Almost every conversation about exosome therapy is a conversation about whether it works. Very few are about how much. A review of the entire registered clinical trial landscape, covering 66 trials of mesenchymal stem cell derived extracellular vesicles, argues that dose is the field’s most consequential unanswered question, and that the answer is not simply “more.”
The variable the field has been quietly skipping
Dose is the first thing a pharmacologist asks about any new agent. It is also the thing exosome medicine has been least able to answer. Two clinics can both tell a patient they are receiving “exosomes” and mean quantities that differ by four orders of magnitude, delivered by different routes, measured in different units, from source cells nobody has characterised the same way.
A 2025 review in Frontiers in Medicine by Wang and colleagues mapped that problem across the whole registered trial landscape rather than a single indication. It remains the most complete dose-focused picture of the field available, and it has become more relevant as the number of clinics offering these products has grown.
What the review actually examined
The authors searched the Cochrane Register of Studies, ClinicalTrials.gov and the Chinese Clinical Trial Registry for trials using MSC-derived extracellular vesicles or exosomes as the primary intervention. After screening, 66 trials met the inclusion criteria, spanning registrations from April 2014 onward.
A few structural facts about that landscape are worth stating plainly, because they set the limits on what anyone can currently claim:
- Roughly 92 percent of the trials sat at preclinical, Phase I or Phase II stage. This is an early field by any reasonable definition.
- Lung disease dominated, with 26 studies, and COVID-19 accounted for 23 of the registrations. The 2020 surge shaped the dataset.
- By route, intravenous infusion led with 24 studies, followed by local administration with 15 and nebulized inhalation with 13.
- Twenty-two studies did not state what tissue the source cells came from at all.
- Of all 66 trials, only six had reported results in enough detail to support any dose-effect analysis.
That last number deserves a pause. Six. The entire evidence base for how much of this material to give a patient rests on a handful of small studies, most of them in respiratory disease during a pandemic.
The dose finding: route changes the number
Nebulized delivery worked at far lower particle counts
The clearest signal in the review is that the effective dose is route-dependent rather than absolute. Trials delivering vesicles directly to the lungs by nebulization reported effects at doses in the region of 108 particles. One trial in healthy volunteers used a single dose in the range of 2 to 16 x 108 particles. Another gave 2 x 108 particles daily for five days in severe COVID-19 pneumonia and reported improvement on CT imaging by day seven.
Intravenous protocols needed considerably more. One Phase II study of an intravenous product used concentrations on the order of 1010 vesicles per millilitre. The authors’ conclusion is that delivering material to the tissue you are actually treating lets you use dramatically less of it, which is exactly what you would expect from any other drug and exactly what a systemic infusion cannot do.
The comparison that gets quoted, and the one that does not
In the Phase II intravenous trial, 60-day survival was 51.6 percent on placebo, 53.4 percent at the lower volume dose and 69.6 percent at the higher volume dose, with ventilator-free days trending in the same direction. That is a dose-response signal in a randomised setting, which is rare in this field and genuinely notable.
It is also, read carefully, a signal that the higher dose performed better. The review’s headline framing is that the optimal window is narrow and that doses above the minimum effective dose may be no better and potentially harmful. That framing is a reasonable caution, but it is not what this particular dataset shows. In a separate dose-escalation trial that was terminated for a fatal serious adverse event, the death occurred in the lowest dose group.
Both things can be true. The honest summary is that dose clearly matters, that route changes the required dose by orders of magnitude, and that nobody yet has enough data to draw the curve. Anyone claiming to know the optimal exosome dose for a given condition today is working ahead of the evidence.
The reporting problem sitting underneath all of it
Characterisation was missing almost everywhere
The review’s most quietly damning detail is in its own results table. For the trials included in the dose-effect analysis, the column recording whether vesicle characterisation was reported reads “not reported” for essentially every entry. Trials were administering a biological product without publishing what was in it.
This is the same argument we made about the exosome certificate of analysis. A particle count on its own is not a dose, because it does not tell you what fraction of those particles are vesicles rather than protein aggregates or lipoproteins, and it does not tell you whether the preparation has any biological activity. Without identity, purity and potency alongside quantity, two products reporting the same particle number are not comparable, and neither are the trials that used them.
The word “exosome” is doing work the data does not support
The authors follow the International Society for Extracellular Vesicles nomenclature and use “extracellular vesicle” as the general term, reserving “exosome” for preparations with actual evidence of endosomal origin. Their assessment is blunt: because most clinical trials do not perform detailed vesicle subtyping, blanket use of the word “exosome” in early-phase studies can be misleading. That applies with considerably more force to marketing copy than it does to trial registrations.
What this means in practice
For practitioners evaluating regenerative products, the review points at a small set of questions that are answerable today and that separate a serious supplier from a promotional one:
- What is the dose, in what unit? Particles, protein micrograms, source cell equivalents and volume are not interchangeable. A supplier who cannot state the unit cannot state the dose.
- What supports the number? A dose figure means little without batch-level documentation of identity, purity, quantity and potency behind it.
- What is the route, and does the dose reflect it? The same number is not appropriate for local and systemic delivery.
- What is the source tissue and the characterisation method? Twenty-two of 66 registered trials could not answer this. A commercial product has less excuse than a trial does.
The regulatory position remains unchanged and worth restating. The authors note directly that no MSC-derived extracellular vesicle product has been approved by the FDA, the EMA or any comparable authority. They also flag the proliferation of unregulated exosome offerings outside formal clinical frameworks as a risk to both patients and the credibility of the science. Our own coverage of the standardization gap and of the current research base reaches the same conclusion from a different direction.
Where OmniGenix sits on this
We supply MSC-derived exosome products to licensed practitioners only, with batch-specific documentation. Dose is not a marketing number. It is a manufacturing claim that has to be traceable to a characterised batch, and our quality standards are written around that. We make no efficacy claims, and we treat the absence of approved indications as a fact to state rather than a detail to soften. That posture is why practitioners work with us.
Sourcing decisions should start with the paperwork
OmniGenix supplies characterised, batch-documented exosome and regenerative products to licensed practitioners. If you want to see what our documentation actually contains before you commit to anything, start here.
Frequently asked questions
Is there a standard exosome dose?
No. The 2025 review found dose reported in at least four incompatible units across registered trials, with only six trials reporting results in enough detail to analyse dose at all. There is no established standard dose for any indication, and no regulator has approved one.
Does a higher particle count mean a stronger treatment?
Not reliably. Particle count measures particles, not vesicles, and says nothing about biological activity. Two preparations with identical particle counts can differ substantially in what they actually contain. This is why identity, purity and potency data matter alongside the number.
Why would an inhaled dose be so much lower than an intravenous one?
Because delivery to the target tissue avoids systemic dilution and clearance. Trials nebulizing vesicles into the lungs reported effects at around 108 particles, while intravenous protocols used concentrations orders of magnitude higher. This mirrors dose behaviour in conventional pharmacology and is one of the review’s more robust observations.
How many exosome treatments are typically given?
Protocols in the reviewed trials varied from a single administration to twice-daily dosing over ten days, and none of these regimens are established standards of care. We cover the practical side of this question in more detail in our guide to how many exosome treatments are usually involved.
Source
Wang Y, Zhu J, Ma Q, Zhou W, Yang L, Sheng S, Zhu F, Xia Z. “Trends in mesenchymal stem cell-derived extracellular vesicles clinical trials 2014 to 2024: is efficacy optimal in a narrow dose range?” Frontiers in Medicine. 2025;12:1625787. Published 18 September 2025. doi:10.3389/fmed.2025.1625787
This article is for educational purposes and is not medical advice. No exosome or extracellular vesicle product is approved by the FDA for the treatment of any disease, and nothing here should be read as a claim of efficacy or a promise of any outcome. Dose figures cited are from registered clinical trials and are not treatment recommendations. OmniGenix products are supplied to licensed healthcare practitioners only. Patients should discuss any regenerative therapy with a qualified clinician.