Three injections, three mechanisms

PRP, hyaluronic acid, and iPAAG are frequently compared when patients research knee OA injections — yet they work through entirely different mechanisms, which is why their effects play out over very different timescales.

Platelet-rich plasma (PRP) is prepared from the patient's own blood, spun to concentrate platelets and the growth factors they carry. Once injected, those growth factors trigger a localised biological response in the joint tissue — stimulating cells rather than physically altering the joint environment.

Hyaluronic acid (HA) mimics a molecule already present in healthy synovial fluid. It temporarily supplements the joint's natural lubrication and may modulate pain signalling, though its effects are time-limited as the substance is gradually metabolised.

Polyacrylamide hydrogel (iPAAG, brand name Arthrosamid) is neither a biologic nor a lubricant. Composed of 97.5% sterile water and 2.5% cross-linked polyacrylamide, it integrates into the synovial membrane as a permanent, non-biodegradable cushioning scaffold — one that does not break down and does not require repeat dosing.

Those mechanistic differences produce a rough duration spectrum: corticosteroid injections act fastest but fade within weeks; HA effects last around six months; PRP tends to persist for roughly a year; and iPAAG, once integrated, may offer multi-year relief from a single injection. Understanding where each option sits on that spectrum helps frame the trade-offs explored in the sections that follow.

PRP: a real signal, but preparation is everything

The strongest evidence for any of the three agents sits with PRP — but reading that evidence carefully reveals a nuance that changes how individual trial results should be interpreted.

The RESTORE trial, published in JAMA in 2021 with 288 participants (KL grade 2–3), is the most rigorous single study: three weekly leukocyte-poor PRP injections against saline placebo, followed over 12 months with MRI assessment of cartilage volume alongside pain outcomes. A 2025 meta-analysis then pooled 18 RCTs involving 1,995 patients and confirmed PRP's superiority over placebo on both VAS (a pain scale) and WOMAC (a composite knee-function score) at every follow-up time point measured.

What makes that finding meaningful rather than merely statistically significant is the MCID — the minimum clinically important difference, the threshold above which patients typically notice a real change in daily life. For WOMAC, that threshold is 6.4 points; for VAS, it is 1.37 points. PRP exceeded the WOMAC MCID at all time points in the meta-analysis, and exceeded the VAS MCID at the 3- and 6-month marks.

The critical caveat is formulation. Not all PRP is the same: the concentration of platelets in the final injection varies considerably between preparation protocols, and the 2025 meta-analysis identified this as the dominant moderator of efficacy. High-platelet preparations — those reaching at least one million platelets per microlitre — consistently crossed the MCID threshold. Lower-platelet preparations did not. A trial reporting modest results may therefore reflect a dilute formulation rather than a failure of the treatment class itself.

Clinical effects tend to last roughly a year, but individual variation is substantial — and likely tied, in part, to the same formulation differences that separate stronger from weaker trial results.

Hyaluronic acid: a well-mapped benefit curve

Of the three agents compared here, HA has the most precisely characterised benefit curve — and that precision is itself clinically useful.

A meta-analysis of 54 trials involving 7,545 participants mapped the trajectory in detail. Measurable benefit appears from week 4 (effect size 0.31), builds to a peak at week 8 (effect size 0.46), then gradually declines to a residual effect size of 0.21 by week 24. To put the week-8 peak in plain terms: it is comparable in magnitude to COX-2 inhibitors (effect size 0.44) — a meaningful analgesic effect, not a marginal one, at that window.

By 12 months, pain scores in HA-treated patients typically return towards baseline — a pattern seen consistently across direct comparison trials. This is not a treatment failure; it reflects how the agent behaves biologically. HA is gradually metabolised, and its effects fade as it clears the joint.

The clinical trade-off is therefore predictable rather than uncertain: strong benefit across roughly a two-month window, with a known fade by six months and little residual effect at a year. For patients who respond well at the 8-week mark, repeating the course is a clinically reasonable strategy — and HA's established safety profile, built over decades of use, makes that a low-risk proposition. Broad clinical familiarity also means it is widely accessible and straightforward to administer.

iPAAG: what makes a permanent scaffold different

Unlike HA or PRP, Arthrosamid is not absorbed or broken down by the body. Composed of 97.5% sterile water and 2.5% cross-linked polyacrylamide, it integrates permanently into the synovial membrane — forming a cushioning elastic scaffold that remains in place rather than metabolising over weeks or months. That distinction in mechanism produces a fundamentally different clinical profile: a single injection, with no scheduled repeat course.

The most complete efficacy data come from a 52-week open-label multicentre study of 49 patients who received a single 6 mL ultrasound-guided injection. At one year, WOMAC pain scores had fallen by 17.7 points (95% CI −23.1 to −12.4; p<0.0001), and 62.2% of participants met the OMERACT-OARSI responder threshold — a composite criterion used in OA trials to define meaningful clinical improvement. A published RCT subsequently compared iPAAG directly to HA and found it non-inferior at one year, with a comparable safety profile. It is worth being precise about what that means: non-inferiority was established, but superiority over HA was not.

A 2025 retrospective cohort of 150 patients (KL grade II–IV) adds a useful comparative frame. All three injectates — iPAAG, HA, and corticosteroid — reduced VAS pain scores from 7 to 3 at three months, demonstrating broadly similar short-term relief. By 12 months, both HA and corticosteroid groups had returned to near-baseline, whereas the iPAAG group maintained slight improvement. The trend favoured iPAAG, but the 12-month between-group difference did not reach statistical significance (p=0.128) — an important caveat when interpreting durability claims.

Longer-term observational registry data, following iPAAG-treated patients for 7–10 years, suggest good tolerability and no serious systemic adverse events, with over half of treated patients avoiding or delaying total knee replacement over that horizon. These are registry-level findings, not RCT-level evidence, and should be weighted accordingly. The evidence base for iPAAG is promising and accumulating rapidly since 2021, but it remains younger and thinner than that supporting HA or PRP.

What head-to-head trials actually show

Pulling the comparative trials together reveals a reasonably clear picture in some areas and genuine uncertainty in others — and it is worth being explicit about which is which.

What the evidence confirms

A 2025 systematic review and meta-analysis of RCTs directly comparing PRP to HA found leukocyte-poor PRP superior to HA on WOMAC total score at both 6 and 12 months. Platelet concentration was again the dominant moderator: the advantage over HA tracked closely with high-platelet formulations, reinforcing that PRP results from one trial do not straightforwardly generalise to a clinical setting using a different preparation protocol or a lower platelet yield.

On the iPAAG-versus-HA comparison, a published RCT established non-inferiority at one year — meaning a single iPAAG injection performed at least as well as HA across that follow-up period, with a comparable safety profile.

What remains genuinely open

IPAAG's superiority over HA at 12 months has not been demonstrated. The retrospective cohort described in the previous section showed a trend in that direction, but the between-group difference did not reach statistical significance (p=0.128). Non-inferiority and a non-significant trend are meaningful findings — but they are not superiority.

No published RCT has placed all three agents in a single head-to-head study. The comparative picture is constructed from separate bilateral comparisons, each with its own protocol, patient population, and follow-up window. The overall evidence is genuinely informative about likely relative positioning, but readers should understand it as a mosaic rather than a single definitive trial.

Which option fits which patient

Choosing between these agents in practice comes down to several factors that vary by individual — disease severity, tolerance for repeated injections, cost, and whether speed or longevity of benefit matters more at this stage.

Disease severity is the starting point. The most robust evidence across all three agents comes from early-to-moderate OA (KL grade I–III); data in severe disease (KL IV) are thinner for every option, and patients in that range may be moving towards a different clinical conversation altogether.

Injection schedule divides the agents sharply. HA typically involves a course of three to five injections, often repeated periodically; PRP is similarly given as a course; iPAAG is a single injection with no planned repeat. For patients who find repeated procedures burdensome — whether for practical, cost, or anxiety-related reasons — that distinction is clinically relevant.

Onset versus durability follows a similar split. Patients managing an acute-on-chronic flare, or those needing predictable short-term relief, may favour HA's well-characterised 8-week peak. Patients who primarily want to reduce how often they return for treatment may weight iPAAG's single-injection profile more heavily. PRP sits between the two on the durability spectrum.

PRP preparation is worth raising directly with any centre: high-platelet formulations (≥1,000,000 platelets/µL) drove most of the evidence reviewed here; not all clinics prepare PRP to the same specification.

Cost and insurance coverage differ substantially between agents and are not uniform across providers. Confirming this before committing to a course is sensible.

None of these factors produces a universal answer — suitability depends on a clinical assessment that weighs imaging, symptom pattern, functional goals, and the patient's own priorities.

1. [1] In Vitro Effects of PRP, Ozonized PRP, Hyaluronic Acid, Paracetamol, and Polyacrylamide on Equine Synovial Fluid-Derived Mesenchymal Stem Cells. (2025). https://doi.org/10.3390/life15101558 https://doi.org/10.3390/life15101558
2. [2] PRP Injections for the Treatment of Knee Osteoarthritis: The Improvement Is Clinically Significant and Influenced by Platelet Concentration: A Meta-analysis of Randomized Controlled Trials. (2025). https://doi.org/10.1177/03635465241246524 https://doi.org/10.1177/03635465241246524
3. [3] Therapeutic trajectory following intra-articular hyaluronic acid injection in knee osteoarthritis – meta-analysis. (2010). https://doi.org/10.1016/j.joca.2010.09.014 https://doi.org/10.1016/j.joca.2010.09.014
4. [4] Platelet rich plasma compared to viscosupplementation in the treatment of knee osteoarthritis: A systematic review and meta-analysis of randomised controlled trials with 6 month and 12 month follow-up. (2025). https://doi.org/10.1002/jeo2.70335 https://doi.org/10.1002/jeo2.70335
5. [5] Effect of Intra-Articular Platelet-Rich Plasma vs Placebo Injection on Pain and Medial Tibial Cartilage Volume in Patients With Knee Osteoarthritis: The RESTORE Randomized Clinical Trial. (2021). https://doi.org/10.1001/jama.2021.19415 https://doi.org/10.1001/jama.2021.19415