Four injections, one decision — what you need to know first

Told you might benefit from a knee injection but unsure which one to ask for? That uncertainty is reasonable — four distinct options now have meaningful clinical evidence behind them, and they work in genuinely different ways.

Corticosteroid (cortisone) suppresses joint inflammation rapidly, often easing pain within days. The effect typically lasts 6–12 weeks, making it useful for acute flare-ups rather than long-term management.

Hyaluronic acid (HA) replaces the lubricating fluid that OA gradually depletes. Relief builds over several weeks and may last 6–12 months in suitable candidates.

Platelet-rich plasma (PRP) concentrates growth factors drawn from the patient's own blood to support tissue healing and dampen inflammation. Onset is slower — two to four weeks — but benefits may persist for 6–12 months or longer.

Polyacrylamide hydrogel (iPAAG) is the newest of the four. It acts purely as a mechanical cushion, integrating into the synovial lining to provide durable shock absorption without triggering a pharmacological response.

No single trial has compared all four head-to-head; the comparative picture comes from network meta-analyses and retrospective cohorts. Choice depends on disease severity, symptom pattern, and whether the priority is short-term rescue or longer-duration relief. The sections that follow examine each option in turn, then offer a practical decision framework.

Cortisone: rapid relief and why it's rarely a long-term solution

Steroid injections are often the first option a GP or rheumatologist reaches for — and there are good reasons for that. Corticosteroids work by suppressing the inflammatory cascade inside the joint: they reduce synovial swelling and the chemical signals that sensitise pain receptors, producing relief that most patients notice within two to three days.

The limitation is durability. Benefit typically peaks somewhere between six and eight weeks, and for most patients the effect has largely faded by twelve weeks. That short window matters, because it shapes what the injection is genuinely useful for — and what it is not.

Cortisone is widely available on the NHS and carries no direct patient cost, which makes it a natural default first step. For someone in significant acute pain, or waiting several months for a procedure such as a knee replacement, a well-timed corticosteroid injection can meaningfully bridge that gap.

The concern with repeated or long-term use is better evidenced than it once was. Corticosteroids can activate matrix metalloproteinases — enzymes that degrade cartilage matrix — raising the possibility of cumulative joint harm with frequent dosing. A 2020 randomised controlled trial published in the New England Journal of Medicine (Deyle et al.) found that patients receiving corticosteroid injections performed worse on functional outcomes at one year than those who followed a structured physiotherapy programme.

For a patient in an acute flare or approaching surgery, cortisone remains appropriate and effective. As a strategy for managing knee OA over months or years, the evidence does not support it.

Hyaluronic acid: lubrication and medium-term symptom relief

Think of hyaluronic acid as topping up the joint's natural lubricant. Healthy synovial fluid is rich in HA, a molecule that gives the fluid its viscous, shock-absorbing quality. In knee OA, that fluid becomes thinner and less effective — HA injection aims to restore it. The mechanism is primarily mechanical rather than pharmacological, which is the clearest practical distinction from cortisone: HA is not suppressing inflammation, it is addressing a lubrication deficit.

The clinical effect unfolds gradually. Most patients notice little change in the first two weeks; benefit tends to build over four to eight weeks and, in good responders, may extend for six to twelve months from a single injection.

The evidence base supports this picture. A 2019 meta-analysis specifically examining single-injection HA products (Vincent, Current Therapeutic Research) confirmed symptomatic efficacy over placebo in knee OA. The debate is not whether HA works — it is whether the benefit is large enough to justify routine commissioning. NICE and OARSI have reached different conclusions on this point, and neither guideline consistently recommends HA as standard care. The disagreement centres on cost-effectiveness, not on the direction of the clinical evidence.

In practical terms for UK patients, this means HA is usually self-funded. NHS prescription is uncommon outside specific pathways.

HA tends to suit patients with mild-to-moderate OA who still have meaningful cartilage preservation, where the primary problem is joint stiffness and friction rather than acute inflammation or severe mechanical loading.

PRP: the biologic case and what the evidence actually shows

Unlike cortisone or hyaluronic acid, PRP is prepared from the patient's own blood. A small sample is centrifuged to concentrate platelets, releasing growth factors — including PDGF, TGF-β, and IGF-1 — that modulate the inflammatory environment inside the joint and may support the health of remaining cartilage tissue. It acts through the body's own signalling biology rather than introducing a synthetic or pharmaceutical agent.

The evidence base for PRP is now the most extensive of any biologic injection for knee OA. A 2025 retrospective cohort by Gökçeoğlu and colleagues (205 knees, Kellgren–Lawrence grade 2–3, indicating moderate disease) found PRP produced the largest sustained improvements across all groups studied. On a standard 0–10 pain scale (VAS), patients receiving PRP averaged a reduction of 3.2 points at six to twelve months — a change of two or more points on this scale is generally regarded as clinically meaningful. On the WOMAC function questionnaire (scored 0–100, where higher scores reflect greater disability), PRP produced an average improvement of 20 points, the most of any intervention in the cohort.

A 2024 systematic review by Bensa et al. (cited more than 30 times) confirmed these findings at population level: PRP was superior to corticosteroid at every follow-up interval examined, and superior to hyaluronic acid at mid-to-long-term horizons, with differences exceeding the minimal clinically important difference — the threshold at which patients actually perceive the benefit.

The key limitation is preparation variability. PRP is not a standardised product. Leukocyte content, platelet concentration, and the method used to activate growth factor release differ between clinics and across published trials. This makes direct cross-study comparisons inherently imprecise — a PRP preparation at one centre may differ meaningfully from one at another.

PRP is least suited to acute inflammatory flares. The strongest evidence applies to moderate OA (Kellgren–Lawrence grade 2–3) in patients with functional goals seeking sustained, biologically active benefit over a six-to-twelve month horizon.

Polyacrylamide hydrogel: a single dose designed to last

Polyacrylamide hydrogel (iPAAG, brand name Arthrosamid) works on entirely different principles from the other three options. It contains no drug, no growth factor, and no biological component — a synthetic gel of 97.5% water and 2.5% cross-linked polyacrylamide, injected as a single 6 ml dose. It integrates into the synovial membrane and remains there permanently, providing mechanical cushioning from within the joint. Nothing is metabolised or absorbed; the effect is structural rather than pharmacological.

The clinical evidence supports durable benefit. Bliddal and colleagues' 2024 RCT in Clinical and Experimental Rheumatology (cited 34 times) showed a single iPAAG injection to be non-inferior to hyaluronic acid at twelve months — a meaningful benchmark, not a superiority claim. Prospective studies by the same group confirmed safety and effectiveness at six and twelve months post-injection. The LUNA trial reported a WOMAC pain improvement of approximately 17 points at its one-year interim analysis, with no serious device-related adverse events. Four-year follow-up data show the effect persisting to that horizon, and ten-year safety data have not identified structural complications, including in patients who subsequently underwent knee replacement.

Access is the principal constraint. iPAAG is currently available only through private clinics in the UK; NHS evaluation is under way but commissioning is not yet established, and it does not hold FDA approval in the United States. Direct head-to-head data comparing iPAAG with PRP are absent, which limits any ranking between the two at present.

The patient most likely to benefit from this route is one seeking a single-dose, long-duration intervention who is able to self-fund — someone whose primary need is sustained mechanical support rather than a biologic effect on the joint's tissues.

Choosing the right injection for your stage of OA

The four options covered here map onto different points in the OA journey rather than competing for the same patient.

Tier 1 — Acute flare, or bridging before surgery: Corticosteroid. NHS-funded, fast-acting within days, and short-lived at 6–12 weeks. The right tool when rapid inflammation control is the priority, not sustained relief.

Tier 2 — Mild-to-moderate OA, lubrication the main deficit: Hyaluronic acid. Gradual onset, benefit lasting around 6–12 months, typically self-funded in the UK. Suited to patients who need more duration than cortisone but are not yet seeking a biologic effect.

Tier 3 — Moderate OA with functional goals: PRP. The comparative evidence at six-to-twelve months now supports using it ahead of corticosteroid for patients in this category — its advantage exceeds clinically meaningful thresholds — rather than treating it as a second-line option.

Tier 4 — Single long-duration intervention, private-pay: iPAAG. Four-year durability data distinguish it from anything else on this list; suitable for patients seeking sustained mechanical support from a single procedure.

No injection reverses OA — each addresses symptoms and function by a different route. The most practically significant shift in the evidence is PRP's demonstrated advantage over corticosteroid at mid-to-long term for moderate disease with functional goals: for that patient profile, the evidence no longer supports cortisone as the default. Arriving at the right tier for your own situation, however, requires a clinical assessment covering disease stage, imaging, activity level, and treatment history — a conversation with a musculoskeletal specialist is the place to start.