Where do gel and biologic injections realistically fit in my plan

Most people considering a “gel” or biologic injection are really asking a practical question: is this likely to change the direction of a painful knee or hip, or is it mainly a way of reducing symptoms for a period of time. Across the published evidence, injections can be very helpful for pain and function in the short to medium term, but none is a guaranteed alternative to joint replacement, and none has firm proof of restoring worn cartilage across an arthritic joint.

Two patterns of joint damage tend to get mixed up online. Diffuse osteoarthritis is the “whole joint” picture—thinning and roughening across a broad surface. A focal cartilage defect is closer to a pothole: a discrete full‑thickness area (often discussed in sizes such as “<3 cm²”) surrounded by otherwise healthier cartilage. That distinction matters because products sold as “gel injections” do not all target the same problem: symptom-modifying joint injections are typically aimed at diffuse osteoarthritis, whereas scaffold-style treatments such as ChondroFiller have been reported in small series as being placed into a prepared, localised defect (for example, a 17‑patient knee series with 12‑month follow‑up, mean age around 31 years, reported improved Lysholm and IKDC scores) ([11], [12]).

A clearer way to place the main options is to compare them by intent, typical time window, and certainty of evidence:

• Fast, short-term anti-inflammatory relief (corticosteroid/cortisone): meta-analytic trial evidence in knee osteoarthritis suggests clinically meaningful benefit versus placebo is mainly confined to the first ~6 weeks, with little proven advantage by 6 months ([1]). A separate 2‑year randomised trial in 140 patients found that triamcinolone 40 mg every 3 months led to greater cartilage volume loss than saline without better pain relief, which is why frequent repeat dosing tends to raise concern where cartilage preservation matters ([2]).
• Viscosupplementation “gel” for osteoarthritis symptoms (hyaluronic acid, HA): umbrella-review evidence in knee osteoarthritis suggests moderate average improvements in pain and function, with the clearest signal in early–moderate disease and a benefit window commonly measured in months rather than weeks; guideline recommendations remain inconsistent ([3]). In the hip, a level‑I systematic review of randomised trials reported that high‑molecular‑weight HA can improve WOMAC and VAS pain, with maximal benefits around 4–6 months ([4]).
• Longer-lasting hydrogel filler (polyacrylamide hydrogel; Arthrosamid-type PAAG): single‑injection studies report symptom improvement lasting to 52 weeks (for example, a 49‑patient open-label study with an ~18‑point WOMAC pain reduction on a 0–100 scale) and cohorts reporting maintained improvements to 24 months ([5], [6]). The key trade-off is that the evidence base is largely uncontrolled, and the material is intended to be long-lasting within the joint.
• Regenerative-leaning autologous biologic support (BMAC): a narrative review and a 175‑patient comparative clinical study report improvements in pain, function and quality-of-life scores at around 12 months in knee osteoarthritis, with substantial variation in preparation and protocols ([7], [8]). However, the multicentre MILES phase‑3 randomised trial in 480 patients found that cell-based injections (including BMAC) were not superior to a corticosteroid injection for pain or MRI osteoarthritis outcomes at 12 months, highlighting that results are mixed and not yet a settled “upgrade” over simpler options ([9]).
• Early-stage products marketed as regenerative (exosomes): a 2025 preclinical meta-analysis in rat models found consistent improvements in cartilage histology and inflammatory markers, but it also notes the lack of completed randomised human trials of isolated exosome products for knee osteoarthritis as of 2025 ([10]).

Instead of ending on a sales-style conclusion, the key takeaway is a simple fit-for-purpose grid: weeks (steroid), months (HA), potentially longer but with more uncertainty and permanence considerations (PAAG), and higher biological ambition but mixed human comparative evidence (BMAC), with exosomes still largely preclinical. The sections that follow focus on the specific knee-versus-hip differences and the most relevant head-to-head data behind these time windows.

Knee gel injections HA and Arthrosamid vs ChondroFiller and steroid

Thinking in time windows helps separate what these knee injections are actually trying to do: the first 6 weeks (rapid anti-inflammatory relief), the 3–6 month window (symptom control that can outlast a flare), and the year‑plus horizon (where permanence and evidence gaps start to matter).

Hyaluronic acid (HA): symptom “smoothing”, not cartilage repair

Hyaluronic acid knee injections are usually framed as viscosupplementation—aimed at “topping up” the joint’s natural lubricant to reduce friction and dampen inflammatory signalling, rather than rebuilding worn cartilage. In published trials, the benefit is typically not immediate; it is often assessed over weeks to months, with the clearest signal in early–moderate osteoarthritis rather than end‑stage disease.

An umbrella review of systematic reviews concludes that HA provides moderate average improvements in pain and function in knee osteoarthritis, but with heterogeneous trials and inconsistent guideline recommendations—a practical reflection of modest effect sizes and variable patient selection (for example, early–moderate radiographic grades tend to do better than advanced “bone‑on‑bone” patterns) ([3]).

Corticosteroid: the “sprinter” for pain flares, with limits for maintenance

Corticosteroid injections are largely an anti-inflammatory strategy, and the clinical pattern is often front‑loaded. In a meta-analysis of 11 randomised trials (842 patients), corticosteroid injections for knee osteoarthritis produced clinically meaningful pain and function improvements versus placebo mainly up to about 6 weeks, with diminishing effects by 6–12 weeks and no clear superiority by 6 months ([1]).

Where HA and steroid are compared directly, the trade-off shows up as speed versus persistence. A prospective comparative study in 60 patients with Kellgren–Lawrence grade II–III knee osteoarthritis reported that steroid improved pain faster at 1 month, whereas HA achieved better pain and function scores at 3 and 6 months, with a higher proportion reaching clinically meaningful pain relief at 6 months (about two‑thirds with HA versus 40% with steroid) ([13]).

The cautionary note is repeated dosing. In a 2‑year randomised trial of 140 patients, triamcinolone 40 mg every 3 months led to greater MRI-measured cartilage volume loss than saline and did not deliver better pain relief, which is why frequent, long-term steroid “maintenance” is generally treated cautiously in joints being actively preserved ([2]).

Arthrosamid (polyacrylamide hydrogel/PAAG): a long-lasting filler with higher commitment

Arthrosamid-type injections are a different category of “gel”: a non-biodegradable synthetic hydrogel intended to remain in the joint long term, aiming to cushion and improve joint mechanics rather than regenerate cartilage. The time course reported in studies is usually slower than steroid and often discussed beyond the 6‑month horizon that frames HA.

Evidence to date is encouraging for symptom duration but limited in design. A multicentre open-label study of 49 people receiving a single 6 mL ultrasound-guided intra-articular polyacrylamide hydrogel injection reported sustained improvement at 52 weeks, including about a 17.7‑point reduction in WOMAC pain on a 0–100 scale and 62.2% meeting OMERACT–OARSI responder criteria, with no new device-related serious adverse events reported between 26 and 52 weeks ([5]). A larger PROMs-based cohort of 314 knees reported improvements maintained to 24 months, with outcomes appearing better in older, non-diabetic patients and in lower radiographic grades; however, the same cohort also recorded a substantial number of complications (reported as 155/314 knees) and—crucially—had no control group, so placebo effects, regression to the mean, and co-interventions cannot be fully separated from any true device effect ([6]).

The permanence point changes the decision-frame. Because PAAG is designed to persist, any unexpected inflammatory response or mechanical issue is not in the same category as a short-acting steroid that simply wears off over weeks. That “high-commitment” nature is one reason it is often discussed more like a long-duration implantable material than like standard viscosupplementation, and why long-term joint effects remain an open question in the absence of robust randomised comparisons.

A practical corollary is cost: UK private price guides published in January 2025 have listed Arthrosamid at roughly £2100–£3000 per knee, versus around £120–£200 for a steroid injection, which tends to position PAAG as a relatively expensive option that is usually considered only after weighing the permanence and the uncertainty in controlled evidence ([14]).

ChondroFiller: a targeted scaffold for a defined cartilage defect (not an OA “whole-joint gel”)

ChondroFiller sits outside the usual HA–steroid–PAAG conversation because it is intended for focal, full-thickness cartilage defects rather than diffuse osteoarthritis. Published descriptions characterise it as a cell-free type I collagen hydrogel scaffold placed into a prepared, localised defect (often discussed at sizes such as <~3 cm²), where it polymerises and acts as an acellular matrix intended to recruit the patient’s own cells and support repair ([11], [12]).

Outcome data are early-stage and selected. A knee case series of 17 patients (mean age around 31, treated between 2012–2023) reported improved Lysholm and IKDC scores at 12 months, but the evidence base is not yet built on large randomised comparisons, and expert commentary has been explicit that ChondroFiller is not a simple office “gel injection” or a cure for generalised osteoarthritis ([11], [12]).

Putting the differences into a single “what tends to happen when” view

• Week 1 to week 6: corticosteroid is the option with the strongest evidence for short-term relief (including placebo-controlled data), but benefits commonly fade after the early weeks ([1]).
• Month 3 to month 6: HA tends to look better than steroid in head-to-head follow-up, particularly in KL II–III disease, but overall effects are typically modest and guideline support is mixed ([13], [3]).
• Year 1 and beyond: PAAG (Arthrosamid-type) has uncontrolled data suggesting many patients maintain improvements at 12–24 months, but the trade-off is a permanent material and weaker comparative evidence; ChondroFiller is aimed at a different problem—a defined focal defect—rather than whole-joint osteoarthritis symptoms ([5], [6], [11]).

In clinical decision-making, the “wrong candidate” is often defined by mismatch: HA or steroid being expected to reverse advanced structural change, PAAG being chosen without accepting the implications of a persistent implant, or ChondroFiller being considered for diffuse osteoarthritis when the real target is a localised defect on imaging.

Hip injections ultrasound guided HA vs steroid

Because the hip joint sits deep in the groin, close to major blood vessels and nerves, accurate needle placement is harder than in a knee. Many modern hip injection trials described in level‑I reviews use image guidance (ultrasound or fluoroscopy) to improve the chance that the injectate is delivered intra‑articularly ([4]).

Why ultrasound guidance changes the “real‑world” experience

Ultrasound guidance is less about making an injection “stronger” and more about making it more reliably intra‑articular. In a joint where millimetres matter, seeing the capsule and needle in real time can reduce the chance of a misplaced injection (which otherwise risks a disappointing response that is really a placement problem), and it allows the clinician to adjust the needle route to avoid sensitive structures during the appointment.

What a typical ultrasound‑guided hip injection visit looks like

In an outpatient setting, the appointment usually starts with a focused clinical check (for example, confirming the pain pattern—often deep groin pain—and any recent infection or blood‑thinner use), followed by an ultrasound scan to visualise the front of the hip joint and plan the safest path.

After skin cleaning, a small amount of local anaesthetic is typically used in the skin and deeper tissues. The needle is then advanced under real‑time ultrasound into the joint space, and either HA or corticosteroid is injected (some protocols include a small volume of local anaesthetic with the steroid). Most centres allow a short observation period afterwards—measured in minutes rather than hours—before normal walking out of clinic.

High–molecular‑weight HA for hip osteoarthritis: what the best trials show

The best direct evidence for HA in the hip comes from randomised trials summarised in a 2025 level‑I systematic review (982 patients across included RCTs). In that review, intra‑articular HA improved pain (VAS) and function (WOMAC) versus placebo or usual care, with maximal benefit commonly reported around the 4–6 month mark; high–molecular‑weight HA showed stronger pain reduction at 4–6 months than medium‑molecular‑weight HA in the pooled comparisons ([4]).

In practical terms, this profile tends to align with a gradual improvement rather than an overnight change—often described as easing of deep ache, better tolerance for walking, or less night discomfort over weeks, rather than hours. Even in RCT‑level evidence, responses vary, and symptom relief should still be seen as symptom control rather than cartilage restoration ([4]).

Hip steroid injections: useful short‑term relief, with a tighter ceiling

For corticosteroid (often combined with lidocaine), systematic review evidence over a 10‑year window (2015–2025) supports a short‑term role, particularly where there is active synovitis or an effusion, and where quick reduction of inflammatory pain is the main objective ([15]).

The caution in the hip is that, beyond the early pain‑relief window, repeated injections are viewed more conservatively than many patients expect. A 2025 systematic review covering 34,367 hips reported variable but concerning incidences across studies of rapidly progressive osteoarthritis (reported ranges 0.2–21.1%), osteonecrosis (0.6–27.1%), and femoral head collapse (3.2–20.4%) after intra‑articular hip corticosteroid injections—data that are not proof of causation in every individual case, but are difficult to ignore when planning repeat injections in a weight‑bearing joint ([16]).

HA vs steroid in the hip: a pragmatic comparison

• Speed of effect: steroid tends to act faster (often days) in inflammatory flares; HA tends to build more gradually over weeks ([15], [4]).
• Duration: HA trial data most consistently point to benefit out to roughly 4–6 months; steroid benefit is usually framed as shorter‑term symptomatic relief (weeks), and may diminish with repetition ([4], [15]).
• Joint‑preservation comfort level: HA does not come with the same published signal for severe structural complications seen in some hip steroid series; in contrast, hip steroid repeat‑use decisions often explicitly weigh the observational complication literature ([16]).
• Aim of treatment: both are best described as pain‑ and function‑modifying tools rather than regenerative treatments; the main difference is the balance of quick relief (steroid) versus steadier medium‑term relief (HA) ([4], [15]).

Where BMAC fits in hip conversations (and why it comes up here)

When steroid relief is short‑lived or repeat steroid is unattractive because of the structural risk signal reported in hip series, discussions often drift towards “regenerative” options such as bone marrow aspirate concentrate (BMAC). A 2024 systematic review found that BMAC for hip osteoarthritis is technically feasible (intra‑articular or intraosseous, usually image‑guided) and appears to have few serious adverse events in published reports—but the evidence is still limited to small, uncontrolled studies with heterogeneous protocols and no robust randomised trials ([17]). A small fluoroscopy‑guided intraosseous case series (5 patients, Tönnis grade 2) reported VAS pain dropping from 7.6 to 2.4, alongside >50% subjective functional improvement, but that scale of evidence remains preliminary ([18]).

At London Cartilage Clinic (Harley Street), hip injection choices are usually framed around imaging findings, the need for fast flare control versus medium‑term symptom smoothing, and how strongly joint preservation is prioritised over repeated short‑acting treatments; appointments can be booked via londoncartilage.com.

Using cortisone without closing the door on regenerative options

A common worry after having one or two cortisone (corticosteroid) injections is whether the joint has been “spoiled” for later, more regenerative-leaning options such as PRP, BMAC, or a scaffold treatment like ChondroFiller. This section deliberately avoids anchoring that question to any particular clinic or location and focuses instead on what “closing the door” usually means in practical knee terms.

Cortisone is best understood as a fast anti-inflammatory pain reliever rather than a cartilage-repair treatment. In placebo-controlled knee osteoarthritis trials, the meaningful benefit is concentrated in the early window (roughly the first 2–6 weeks), with a clear taper thereafter and no consistent signal by 6 months ([1]). That time course matters because it encourages a “reset the flare” mindset, rather than a long-term maintenance plan.

“Closing the door” is usually less about a past injection and more about the joint’s current structure on imaging. Options that aim to support endogenous repair tend to be most relevant when there is still enough preserved architecture to work with. For example, published descriptions stress that ChondroFiller is designed for small, focal full‑thickness cartilage defects (often discussed as around <3 cm²) and is not positioned as a simple fix for generalised osteoarthritis ([12], [11]). In the same vein, clinical summaries of BMAC for knee osteoarthritis describe symptomatic improvements in some series, but repeatedly highlight wide variation in protocols and the importance of patient selection—often meaning earlier-stage disease rather than end-stage, diffuse loss ([7]).

The reason repeated steroid use raises eyebrows in joint-preservation planning is that there is a structural signal in one well-known long-duration trial. In a 2‑year randomised study of 140 people given triamcinolone 40 mg every 3 months, the steroid group had greater MRI-measured cartilage volume loss than saline and did not have better pain outcomes over time ([2]). That pattern—high-frequency injections over years—maps closely to the sort of “maintenance” cycle that can be hard to step away from.

Reassurance is still warranted: there is no direct evidence that a small number of well-spaced steroid injections permanently prevents later PRP, BMAC, or ChondroFiller. The concern is more cumulative and strategic: repeated exposure may (in some circumstances) push the joint further towards widespread cartilage loss, where focal scaffold approaches are less relevant by definition, and where biologic injections are less predictable ([7]).

It also helps to keep expectations about “regenerative” injections realistic. In the MILES phase 3 trial (480 patients with KL II–IV knee OA), several cell-based injections (including BMAC) were not superior to a single corticosteroid injection for pain or MRI osteoarthritis scores at 12 months, and none showed structural improvement compared with baseline ([9]). That finding does not make steroid a long-term solution; it does underline that escalating to increasingly complex injections is not automatically a shortcut to cartilage restoration.

Pragmatically, the most “door‑preserving” way steroid tends to be used is as a short-term tool with a defined purpose and an exit plan, rather than as repeat treatment every few months:

• Acute inflammatory flare control: a short-term bridge when pain is preventing progress (the same “weeks not months” window seen in placebo-controlled evidence) ([1]).
• Diagnostic clarification: when a clear, time-limited reduction in pain helps confirm the joint as the main pain generator (often considered alongside imaging).
• Time-critical function: an “important event” scenario, with a deliberate decision not to roll straight into a quarterly routine.
• Rehab and mechanics as the backbone: pairing the injection with a plan for strengthening, load management and mechanical optimisation, so that symptom relief is used to change the trajectory rather than simply to repeat the shot ([7]).

When symptoms return quickly—within the same 6–12 week period where steroid effects often fade in trials—it commonly triggers a different question: not “another injection?”, but “what does the imaging show now, and is the pattern focal or diffuse?” That is the point at which scaffold-based approaches for a discrete defect (as described for ChondroFiller) or biologic options such as BMAC are usually assessed against the joint’s current structural picture rather than against the fact that a prior cortisone injection happened ([11], [7]).

BMAC injections for knee and hip how experimental are they

BMAC is often described online as a “stem cell injection”, but in practical terms it is a same‑day, two‑part outpatient procedure: a needle-based marrow harvest (most commonly from the iliac crest of the pelvis) followed by an image‑guided injection back into the target joint or the bone just beneath it. The aim is not to “replace” missing cartilage directly, but to deliver a concentrated mix of marrow‑derived cells and signalling molecules that may support the joint’s own repair responses ([7]).

What actually happens on the day (and why it’s more involved than a gel or steroid shot)

The typical pathway described in clinical reviews starts with local anaesthetic at the pelvic harvest site, then bone marrow aspiration using a needle, followed by on‑site processing (centrifugation) to create the “concentrate” (BMAC). The prepared BMAC is then delivered with ultrasound or X‑ray guidance into the knee joint, hip joint, or—in some protocols—intraosseously into subchondral bone ([7], [17]).

A useful way to remember the experience is that there are two treatment sites on one date: the pelvis (harvest) and the painful joint (injection). Even when everything is done with needles and local anaesthetic rather than incisions, that “two sites” reality is one reason BMAC is usually positioned as a step up in intensity compared with a single intra‑articular hyaluronic acid or corticosteroid injection.

Knee osteoarthritis: encouraging signals, then a reality-check in larger trials

Across published summaries, BMAC for knee osteoarthritis is mainly supported by short‑ to mid‑term symptom improvement in selected patients, but with major variability in how marrow is collected, processed, and injected ([7]). One single‑centre comparative study in 175 people with Kellgren–Lawrence II–IV knee OA reported greater improvements over 12 months in WOMAC and SF‑36 physical components after BMAC than after PRP or hyaluronic acid in that specific protocol ([8]).

Set against that, the multicentre phase 3 MILES randomised trial ( 480 patients, KL II–IV knee OA) found that several cell‑based injections—including BMAC—were not superior to a single corticosteroid injection for pain outcomes or MRI osteoarthritis scores at 12 months, and none of the injection groups showed structural improvement compared with baseline ([9]). Taken together, this is why BMAC in knee OA still sits in a “promising but not settled” category: it may help some patients, but the strongest comparative trial data do not yet show a clear advantage over simpler injections in a broad OA population.

Focal cartilage defects: where BMAC is used as part of a combined repair strategy

BMAC is also used in focal full‑thickness cartilage lesions, particularly in younger adults, but commonly as one component of a broader, one‑step repair construct rather than a stand‑alone joint injection. In a series of 165 patients (mean age 26.4 years) treated with BMAC loaded onto a hyaluronic‑acid scaffold plus an intra‑articular PRP injection, KOOS scores improved significantly at 1 year, and MRI repair quality was generally favourable (mean MOCART‑2.0 71) ([19]). The limitation is attribution: when BMAC, a scaffold, and PRP are delivered together, it is difficult to separate what proportion of the clinical or MRI signal belongs to the marrow concentrate itself ([19]).

This is also the point where comparisons with ChondroFiller need to stay precise. Published descriptions frame ChondroFiller as a cell‑free type I collagen scaffold intended for small, localised full‑thickness defects (often discussed around <3 cm²) rather than generalised osteoarthritis ([12]). A knee case series of 17 patients (treated across 2012–2023, mean age about 31) reported improvements in Lysholm and IKDC scores at 12 months after ChondroFiller implantation for chondral lesions, but that evidence base remains early and small ([11]). In simple terms, BMAC is a biologic “signal” approach that may influence the joint environment, whereas ChondroFiller is a defined scaffold aimed at a defined defect—so they often apply to different structural problems even if both are discussed under the “regenerative” umbrella.

Hip BMAC: technically feasible, but still low-evidence and small-number

For hip osteoarthritis, the clinical literature remains dominated by small, heterogeneous, uncontrolled studies. A 2024 systematic review concluded that both intra‑articular and intraosseous BMAC approaches appear technically feasible with few serious adverse events reported, but graded the overall evidence as low level and insufficient to define optimal protocols or long‑term durability ([17]). A fluoroscopy‑guided intraosseous case series in 5 patients with Tönnis grade 2 hip OA reported meaningful pain and function improvements, but the sample size and lack of a comparator limit how confidently that translates into expected results for broader groups ([18]).

So, how experimental is BMAC in 2025?

A clean way to answer the headline question is to place BMAC on an “evidence and complexity” spectrum.

• Standard, well‑defined symptom injections: hyaluronic acid and corticosteroid have larger bodies of comparative evidence in OA and are simpler single‑site injections (covered earlier for knee and hip).
• Emerging but not proven in broad OA: BMAC has credible signals in some cohorts (for example the 175‑patient knee study) but has not shown clear superiority to steroid in a large multicentre RCT at 12 months ([8], [9]).
• Most experimental in routine practice: hip BMAC sits closer to “experimental” because the published human data remain small and uncontrolled, even when early results look encouraging ([17], [18]).

In joint‑preservation planning, the most important dividing line is usually structural: diffuse osteoarthritis pain versus a discrete focal cartilage defect (where scaffold approaches such as ChondroFiller are conceptually better matched to the problem described in case series), and that distinction often matters more than the label “stem cells” on an advert ([12], [11]).

Exosome injections vs PRP and BMAC why most people should wait

Exosome injections sit right at the “next wave” end of regenerative medicine in 2025: biologically plausible, heavily marketed, but not yet backed by completed randomised human trials for knee osteoarthritis or focal cartilage damage.

What exosomes are (and what they are not)

Exosomes are tiny membrane-bound packets released by cells. They carry signalling cargo—proteins and genetic material—that can influence inflammation and tissue behaviour. In orthopaedics, most of the excitement centres on exosomes derived from mesenchymal stromal/stem cells (MSCs), because MSCs are known to produce many of the “calming” signals researchers want in an arthritic joint ([10]).

What the animal studies suggest

The strongest evidence base so far is preclinical. A 2025 systematic review and meta-analysis pooling 28 rat models of knee osteoarthritis found MSC-derived exosomes were consistently associated with “healthier” cartilage on histology and a shift towards a less inflammatory joint environment ([10]). The practical meaning—if it translated to people—would be a treatment that not only reduces pain signalling, but also nudges cartilage towards maintenance rather than breakdown.

Mechanistic work in mice adds detail about how that could happen. In a 2023 study using synovial MSC-derived exosomes carrying a specific cartilage-related protein (MATN3), treated animals showed less cartilage destruction, with signalling changes involving the PI3K/AKT/mTOR pathway ([20]). The takeaway is not the pathway name; it is that exosomes can act like targeted “instruction sets” that change inflammatory and repair behaviour in a living joint.

Why PRP is the most useful bridge to human reality

A key reason exosome therapy generates confusion is that many patients may already be getting some exosome-related effects indirectly through better-studied biologics—particularly PRP. In a prospective study of 60 people with knee osteoarthritis treated with intra-articular PRP, 73% met a responder definition (≥20% pain reduction) at 12 weeks, and the exosome characteristics within the PRP differed between responders and non-responders ([21]). That does not mean “PRP equals exosomes”, but it supports a more grounded idea: exosomes may be one reason PRP sometimes works, without needing a separate, purified exosome product.

The missing piece: completed randomised human trials of purified exosomes

Despite the strong animal signal, the same 2025 preclinical review highlights a critical gap: no completed randomised human trials of isolated/purified exosome injections for knee osteoarthritis were identified as of 2025 ([10]). That absence matters because it leaves several practical questions unanswered in real clinics—dose, durability beyond a few months, meaningful safety rates, and whether a given commercial product resembles what was tested in the lab.

A simple “wait” rule that holds up in 2025

A practical rule of thumb is: exosomes become a sensible mainstream option only once there are human randomised trials using a clearly defined exosome product (source, manufacturing, dose) with 12‑month outcomes and transparent adverse-event reporting, ideally replicated by more than one research group. Until then, confident claims of “cartilage regrowth” or guaranteed long-term benefit are ahead of the evidence ([10]).

Regulation and why “clinic products” are a red flag

Regulators have repeatedly flagged concerns about regenerative products being sold before adequate approvals and trials. The section brief references an FDA consumer alert, but that document is not included in the source set provided here, so its wording cannot be quoted directly. Even without that specific document, the decision logic remains the same in both the US and UK: if the product is not supported by completed randomised human trials and is being sold as a proven osteoarthritis treatment, governance and consent need to be treated as especially high-stakes.

In practice, this is why exosomes currently sit in a different category from PRP and BMAC. PRP has direct human outcome studies (including the 12‑week PRP cohort above) and a more established “what is being injected” story; exosome injections are still largely a laboratory concept being pushed into clinics faster than the human evidence base ([21], [10]). Within a specialist joint-preservation consultation in London, discussion therefore tends to focus on options with published human data, while exosomes are usually considered research-adjacent rather than routine care.

1. [1] Cell-based versus corticosteroid injections for knee pain in osteoarthritis: a randomized phase 3 trial. (2023). https://doi.org/10.1038/s41591-023-02632-w https://doi.org/10.1038/s41591-023-02632-w