Collection mapClinical judgment
Stem Cells & Perinatal Therapies · Professional series · EN
Safety, uncertainty, and follow-up: compare the right risks
Neither “natural” nor “experimental” is a complete risk assessment.
On this page
Avoid both reassurance by slogan and fear by association
Autologous origin does not eliminate procedural or processing risks. Donor origin does not make a treatment uniformly unsafe. A severe complication reported with one uncontrolled product or unusual route should not be converted into the expected risk of every MSC trial. The converse is also true: a small study with no treatment-attributed serious events cannot establish universal safety. R05 R07
The correct unit of evaluation is the defined intervention and context.
Separate categories of risk
For consultation, it helps to distinguish product-related risks, administration-related risks, disease-related risks, and risks arising from delay or displacement of other care. Product issues may involve contamination, identity, composition, or consistency. Administration issues depend on route and setting. Disease-related assessment depends on the individual’s medical situation. Opportunity cost includes financial burden and time lost from a more appropriate evaluation.
This is a practical classification, not a numerical risk calculator. It helps keep a discussion complete without pretending that all categories have the same probability or severity.
The route changes the safety question
An intra-articular procedure, an intravenous infusion, a skin injection, a surgical graft, and an intrathecal or ocular intervention are different procedures. Safety information from one route is not a substitute for safety information from another. When a provider cites a favorable safety profile, ask whether the cited study used the same route and formulation. R05
The same caution applies to cell types. Risks associated with intensive hematopoietic transplantation or with pluripotent-cell-derived products should not simply be assigned to an MSC preparation. Each requires its own analysis. R07 R09
Individual factors to review
A proposed intervention may require assessment of active infection, malignancy history, clotting history, organ function, immune suppression, pregnancy or reproductive plans, prior reactions, and the medications currently being used. These are review domains, not a universal contraindication list. A competent team should explain which factors actually change eligibility or monitoring for the particular product.
Do not independently stop prescribed treatment to make room for an infusion. Medication management must be coordinated with the clinicians responsible for the underlying disease and, when applicable, the research protocol.
An event is not automatically a treatment effect
When reading follow-up data, distinguish an adverse event that occurred during observation from an event judged to be caused by treatment. Both should be reported; neither should be silently discarded or misrepresented. Small uncontrolled cohorts may be unable to resolve causality or detect uncommon harms. The 2026 RA follow-up illustrates why actual event tables and attribution matter more than a generalized reassurance statement. R10
What a follow-up plan should specify
Before treatment, identify who receives urgent calls, where complications will be evaluated, who has access to the product and lot information, how later events will be recorded, and how communication with the treating specialist will occur. The plan should survive the patient’s return home from another city or country.
Clinical deterioration after a procedure should not be dismissed as a required “healing response” without appropriate evaluation. Severe symptoms such as breathing difficulty, chest pain, fainting, new neurologic symptoms, or a rapidly worsening painful swollen joint require urgent medical assessment; a website or routine portal message is not the appropriate pathway.
The central principle
Good safety practice does not oppose innovation. It makes the result of innovation interpretable and protects the people whose trust allows the field to develop.
Sources
R05. FDA. Important Patient and Consumer Information About Regenerative Medicine Therapies. 2021.
Historical consumer safety guidance. Its older statement about the only approved stem-cell products must NOT substitute for the current FDA register. Access: Official FDA consumer page.
R07. ISSCR. Guidelines for Stem Cell Research and Clinical Translation: Clinical Translation of Stem Cell-Based Interventions.
https://www.isscr.org/guidelines-online/clinical-translation-of-stem-cell-based-interventions
Primary professional framework for translation, manufacturing, research oversight, and informed consent. Access: Online primary guidelines.
R09. Alexander T, et al. Autologous haematopoietic stem cell transplantation for rheumatic diseases: best practice recommendations from the EBMT Practice Harmonization and Guidelines Committee. Bone Marrow Transplant. 2025;60:1451–1464. doi:10.1038/s41409-025-02695-y.
https://www.nature.com/articles/s41409-025-02695-y
Published August 20, 2025. Center selection and multidisciplinary recommendations; underlying literature search through September 2024. Access: Publisher full text.
R10. Park EH, et al. Long-term safety of intravenous human umbilical cord blood-derived mesenchymal stem cell therapy in rheumatoid arthritis: a 5-year follow-up study. Stem Cells Transl Med. 2026;15(8):szag062. doi:10.1093/stcltm/szag062.
https://academic.oup.com/stcltm/article/15/8/szag062/8752872
Published August 5, 2026. Nine-patient observational follow-up; full results read, not just the optimistic significance statement. Access: Publisher full text.