Plain-language overview
A treatment's mechanism of action is the biological process through which it is thought to produce an effect. For MSC and extracellular-vesicle treatments, there may not be one mechanism that applies to every product, disease, dose, patient, or point in time.
Two products both described as 'MSCs' or 'exosomes' can differ substantially. The same product can also behave differently in different diseases or patients.
Mechanistic evidence is useful for understanding and developing a treatment, but a biologically plausible explanation is not proof that the treatment improves meaningful outcomes in people.
Product-specific mechanisms
Product-specific mechanisms matter because broad labels such as MSC therapy or exosome therapy do not identify one uniform product.
Cell or tissue source, donor, culture conditions, manufacturing, purification, storage, viability, EV composition, formulation, dose, and route can all affect biological activity. Product identity therefore needs to be defined before evidence from one product is applied to another.
A mechanism demonstrated for one carefully characterized product should not automatically be assigned to a different product simply because both use the same general name.
Figure 1 placeholder — Product-specific mechanisms
Disease-specific mechanisms
Disease-specific mechanisms matter because different diseases have different biology.
Acute injury, chronic fibrosis, infection, autoimmune inflammation, degenerative disease, and ischemia can involve very different cells, signals, tissue structure, blood supply, and repair processes. Disease stage can also change which biological pathway is important.
A mechanism that is useful in one disease may be irrelevant or even undesirable in another. Evidence should therefore match the exact condition being treated.
Figure 2 placeholder — Disease-specific mechanisms
Dose dependence
Dose dependence means that the biological effect can change as the amount of a treatment changes.
A dose-response relationship may increase steadily, reach a plateau, or behave as a nonlinear dose response. More is not necessarily better. Larger cell doses can alter vascular trapping or blood interactions, while larger EV doses can change biodistribution and exposure of other tissues.
A useful mechanistic study defines the dose clearly and tests whether changes in the proposed pathway occur across a scientifically justified range.
Figure 3 placeholder — Dose dependence
Timing
Timing can determine whether a biological mechanism is relevant.
The environment after injury or during disease changes over time. Inflammation, cell death, vascular permeability, fibrosis, and tissue repair may dominate at different stages. The treatment window that is useful early in disease may not be useful later.
Results from an experiment in which treatment is given immediately before or after an induced injury may have limited translational relevance to a patient with long-standing established disease.
Figure 4 placeholder — Timing
Patient heterogeneity
Patient heterogeneity means that patients differ biologically in ways that can affect treatment response.
Age, genetics, medications, comorbidities, immune status, disease severity, previous treatment, and organ function can influence recipient response, biodistribution, clearance, inflammation, and tissue repair.
A responder subgroup may exist even when an average study result is modest. However, stratified analysis needs adequate study design and independent validation so that chance findings are not mistaken for real biological differences.
Figure 5 placeholder — Patient heterogeneity
Why biological plausibility is not clinical proof
Biological plausibility means that a proposed explanation makes sense based on known biology and available experimental evidence. Clinical proof requires evidence that patients actually benefit.
A mechanism can look convincing in cells, animals, or biomarker studies and still fail to improve meaningful human outcomes. A surrogate endpoint or mechanistic biomarker may support the proposed pathway, but it should not be confused with a clinical endpoint.
Clinical efficacy requires appropriate human studies for the specific product and indication. Mechanism can strengthen the explanation for a result, but it cannot substitute for evidence of patient benefit.
Figure 6 placeholder — Why biological plausibility is not clinical proof
Mechanistic evidence has levels
Evidence can progress from a molecular interaction to a cellular response, tissue effect, whole-organism physiology, biomarker change, and finally a meaningful clinical outcome.
Evidence at one level should not automatically be presented as proof at the next. A laboratory signal can be scientifically important without proving that patients will feel, function, or survive better.
Models have limits
A preclinical model can help researchers isolate a mechanism under controlled conditions. Model validity depends on how closely the experiment represents the human product, disease, dose, route, timing, and outcome.
Animal and cell-culture studies can be essential for mechanistic research, but they do not eliminate the need for human clinical evidence.
How stronger causal evidence is built
Causal inference becomes stronger when changing or blocking the proposed pathway changes the biological effect and when an alternative mechanism has been considered.
Target engagement can show that a product reaches or affects its intended biological target. Independent methods that point to the same mechanism provide stronger evidence than one assay alone.
Negative results matter
A null finding means that an expected effect was not demonstrated. This can be scientifically useful.
Negative or null mechanistic results can reveal that the dose was inadequate, the treatment window was wrong, the target was irrelevant, the product differed from expectations, or only a particular subgroup responds.
Regulatory status is a separate question
Regulatory status describes how a product or use is classified or authorized in a particular jurisdiction.
It should not be used as a substitute for mechanistic evidence or clinical evidence. A regulatory decision and a scientific conclusion about mechanism or effectiveness answer different questions.
Questions patients should ask
What exact product was studied? Is the proposed mechanism demonstrated with that same product and disease? What dose and route were tested? At what disease stage was treatment given? Were patient differences examined? Is the evidence based on laboratory models, biomarkers, or meaningful clinical endpoints? Does the study show target engagement or causal evidence? Has clinical efficacy been demonstrated separately from the proposed mechanism?
Bottom line
The most defensible mechanism-of-action claim is specific rather than universal.
It should identify the exact product and disease, account for dose, timing, and patient differences, test whether the proposed pathway actually causes the observed effect, and clearly separate biological plausibility from proof of clinical benefit.
How this review was prepared
This Patient Edition was derived directly from the explicitly approved 0708 Professional Edition and preserves the same six controlling Points.