Why OA stage drives injection choice

When a consultant mentions corticosteroid, hyaluronic acid, PRP, and Arthrosamid in the same conversation, the natural question is: which one is actually right for my knee? The honest answer depends less on how much pain is present and more on how far the arthritis has progressed — because each injection addresses a different aspect of the disease, and the tissue available to treat changes as OA advances.

Clinicians typically use the Kellgren-Lawrence (K-L) radiographic grading scale — from grade I (minor joint-space changes) to grade IV (severe narrowing and bone-on-bone contact) — as the primary staging tool. At grade I or II, the joint retains enough cartilage and synovial tissue to respond to lubrication or short-term symptom suppression. The distinction is consequential: cartilage loss is largely irreversible, so an injection suited to early disease may offer little to the same knee a few years later.

Beyond K-L grade, patient factors further shape the decision — BMI, activity level, how many previous injection courses have been tried, and how close a patient is to considering surgery. No large trial has yet tested all four agents head-to-head at matched K-L grades; the available guidance rests on indirect comparisons and meta-analyses, which is why individual clinical assessment remains central to any choice.

Corticosteroid injections — fast relief for acute flares

Corticosteroids earn their place in knee OA management through speed. When synovitis flares — the knee swells, warms, and becomes acutely painful — an intra-articular corticosteroid can bring meaningful relief within days, a response no other injection in this class matches. That rapid anti-inflammatory effect makes them a practical tool for acute flare management rather than a standing maintenance strategy.

The trade-off is durability. Relief typically lasts weeks to a few months — the shortest window of all four options. The 2024 Bensa network meta-analysis (PMC11457815) confirmed that corticosteroids produce results comparable to hyaluronic acid and PRP only at short-term follow-up; at 6 and 12 months, PRP consistently outperforms them on pain and function scores. Multiple RCTs in K-L grade II–III patients have found that PRP produces significantly lower VAS and WOMAC scores than corticosteroid at both those time points, with a more favourable adverse-effect profile.

The stronger caution around corticosteroids relates to frequency rather than a single injection. Repeated administration carries a documented risk of chondrotoxicity — accelerated cartilage breakdown — which limits their role to episodic use when the knee flares acutely, not as a recurring substitute for a longer-acting treatment. For K-L grade 1–3 with an identifiable inflammatory episode, that role is well-supported by the evidence. Once flares become frequent, or OA has progressed to K-L grade 3 or beyond without a clear acute component, other options are more appropriate.

Viscosupplementation — lubrication across early-to-moderate OA

Hyaluronic acid (HA) viscosupplementation has the longest clinical history of any intra-articular knee injection, with the first formal assessment published by Peyron and Balazs in 1974. Its mechanism is straightforward: synovial fluid in an arthritic knee loses its normal viscoelastic properties, reducing lubrication and shock absorption. HA injections top up and temporarily restore that fluid environment. The effect is mechanical support for the joint, not cartilage repair.

The evidence base is strongest for Kellgren-Lawrence grade I–III disease. A 2019 systematic review by Nicholls et al. found that pooled outcomes improve substantially when end-stage patients are excluded from HA trials — a clear signal that K-L grade 4 knees respond poorly, and that meaningful residual cartilage and synovial tissue are prerequisites for a useful response.

The 2025 EUROVISCO expert consensus provides the most detailed patient-selection guidance to date, drawing on a Delphi process across seven European countries. On the positive side, HA can be used regardless of patient age and is supported across a range of comorbidities: diabetes, moderate-to-severe obesity, gout, meniscocalcinosis, and mild-to-moderate varus or valgus malalignment. Two factors predict poor response most strongly — obesity and advanced radiological OA (K-L grade 4). The 2024 EUROVISCO guidelines also advise against injecting during an active OA flare, and recommend avoiding HA in pregnancy.

Durability data from a cross-sectional study of 149 knees found mean efficacy lasting 48.2 weeks, though this shortens in patients with BMI above 27.5, multicompartmental involvement, more than three prior injection courses, or a sedentary lifestyle. HA sits most comfortably with patients in the mild-to-moderate OA range whose joint environment can still respond to improved lubrication.

PRP — a biologic option for mild-to-moderate OA

PRP takes a different approach to the problem. Rather than adding lubrication (as HA does) or suppressing inflammation directly (as corticosteroids do), it concentrates growth factors from the patient's own blood and re-introduces them into the joint environment. The aim is to support the knee's natural tissue responses — not to regenerate cartilage, but to create conditions more conducive to it. As an autologous preparation, it carries no rejection risk and shows a favourable adverse-event profile in published series.

Its position on the disease-stage spectrum is Kellgren-Lawrence grade II–III. Onset is slower than corticosteroid — benefits typically emerge over weeks — with durability estimated at 6 to 12 months. A 2025 systematic review and meta-analysis of RCTs found PRP superior to HA for patient-reported pain and function at both 6 and 12 months, with platelet concentration identified as a modifier of that relative benefit. The medium-term advantage over corticosteroid — already apparent at the 6-month mark in multiple 2021–2026 RCTs — is consistent across those trials, as is the more favourable adverse-event profile.

The meaningful caveat is preparation. Unlike HA or corticosteroid — both standardised pharmaceutical products — PRP is produced from whole blood by in-clinic centrifugation, and platelet concentration varies considerably between protocols and machines. That variability makes direct trial-to-trial comparison difficult and has led guideline bodies to hold back from strong recommendations. A 2026 RCT in K-L grade II–III patients found PRP plasma gel outcomes broadly comparable to HA alone, while a combination of PRP, HA, and corticosteroid yielded the most consistent results — a sign that preparation and combination protocols are still being refined.

For patients at K-L grade II–III who have moved beyond acute flare management, PRP is currently the best-supported single-agent biologic choice at medium-term follow-up.

Arthrosamid — a single injection for moderate-to-severe OA

For patients at Kellgren-Lawrence grade 3–4 whose relief from corticosteroid or HA injections has worn off but who are not yet candidates for joint replacement, Arthrosamid (iPAAG) addresses a gap the other options cannot easily cover.

The product is a hydrogel comprising 97.5% water and 2.5% cross-linked polyacrylamide, CE-marked by Contura International Ltd as a class IIb medical device. It is administered as a single 6 mL intra-articular injection. Unlike HA, which disperses and is absorbed, iPAAG is non-biodegradable: it integrates into the synovial membrane and remains as a permanent mechanical cushioning layer. The action is physical rather than regenerative — it does not repair cartilage.

Evidence for sustained efficacy comes from a 2022 PRISMA systematic review across 463 patients showing statistically significant WOMAC improvements at 52 weeks and beyond, and a 2026 systematic review confirming those gains to 2–3 years from a single injection, with WOMAC pain reductions ranging from −15.7 to −20.8 across the follow-up period. In head-to-head data, iPAAG showed no significant difference versus HA at 3 months but outperformed intra-articular corticosteroid at 6 months (median WOMAC 47.5 vs 57; P=0.008). Transient post-injection pain affected around 16% of patients; no serious complications were reported. A prospective multicentre study presented at EULAR 2023 — with three-year extension data already published in 35 participants — is continuing to track WOMAC and Patient Global Assessment outcomes to five years.

Two uncertainties are worth noting. Biomarker studies have identified anti-inflammatory changes in synovial fluid at 3 months, suggesting the mechanism may extend beyond mechanical cushioning alone, but this remains under active NHS-supported investigation. Arthrosamid is also not appropriate when the knee is severely unstable or deformed, or when arthroplasty is already the clinically indicated next step — individual assessment determines suitability in those scenarios.

Stage-by-stage decision framework and when to seek assessment

The framework below maps Kellgren-Lawrence grade to the most evidence-supported injection at each level — but disease stage is the starting point, not the whole decision. Individual patient factors consistently shift the choice within any grade, and the mapping should be read as a clinical orientation, not a formula.

K-L grade 1–2: Corticosteroid first, for acute inflammatory flares — onset within days, relief lasting weeks. HA viscosupplementation follows for medium-term lubrication once the acute phase settles, with a mean efficacy duration of around 48 weeks in unselected patients.

K-L grade 2–3: For persistent, non-acute symptoms, HA and PRP are the principal options. The 2025 systematic review and meta-analysis of RCTs found PRP superior to HA at 6 and 12 months, making it the better-supported choice where medium-to-longer biologic activity is the goal. HA remains appropriate where PRP preparation quality cannot be assured or the clinical picture favours a standardised product.

K-L grade 3–4, prior injections insufficient, surgery deferred: Arthrosamid — a single 6 mL injection with documented WOMAC improvements sustained to three years — is the option for patients who have not found durable relief from shorter-acting therapies and are not yet surgical candidates. It is less suitable when arthroplasty is already the indicated step.

Patient-level modifiers

Four factors consistently shift efficacy within any grade:

• BMI above 27.5 — the strongest single predictor of poor HA response, and it shortens duration of benefit
• Multicompartmental involvement — reduces the yield of any single-compartment-targeted injection
• More than three prior HA courses — independently associated with diminishing returns
• Sedentary lifestyle — shortens HA durability in cross-sectional data

Proximity to the surgical threshold also limits Arthrosamid's suitability: published evidence cannot reliably predict individual benefit when arthroplasty is already clinically close.

Evidence gaps

The stage mapping above rests on individual-therapy trial data rather than direct cross-agent comparisons at the same K-L grade — that matched head-to-head evidence remains absent. Long-term cartilage-structural outcomes for iPAAG also extend only to three years in the current dataset. These gaps mean the framework will likely sharpen as larger comparative trials report.

Because the right injection depends on grade, prior response, imaging findings, and individual anatomy, a consultant-led assessment that includes imaging review is the appropriate next step for anyone uncertain which option fits their situation. Lincolnshire Knee, part of the MSK Doctors group, provides that assessment at Sleaford (NG34) and Grantham (NG31) without the need for a GP referral.

1. [1] Efficacy of novel polyacrylamide hydrogel injection for osteoarthritis. (2026). https://doi.org/10.1016/j.knee.2026.104492 https://doi.org/10.1016/j.knee.2026.104492
2. [2] A Cross-Sectional Study of Factors Predicting the Duration of the Efficacy of Viscosupplementation in Knee Osteoarthritis. (2024). https://doi.org/10.3390/jcm13071949 https://doi.org/10.3390/jcm13071949
3. [3] PRP Plasma Gel versus Hyaluronic Acid versus Corticosteroid for Knee Osteoarthritis: A Prospective Randomized Clinical Trial. (2026). https://doi.org/10.63403/re.v33i2.480 https://doi.org/10.63403/re.v33i2.480
4. [4] COMPARISON OF CORTICOSTEROID AND PLATELET-RICH PLASMA (PRP) INJECTIONS FOR SYMPTOMATIC RELIEF IN KNEE OSTEOARTHRITIS: A RANDOMIZED CONTROLLED STUDY. (2025). https://doi.org/10.36106/ijsr/2500538 https://doi.org/10.36106/ijsr/2500538
5. [5] A comparative study between role of platelet rich plasma (PRP) and corticosteroid injection in the treatment of osteoarthritis knee. (2021). https://doi.org/10.22271/ORTHO.2021.V7.I1E.2494 https://doi.org/10.22271/ORTHO.2021.V7.I1E.2494
6. [6] Efficacy and Safety of Intra-articular Platelet-Rich Plasma (PRP) Versus Corticosteroid Injections in the Treatment of Knee Osteoarthritis: A Systematic Review of Randomized Clinical Trials. (2025). https://doi.org/10.7759/cureus.80948 https://doi.org/10.7759/cureus.80948
7. [7] EUROVISCO Good Medical Practice Recommendations for the Use of Viscosupplementation with Hyaluronic Acid in the Management of Knee Osteoarthritis. (2025). https://doi.org/10.1177/19476035241286578 https://doi.org/10.1177/19476035241286578
8. [8] 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
9. [9] AB1185 POLYACRYLAMIDE HYDROGEL FOR THE TREATMENT OF KNEE OSTEOARTHRITIS: 3 YEAR FOLLOW UP RESULTS OF A PROSPECTIVE CLINICAL STUDY. (2023). https://doi.org/10.1136/annrheumdis-2023-eular.6038 https://doi.org/10.1136/annrheumdis-2023-eular.6038
10. [10] EUROVISCO Consensus Guidelines for the Use of Hyaluronic Acid Viscosupplementation in Knee Osteoarthritis Based on Patient Characteristics. (2024). https://doi.org/10.1177/19476035241271970 https://doi.org/10.1177/19476035241271970