Stem cells can rebuild blood and immune-cell production, and certain stem-cell-derived products can repair damaged skin or the eye surface or treat a specific immune complication after transplantation. Gene-modified blood stem cells can also treat selected inherited blood disorders. These are product-specific treatments, not evidence that a general stem cell infusion repairs any damaged organ.
What is still uncertain? Japan’s newer iPS cell products for Parkinson’s disease and ischemic heart failure have conditional approval, with benefit still to be confirmed. Stem-cell-derived islets for type 1 diabetes and neural cells for spinal cord injury have encouraging early results, not established cures. No stem cell treatment is FDA-approved for anti-aging.
The practical distinction is between replacing a defined cell population and rebuilding an entire organ. A manufactured cell sheet, a blood stem cell transplant, and a clinic’s MSC infusion are not interchangeable. MSC stands for mesenchymal stromal cell; iPS cells are induced pluripotent stem cells, made by reprogramming mature cells.
Common question
Status
What the evidence supports
Can stem cells rebuild blood?
Established
Blood stem cell transplantation treats selected cancers and blood disorders; some approved therapies use gene-modified cells.
Can they repair skin or the eye surface?
Approved products for specific defects
Cultured epidermal and corneal epithelial sheets are approved in Japan for selected patients.
Can they calm a dangerous immune reaction?
Approved product for a narrow indication
One MSC product is approved in the U.S. for a specific pediatric transplant complication; a similar product is approved in Japan.
Can they improve Parkinson’s symptoms?
Conditionally approved in Japan, 2026
One iPS cell–derived dopaminergic progenitor product is approved with a time limit; confirmatory data are still required.
Can they replace heart muscle?
Conditionally approved in Japan, 2026
RiHeart uses iPS-derived heart-muscle sheets, but its small study does not establish a survival benefit.
Can they cure diabetes, spinal cord injury, or aging?
No established general cure
Stem-cell-derived islets and neural cells remain investigational. Japan’s historical conditional approval of Stemirac is a separate product-specific decision. Anti-aging claims are unsupported.
What stem cells do in the body
Stem cells have two defining abilities: self-renewal and differentiation into specialized cells. A division can produce two stem cells, one stem cell and one specialized cell, or two specialized cells. In tissues such as blood, skin, and the gut lining, this balance helps replace cells as they wear out Stc basics.
Adult stem cells generally produce the cell types associated with their tissue. Blood-forming stem cells make blood cells; that does not mean an infusion of them will become kidney cells or rebuild a heart Stc basics.
They are also in short supply, and once removed from the body they do not keep dividing forever. So growing enough of them for treatment is hard Stc basics. Pluripotent stem cells, which include embryonic stem cells and induced pluripotent stem cells (iPSCs), are easier to grow in bulk in the lab. But expansion is not the same as control. The cells still have to be turned into the right type, purified, and shown to survive after transplant Stc basics.
Some cell products may act through signals that affect immune activity rather than by replacing tissue. The term “MSC” alone does not identify the final product, its dose, or its clinical evidence. The approved MSC example below treats a transplant complication, not knee arthritis.
What stem cells can do today
Established blood transplantation, approved manufactured products, and conditionally approved products have different evidence bases. Country, diagnosis, cell source, manufacturing, and delivery route all matter.
Rebuilding blood after high-dose treatment
Blood stem cell transplantation can restore blood production after high-dose chemotherapy, sometimes with radiation. The treatment can destroy blood-forming cells as well as cancer cells. Donated or previously collected stem cells are infused into a vein, settle into the marrow (engraftment), and produce red cells, white cells, and platelets Stem cell transplant.
It applies to selected patients: leukemia, lymphoma, multiple myeloma, myelodysplastic syndromes, aplastic anemia, sickle cell disease, and some immune disorders. Most of the time the transplant is not attacking cancer itself. It rescues blood production after intense treatment. One exception is leukemia, where donor white cells can hunt down leftover cancer cells. Doctors call that graft-versus-leukemia Stem cell transplant.
The infusion takes 1 to 5 hours, but recovery is much longer. Immune recovery can take several months with your own cells and one to two years with donor cells. Treatment requires conditioning and specialist follow-up; some care may be inpatient and some outpatient. Risks include infection, bleeding, donor-cell graft-versus-host disease, infertility, and later cancers Stem cell transplant.
Treating inherited blood disorders
Gene-modified blood stem cells offer another capability. Casgevy uses a patient’s own blood stem cells, edited with CRISPR/Cas9, to increase fetal hemoglobin and reduce red-cell sickling. Its FDA approval for selected patients with sickle cell disease is not an approval for a generic stem cell infusion. Treatment includes high-dose conditioning chemotherapy, reinfusion, and long-term monitoring, with risks including low blood counts and infection Fda approves first gene therapies treat patients sickle cell disease.
Repairing skin and eye surfaces
Japan has approved two cell sheets that go straight onto the damaged spot: JACE (2007) for major burns when there is not enough donor skin, and Nepic (2020) for limbal stem cell deficiency, where the eye surface cannot renew itself 000223079.pdf000237511.pdf.
But a win on burned skin says nothing about whether a drip into your arm will fix a knee, a brain, or a heart.
Changing an immune reaction
Ryoncil is made from mesenchymal stromal cells donated from healthy adults’ marrow. On December 18, 2024, the FDA approved it for children 2 months and older with steroid-refractory acute graft-versus-host disease, a life-threatening complication that can follow a donor transplant FdaDrugXsl.cfm?setid=a4e0918f 7444 4694 adeb d38d98345659&type=display.
The approval rested on a single-arm study of 54 children. At day 28, 38 of 54 (70%) had a complete or partial response; 16 (30%) had a complete response. Median response duration was 54 days, not a measure of permanent cure. The initial course was eight intravenous infusions over four weeks, with additional treatment possible under the label FdaDrugXsl.cfm?setid=a4e0918f 7444 4694 adeb d38d98345659&type=display. Japan approved Temcell, another donor bone-marrow MSC product, for acute graft-versus-host disease after blood stem cell transplantation in 2015 000215658.pdf.
The label says the mechanism is unclear and may involve immune modulation. Laboratory findings of reduced T-cell activation do not establish clinical benefit. The pivotal study had no control group. Serious adverse reactions were reported in 35 of 54 patients (65%); the label also warns about infusion reactions, transmission of infection, and unwanted tissue formation. These were seriously ill transplant patients, so the reported safety figures should not be interpreted as a controlled estimate of risk caused by the cells alone FdaDrugXsl.cfm?setid=a4e0918f 7444 4694 adeb d38d98345659&type=display.
Replacing missing cell types
On March 6, 2026, Japan’s health ministry gave conditional, time-limited approval to two products grown from induced pluripotent stem cells. AMED, the country’s medical research agency, described it as the world’s first practical use of iPS-based treatments 20260220.
Amchepry uses lab-made dopamine precursor cells from donor iPS cells for the OFF-state in advanced Parkinson’s, those stretches when levodopa wears off and symptoms return. Surgeons place about 5.4 million cells into each side of the putamen, a movement hub deep in the brain, using pinpoint stereotactic surgery 000281617.pdf.
Seven people received treatment and six were evaluated for efficacy. Four of six had a reduction of at least 5 points in the OFF-state MDS-UPDRS Part III motor score at 24 months, the study’s threshold for a marked response 000281617.pdf. The mean reduction was 9.5 points, about 20%; lower scores indicate fewer motor symptoms. This was a phase I/II, open-label study without a control group, not proof of a cure S41586 025 08700 0.
All seven participants had adverse events: 73 in total, comprising 72 mild and one moderate event, with none serious. An adverse event is not necessarily caused by treatment. Dyskinesia, or involuntary movement, worsened by an average of about 12.3 points on its rating scale. Trial participants received tacrolimus to suppress immune rejection, tapered after 12 months and stopped at 15 months S41586 025 08700 0. The approval lasts seven years, and PMDA calls the efficacy evaluation “highly limited,” requiring post-marketing reassessment 000281617.pdf.
RiHeart uses sheets of heart-muscle cells grown from iPS cells for severe heart failure that standard drugs and procedures could not control, specifically when years of poor blood flow (ischemic cardiomyopathy) damaged the muscle. Surgeons lay three sheets on the heart’s surface through an open-chest operation, then the patient takes three immune-dampening drugs for 90 days 000281693.pdf.
In the eight-patient, open-label study, two had an increase of at least five percentage points in left ventricular ejection fraction, a measure of pumping function, at week 26. Ten serious adverse events occurred in five patients; the review discusses infections potentially associated with immunosuppression and cardiac ventricular thrombosis as safety concerns. The small, uncontrolled study does not establish a survival benefit 000281693.pdf.
Conditional approval has a specific meaning. Regulators accepted a reasonable prediction of benefit, not proof. The companies must gather real-world data and return for full approval within seven years 20260220000281617.pdf000281693.pdf.
Diabetes and spinal cord injury
Zimislecel consists of fully differentiated islet cells made from allogeneic stem cells and infused into the liver’s portal vein. A phase 1–2 report described 14 participants with type 1 diabetes who completed at least 12 months of follow-up. Twelve received the full dose. Ten of those 12 (83%) were insulin-independent at day 365. All 12 were free of severe hypoglycemic events during days 90 through 365 and had HbA1c, a measure of average blood sugar, below 7% 40544428.
All participants received immunosuppressants. Serious neutropenia, a low level of infection-fighting white cells, occurred in three participants. Two deaths occurred: one from cryptococcal meningitis and one from severe dementia associated with preexisting neurocognitive impairment. The analyses were interim and not prespecified, and the report does not establish a durable cure. Zimislecel remains investigational 40544428.
This is different from Lantidra, an FDA-approved islet product for a narrowly defined group of adults with type 1 diabetes and recurrent severe hypoglycemia despite intensive management. Lantidra uses islets from deceased donor pancreases, not islets manufactured from stem cells, and requires immunosuppression. Its approval does not validate stem cell products sold for diabetes Fda approves first cellular therapy treat patients type 1 diabetes.
At Keio University Hospital, four men with subacute complete cervical spinal cord injuries received iPS-derived neural stem/progenitor cells injected into the spinal cord in a phase 1 safety study. No tumors or graft-related adverse events were observed during two to four years of follow-up. Median motor-score improvement from baseline to week 52 was 13 points. Two participants moved from AIS grade A (complete injury) to C or D (motor-incomplete injury) S41591 026 04549 6.
The Keio study had no concurrent control group and was not designed to establish efficacy; natural recovery and rehabilitation complicate interpretation S41591 026 04549 6. Separately, Japan granted Stemirac, an autologous bone-marrow MSC product for traumatic spinal cord injury, conditional, time-limited approval in 2018. That historical decision relied on 13 patients, 12 of whom improved by at least one AIS grade, with confirmation required after marketing; it is not a statement about the product’s present availability 000231946.pdf.
The evidence above does not show that a general-purpose stem cell injection regrows a whole kidney, heart, or spinal cord. Replacing selected cells is not the same as recreating an organ’s blood supply, structure, and function. Small improvements on a symptom scale are not evidence of whole-organ regeneration.
On the FDA approved-products list checked 27 October 2026, no stem cell product is listed for anti-aging or “rejuvenation.” A claim that an infusion reverses biological aging needs direct human evidence, not a reference to cell renewal in a laboratory Approved cellular and gene therapy products.
Approvals do not transfer across diseases. A graft-versus-host disease product is not a knee-arthritis, ALS, or Alzheimer’s treatment. The FDA’s consumer warning addresses unapproved regenerative products promoted for conditions including orthopedic disorders, neurological disorders, chronic pain, fatigue, cardiovascular or lung disease, autism, and macular degeneration. It should not be read as a current inventory of every approved cell or gene therapy Important patient and consumer information about regenerative medicine therapies.
For cancer, a transplant is part of a treatment plan that may include chemotherapy, sometimes radiation, and follow-up, not a replacement for all other care Stem cell transplant. Do not delay established treatment to try an unproven cell product.
Cell counts are not a universal measure of effectiveness. Ryoncil is dosed at 2 million MSCs per kilogram per infusion, while Amchepry’s approved regimen targets about 5.4 million cells per side of the brain. Different cell types and routes make these numbers unsuitable for comparing which treatment is “stronger” FdaDrugXsl.cfm?setid=a4e0918f 7444 4694 adeb d38d98345659&type=display000281617.pdf.
A useful comparison starts with the exact product, not the total number of cells advertised.
What varies
Why it matters
Cell type
Embryonic stem cells and iPSCs are pluripotent, meaning they can generate many cell types. Adult stem cells generally have a more restricted range. A treatment may contain differentiated cells made from stem cells, not the stem cells themselves Stc basics.
Donor or patient
Cells from the patient are autologous, while cells from another person are allogeneic and may require matching and immune suppression. Marrow, fat, cord blood, skin, and cornea cannot replace each other Stem cell transplant.
Processing
A same-day fat graft, a lab-grown cell sheet, and a manufactured cell product involve different steps and controls. Sharing living cells does not make them equivalent.
What available in Japan actually means
Japan uses a separate framework for cell treatments offered at hospitals and clinics. Under the Act on the Safety of Regenerative Medicine, a clinic submits a provision plan to the health ministry, and a certified committee reviews it in advance. The law assigns plans to class I, II, or III by risk, with stated goals of safety and bioethics En.
That filing is not a product approval or a finding that the treatment is effective. It does not guarantee safety, and a plan number is not a success rate En. Marketed cell products undergo a separate regulatory review, as illustrated by the Temcell and Stemirac reports 000215658.pdf000231946.pdf.
Patients researching treatment in Japan may face four distinct options: an approved product, a conditionally approved product, a clinical trial, or a privately paid treatment under a provision plan. “Available in Japan” alone does not identify which option is offered.
How to test what a clinic promises
Consider a hypothetical knee-arthritis consultation. Before agreeing to treatment, ask for the proposed product’s name, cell source, injection route, and a human study in comparable knee-arthritis patients. A Ryoncil response rate in children with graft-versus-host disease would not answer that question. Bring the written proposal and study to your treating clinician rather than relying on a testimonial.
Question
Why it matters
What is the exact diagnosis and stage?
“Knee pain” or “neurological disease” is not a diagnosis. The evidence for one condition does not transfer to another.
What cells are used, from where, and how are they processed?
Source and manufacturing change the risk of infection, immune reaction, and unwanted tissue formation.
How are the cells delivered?
An intravenous infusion, a joint injection, an intrathecal injection, and a surgical transplant are different procedures with different evidence.
What human data exist?
Laboratory and animal studies can support a hypothesis. They cannot establish benefit in people.
Was there a control group, and what was measured?
Open-label improvements can reflect placebo, natural recovery, or observer bias. The endpoint should match the benefit you care about.
How long were patients followed?
A result at 28 days is not a result at five years.
What is approved, where, and for which indication?
Ask for the product name and the approved indication. “FDA approved” or “PMDA approved” without those details is not enough.
What are the risks, alternatives, and who is accountable?
Informed consent should cover failure, complications, and what happens if the treatment does not work.
Watch for one treatment pitched for many unrelated diseases, guaranteed success, “no side effects,” approval claims with no product name, or testimonials substituted for trial data. Do not conceal treatment from your usual clinician or change prescribed medication without discussing it with the responsible medical team.
A clinical trial registration number is not proof of efficacy. A registry record may describe study design, planned enrollment, and posted results; assess what was actually completed and reported. PubMed indexing likewise does not establish that a study is well controlled or that its conclusions are reliable.
So what can stem cells actually do
Stem cells can restore blood formation, support gene-modified treatments for inherited blood disorders, and supply specific tissue-repair or immune-modulating products. Those capabilities are real but diagnosis- and product-specific. Japan’s two 2026 iPS approvals remain conditional; the six efficacy-evaluable Parkinson’s patients and eight heart-failure patients do not establish benefits for unrelated conditions.
The useful question is not “what can stem cells do” in general. It is what this exact product has been shown to do in this exact disease, at this dose and route, and for how long.
Have questions about a proposed treatment in Japan?
Ask the clinic for the exact product or provision-plan details, the proposed indication, and the evidence to discuss with your treating clinician.
This article is general information, not medical advice. Treatment decisions should be made with a clinician who knows your diagnosis and medical history.