CRISPR and Genome Editing: Treatment, Heredity and the Boundary of Intervention
CRISPR can treat disease, alter an individual's cells, or intervene in a heritable line. These are not the same thing. Where does treatment end and an intervention with societal consequences begin?
When CRISPR is discussed, it is easy to picture molecular scissors that simply cut out a faulty piece of DNA and replace it with the correct one. That metaphor is useful as a first approximation, but it is too narrow. CRISPR-Cas systems are programmable tools that can be directed to specific DNA sequences, while the outcome depends on the particular editor, the genomic site, and the way a cell repairs or interprets the change.
That distinction matters ethically as well. Editing blood-forming stem cells from an adult patient and returning them to that same person is not the same as altering an embryo in a way that could be transmitted to descendants. Both involve the genome, but the risks, consent, scale of consequences, and responsibilities are very different.
By 2026 CRISPR is no longer only a laboratory promise. The U.S. FDA first approved Casgevy in December 2023 as the first FDA-approved therapy to use CRISPR/Cas9; in July 2026 it expanded approval to patients as young as 2 years with specified indications of sickle cell disease or transfusion-dependent beta-thalassemia. Heritable editing of the human germ line, however, remains a domain in which scientific possibility does not amount to social or clinical acceptability.
The most useful question is therefore not whether genome editing is intrinsically good or bad. We should ask which cells are being changed, for what purpose, with what degree of certainty, who bears the risk, who can consent, and whether the consequences end with one patient or may extend into future generations.
CRISPR is not just one kind of 'genetic scissors'
The classic CRISPR-Cas9 system uses a guide RNA to direct Cas9 toward a selected DNA sequence next to an appropriate adjacent motif. Cas9 then cuts both strands of DNA. Much of the final change arises because the cell must repair that break. If it uses a rapid error-prone pathway, small insertions or deletions can appear; if a suitable template is available, more precise changes are possible in some settings.
Modern genome editing is broader than this model. Base editors can chemically convert particular bases without a conventional double-strand break, while prime editing combines targeting with reverse transcription and can support some precise substitutions, insertions, or deletions. CRISPR increasingly names a family of programmable tools rather than one single technique.
The practical consequence is important: 'editing' is not one intervention with one risk profile. Different platforms have different strengths, limitations, delivery problems, and possible errors.
Somatic and heritable editing are not the same ethical category
Somatic editing changes cells in a person who has already been born. If blood-forming stem cells are edited and returned to that patient, the change is generally not intended to reach reproductive cells and is not designed to be inherited. The ordinary framework of clinical medicine — expected benefit, risk, patient consent, and follow-up — therefore remains the basic reference point.
Heritable or germline editing changes eggs, sperm, early embryos, or their precursors in a way that could place the alteration into the germ line and transmit it to descendants. Risk to the first person born is therefore joined by uncertainty for future descendants who cannot participate in the decision as research subjects.
This does not mean that every reproductive decision is impermissible because future children cannot consent. It does mean that the ordinary informed consent of one adult cannot exhaust the ethical analysis when an intentional, inherited genomic change may have long-term and partly unpredictable consequences.
Treatment is no longer only theoretical
Early clinical successes showed that genome editing can produce effects that were previously unavailable for some diseases. In sickle cell disease and transfusion-dependent beta-thalassemia, researchers edited patients' own blood-forming stem cells outside the body to increase fetal hemoglobin. This approach led to Casgevy, which the FDA approved in December 2023 as its first approved therapy using CRISPR/Cas9; in July 2026 the FDA expanded approval to patients as young as 2 years with specified indications of both diseases.
Other approaches edit cells directly inside the body. In 2021 an early clinical study in transthyretin amyloidosis used lipid nanoparticles to deliver Cas9 mRNA and a guide RNA to the liver, reducing transthyretin production. This was an important proof that programmable editing could also be performed in vivo.
These examples also clarify the distinction between treating an individual and changing a future population. Therapeutic success in somatic editing is not by itself an argument for heritable embryo editing.
Precise targeting is not perfect precision
CRISPR can be directed to a highly specific genomic region, but that does not guarantee that the outcome is confined to one intended DNA letter. Off-target effects occur when an editor modifies a sufficiently similar sequence elsewhere in the genome. Their likelihood depends on the guide RNA, nuclease, cell type, dose, and delivery method, so increasingly sensitive detection methods continue to be developed.
Unwanted changes can also occur at the target itself. Repair of a double-strand break can produce large deletions, rearrangements, or outcomes different from those intended. In embryos there is an additional problem of mosaicism: different cells in the same developing organism may end up carrying different versions of the edit.
Safety therefore means more than asking whether the correct gene was targeted. It includes what actually happened at the target, whether changes appeared elsewhere, which cells were edited, and how stable the effect will be over time.
A gene is not an independent dial for a trait
For some monogenic diseases, the relationship between a pathogenic variant and disease is direct enough that a targeted intervention can meaningfully reduce risk. Even then, however, a gene functions inside a broader biological system. The same protein can matter in multiple tissues, and the same variant can have several effects.
For complex traits the picture is much harder. Height, most common diseases, cognitive traits, and many behavioral characteristics are not products of one gene but of large numbers of genetic variants, developmental processes, and environment. A single variant may also affect several traits, creating pleiotropic trade-offs.
This is an important boundary for enhancement. A more powerful DNA editor does not turn complex biology into a menu of independent settings. Technical ability to change a variant is not the same as reliable prediction of all consequences of that change.
Heritable changes extend risk through time
With somatic therapy, clinicians can monitor a patient and adjust later care if needed. Heritable editing is different. If a change enters the germ line, it can become part of the genomes of future descendants and spread through a family line before decades of long-term outcome data exist.
The international commission convened by the U.S. National Academies and the U.K. Royal Society therefore stressed that clinical heritable editing should not proceed until intended genomic changes can be made sufficiently reliably without undesired alterations and until stringent scientific, medical, and governance conditions are met. The report also emphasized that technical safety alone would not settle the social question.
The issue is therefore not merely the risk of one procedure. It is how much uncertainty society is prepared to introduce into reproduction and future generations, and who has legitimate authority to set that threshold.
Treatment and enhancement do not always have a sharp border
The distinction is easiest at the extremes. Correcting a variant that causes a severe monogenic disease is more clearly therapeutic than altering height or athletic capacity in a healthy person. Between those cases lies a wide gray zone: reducing disease susceptibility, increasing resistance to infection, or changing traits that may be advantageous in one environment and costly in another.
The label 'treatment' therefore does not settle the ethical issue. Severity of disease, safer alternatives, benefit-risk balance, the patient's capacity to consent, fair access, and whether the change affects one body or an inherited lineage all matter.
Nor is it useful to label every form of genetic treatment as eugenics. Eugenics was historically a political and social project of selection, coercion, and population management. Some future forms of heritable enhancement could create related problems, but treating a particular patient for a genetic disease is not automatically the same thing.
The CRISPR-babies case is mainly a lesson about governance
In 2018 He Jiankui announced the birth of children from embryos in which he had attempted to alter CCR5 using CRISPR. The experiment drew international condemnation because of scientific immaturity, problems with informed consent, the choice of target, inadequate oversight, and the fact that the intervention was intended to affect the germ line.
The case is not evidence that every use of CRISPR is unethical. It shows why technology must be separated from a particular application. The same underlying tool can be studied in cultured cells, used to treat an adult patient, or applied to an embryo with heritable consequences — and the ethical framework changes dramatically between those cases.
It also exposed the weakness of systems that rely on the personal judgment of one investigator. A technology whose consequences can extend beyond one laboratory or one generation requires transparent registries, independent oversight, and the capacity for international response to unsafe or unregistered work.
Governance is more than permission or prohibition
In 2021 the WHO proposed a global governance framework for human genome editing. Its recommendations include research registries, international collaboration, responses to unregistered or unsafe work, public education and engagement, and mechanisms for dealing with research and medical travel. The idea is not that one global body should approve every experiment, but that dangerous applications should not simply migrate to jurisdictions with the weakest oversight.
Governance also includes equity. A therapy can work scientifically while remaining available to very few people because of cost, infrastructure, or geography. If societies invest in genome-based medicine, the question of benefit therefore does not end with laboratory efficacy.
With heritable editing the issue becomes even broader: who decides which traits count as disease, which as acceptable human variation, and which as desirable enhancement? The answer cannot be purely technical because choosing the target already contains a value judgment.
The boundary of intervention is a set of thresholds, not one line
CRISPR is powerful because it has made genome intervention sufficiently programmable that we can select particular genomic sites. But the ability to intervene does not tell us when intervention is justified. That requires several separate questions: how well we understand the biological consequence, how reliably we can make the change, whether safer alternatives exist, who bears the risk, who can consent, and whether the effect is limited to one patient or is heritable.
For severe disease and somatic treatment, the balance may strongly favor intervention when benefits are large and risks acceptable. For heritable embryo editing the threshold is higher because uncertainty is greater, consequences may be more durable, and ordinary consent from future descendants cannot be obtained. Enhancement in healthy people adds questions about inequality, social pressure, and who gets to define a 'better' human being.
The most defensible conclusion is therefore neither 'the genome must never be touched' nor 'if we can do it, we should.' Genome editing is already part of medicine, but the boundary of responsible use moves more slowly than technical capability. That gap between what we can do and what we can justify doing is the central problem of human genome editing.
Sources and further reading
- THY-REALITY — Transhumanizem: od zdravljenja do preoblikovanja človeka / Transhumanism: From Healing to Transforming the Human Being (LOCKED): therapy–enhancement boundary and human modification context.
- THY-REALITY — Izboljševanje človeka: terapija, nadgradnja in nova neenakost / Human Enhancement: Therapy, Upgrade and New Inequality (LOCKED): enhancement, access and inequality.
- THY-REALITY — Evgenika: kako je »izboljševanje človeštva« postalo politični projekt / Eugenics: How 'Improving Humanity' Became a Political Project (LOCKED): historical boundary between medical treatment and coercive population projects.
- THY-REALITY — Informirano soglasje: komu pripada odločitev o človeškem telesu? / Informed Consent: Who Owns the Decision About the Human Body? (LOCKED): consent, bodily autonomy and limits of proxy decision-making.
- Jinek, M. et al. — A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science 337 (2012): 816–821.
- Doudna, J. A. — The promise and challenge of therapeutic genome editing. Nature 578 (2020): 229–236.
- Frangoul, H. et al. — CRISPR-Cas9 Gene Editing for Sickle Cell Disease and β-Thalassemia. New England Journal of Medicine 384 (2021): 252–260.
- Gillmore, J. D. et al. — CRISPR-Cas9 In Vivo Gene Editing for Transthyretin Amyloidosis. New England Journal of Medicine 385 (2021): 493–502.
- U.S. Food and Drug Administration — FDA Approves First Gene Therapies to Treat Patients with Sickle Cell Disease (8 Dec 2023): Casgevy as the first FDA-approved therapy using CRISPR/Cas9.
- U.S. Food and Drug Administration — CASGEVY product and approval history (current regulatory status, including later indication expansions).
- U.S. Food and Drug Administration — FDA Approves First Gene Therapy for Young Children with Sickle Cell Disease (1 Jul 2026): supplemental Casgevy approval for patients aged 2 years and older with specified SCD or transfusion-dependent beta-thalassemia indications.
- Kosicki, M.; Tomberg, K.; Bradley, A. — Repair of double-strand breaks induced by CRISPR–Cas9 leads to large deletions and complex rearrangements. Nature Biotechnology 36 (2018): 765–771.
- Zhu, M. et al. — Tracking-seq reveals the heterogeneity of off-target effects in CRISPR–Cas9-mediated genome editing. Nature Biotechnology (2024).
- Ma, H. et al. / related Nature analysis — Human embryo genome editing and the problem of mosaicism; clinical relevance requires avoiding mosaic outcomes.
- National Academy of Medicine; National Academy of Sciences; Royal Society — Heritable Human Genome Editing. National Academies Press (2020).
- World Health Organization — Human genome editing: recommendations (2021).
- World Health Organization — Human genome editing: a framework for governance (2021).
- Cyranoski, D.; Ledford, H. — Genome-edited baby claim provokes international outcry. Nature 563 (2018): 607–608; historical governance case.
- Murray, J. B.; Harrison, P. T.; Scholefield, J. — Prime editing: therapeutic advances and mechanistic insights. Gene Therapy 32 (2025): 83–92.