From one default to a real menu of choices

If you have had a corticosteroid injection in the past and wondered whether anything better exists, the short answer in 2026 is: quite a lot has changed. The standard 'cortisone first, surgery later' pathway has given way to at least five mechanistically distinct injectable options — each with a different rationale, a different ideal patient, and a different expected duration of benefit.

Choosing between them now turns on three practical axes. First, how long does it last — from weeks with a steroid to potentially years with a hydrogel or regenerative therapy? Second, what does it actually do inside the joint — lubricate, dampen inflammation, scaffold tissue, or support repair? Third, who is it best suited to — which combination of OA severity, age, and comorbidities tips the balance toward one option over another?

Guideline bodies — NICE, EULAR, and the ACR — have not yet incorporated several of these options into formal recommendations, meaning clinical practice in 2026 is running ahead of the guidance your GP may have to hand. The sections below map that gap.

PAAG hydrogel — the durability case

Polyacrylamide hydrogel — sold under the brand name Arthrosamid — works differently from every other option on the current menu. A single 6 mL injection of 2.5% iPAAG integrates permanently into the synovial lining of the joint rather than dissolving, lubricating, or delivering a drug. The result is a soft, water-rich cushion that alters the mechanical environment of the joint without needing to be repeated or topped up.

The durability data behind this mechanism are now the strongest in the injection field. Two 2025-published five-year studies — one open-label (27 completers) and one RCT extension cohort (58 completers) — both show statistically significant sustained improvements in WOMAC pain, stiffness, and physical function across the full follow-up period, with no serious adverse device effects. A separate 2025 retrospective cohort of 150 patients adds a practical comparison: at 12 months, hyaluronic acid and corticosteroid groups had returned to near-baseline pain scores, while the iPAAG group maintained stable improvement.

Who tends to benefit

A 2025 real-world PROMs cohort of 314 knees has begun to define the patients most likely to reach meaningful clinical improvement: older age, absence of diabetes, lower Kellgren–Lawrence (KL) grade, and bilateral OA involvement all increased the odds of a good outcome. Among those with higher-grade disease, a notable proportion proceeded to total knee replacement within two years — reinforcing that iPAAG sits within the mild-to-moderate OA indication, not as an alternative to surgical planning.

The same cohort recorded complications in 155 of 314 knees; the nature and severity of these are still being characterised in the literature, and individual risk should be discussed with a specialist. Longer-lasting does not mean universally appropriate.

PRP — a cleaner evidence signal, with one important caveat

The evidence base for platelet-rich plasma has sharpened considerably. A 2025 meta-analysis pooling 12 RCTs found that leukocyte-poor PRP (LP-PRP) outperformed hyaluronic acid on every key patient-reported measure — VAS pain, WOMAC total score, WOMAC physical function, and IKDC — at 3, 6, and 12 months, with no meaningful difference in adverse events between the two. Two years ago, the PRP literature was a tangle of conflicting results; the 2025 analysis helps explain why.

The main culprit behind earlier inconsistency was protocol heterogeneity. Studies mixed leukocyte-rich and leukocyte-poor preparations, varied centrifugation speeds, and used different injection frequencies — producing results that could not sensibly be pooled. LP-PRP, which removes most white cells before injection, now has the more reliable signal. LR-PRP evidence remains messier and the two should not be treated as equivalent.

The caveat patients deserve to hear

A stronger symptomatic signal is not the same as evidence of disease modification. The RESTORE trial, published in JAMA, enrolled 288 patients with Kellgren–Lawrence grade 2–3 knee OA and compared three weekly LP-PRP injections against placebo. Pain improved meaningfully — but MRI cartilage volume showed no significant difference between groups at 12 months. This is an important distinction: PRP may reduce pain and improve function, but it has not been shown to slow the structural progression of OA. Patients who have been told PRP 'rebuilds cartilage' should be aware that current structural evidence does not support that claim.

None of this makes PRP a poor choice. For patients seeking a biologically active option with a well-characterised 12-month benefit profile and equivalent safety to HA, LP-PRP is a reasonable pathway — particularly where the longer single-injection durability of iPAAG is not the priority, or where cost or access is a factor.

Hyaluronic acid — reformulated, not retired

Hyaluronic acid has a reputation problem it does not entirely deserve. After years of mixed guideline signals and the arrival of newer modalities, many patients assume it has been quietly retired. The evidence says otherwise.

A 2026 evidence-based review drawing on 43 RCTs and 5,554 patients found that intra-articular HA significantly reduced pain versus placebo (SMD 0.94, 95% CI 0.09–1.78). That is a substantial evidence base by any standard. The live clinical question is no longer whether HA works but which formulation works best and for whom.

The single-injection shift

Traditionally, HA treatment meant three to five weekly clinic visits for a course of low-molecular-weight injections. The same 2026 review found that a single cross-linked high-molecular-weight HA injection produces comparable pain and function outcomes to those multi-injection regimens. For a patient who works full-time or travels to reach a clinic, that is a meaningful practical difference — equivalent benefit with fewer appointments.

Combination with PRP

HA+PRP protocols are attracting interest based on the rationale that the two work through complementary mechanisms: HA restores viscosity and joint lubrication, while PRP contributes anti-inflammatory growth factors. The biological logic is coherent, but head-to-head trial data comparing the combination with either agent alone remain limited. Clinicians considering this route are working ahead of the formal evidence.

Why guidelines still diverge

NICE, EULAR, and ACR hold varying positions on HA — a fact sometimes misread as evidence of inefficacy. The disagreement sits primarily in methodology: how trials were pooled, which comparators counted, and what outcome threshold constitutes a clinically meaningful difference. Forty-three randomised trials are not being dismissed; the debate is about how to interpret them consistently.

For patients who want a well-tolerated, widely available option with a long safety record — particularly those who are ineligible for or not yet ready to consider newer modalities — HA remains a clinically reasonable choice.

Corticosteroids and emerging biologics — where each now fits

Steroids have not disappeared from the injection menu — their role has simply become more precisely defined. For acute inflammatory flares, where rapid symptom control matters, a corticosteroid remains clinically appropriate. What has changed is the formulation. Extended-release triamcinolone acetonide microspheres (FX006) sustain measurable synovial drug concentrations through Week 12, whereas standard crystalline suspension becomes undetectable in most patients by Week 6. More significantly, the extended-release formulation dramatically reduces peak systemic plasma exposure — 836 pg/mL versus 9,629 pg/mL with the standard form — addressing a longstanding patient concern about systemic corticosteroid effects. Steroids used in this way function best as a fast-onset bridge or flare-management tool rather than a primary long-term strategy.

Orthobiologics — promising, not yet routine

The most scientifically compelling developments in 2025–2026 sit in the orthobiologic category, but they require honest framing. A double-blind RCT of allogeneic bone marrow–derived mesenchymal stem cells (BM-MSCs) in 24 knee OA patients showed between-group improvements in WOMAC (p=0.02) and KOOS (p=0.028) at 9 months, with T2 MRI mapping suggesting a cartilage-protective effect at 12 months — the strongest disease-modification signal produced by any injection modality to date. The caveat is unavoidable: 24 patients is a small sample, and the finding, however encouraging, awaits replication at scale.

Adipose-derived stromal vascular fraction demonstrated 12-month benefit over rehabilitation alone in a separate 2025 RCT (62 completers), further adding to the regenerative case. At the furthest frontier, human umbilical cord MSC–derived exosomes have cleared first-in-human safety data with early MRI and clinical score improvements — but sample sizes and follow-up remain insufficient for any adoption guidance.

For patients, the honest message is this: orthobiologics represent an active and credible research direction, but they are not yet guideline-ready options. Monitoring this space is warranted; acting on current data alone is premature.

Matching the right injection to the right patient

Running through all of this evidence, a practical question remains: how does a specialist actually decide which option fits which patient?

The first sorting factor is OA grade. Kellgren–Lawrence grading does real work here. Lower-grade disease (KL 1–2) opens up the widest menu — PAAG, PRP, HA, and extended-release corticosteroid are all in scope depending on other factors. Higher-grade disease narrows the field considerably; the 2025 PAAG cohort data showed patients with advanced OA were more likely to proceed to total knee replacement within two years, making surgical consultation the more appropriate next step than any injection.

The second filter is symptom character. An acute inflammatory flare — warm, swollen joint, pain at rest — points toward corticosteroid for rapid control. Mechanical joint-space discomfort without significant inflammation is where PAAG and HA are better positioned. Where biological activity and longer-term symptom relief are the priority, and the patient has realistic expectations about structural outcomes, LP-PRP is a reasonable discussion.

Health factors add further nuance. Diabetes reduces the likelihood of a favourable PAAG response, as the 2025 cohort evidence confirmed. Prior injection history, overall fitness, and treatment goals all inform a shared decision — there is no universal correct answer.

One point worth stating clearly: several 2025-supported options, including PAAG and orthobiologics, remain outside current NICE and EULAR recommendations and are accessed through specialist private practice in the UK. Guideline lag is real, and patients researching these options should expect to have that conversation with a specialist rather than a GP.

No injection is appropriate without a structured assessment first — history, examination, and imaging review are the minimum. If you are unsure where you sit in this picture, an eligibility assessment is the right starting point.

1. [1] Comparative efficacy of polyacrylamide hydrogel versus hyaluronic acid and corticosteroids in knee osteoarthritis: A retrospective cohort study (Medicine, 2025). (2025). https://doi.org/10.1097/MD.0000000000044655 https://doi.org/10.1097/MD.0000000000044655
2. [2] Intraarticular leukocyte-poor PRP injection is more effective than HA injection in knee OA: systematic review and meta-analysis of 12 RCTs (Knee Surg Relat Res, 2025). (2025). https://doi.org/10.1186/s43019-025-00266-5 https://doi.org/10.1186/s43019-025-00266-5
3. [3] Effect of Intra-articular PRP vs Placebo on Pain and Medial Tibial Cartilage Volume: The RESTORE Randomized Clinical Trial (JAMA, 2021). (2021). https://doi.org/10.1001/jama.2021.19415 https://doi.org/10.1001/jama.2021.19415
4. [4] Sustained symptom relief and safety over five years following a single intra-articular injection of 2.5% polyacrylamide hydrogel (Clin Exp Rheumatol, 2025). (2025). https://doi.org/10.55563/clinexprheumatol/bsper8 https://doi.org/10.55563/clinexprheumatol/bsper8
5. [5] A prospective, open-label clinical investigation of a single intra-articular polyacrylamide hydrogel injection: a 5-year extension study (J Orthop Surg Res, 2025). (2025). https://doi.org/10.1186/s13018-025-06526-0 https://doi.org/10.1186/s13018-025-06526-0
6. [6] Intra-Articular Injection of Human Bone Marrow–Derived Mesenchymal Stem Cells in Knee OA: A Randomized, Double-Blind, Controlled Trial (Cell Transplantation, 2025). (2025). https://doi.org/10.1177/09636897241303275 https://doi.org/10.1177/09636897241303275
7. [7] Polyacrylamide hydrogel injections in knee osteoarthritis: A PROMs-based 24-month cohort study (J Clin Orthop Trauma, 2025). (2025). https://doi.org/10.1016/j.jcot.2025.103136 https://doi.org/10.1016/j.jcot.2025.103136
8. [8] Synovial and systemic PK of triamcinolone acetonide (FX006) vs crystalline suspension in knee OA (J Orthop Res, 2018). (2018). https://doi.org/10.1016/j.joca.2017.10.003 https://doi.org/10.1016/j.joca.2017.10.003
9. [9] Intraarticular injection of the stromal vascular fraction for knee osteoarthritis: a prospective randomized controlled clinical trial (Sci Rep, 2025). (2025). https://doi.org/10.1038/s41598-025-09398-w https://doi.org/10.1038/s41598-025-09398-w