Bone-on-bone diagnosis, but replacement is not automatic

A grade IV osteoarthritis, bone-on-bone finding on an X-ray is unambiguous: cartilage has been largely or entirely lost, bone is contacting bone, and the joint space visible in earlier disease has gone. It is the most advanced radiographic classification of knee osteoarthritis, and when a surgeon mentions total knee arthroplasty at this stage, the recommendation is clinically grounded — for many patients, replacement is the right answer and declining it carries its own risks.

But the X-ray does not tell the whole story. A 2025 case-control study of 87 knee osteoarthritis patients found that those assigned to total knee replacement carried significantly worse WOMAC function scores, higher anxiety and depression ratings, and greater pain-catastrophising than patients managed conservatively — despite radiographically similar disease grades. Two people can present with the same imaging severity and have materially different functional lives, psychological profiles, and surgical candidacy. Grade IV on a scan describes the joint; it does not, by itself, determine what a particular patient needs next.

The clinical question worth asking — before any irreversible decision — is a more precise one: given this patient's symptoms, functional capacity, comorbidities, and goals, what non-surgical options remain worth evaluating, and what should realistic expectations look like? The sections that follow address that question directly, working through the evidence for each available approach and the patient selection criteria that determine where it applies.

Multimodal conservative management — the non-negotiable baseline

Before any procedural intervention is considered, three established pillars deserve proper attention: physiotherapy, weight optimisation, and pharmacotherapy. A 2025 AI-assisted consensus of 30 orthopaedic surgeons confirmed high agreement that this multimodal combination is the appropriate non-surgical starting framework for knee osteoarthritis — not a stopgap pending surgery, but the clinically affirmed baseline.

Physiotherapy at this stage concentrates on quadriceps and hamstring strengthening, proprioceptive retraining, gait mechanics, and load redistribution. Weight management carries sound mechanistic rationale — each kilogram of body weight exerts roughly four kilograms of force across the knee joint — though most trial data derive from broader OA populations rather than confirmed grade IV cohorts, and that distinction matters when setting expectations.

Pharmacotherapy requires careful calibration. Corticosteroid injections can provide meaningful short-term relief during acute inflammatory flares, but the conventional guidance limits them to approximately three administrations per year. Cumulative chondrotoxicity risk is the reason; at a stage where cartilage is already absent, repeated high-dose intra-articular steroids warrant particular caution rather than routine use.

Intra-articular hyaluronic acid carries a structural evidence limitation that is worth stating plainly: a systematic review confirmed that end-stage patients are routinely excluded from HA trials, so published efficacy data cannot be reliably extrapolated to grade IV disease. Prescribing on the basis of those results means reasoning beyond what the evidence actually covers.

On surgical alternatives at this tier, a 2024 meta-analysis of 10 RCTs found arthroscopic surgery provides no advantage over conservative treatment across pain, function, or quality-of-life measures — confirming that the non-operative pathway is the clinically appropriate direction, not merely the default when nothing else is available.

Genicular nerve and artery procedures — what the 2025 evidence shows

Two interventional procedures stand out for having the strongest direct evidence in grade IV knee osteoarthritis specifically — and understanding what distinguishes them matters as much as knowing what they share.

Genicular nerve ablation (GNA) works by interrupting pain signals from the sensory nerves surrounding the knee joint, using either pulsed radiofrequency or phenol. A 2025 prospective RCT enrolled 35 patients with Kellgren-Lawrence grade IV disease and found both techniques reduced NRS pain scores by approximately 30% from baseline at three months, with equivalent improvements on timed walk and 30-second chair-stand tests. A separate retrospective analysis of 6,035 GNA patients found that roughly 87% had not undergone knee arthroplasty within five years — a meaningful surgery-deferral signal, though the figure is subject to selection effects: patients who chose GNA may have been less surgically suitable or less willing to proceed to replacement from the outset.

Genicular artery embolisation (GAE) operates through an entirely different mechanism, selectively targeting the abnormal neovascularity feeding an inflamed synovium. By reducing synovial blood supply, GAE addresses the inflammatory environment rather than the pain pathway. Clinical data report significant improvements in VAS and WOMAC scores, with pain relief sustained in selected patients for up to 24 months — considerably longer than the 6–12 months typically associated with GNA. They are distinct procedures addressing different contributors to pain, not interchangeable options.

Patient selection is central to both. A 2025 study using pre-procedural ultrasound phenotyping found that GNA patients with minimal synovial hypertrophy achieved a WOMAC reduction of approximately 33 points, compared with roughly 18–19 points in those with moderate or severe synovial thickening or effusion — approximately double the functional gain. This is a clinically significant difference that argues strongly against treating either procedure as a blanket offering.

One distinction must be made explicit to any patient considering this pathway: neither GNA nor GAE is disease-modifying. They do not restore cartilage, slow structural deterioration, or alter the underlying biology of the joint. For the appropriately selected patient, they can meaningfully reduce pain and defer surgical decision-making — sometimes for years — but the grade IV pathology remains.

Arthrosamid and emerging injectable options

Polyacrylamide hydrogel (Arthrosamid®, iPAAG) takes a different approach to the problem: a single intra-articular injection that integrates with the synovial membrane, providing mechanical cushioning without a repeat-dosing cycle. CE-approved for knee osteoarthritis, it has been studied in advanced disease — a 2022 case series co-authored by myself reported reduction in patellofemoral bone marrow lesions following a single injection in selected patients with advanced OA. An NHS mechanistic study is ongoing, and that institutional involvement reflects genuine clinical interest in understanding how it works. The positioning is deliberately clear: Arthrosamid® is a surgery-delay option for appropriately selected patients — those for whom timing, commitments, or clinical suitability makes a non-surgical attempt reasonable — not a substitute for arthroplasty where replacement is the right answer. After 1200 injections of Arthrosamid, I have further refined our injection protocol to better help or patients.

Platelet-rich plasma (PRP) draws on the same growth-factor biology used in more complex regenerative orthobiologic protocols. In earlier-stage OA there is reasonable trial evidence for symptomatic benefit; at grade IV specifically, end-stage-focused data remain limited, and results from those earlier cohorts cannot be directly extrapolated. In practice, PRP features as a supporting adjunct in some protocols — providing short-term biological signalling — rather than a primary intervention at this stage.

Both options sit within the preserve-before-replace sequence: they can buy useful time, with minimal procedural burden, for patients who are not yet clinically or personally ready for surgery — and that is a legitimate and sometimes significant clinical contribution.

Cartilage regeneration and NanoACi — the case for repair before replacement

Surgical replacement sits at the far end of a clinical spectrum, not the beginning of it. The more useful question, for many patients, is where on that spectrum they currently sit and what options remain available before reaching that endpoint.

Cartilage regeneration occupies the middle ground between pain management and replacement. Marrow-stimulation techniques such as microfracture produce fibrocartilage — structurally inferior to native hyaline cartilage. Cell-based approaches, including autologous chondrocyte implantation (ACI), aim for something closer to hyaline tissue; no current technique fully restores what was originally there. Honesty about that ceiling is part of setting realistic expectations from the outset.

NanoACi — Non-Arthroscopic Needle-Delivered One-Stage Autologous Chondrogenic Injection — is a surgeon-led clinical technique developed by me -Professor Paul Lee. It combines three autologous components: ear-cartilage micrografts (three 2.5mm concha punches taken under local anaesthetic), a collagen scaffold (ChondroFiller), and platelet-rich fibrin prepared from the patient's own blood. All three are delivered by needle in a single outpatient session — no general anaesthetic, no theatre admission, no laboratory waiting period. The biological intent draws on the same autologous-cell rationale as ACI; the procedural burden is fundamentally different.

Each component carries its own peer-reviewed evidence: the collagen scaffold has been compared against microfracture in trials, platelet-rich fibrin has established biological credentials for sustained growth-factor signalling, and the micrograft science draws on neural-crest research. The full three-part combined protocol, however, has not yet completed long randomised controlled trials. That is a genuine limitation of the current evidence base, but it will soon catch up.

Patient selection is a matter of clinical judgement rather than a checklist. NanoACi is designed for appropriate candidates, and the principle of doing less harm where that better serves the patient does not make it suitable for all grade IV presentations. It also does not prevent future surgery — a material consideration for anyone weighing the decision carefully.

Who is actually a candidate — the role of selection and timing

Imaging grade tells one story; the clinical consultation tells another. The divergence in psychological burden, functional capacity, and pain catastrophising that separates conservatively managed patients from those heading to surgery — even at comparable radiographic grades — confirms that grade IV on an X-ray does not produce a uniform clinical picture. The same scan finding can describe patients with very different symptoms, reserves, and trajectories.

Candidacy for the options discussed in this article depends on factors no radiograph captures: symptom burden relative to functional demands, comorbidities affecting procedural risk or recovery, psychological readiness, and what the patient is trying to preserve over what timeframe. Some approaches — genicular nerve ablation, multimodal conservative management — are appropriate at any point in the stage 4 trajectory. Regenerative options involve a biological window and a tissue environment that must be assessed individually; not every grade IV presentation is a suitable substrate, and no honest clinical conversation pretends otherwise.

Timing is its own variable. A patient in their late forties with a physically demanding occupation and some residual tissue asks a fundamentally different question than one in their seventies with multiple comorbidities carrying the same imaging grade. These calibrations accrue from practice — the pattern recognition that Professor Lee, with over 20 years of NHS consultant orthopaedic work and published research across knee OA, orthobiologics, and motion-capture diagnostics for joint disease, brings to a first consultation alongside formal assessment tools.

Total knee arthroplasty remains the right answer for a significant proportion of stage 4 patients if the timing is right. The preserve-before-replace framework is not an argument against surgery; it is a commitment to reaching that decision at the right time and for the right reasons. The approximately 87% five-year arthroplasty-deferral rate observed across 6,035 patients in a national database suggests that, for carefully selected individuals, meaningful deferral is achievable. The clinician's responsibility is to distinguish who benefits from that time — and who is better served by regeneration or replacement now.

1. [1] The Role of Sex, Age, and BMI in Treatment Decisions for Knee Osteoarthritis: Conservative Management versus Total Knee Replacement. (2025). https://doi.org/10.1186/s13018-025-05552-2 https://doi.org/10.1186/s13018-025-05552-2
2. [2] Genicular Nerve Ablation in Knee Osteoarthritis: A Randomized Prospective Study. (2025). https://doi.org/10.1590/1413-785220253303e289218 https://doi.org/10.1590/1413-785220253303e289218
3. [3] Can Artificial Intelligence Help Orthopaedic Surgeons in the Conservative Management of Knee Osteoarthritis? A Consensus Analysis. (2025). https://doi.org/10.3390/jcm14030690 https://doi.org/10.3390/jcm14030690
4. [4] Conservative and Emerging Rehabilitative Approaches for Knee Osteoarthritis Management. (2024). https://doi.org/10.31579/2694-0248/082 https://doi.org/10.31579/2694-0248/082
5. [5] Arthroscopic Surgery Is Not Superior to Conservative Treatment in Knee Osteoarthritis: A Systematic Review and Meta-analysis of Randomized Controlled Trials. (2024). https://doi.org/10.1186/s12891-024-07813-3 https://doi.org/10.1186/s12891-024-07813-3