rather than filling a discrete structural hole. Established, advanced osteoarthritis is a contraindication for ChondroFiller: the hip cohort data showed consistently poor results in that group, making baseline OA grade a decision gate rather than a minor variable.

For patients who fit neither profile cleanly, combining a scaffold with cell loading is a scientifically credible direction — the 2025 ex vivo model found greater collagen and GAG production from ChondroFiller plus MSCs than either alone — but no human RCT has yet validated this approach, so it remains a horizon-level possibility rather than an established pathway.

An MRI to characterise defect size, OA grade, and joint alignment is the starting point for any injection decision.

1. [1] Development of an Ex Vivo Osteochondral Biomimetic Platform for Mechanistic Investigation of Cartilage Regeneration. (2025). https://doi.org/10.3390/ijms262311759 https://doi.org/10.3390/ijms262311759
2. [2] Influence of cartilage defects and a collagen gel on integrity of corresponding intact cartilage: a biomechanical in-vitro study. (2024). https://doi.org/10.1007/s00402-024-05530-z https://doi.org/10.1007/s00402-024-05530-z
3. [3] Cartilage reconstruction using Chondrofiller in intra-articular distal radius fractures. (2025). https://doi.org/10.1186/s42836-025-00333-y https://doi.org/10.1186/s42836-025-00333-y
4. [4] Arthroscopic utilization of ChondroFiller gel for the treatment of hip articular cartilage defects: a cohort study with 12- to 60-month follow-up. (2021). https://doi.org/10.1093/jhps/hnab002 https://doi.org/10.1093/jhps/hnab002
5. [5] Contemporary Management of Knee Chondral Defects, Part III: Cell-Based Cartilage Repair Including ACI, Orthobiologics, and Tissue-Engineered Constructs. (2026). https://doi.org/10.1055/a-2865-1851 https://doi.org/10.1055/a-2865-1851
6. [6] One-Step Cartilage Repair of Full-Thickness Knee Chondral Lesions Using a Hyaluronic Acid–Based Scaffold Embedded With Bone Marrow Aspirate Concentrate: Long-term Outcomes After Mean Follow-up Duration of 14 Years. (2024). https://doi.org/10.1177/03635465241287524 https://doi.org/10.1177/03635465241287524
7. [7] Controlled, randomized multicenter study to compare ChondroFiller liquid with microfracturing of patients with focal cartilage defects of the knee joint. (2016). https://doi.org/10.5348/VNP05-2016-1-OA-1 https://doi.org/10.5348/VNP05-2016-1-OA-1