Cell therapies

Embryonic Stem Cells

Pluripotent, tightly regulated, and used in a very small number of tightly controlled trials — not in general practice.

Legally restricted

Medical review statusPending clinical sign-offLast evidence update: 2026-08-05Methodology and editorial standards
Clinical review team: Dr Kamelia Milcheva, Hematologist · Dr Vadym Uvarov, Board-Certified Physician · Hepatobiliary Surgeon · Candidate of Medical SciencesEditorial responsibility: StemCellAtlas research teamEducational information only. This page does not provide medical advice, diagnosis or a treatment recommendation.

What it actually is

Human embryonic stem cells are pluripotent: they can in principle become any cell in the body, and they can be expanded almost indefinitely. That is exactly what makes them scientifically powerful and clinically demanding. Pluripotency is not a therapeutic property in itself — the therapy is always a differentiated derivative, and the undifferentiated cell is the hazard to be removed.

Where it comes from

Derived from the inner cell mass of surplus blastocysts from in-vitro fertilisation, donated with consent. Derivation destroys the blastocyst, which is why national law differs sharply across Europe and why EU research funding carries specific restrictions on this step.

How it is made — and what can go wrong

Lines are established, banked and characterised for karyotype and identity, then directed down a specific lineage. The release-critical question is residual undifferentiated cells: any that survive into the final product carry teratoma risk. Purification, differentiation efficiency and sensitive assays for residual pluripotent cells are the core of the manufacturing problem.

Regulatory status

Legal status varies by member state. Clinical programmes exist — retinal pigment epithelium for macular degeneration is the most advanced example — and they run as formal, regulator-authorised trials at specialist centres, not as private services. A private clinic offering embryonic stem cells as a treatment is operating outside the framework these programmes work within.

What the evidence shows

Early-phase trials in a small number of indications, chiefly ophthalmology and some neurological programmes, have reported safety findings and preliminary signals. There is no approved embryonic-stem-cell medicine for general use.

Limits worth knowing

Teratoma risk from residual undifferentiated cells. Allogeneic by definition, so immunosuppression or immune-privileged delivery sites matter. Ethical and legal constraints are substantive, not procedural. The gap between 'can become any cell' and 'is a treatment for my condition' is the entire remaining problem.

What this could actually be used for

Pluripotent: they can become any cell type in the body, and they divide indefinitely without ageing out of culture. No adult cell does both. Everything that makes them useful and everything that makes them dangerous comes from those same two properties.

They are the only realistic source when the goal is to manufacture a specific missing cell type in quantity, to a consistent specification, independent of any donor. Which is also why controlling differentiation — making certain every cell in a batch became what it was supposed to become — is the whole engineering problem.

Dopaminergic neurons 8 studies registered

Parkinson disease destroys one identifiable cell population in one anatomical location. That makes it the cleanest target in the field: manufacture the missing neuron type, place it where it belongs, and it has to form correct connections locally rather than rewire a whole system. The unsolved parts are graft survival and how much of the disease has already spread elsewhere by the time symptoms appear. All cell therapy research registered for this condition →

Retinal pigment epithelium

The eye is small, immune-privileged, and the graft can be watched through a lens for years afterwards. Cells are delivered as a thin sheet beneath the retina to support photoreceptors that are starving rather than dead. Low cell numbers, direct visual confirmation and a contained site make this the most advanced application of the whole family.

Insulin-producing beta cells 3 studies registered

Type 1 diabetes is the destruction of a single cell type by the immune system. Manufacturing replacements is now the solved half; keeping them alive against the same immune attack that destroyed the originals — by encapsulation, by gene editing, or by immunosuppression — is the half that decides whether any of it lasts. All cell therapy research registered for this condition →

Cardiac muscle 1 study registered

Heart muscle does not regenerate after an infarction; it scars. Manufactured cardiomyocytes are studied for repopulating that scar, and the specific obstacle is electrical: a patch of muscle beating out of time with the rest of the heart causes arrhythmia rather than recovery. All cell therapy research registered for this condition →

Myelin and the spinal cord 1 study registered

Rather than replacing neurons, the aim is to supply the support cells that insulate surviving axons, so signals blocked by lost myelin can pass again. All cell therapy research registered for this condition →

Where it stops. Two constraints shape everything here. Undifferentiated cells left in a batch can form teratomas, so purity is a safety specification rather than a quality nicety. And the cells come from a donor embryo, so outside immune-privileged sites rejection has to be managed for as long as the graft is meant to last.

What the registry actually shows for this cell type

Our own count of every study registered on ClinicalTrials.gov under this cell type, retrieved 2026-09-04. A registration is a declaration of intent, not a result — which is why the status column matters more than the total does.

Status in the registryStudiesShare
Completed1529%
Status not updated by sponsor1325%
Active, closed to entry815%
Recruiting now713%
Not yet recruiting48%
Withdrawn before enrolling24%
Enrolling by invitation only24%
Terminated early12%
All registered studies52100%

Reading it: 3 studies stopped before finishing — terminated, withdrawn or suspended, 6% of the total; 11 studies carry no assigned trial phase, meaning they sit outside the phased development path that leads to an approved medicine; 4 studies reached Phase 2 or later, against 37 studies still at Phase 1 or earlier; 13 records have not had a status update from the sponsor and may be dormant.

The directions above describe where the biology points and where research has been registered. They are not claims that any of it works, not evidence of benefit, and not a treatment recommendation. Counts are our own extraction from the ClinicalTrials.gov API, retrieved 2026-09-04; the queries behind them are published with the dataset so that anyone can repeat the count.

What to ask before agreeing to anything

  1. Is this a registered clinical trial with a named sponsor and an identifier, or a private service?
  2. Which differentiated derivative is being administered — not which parent line?
  3. What assay demonstrates absence of residual undifferentiated cells, and to what detection limit?
  4. Under which national legal basis is the material derived and used?
A cell label is not a treatment recommendation. Composition, manufacturing and authorisation differ between providers using the same word. Take the exact written proposal to a physician who is independent of the clinic selling it.

Sources behind the numbers on this page

Study counts above are our own extraction from ClinicalTrials.gov; the links let you reproduce them. Not medical advice, and not an assessment of whether any treatment is appropriate for you.

Compare cell-therapy evidence, registered studies and published price observations.

StemCellAtlas is a source-first research and cost-planning guide. It separates registry and regulator evidence from heterogeneous commercial observations.

Request written information