Regrowth, cell replacement, gene therapy and assistive devices can aim at a similar human outcome while doing very different things biologically. A person communicating again, producing insulin after a transplant, or improving a measured function has not necessarily regenerated the original damaged tissue.
This guide organizes the selected research in this edition by what is being changed. It is an evidence-reading framework, not a complete taxonomy of regenerative medicine or a recommendation to obtain a procedure.
Regrowth: generating a biological structure
The tooth-regrowth program followed here investigates an approach related to tooth development. Its currently reported research indication is not every possible cause of adult tooth loss. 1 2
A meaningful regrowth claim would need to establish what formed and whether it performed the relevant function. A process milestone or an image of a structure would not, by itself, demonstrate the whole clinical outcome.
For the actual status and limits, read the tooth-regrowth guide. For an important boundary, see enamel remineralization versus whole-tooth regrowth. The same everyday word “repair” should not erase that difference.
Replacement: supplying cells that perform a missing job
The islet-cell programs in this edition aim to provide insulin-producing function using supplied cells. The source of those cells and the strategy for dealing with immune responses differ by program. 3 4 5
The clinical questions include how much function is supplied, how long it lasts and what additional treatment is required. That is different from showing that a person’s original pancreatic cells have regenerated and the underlying disease process has disappeared.
Our cell-therapy comparison places those differences beside one another. A positive signal such as measurable C-peptide and an outcome such as insulin independence must remain distinct.
Gene therapy: changing the information available to cells
Otarmeni’s FDA record concerns a specific genetic form of hearing loss and includes requirements about preserved outer hair-cell function. It should not be described as proof that the treatment regenerates all destroyed hearing cells or treats every cause of deafness. 6
The distinction is not semantic. Correcting a defined biological deficiency and replacing cells that are no longer present require different evidence. The OTOF explanation and hair-cell research article separate those questions.
A result in one genetic population can be important without automatically applying to acquired hearing loss from another cause.
Assistance: enabling an outcome without claiming tissue replacement
A speech BCI can decode signals into communication. A brain–spine interface can connect decoded intentions with stimulation. The studies covered here do not establish that these systems regrow the original speech pathway or anatomically repair a spinal lesion. 7 8
An assistive result does not become less useful because it is not regeneration. The relevant question is what a person can do with the system, under what conditions, and with what burden or risk.
The speech research guide and spinal research guide describe those outcomes without assigning a single restoration score to different tasks.
Use a three-part description
| Part of the description | Example of a precise question |
|---|---|
| Biological action | Were cells added, information delivered, a structure generated or an existing circuit stimulated? |
| Observed outcome | Was a biomarker detected, a task completed or an important clinical event reduced? |
| Conditions and limits | Was the result temporary, device-dependent, treatment-dependent or demonstrated in a restricted population? |
An article that answers all three is more useful than one that says “regenerative breakthrough.” It also exposes what the source has not established without requiring an arbitrary scientific-quality score.
A research label is not a marketing authorization
FDA’s consumer warning explicitly distinguishes trial listings and company registration from proof that a regenerative product is legally marketed. It also describes harms reported with unapproved products. The warning’s historical publication date does not turn its old category counts into a complete current approval inventory. 9
Consequently, the evidence for a named research program cannot be transferred to an unrelated clinic’s product because both use “stem cells,” “regenerative” or “gene therapy” in their advertising. Product identity, manufacturing, indication and oversight matter.
What counts as progress?
Progress is a more reliable answer to a defined question: functional evidence where only a mechanism was known, longer observation where durability was uncertain, or a controlled result where only an uncontrolled association existed.
It does not require pretending that all approaches are biologically identical or close to availability. The 2026 developments collection uses these distinctions to show what actually changed across different kinds of research.
A specific approval is not an industry-wide authorization
The Ryoncil report examines an actual MSC-product approval and its indication. The gene/cell/editing comparison separates delivery, mechanism and manufacturing rather than treating all engineered therapies alike.