The three signals that prompt escalation

Knowing when to escalate is, for many people with knee osteoarthritis, the most pressing practical question they face. Exercise and physiotherapy remain the cornerstone of management — but there are three specific clinical signals that indicate injection therapy should enter the plan.

The first is persistent, inadequately controlled pain despite a sustained and structured exercise programme. This is not about finding exercise difficult or uncomfortable; it means that pain levels remain high enough to impair daily function even after a genuine period of quadriceps strengthening, low-impact aerobic activity, and, where relevant, weight management.

The second is an acute inflammatory flare. When a joint becomes acutely swollen and painful, it can make participation in physiotherapy physically impossible. Here, an injection may reduce inflammation quickly enough to allow rehabilitation to resume — the goal being to restore exercise capacity, not to bypass it.

The third signal is a surgical timeline. Some patients are not yet ready for joint replacement, or wish to avoid it altogether; others are waiting for an operation. In both situations, injection therapy may provide a period of meaningful relief that preserves function and quality of life in the interim.

It is worth stating plainly: reaching this point is not a failure. OA follows a progressive course in a significant proportion of patients, and escalation to injection therapy reflects where the condition has moved, not how well the patient has tried. Both NICE (NG226, 2022) and the 2026 expert consensus by Huang and colleagues frame this within a stepwise model in which conservative care continues alongside injection therapy rather than being set aside.

Injections as pain enablers, not replacements for exercise

The Liles 2024 scoping review of 11 studies produces a finding worth sitting with: in most cases, adding a knee injection to a structured exercise programme did not produce superior pain outcomes compared to exercise alone. Corticosteroid plus exercise matched exercise-only for pain improvement in the majority of studies. Hyaluronic acid results were mixed — two studies found superiority over exercise alone, two did not.

Rather than undermining injections, this finding clarifies what they actually do. When pain levels are high enough to prevent meaningful participation in rehabilitation, a patient cannot access the functional benefits that exercise offers. An injection that brings pain down to a workable threshold restores that access — and it is the sustained engagement with movement that then delivers durable improvement. The injection removes the barrier; the exercise produces the result. This is why clinical guidance, including NHS England's multi-modal framework, consistently places injection therapy in support of rehabilitation, not as a parallel or alternative track to it.

Patients beginning injection treatment are generally best served by understanding this from the outset. The goal is not injection-led symptom control as an end in itself, but sufficient pain relief to keep rehabilitation going — and, where possible, to intensify it. Reframing the injections this way also sets realistic expectations: if ongoing exercise is not part of the plan, the benefit of the injection alone is likely to be limited and short-lived.

Corticosteroid injections: rapid relief with firm limits

Corticosteroid injections have been used in knee OA management for decades, and for good reason: they act quickly, reliably reduce inflammation, and can produce meaningful pain relief within days. For patients in the acute inflammatory phase — where the joint is swollen and reactive — they remain a well-established, guideline-supported first choice.

NICE (NG226, 2022) recommends corticosteroids specifically for short-term symptom relief. The guidance notes that restricting their use to this indication may reduce how frequently they are given in current practice, which implies some over-prescription exists — but that is a systems-level observation, not a reason to view the treatment itself with suspicion. Many patients benefit substantially from a well-timed corticosteroid injection at the right stage of their condition.

The practical constraint is frequency. Clinical bodies advise a cap of roughly 3–4 injections per year; beyond this, cumulative exposure may contribute to cartilage degradation and progressive joint deterioration. A 2025 BMJ guideline (Benzon) examined the safety and efficacy of corticosteroid joint injections in adults, reinforcing the importance of considered use within defined limits. A 2026 multidisciplinary expert panel (Huang and colleagues) went further, producing 22 standardised recommendations covering agent selection, patient eligibility, dosing intervals, and contraindications — bringing greater consistency to an area that has historically varied in practice.

One additional consideration arises for patients approaching a surgical threshold. Evidence suggests that corticosteroid injections delivered in the period before total knee replacement may increase the risk of post-operative infection. This does not mean injections should be withheld from patients who need them, but it does prompt a timely clinical question: when the frequency limit is approaching and surgery appears likely, whether a longer-acting alternative might better serve the patient's remaining pre-operative window.

Hyaluronic acid and PRP: the intermediate tier

Beyond corticosteroids, two options occupy a distinct intermediate tier — hyaluronic acid (HA) and platelet-rich plasma (PRP) — and they work through fundamentally different mechanisms.

HA, sometimes called viscosupplementation, supplements the natural joint fluid that thins and degrades as OA progresses. It acts as a lubricant rather than an anti-inflammatory, making it biologically suited to early-to-moderate disease where joint friction and stiffness are the dominant complaints. Clinical guidelines conditionally recommend it for patients with moderate-to-severe OA who have not responded adequately to exercise or oral analgesics, though the evidence base remains uneven. As noted in s2, the Liles 2024 scoping review found that HA's benefit over exercise alone was inconsistent across studies — a finding that reflects genuine heterogeneity in the HA literature rather than a settled verdict either way.

PRP operates differently. Prepared from the patient's own blood, it delivers a concentrated mix of growth factors into the joint, modulating the inflammatory environment and, in some studies, supporting tissue-level changes that HA does not target. The comparative evidence here is notably consistent. A 2021 systematic review and meta-analysis by Belk and colleagues — since cited by over 500 published studies — found that patients treated with PRP for knee OA experienced meaningfully better clinical outcomes than those receiving HA. A 2024 review by Murali and colleagues reached the same conclusion, confirming PRP's advantage for both pain relief and function.

For patients who have exhausted or are unsuitable for repeated corticosteroid use, this evidence positions PRP as the stronger intermediate option. Neither treatment is curative, and neither replaces the exercise foundation established at earlier stages — but PRP's biological rationale and comparative track record make it the more defensible choice at this point in the pathway.

Arthrosamid® (iPAAG): one injection, sustained relief

Polyacrylamide hydrogel — marketed as Arthrosamid® and abbreviated iPAAG — sits in a different category from everything discussed so far. It is not an anti-inflammatory, not a lubricant, and not derived from blood. A single 6 ml injection integrates with the synovial tissue surrounding the joint, forming a stable cushioning matrix that is not absorbed or broken down over time in the way corticosteroids or HA are. That permanence within the joint is the mechanism behind its duration of effect.

The published evidence is still building but the early data are notable. A 2024 open-label study by Bliddal and colleagues followed 49 patients for 52 weeks after a single ultrasound-guided iPAAG injection. Mean WOMAC pain scores fell by 17.7 points from baseline (95% CI −23.1 to −12.4; p<0.0001), and the reduction was sustained across the full 12-month follow-up — not a peak-and-fade profile. At 52 weeks, 62.2% of participants met the OMERACT-OARSI responder threshold, which is a recognised standard for clinically meaningful improvement in OA trials. No device-related serious adverse events were recorded.

Larger real-world data followed. The LUNA multicentre observational trial, with 199 participants with moderate-to-severe knee OA, presented its 12-month results in November 2025 at ISIAT in Bucharest. A single 6 ml injection again produced sustained pain relief at one year across a considerably broader patient population. Five-year follow-up monitoring is ongoing — meaning longer-term durability data are being collected but are not yet available. The current evidence base is encouraging; it is not yet complete.

The practical appeal of this option is specific: it is suited to patients who want to avoid the cycle of repeated injections, those approaching the corticosteroid frequency limit, and those seeking to delay surgery without adding further procedural burden. For that profile, a single injection with a demonstrated 12-month signal offers a meaningfully different proposition.

What the evidence still doesn't tell us

Four injection types, four different mechanisms, and no single one that fits every patient — that is the honest read of the evidence laid out above. What the research does support is a rough clinical logic: corticosteroids suit acute inflammatory flares and short-term relief; PRP is the stronger evidence-based option once the corticosteroid frequency ceiling becomes a constraint; HA remains useful where stiffness and lubrication are the dominant complaint; iPAAG offers the longest projected duration from a single procedure, most relevant for patients managing a surgical timeline.

What is still absent is a head-to-head RCT comparing these options at the specific point of exercise failure — the comparison that would most directly answer the question most patients arrive with. The five-year durability endpoint for Arthrosamid® from the LUNA trial is also still outstanding. Current practice navigates these gaps using OA severity, inflammatory pattern, proximity to surgical threshold, and the patient's prior response to treatment.

Those are precisely the variables a specialist assessment is designed to weigh. The pathway is evidence-based; matching the right option to a specific clinical profile is where individual review adds what published trials cannot.