Gene therapy, gene editing and cell therapy are not three mutually exclusive choices. A treatment can consist of living cells that were genetically edited before being returned to a patient. Describing it only as “cell therapy” or only as “CRISPR” leaves out part of the mechanism. FDA's gene-therapy explanation explicitly includes approaches involving genetically modified patient cells. 1
A more useful comparison asks three questions: what reaches the patient, what was changed, and where that change happened.
Separate the delivery from the biological change
| Question | Examples of the distinction | Why it helps |
|---|---|---|
| What is delivered? | A vector or a prepared population of living cells | Identifies the actual therapeutic material |
| What changes? | A gene is added, gene activity is altered, or a DNA sequence is edited | Avoids treating every genetic intervention as the same operation |
| Where is the change made? | In the body, or in cells handled outside it | Distinguishes in vivo and ex vivo workflows |
FDA describes gene therapy as changing gene expression or cellular properties for a therapeutic purpose through several approaches. Genome editing uses tools that can make targeted changes to DNA; CRISPR-Cas9 is one such approach, not a synonym for all gene therapy. 1 2
The table is a mechanism map. It is not a prediction of benefit, safety, price or approval for an unnamed treatment.
Two real products show the overlap
In its December 2023 announcement, FDA described Casgevy as a treatment using a patient's blood stem cells edited with CRISPR-Cas9, enabling increased fetal-hemoglobin production. Lyfgenia also uses the patient's blood stem cells but uses a lentiviral approach to introduce genetic material that produces a modified form of hemoglobin. 3
Both therefore involve cells and genetic modification, but the modification is not identical. The shared word “gene” does not mean both use CRISPR. The shared word “stem cell” does not mean they are unmodified donor-cell transplants.
FDA's same announcement describes preparative myeloablative chemotherapy for these treatments. It also notes a boxed warning about hematologic malignancy for Lyfgenia and the need for longer-term monitoring. The mechanism comparison must not omit the surrounding treatment burden. 3
These are historical regulatory examples, not a head-to-head comparison or a current guide to choosing between the products.
Ex vivo and in vivo describe location, not a quality score
An ex vivo description says cells are handled outside the body before administration. In vivo delivery acts within the body. That distinction helps explain the process, but it does not by itself establish which approach is safer or more effective. The target, delivery system, product and study still matter. 1
A reader should be suspicious of a comparison that assigns a universal winner based only on those labels. A workflow can be convenient in one respect and demanding in another. Without matched evidence, a simple ranking hides rather than resolves the differences.
Cell replacement does not always involve gene editing
The edition's islet-therapy comparison separates programs by cell source and immune strategy. It is a reminder to preserve product-specific details rather than importing CRISPR into every stem-cell headline.
Similarly, a treatment that helps the immune system recognize a cancer is not automatically a DNA-editing treatment. See the personalized cancer-vaccine explanation for the distinct question being asked in those programs. Shared terminology such as “personalized” or “engineered” is not a complete mechanism.
Four claims that require separate evidence
Mechanism: what the intervention is intended to change.
Observed result: what actually happened in the studied population.
Durability and safety: what was monitored and for how long.
Authorization and access: what a regulator permitted, for which use and jurisdiction, and what practical access has been established.
These are this publication's reporting categories. An elegant explanation of the first does not prove the other three. A treatment can have a well-described mechanism and still lack evidence of the clinical outcome a reader hopes for.
Read the complete intervention sentence
A useful description sounds like this: “Patient-derived blood stem cells were modified outside the body using a specified method, followed by the named conditioning and administration process.” It supplies more information than “a gene cure,” even before the outcome is described.
The next sentence should name the population, measured benefit and limits. Our source-comparison tool can preserve those details side by side without pretending that different studies are automatically comparable.
Bottom line: classify the intervention on more than one axis. The overlap between cells, gene addition and editing is part of the science—not an inconsistency to be removed from the headline.