On this page
- Follow-up is the largest gap
- The comparison that is almost never made
- Heterogeneity makes pooling fragile
- Blinding is structurally difficult
- Publication and funding pressures
- What is not known about mechanism
- The questions unlikely to be answered
- What this does not mean
- Frequently asked questions
- Requesting a medical evaluation
Most pages in this field describe what might be possible. This one describes where the research stops — the questions nobody has answered, the study designs that have not been run, and the structural reasons several of those gaps are unlikely to close soon.
Follow-up is the largest gap
Almost every trial cited anywhere on this site followed patients for months. Six, twelve, occasionally twenty-four.
The conditions being treated run for decades. Osteoarthritis progresses over twenty or thirty years; multiple sclerosis over a lifetime; diabetes indefinitely.
So the honest statement about durability is not that benefit fades — it is that nobody looked. Any claim about lasting effect beyond the follow-up window is extrapolation, and should be described as such.
This gap is also expensive to close, which is why it persists. A decade-long follow-up study costs more than the treatment generates and benefits no sponsor, so the funding does not appear and the question stays open indefinitely.
The comparison that is almost never made
Most trials compare cells against placebo, or against nothing, or occasionally against an active injection.
Very few compare cell therapy against the best available standard care delivered properly — a completed exercise programme, optimised medication, a multidisciplinary approach.
That is the comparison a patient actually faces, and its absence means the practical question “is this better than doing the cheap thing well?” has largely not been tested.
Where it has been tested in adjacent fields, the cheap thing delivered properly has often performed better than expected — see osteoarthritis.
The difficulty is partly practical. Standard care delivered properly is hard to standardise across a trial, because it depends on the patient turning up and doing the work. That is a real methodological obstacle and it is also, conveniently, a reason the comparison keeps not being made.
Heterogeneity makes pooling fragile
Trials differ in cell source, donor age, passage number, dose, whether cells were fresh or frozen, route, number of treatments and patient selection.
Meta-analyses pool them anyway, because that is the only way to get numbers large enough to say anything. Every one of them declares this limitation, and summaries downstream reliably drop it.
A pooled effect across genuinely different interventions is a real number describing an average of things that are not the same — see MSC dosage research.
Blinding is structurally difficult
A patient who has had a liposuction knows they were in the treatment arm. So does one who spent six weeks waiting for cells to be cultured.
Sham-controlled trials exist but are harder to run, more expensive, and raise ethical questions about sham procedures.
Since the main outcome measures are subjective pain and function scores, and placebo response in injection studies is large, this is not a minor methodological quibble — it is close to the centre of the problem.
Objective endpoints would help, and they are harder. Cartilage thickness on MRI, biomarkers, functional testing — each is more expensive, less sensitive to short-term change, and less likely to show the result a sponsor hoped for.
Publication and funding pressures
Positive results are more likely to be published than negative ones, in every field. Here the effect is amplified by commercial interest in the outcome.
Notable exceptions exist, and they are worth respecting — a large commercially sponsored disc trial that did not support its product, and a rotator cuff study finding no benefit from added bone marrow stimulation.
A field that publishes its negatives is more trustworthy than one that does not, and this field does so more than its marketing suggests.
The asymmetry sits further downstream. The trials get published; the clinic pages summarising them cite the favourable ones. The selection happens in the marketing rather than in the literature, which is a more tractable problem for a reader who checks the sources — see what the evidence levels mean.
What is not known about mechanism
Whether the cells need to be alive on arrival. Whether the effect is from the cells, their secreted factors, or the injection itself.
How long any effect persists at a cellular level, and whether repeat dosing produces diminishing returns or immune sensitisation.
Why some patients respond and others do not, and whether anyone can be identified in advance as a likely responder. That last question would change the economics of the whole field if answered — see MSC immunomodulation.
The questions unlikely to be answered
Head-to-head comparisons between cell sources would be expensive, and no commercial party benefits from the result.
Long-term durability requires decade-scale follow-up that nobody funds.
Comparison against properly delivered standard care is unattractive to anyone selling either.
Knowing which questions are structurally unlikely to be answered is itself useful, because it tells you which gaps will still be there when you look again in five years.
It also suggests where genuine progress will come from: publicly funded trials, disease registries and academic consortia rather than from the commercial sector, whose incentives point toward selling the current product rather than testing it against the alternatives.
What this does not mean
Not that none of this works. Knee osteoarthritis has randomised trials and meta-analyses, and dismissing the whole field would be as unwarranted as embracing it.
Not that early evidence is worthless. Most established medicine passed through exactly this stage, and a treatment graded early here may be graded higher in ten years or abandoned. Both happen.
It means the field is younger than its marketing, and that a decision made now is made under genuine uncertainty rather than despite a settled answer — see what the evidence levels mean and the research library.
Biological age reversal and psoriasis are the clearest worked examples on this site of evidence being borrowed from somewhere it does not belong.
Frequently asked questions
What is the biggest weakness in the evidence?
Follow-up length. Almost everything runs for months against conditions lasting decades, so durability is unknown rather than disproved.
Has this been compared with standard care?
Rarely, and that is the comparison patients actually face. Cells against placebo is a different question from cells against doing the cheap thing properly.
Why are meta-analyses limited here?
They pool trials differing in source, dose, route and patients. They say so; summaries downstream tend not to.
Do negative results get published?
More than the marketing suggests. A large sponsored disc trial and a rotator cuff study both reported findings that did not support the intervention.
Can anyone predict who will respond?
No. That question would change the economics of the field if answered, and it has not been.
Does this mean none of it works?
No. Knee osteoarthritis has genuine randomised evidence. It means the field is younger than it is presented as being.
Requesting a medical evaluation
Nothing on this page establishes whether any treatment is appropriate for you. That needs your history, your imaging and your current medications read by a clinician. If you would like that review, send your records and we will tell you honestly whether there is anything worth discussing — including when the answer is no.
This page is general information, not medical advice, and does not create a doctor–patient relationship. Regenerative treatments discussed here are in most cases investigational. Discuss any treatment with a clinician who knows your history.