Japan’s Regenerative Medicine System, Broken Down
How to Choose a Stem Cell Hospital in Japan
Before you compare prices or book a flight, match your diagnosis, disease stage, and the cell route that’s actually on the table against Japan’s regulatory record — and the human data behind it.
Four Things to Check First
Start with four facts you can verify.
A high cell count, a polished clinic, or a glowing patient story — none of those matter more than how well your case lines up with the published evidence.
Source basis: PMDA regulatory framework, updated 20 May 2026.
Japan uses two different legal routes. The Safety Act covers medical care or research using processed-cell technologies; the PMD Act covers commercial product marketing authorization. A Safety Act record is not PMDA product approval.
Compare diagnosis and cause, disease stage, time since onset, prior standard treatment and baseline function. A study in acute disease cannot be used to predict a chronic case.
Confirm the cell source, autologous or allogeneic use, processing, dose and route. Intravenous infusion, joint administration, brain implantation and a heart-surface sheet are not interchangeable treatments.
Ask what will be measured, when it will be measured, what rehabilitation is expected and how complications or non-response will be handled after you return home.
Does the Evidence Actually Match Your Case?
Check whether your condition matches a real Japanese data set.
The question most people skip — and the one that actually matters — isn’t whether Japan uses stem cells. It’s whether the exact disease, stage, route, and product you’re considering line up with evidence that exists.
Traumatic spinal cord injury
STEMIRAC received conditional and time-limited approval in 2018 for a defined traumatic spinal cord injury setting. The pivotal efficacy analysis had only 13 patients. Because the approval was time-limited, verify the current product status rather than relying on the 2018 headline.
Product-specific evidenceChronic traumatic brain injury
Akuugo was studied with intracranial implantation and sham surgery. The PMDA review contains randomized motor-score data, but PMDA also noted that the link between the measured motor change and daily-life improvement remained unclear.
PMDA-reviewed dataParkinson’s disease
AMCHEPRY received conditional and time-limited approval in March 2026 for selected patients with motor symptoms inadequately controlled by existing drug therapy. It is an iPS-derived cell product placed into the brain by stereotactic surgery, not an IV infusion.
Conditional approvalSevere ischemic cardiomyopathy
RiHEART received conditional and time-limited approval in March 2026 for severe ischemic cardiomyopathy with inadequate response to standard therapy. It is an iPS-derived cardiomyocyte sheet placed on the heart through surgery.
Conditional approvalAcute ischemic stroke
The Japanese TREASURE randomized trial enrolled 206 patients. Intravenous MultiStem did not significantly improve the 90-day primary efficacy outcome versus placebo. Controlled data like this are more useful than a testimonial.
Randomized trial negativeKnee and meniscal disease
Savyscus was approved in May 2026 for a specific meniscal injury setting after arthroscopic repair. JACC has a 2025 knee osteoarthritis indication for cartilage defects of at least 2 cm² after conservative treatment fails, but JACC is cultured cartilage, not a stem-cell injection. Do not group these products together.
Indication-specificDiabetes, kidney, liver, ED and anti-aging
PMDA’s approved-product list through May 2026 does not show a disease-specific stem-cell product approved for these uses. A private regenerative medicine plan may follow the Safety Act route instead, which is different from PMDA product approval and does not by itself establish effectiveness.
Evidence must be checkedSource: PMDA, Japan’s Regulatory Framework and Recent Updates on Regenerative Medical Products, ARM APAC Forum Meeting, 20 May 2026. Counts above distinguish total approvals from the entries explicitly marked expired or withdrawn on the PMDA slide.
Already have a treatment description or quote? Line up the diagnosis, cell source, route, dose, and regulatory wording against the published record before you start comparing prices.
Walk Me Through It
What Counts as Approval in Japan
An approval applies to a specific product and indication.
Original conditional and time-limited approval for a defined traumatic spinal cord injury setting. Verify current status because time-limited approval is not permanent approval.
For knee osteoarthritis with a cartilage defect of at least 2 cm² when symptoms do not improve with conservative treatment. JACC is autologous cultured cartilage, not a stem-cell injection.
Both received conditional and time-limited approval in March 2026 for tightly defined Parkinson’s disease and severe ischemic cardiomyopathy populations.
Approved in May 2026. It uses autologous synovium-derived MSCs for meniscal injury indicated for meniscectomy and is administered after arthroscopic repair and suturing.
PMDA’s 2026 framework separates medical care or academic research under the Safety Act from commercial product marketing authorization under the PMD Act. Ask which route the exact intervention follows.
Don’t Let the Percentages Fool You
Read the patient numbers before the percentage.
These numbers come straight from PMDA review reports and a Japanese randomized trial. What stands out isn’t the headline — it’s the study design behind it.
STEMIRAC
Traumatic spinal cord injury
13
By day 220, 12 of 13 patients improved by at least one AIS grade and 3 of 13 improved by at least two grades.
Very small study without a randomized comparator. Do not use 12 of 13 as a general personal success rate.
Akuugo
Chronic traumatic brain injury
63 randomized
Sixty-one patients entered the modified analysis: 46 cell-treated and 15 sham-surgery patients. At week 24, the Fugl-Meyer motor score changed by 8.3 versus 2.3 points; difference 6.0 points, 95% CI 0.3 to 11.8, P=.0401.
Randomized sham-controlled evidence is stronger, but PMDA noted that the link to improvement in daily activities was not clear.
AMCHEPRY
Parkinson’s disease
7 treated
Six patients were in the efficacy analysis. Four of 6 improved the key OFF-state motor score at 24 months.
Single-center, open-label and uncontrolled. PMDA called the efficacy evaluation highly limited and required post-marketing confirmation.
RiHEART
Severe ischemic cardiomyopathy
8
All 8 were NYHA class III before treatment. At week 52, 4 were class I and 4 were class II; 4 of 8 improved six-minute walking distance by at least 45 m.
Open-label study with no control group and a surgical procedure. The small sample cannot provide a reliable individual probability.
TREASURE
Acute ischemic stroke
206
104 received MultiStem and 102 placebo. Excellent outcome at day 90 was 11.5% versus 9.8%; P=.90. There was no significant difference in the primary efficacy result.
A larger randomized, double-blind, placebo-controlled trial did not show short-term efficacy. This is important negative evidence.
Sources: PMDA review reports for STEMIRAC, Akuugo, AMCHEPRY and RiHEART; Houkin et al., JAMA Neurology, 2024, TREASURE randomized clinical trial.
Is This Study About Someone Like You?
Five details can change the answer.
A study result means something only when the patients in it look enough like you. Skip the diagnosis, stage, timing, or prior treatment — and the comparison falls apart.
Traumatic spinal cord injury, chronic traumatic brain injury, ischemic stroke and degenerative disease are different conditions even when symptoms overlap.
Check time since injury or diagnosis, baseline severity and whether the study treated acute, subacute or chronic disease.
Some approved indications require an inadequate response to standard therapy. Your previous treatment history can be part of eligibility.
Match the cell source, dose, route and procedure. Evidence from IV infusion does not validate brain implantation, joint administration or another product.
Check whether you can complete the required rehabilitation, imaging, laboratory tests and local follow-up after returning home.
What Outcomes Actually Mean
A percentage needs context.
A percentage with no context doesn’t tell you much. Every result needs four things to make sense: how many patients were studied, what it was compared against, what was measured, and when.
In TREASURE, 206 patients were randomized. The day-90 excellent-outcome rate was 11.5% with MultiStem and 9.8% with placebo, P=.90. The difference was not significant.
Source: Houkin et al., JAMA Neurology, 2024.
For spinal cord injury, traumatic brain injury, Parkinson’s disease and knee disease, function may be more informative than survival. In severe heart disease, cancer or transplant settings, hospitalization and survival can be relevant. Ask for the endpoint used in the evidence for your exact condition.
Let’s Talk Real Numbers
Use published prices only as context, not as a quote.
MSC price in one 2024 Japanese obstetrics and gynecology analysis, with wide variation. Use it to understand price spread, not to predict what your case should cost.
Source: Hosoya et al., Regenerative Therapy, 2024. MSC mean ¥1,832.1 ± ¥1,139.8 thousand; n=14 with price information.
How Cells Get Collected and Prepared
The route changes the risk and the recovery plan.
“Stem cell treatment” is a label that covers wildly different procedures. Before you compare safety, recovery times, or study results, you need to know the exact route — it changes everything.
TREASURE used a single IV infusion after acute ischemic stroke. Its evidence should not be transferred to a surgical or joint procedure.
AMCHEPRY is transplanted into the putamen by stereotactic brain surgery. The PMDA review also describes concomitant tacrolimus immunosuppression.
All 8 patients in the RiHEART clinical study underwent thoracotomy and transplantation onto the heart surface, followed for 52 weeks.
Savyscus is administered to a meniscal injury site after arthroscopic repair and suturing. It is not a general-purpose IV or knee OA injection.
Source and identity
Ask where the cells come from, whether they are your own or donor-derived and how identity is controlled.
Release criteria
Ask which final checks are used for sterility, contamination, viability, identity and any product-specific release criteria.
Procedure risk
Ask which risks come from the cells, which come from surgery or administration, and which come from medicines used with the procedure.
Complication plan
Know who handles urgent problems and what information your local physician will receive after you return home.
After the Procedure
Use the study timeline to judge the follow-up plan.
There’s no one-size-fits-all recovery after a stem-cell procedure. The Japanese studies we’ve reviewed measured their key outcomes over very different time windows — from 24 weeks out to 24 months.
STEMIRAC spinal cord injury study
AIS improvement was assessed at 220 days after injury. Rehabilitation and neurological function remain part of the outcome picture.
Akuugo traumatic brain injury
The primary motor-score comparison was at week 24; the study also followed patients through week 48.
AMCHEPRY Parkinson’s disease
The key motor result was assessed at 24 months after brain transplantation. The PMDA review requires post-marketing confirmation of efficacy.
RiHEART ischemic cardiomyopathy
The eight-patient study observed outcomes for 52 weeks after thoracotomy and heart-surface transplantation.
Before you travel, make sure you’ve asked: exact review dates and tests, the rehabilitation schedule, warning signs to watch for, who handles complications, and what discharge records your doctor back home will need.
Help Me Figure Out What I’ll NeedSource: PMDA review reports for STEMIRAC, Akuugo, AMCHEPRY and RiHEART.
When Something Doesn’t Feel Right
Six reasons to stop and ask for more detail.
Different tissues and diseases do not share the same evidence, dose, route or target outcome.
Ask how many patients were treated, how many were followed, for how long and what counted as success.
A Safety Act provision record and PMDA product approval are different. A registry entry does not replace disease-specific human outcome data.
You should be able to understand what cells are used, where they are processed and what is checked before administration.
A serious decision should allow time for document review, questions, alternatives and a written cost breakdown.
Travel is only one part of care. Follow-up tests, rehabilitation, complication management and local handover need to be clear.
How We Can Help
A clear case-to-evidence review package.
Japan Medical is an independent information and coordination service. We don’t diagnose, prescribe, rank hospitals, or promise results — that’s not our role. Our job is to help you understand the evidence and organize the questions to ask.
Your diagnosis, timing, stage, imaging, laboratory results, prior care and current medicines organized into one concise record set.
Published human data compared with your condition, disease stage, proposed product and route, with obvious mismatches identified.
Relevant PMDA, jRCT and peer-reviewed source links gathered so you can check the wording yourself.
Direct questions covering product, dose, route, expected outcome, risks, price, rehabilitation and follow-up.
Translation needs, record organization, visit timing and the documents to keep for your local clinician after you return home.
How This Works in Practice
Four practical steps.
Send what you already have
Diagnosis summary, recent imaging and labs, current medicines, prior treatment history, plus any proposal or quote.
We organize the gaps
We identify missing facts about evidence, regulatory status, cell product, route, cost and follow-up.
You get a question set
Questions are grouped so answers can be compared without mixing marketing claims with documented facts.
Prepare the next step
If you continue, we help organize documents, translation needs, visit timing and the records to take home.
Jump to Your Condition
Read the evidence for the condition you are actually dealing with.
Timing, product-specific evidence, functional endpoints and follow-up.
Pain, function, injection route, imaging and evidence limits.
Cell-based research, realistic endpoints and evidence stage.
Stroke evidence, heart failure pathways and outcome measures.
Current research questions, eligibility limits and monitoring.
Study endpoints, route, evidence limits and patient questions.
How to separate biomarker claims from clinically meaningful outcomes.
Records, translation, timing, travel preparation and follow-up handover.
The Questions We Get Most
Questions worth answering before payment or travel.
Does a regenerative medicine plan under Japan’s Safety Act mean the treatment is PMDA approved?
No. Japan uses different legal routes for regenerative medicine services and commercial products. A Safety Act provision record is not the same as PMDA marketing approval under the PMD Act.
How do I know whether published stem cell results apply to my case?
Match the diagnosis and cause, disease stage and timing, previous standard care, cell product, dose, route and follow-up. A result from a different disease, route or stage should not be treated as your personal prediction.
Is there one success rate for stem cell treatment?
No. Read any percentage together with the number of patients, comparator, outcome and follow-up time. Survival, motor function, disability, pain and imaging are different outcomes.
How much does stem cell treatment in Japan cost?
There is no single national private-care price. One 2024 Japanese specialty analysis reported a mean listed MSC price of ¥1.832 million with wide variation. It should not be used as a nationwide quote.
How long does recovery and follow-up take?
There is no universal timeline. Japanese studies range from IV infusion to joint procedures, stereotactic brain surgery and thoracotomy, with key assessments from weeks to 24 months. Ask for the route-specific recovery and follow-up schedule.
What records should I prepare first?
Start with the diagnosis summary, recent imaging and laboratory results, operative or pathology reports when relevant, current medicines, previous treatment history and any proposal or quotation already received.
Check the Sources Yourself
The main numbers above can be checked at the source.
Source review updated 11 August 2026.
Two legal routes and current approved-product statistics as of 20 May 2026.
Regenerative medical products approved from April 2015 through May 2026, including 2025 and 2026 indication details.
Thirteen-patient spinal cord injury efficacy analysis and day-220 AIS results.
Sham-controlled traumatic brain injury study, motor-score result and PMDA limitations.
Parkinson’s disease study population, 24-month motor result, surgery, immunosuppression and evidence limitations.
Eight-patient ischemic cardiomyopathy study, 52-week function data and surgical route.
206 patients, placebo-controlled design and day-90 efficacy result published in JAMA Neurology in 2024.
Peer-reviewed specialty-specific analysis of regenerative medicine provision plans and listed MSC prices.
Japanese clinical research and regenerative medicine records used to verify current study or provision information.
Independent patient guidance on evidence, risks and questions to ask about stem-cell interventions.
Before You Go
Send what you already have. See what is clear and what still needs an answer.
Send over your main reports, any proposal or quote you’ve received, and your list of questions. We’ll help you spot what’s clear — and what still needs an answer.
We provide health information and coordination support — that’s it. Japan Medical isn’t a hospital or clinic, and we don’t diagnose conditions, prescribe treatments, rank providers, or promise outcomes. Whether any regenerative medicine procedure is right for you is something only a qualified clinician can decide, after they’ve reviewed your full medical history, standard treatment options, procedure-specific risks, and the current regulatory status of the product you’re considering.