Why KL Grade III/IV patients are hard to treat with a single injection

For many patients reaching Kellgren-Lawrence Grade III or IV knee osteoarthritis, the conversation with a clinician arrives at an uncomfortable middle ground: surgical cartilage repair — techniques such as ACI or OATS — is typically off the table because the damage is too diffuse, yet the standard injection options often feel like sticking plasters on a structural problem.

KL Grade III/IV is not simply a description of worn cartilage. By this stage, joint-space narrowing is substantial, osteophytes have formed along the joint margins, and the subchondral bone has begun to remodel in response to years of altered load-bearing. Two distinct pathologies tend to co-exist and reinforce one another: progressive erosion of the articular surface, and persistent inflammation of the synovial membrane that lines the joint capsule.

That second element — synovitis — matters more than is sometimes appreciated. Inflammation of the synovial tissue appears to be strongly linked with pain levels in knee osteoarthritis, more so than cartilage volume loss alone. A patient at this grade may be experiencing most of their daily pain not primarily from bone-on-bone contact, but from an inflamed, thickened synovium producing pain-mediating cytokines.

Standard single-modality injections were not designed to address both problems simultaneously. Hyaluronic acid targets joint lubrication; corticosteroid suppresses inflammation short-term but does not touch the articular surface; PRP provides biologic support but leaves the mechanical environment of the synovium largely unchanged. Each works on one part of the problem. When both components are active, treating only one of them frequently produces incomplete or short-lived relief.

It is this anatomical duality — structural surface wear alongside synovial inflammation — that provides the clinical rationale for a two-component, staged injection protocol. The approach does not position itself as a cure or a proven means of halting OA progression; rather, it aims to address both pathological targets through products with distinct mechanisms, offering a non-surgical option for patients who currently fall between the two stools of surgical eligibility and adequate palliative care.

How the two injections target different structures in the same joint

Each injection in the CFI+ protocol has a precise anatomical address — and the two addresses do not overlap.

ChondroFiller injection is a Type I collagen hydrogel (2.3 mL) delivered under ultrasound guidance onto the worn articular surface of the knee. Once placed, the scaffold gels in situ across the damaged bone end, providing a top-down cushioning layer without any need for surgical debridement beforehand. Over the following months, the patient's own progenitor cells migrate into the matrix — a process called acellular matrix-induced chondrogenesis — and begin laying down new cartilage-like tissue. ChondroFiller injection is, in other words, the regenerative component of the protocol: a scaffold designed to support the body's own repair processes at the joint surface.

Arthrosamid works at a different anatomical level entirely. It is a 2.5% crosslinked polyacrylamide hydrogel (6 mL) that integrates into the synovial membrane lining the joint capsule. Its role is mechanical rather than regenerative — once integrated, the hydrogel provides cushioning support within the synovial tissue and helps to address the inflammation that drives pain at this grade. It does not act on the cartilage surface, and it does not promote cell-mediated repair.

Because the two products act on structurally distinct targets — articular surface versus synovial lining — they should not be grouped into a single generic 'filler' category. The clinical rationale for using both rests precisely on the fact that they address two co-existing pathologies through two independent mechanisms.

Both are delivered as ultrasound-guided outpatient injections, not surgical procedures. Staging them at separate appointments, each requiring its own suitability assessment, is a deliberate protocol-level safeguard: clinical justification for each component is established independently before any injection is given.

What the ChondroFiller injection evidence shows — and where it was tested

The peer-reviewed evidence for ChondroFiller injection is directionally positive but modest in scale — and the source population matters as much as the results themselves.

The most controlled data come from an RCT comparing ChondroFiller liquid with microfracture (n=13 in the ChondroFiller arm), which found significant IKDC improvement at three and six months (p<0.05), good immediate MRI filling of all treated defects, and impressive maturation of the reconstructed cartilage at 52 weeks, with no adverse events reported. A prospective knee study (n=17, mean age 31) replicated those functional gains: Lysholm and IKDC scores were significantly improved at 3, 6, and 12 months post-implantation (p<0.05), with scores plateauing between six and twelve months — suggesting the repair process consolidates rather than continues advancing after that point.

Both studies enrolled patients with focal, contained chondral lesions — not the diffuse, multi-compartment wear characteristic of KL Grade III/IV osteoarthritis. That distinction is pivotal when applying these results to the dual-protocol population.

The most instructive evidence on this point comes not from a knee study but from a hip arthroscopy cohort (n=26, 12–60 month follow-up): 17 of 21 available patients achieved good or excellent results overall, but those with pre-existing osteoarthritis at Tönnis grades 2–3 had poor results. Tönnis 2–3 in the hip maps to a broadly similar articular environment as advanced knee OA. The implication: ChondroFiller injection performs best where the joint retains enough biological capacity to support repair — a condition that becomes progressively harder to meet as OA advances.

ChondroFiller injection's published track record, then, is in focal lesions in a largely younger patient cohort. Whether those outcomes translate to the diffuse KL III/IV wear the CFI+ protocol addresses remains a question the existing peer-reviewed literature has not yet answered directly.

What the Arthrosamid evidence shows — including for higher-grade OA

Arthrosamid's evidence base is larger in volume than ChondroFiller's, and it spans a longer follow-up horizon — but it contains an internal tension that matters directly to this patient population.

A 5-year extension study (n=49 enrolled; 27 completers) recorded sustained, statistically significant WOMAC improvements throughout the observation period: pain −14.6 (p=0.0002), stiffness −19.6 (p=0.0006), and physical function −12.5 (p=0.0015), with no serious adverse device effects — the longest published follow-up available for this product class. A systematic review across 463 patients confirmed efficacy at 52 weeks and 13 months, with one included RCT showing Arthrosamid numerically superior to hyaluronic acid, and a 12-month comparative cohort (n=150, KL II–IV) finding Arthrosamid outperformed corticosteroid at six months but was not statistically superior to HA at any timepoint.

The critical caveat comes from a 24-month PAAG cohort of 314 knees (269 patients): higher KL grade was the strongest predictor of both needing total knee replacement (b=0.97; p<0.001) and failing to reach MCID across all outcome scores (Exp B=0.576; p=0.001). In plain terms, Arthrosamid's own evidence suggests it is less effective precisely in the KL III/IV patients this protocol is designed for. That finding is not grounds for dismissing the protocol — it is the context within which realistic expectations should be set.

Before leaving the clinical picture, one safety question warrants a direct answer. Because Arthrosamid originates from polymerisation involving acrylamide monomer — a known neurotoxin in its unpolymerised form — the question of residual toxicity is a reasonable one. An in vitro study applied 2.5% iPAAG to human iPSC-derived GlutaNeurons at concentrations up to 20% for 96 hours and found no statistically significant effects on cell survival, apoptotic cell death, or neurite network area, supporting the interpretation that the crosslinked product does not exhibit in vitro neurotoxic or cytotoxic effects.

Within the CFI+ protocol, Arthrosamid's role remains specifically synovial: it addresses the inflammation and mechanical instability within the joint lining that drives pain at this grade. It does not repair articular cartilage — that is the territory of ChondroFiller injection.

The case for combining them — and what evidence currently supports

The mechanistic case for combining the two injections is straightforward: advanced knee OA typically presents two co-existing problems — structural failure at the articular surface and inflammatory activity in the synovial lining — and the two products are designed for anatomically distinct targets. That coherence is the protocol's strongest argument.

What the evidence does not yet supply is a controlled study of the combination itself. The clinic's own documentation states this plainly: direct combination evidence is limited, and neither ChondroFiller injection nor Arthrosamid has been shown to slow OA progression or delay knee replacement when used together. A combined plan is considered only where clinical assessment independently supports each component — the staged scheduling and separate suitability reviews are structured around that absence, not in spite of it.

Individual-product data compound the picture rather than resolve it. Both ChondroFiller injection and Arthrosamid show reduced efficacy in the higher-grade articular environments the protocol is specifically designed to address — the hip cohort's poor results in Tönnis 2–3 patients and the PAAG cohort's finding that higher KL grade predicts both TKR need and failure to reach MCID point in the same direction. Advanced degeneration limits biological response regardless of how precisely targeted the intervention is.

The Tri-Active tier adds autologous mesenchymal stem cells — sourced from bone marrow, fat, or ear cartilage, with the last representing a harvesting site that is relatively accessible and rich in chondrogenic progenitor cells — to the base protocol at £11,000. The cellular signalling rationale is that MSCs may modulate the inflammatory joint environment in which the ChondroFiller scaffold has to function, but the evidence base for this combination is thinner still.

The plain bottom line: the mechanistic rationale is coherent, and individual-product evidence is encouraging for selected patients. At KL Grade III/IV, however, both components show reduced efficacy in their own published data — and that is the most important factor to weigh with a clinician before committing to the protocol.

Assessment and the treatment pathway at London Cartilage Clinic

At this stage of the pathway, the most consequential appointment is the first one.

Before any injection is planned, suitability is mapped against imaging — Kellgren-Lawrence grade, the distribution of cartilage loss, synovial involvement, and functional history. The assessment is not a formality: as both components of the CFI+ protocol show reduced efficacy in more severely degenerated joints, the consultation is where that clinical picture translates into realistic expectations rather than a default treatment offer. Patients who have read this far should arrive with specific questions: What does my imaging actually show? Which component — if either — is independently supported by my clinical profile? What does a realistic outcome look like at my KL grade?

Professor Paul Y. F. Lee delivers Liquid Cartilage™ in the UK, with particular focus on ultrasound-guided scaffold placement — a technique-sensitive step where delivery precision influences how well the collagen matrix seats within the joint. The full protocol is conducted on an outpatient basis: no general anaesthetic, no theatre, no arthroscopy.

London Cartilage Clinic on Harley Street is the UK certified delivery centre for Liquid Cartilage™ / ChondroFiller injection. Assessment enquiries can be made via londoncartilage.com.

1. [1] A Systematic Review of the Novel Compound Arthrosamid Polyacrylamide (PAAG) Hydrogel for Treatment of Knee Osteoarthritis. (2022). https://doi.org/10.18103/mra.v10i8.2950 https://doi.org/10.18103/mra.v10i8.2950
2. [2] Polyacrylamide hydrogel injections in knee osteoarthritis: A PROMs-based 24 month cohort study. (2025). https://doi.org/10.1016/j.jcot.2025.103136 https://doi.org/10.1016/j.jcot.2025.103136
3. [3] 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
4. [4] Implantation of ChondroFiller Liquid® as a Scaffold Material for the Treatment of Chondral Lesions of the Knee Joint. (2024). https://doi.org/10.5272/jimab.2024304.5936 https://doi.org/10.5272/jimab.2024304.5936
5. [5] Controlled, randomized multicenter study to compare compatibility and safety of ChondroFiller liquid with microfracturing. (2016). https://doi.org/10.5348/VNP05-2016-1-OA-1 https://doi.org/10.5348/VNP05-2016-1-OA-1
6. [6] A prospective, open-label, clinical investigation of a single intra-articular polyacrylamide hydrogel injection in participants with knee osteoarthritis: a 5-year extension study. (2025). https://doi.org/10.1186/s13018-025-06526-0 https://doi.org/10.1186/s13018-025-06526-0
7. [7] An injectable 2.5% cross-linked polyacrylamide hydrogel (2.5 iPAAG) demonstrates no neurotoxicity in human induced pluripotent stem cells-derived iCell® GlutaNeurons. (2025). https://doi.org/10.3389/ftox.2025.1585430 https://doi.org/10.3389/ftox.2025.1585430