How Long Does Gene Therapy Last? Is It Permanent?

The Million-Dollar Question

If you or someone you love is considering gene therapy, one of the first questions you will ask is: How long will this actually last?

It is a fair question, especially when many of these treatments come with price tags north of one million dollars. The honest answer is that it depends. Some gene therapies appear to be genuinely permanent. Others may fade over time. And for many newer treatments, we simply do not have enough years of follow-up data to know for certain.

This article breaks down why durability varies so widely, what the real-world data shows for today's approved therapies, and what patients should realistically expect when planning for the long term.

Why Some Gene Therapies Are Permanent and Others Are Not

The durability of a gene therapy comes down to a surprisingly simple question: what happens to the new genetic material once it enters your cells?

There are two fundamentally different approaches, and they behave very differently over time.

Approach 1: Integrating Vectors — The Permanent Rewrite

Some gene therapies use delivery vehicles (called vectors) that physically insert new genetic material into your chromosomes. The most common integrating vector is a lentivirus, a modified version of HIV that has been stripped of its ability to cause disease and repurposed as a delivery truck for therapeutic genes.

When a lentiviral vector delivers its payload, the new gene becomes a permanent part of your DNA. It is woven into the fabric of your chromosomes, just like the genes you were born with. Every time that cell divides, the new gene is copied along with everything else.

This matters enormously for therapies that target stem cells — the self-renewing cells in your bone marrow that produce all of your blood cells for your entire life. If you successfully edit or add a gene to a stem cell, every blood cell that stem cell produces going forward will carry that change. In theory, this is a one-and-done fix.

Casgevy (exagamglogene autotemcel) and Lyfgenia (lovotibeglogene autotemcel) both take this approach. They modify a patient's own bone marrow stem cells outside the body, then infuse those modified cells back after chemotherapy clears out the old marrow. Because the changes are in long-lived stem cells and integrated into the chromosomes, the correction is expected to be lifelong.

Approach 2: Episomal Vectors — The Guest That May Eventually Leave

Other gene therapies use adeno-associated viruses (AAV) as their delivery vehicle. AAV vectors are extremely popular in gene therapy because they are efficient at getting into many types of cells, they trigger relatively mild immune responses, and they can target specific tissues like the liver, muscles, or nervous system.

But AAV vectors have an important limitation: in most cases, the therapeutic gene they deliver does not integrate into your chromosomes. Instead, it sits in the nucleus of the cell as a separate, small circle of DNA called an episome. Think of it as a sticky note placed on top of a textbook — the information is there and the cell can read it, but it is not stitched into the binding.

This has two consequences for durability:

Cell division dilutes the therapy. When a cell divides, the episome is not reliably copied to both daughter cells. Over many rounds of division, fewer and fewer cells retain the therapeutic gene. This is a bigger problem in tissues that turn over quickly (like the liver) and less of a problem in tissues that rarely divide (like neurons or muscle fibers).
Cell death removes the therapy. If the cell containing the episome dies — whether from natural turnover, injury, or disease — the therapeutic gene is lost with it.

For tissues like the brain and spinal cord, where neurons last a lifetime and almost never divide, episomal delivery can be remarkably durable. For the liver, which regenerates and turns over more actively, episomal gene therapy may gradually lose potency over years or decades.

What the Data Actually Shows: Therapy by Therapy

Let us look at the real-world evidence for some of the most important approved and late-stage gene therapies.

Zolgensma (Onasemnogene Abeparvovec) — Spinal Muscular Atrophy

Vector: AAV9 (episomal) Target tissue: Motor neurons Approved: 2019

Zolgensma delivers a functional copy of the SMN1 gene to motor neurons in infants and young children with spinal muscular atrophy (SMA), a devastating disease that destroys the nerve cells controlling muscle movement.

The durability news here is encouraging. Motor neurons are long-lived, post-mitotic cells — they do not divide. That means the episomal AAV-delivered gene stays put. Follow-up data published through 2025 shows that children treated with Zolgensma continue to maintain motor milestones at 5+ years post-treatment. Many children who would have required permanent ventilator support are sitting, standing, and in some cases walking.

While we do not yet have 20- or 30-year data, the biology is favorable. Because motor neurons last a lifetime and do not divide, there is good reason to expect the benefit to be very long-lasting — potentially permanent in the cells that survive.

The caveat: Zolgensma cannot rescue motor neurons that have already died before treatment. This is why early treatment (ideally identified through newborn screening) produces the best outcomes.

Roctavian (Valoctocogene Roxaparvovec) — Hemophilia A

Vector: AAV5 (episomal) Target tissue: Liver Approved: EU 2022, US 2023 (withdrawn from EU market 2024)

Roctavian was designed to give people with severe hemophilia A the ability to produce their own Factor VIII, a clotting protein their bodies cannot make. The AAV5 vector delivers a functional Factor VIII gene to liver cells (hepatocytes).

This is where the durability story gets complicated — and serves as an important cautionary tale.

In the pivotal clinical trial, patients initially showed strong Factor VIII production after treatment. Many were able to stop or dramatically reduce their prophylactic Factor VIII infusions. But over the following months and years, Factor VIII levels declined significantly in most patients.

By year two, the median Factor VIII activity had dropped considerably from its peak. By year three and beyond, some patients had Factor VIII levels that had fallen back to a range where bleeding protection was uncertain. A subset of patients maintained clinically meaningful levels, but the therapy did not deliver the sustained, stable correction that many had hoped for.

This decline was a major factor in BioMarin's decision to withdraw Roctavian from the European market in 2024, citing limited commercial uptake. The product remains available in the United States, but the durability concerns have made both physicians and patients cautious.

Why does it fade? The most likely explanation involves the liver's biology. Hepatocytes do turn over, albeit slowly. Over time, cells carrying the episomal Factor VIII gene are gradually replaced by new cells that do not carry it. There may also be a low-level immune response against the transduced cells that accelerates this loss. Researchers continue to study the exact mechanisms.

Casgevy (Exagamglogene Autotemcel) — Sickle Cell Disease and Beta-Thalassemia

Vector: None (ex vivo CRISPR editing of stem cells) Target tissue: Bone marrow stem cells Approved: 2023

Casgevy uses CRISPR-Cas9 to edit the BCL11A gene in a patient's own bone marrow stem cells, reactivating fetal hemoglobin production to compensate for defective adult hemoglobin in sickle cell disease (SCD) and transfusion-dependent beta-thalassemia (TDT).

The durability data here is very encouraging. At three or more years of follow-up, patients treated with Casgevy have shown sustained, high levels of fetal hemoglobin. Sickle cell patients have remained free of vaso-occlusive crises (the painful episodes that define the disease), and beta-thalassemia patients have remained transfusion-free.

Why is durability expected to hold? Because Casgevy edits hematopoietic stem cells — the long-lived, self-renewing cells at the top of the blood cell hierarchy. When you edit the DNA of a stem cell, every blood cell it produces for the rest of the patient's life carries that edit. The modification is chromosomal, not episomal. It is copied faithfully every time the stem cell divides.

The key question for very long-term durability is whether the edited stem cells maintain their engraftment and self-renewal capacity over decades. The conditioning chemotherapy (myeloablative busulfan) that patients receive before infusion is harsh, and the long-term health of the transplanted stem cell pool is something that will be monitored for years to come. But the fundamental biology — a permanent edit in a self-renewing cell — supports lasting benefit.

Lyfgenia (Lovotibeglogene Autotemcel) — Sickle Cell Disease

Vector: Lentiviral (integrating) Target tissue: Bone marrow stem cells Approved: 2023

Lyfgenia takes a different approach to the same problem. Instead of editing an existing gene with CRISPR, it uses a lentiviral vector to add a modified beta-globin gene (called betaA-T87Q-globin) to the patient's stem cells. This anti-sickling hemoglobin mixes with the patient's own hemoglobin and reduces sickling.

Because lentiviral vectors integrate into chromosomes, the added gene is permanent in the transduced cells. Clinical data through multiple years of follow-up shows sustained production of anti-sickling hemoglobin and significant reductions in vaso-occlusive events.

Lyfgenia does carry a boxed warning about a potential risk of blood cancer (hematologic malignancy) associated with lentiviral integration — a concern that comes with any integrating vector, since the new gene could theoretically land in or near a cancer-related gene. This risk appears to be very low based on available data, but it is part of the long-term monitoring picture.

Intellia's NTLA-2001 — Transthyretin Amyloidosis (In Vivo CRISPR)

Vector: Lipid nanoparticle (non-viral delivery of CRISPR components) Target tissue: Liver Not yet approved (Phase 3 as of early 2026)

NTLA-2001 is one of the most closely watched gene therapies in development. It uses CRISPR-Cas9 delivered via lipid nanoparticles to knock out the TTR gene directly in the liver, reducing production of the misfolded transthyretin protein that causes hereditary transthyretin amyloidosis (hATTR), a progressive and fatal disease.

What makes NTLA-2001 remarkable from a durability standpoint is that it is a one-time intravenous infusion that edits liver cells in the body (in vivo) — no surgery, no stem cell transplant, no ex vivo manipulation. And the results have been striking: at three or more years of follow-up, patients have shown sustained TTR protein knockdown of approximately 90% or greater.

Why does this in vivo liver editing last when AAV-based liver gene therapy (like Roctavian) fades? The critical difference is that CRISPR makes a permanent change to the cell's own DNA. Even though the CRISPR components themselves are transient — the lipid nanoparticle delivers the Cas9 protein and guide RNA, which do their work and are then degraded — the edit they leave behind is permanent. When an edited hepatocyte divides, both daughter cells carry the knockout.

There is, however, a nuance: the initial single dose of NTLA-2001 edits a large fraction of hepatocytes, but not 100%. Over many years, if there is significant liver turnover, the fraction of edited cells could theoretically decrease slightly as un-edited progenitor cells contribute new hepatocytes. The long-term data so far is reassuring, but this will be tracked closely in ongoing trials and post-approval monitoring.

CRISPR Editing Is Permanent by Definition

It is worth pausing to make a fundamental point that applies to all CRISPR-based therapies: the DNA edit itself is permanent.

When CRISPR-Cas9 cuts a specific location in the genome and the cell's repair machinery fixes the break (either by disrupting the gene or inserting a new sequence), that change is written into the cell's chromosomes forever. It will be copied every time the cell divides. It cannot "wear off." There is no molecular clock counting down.

This is fundamentally different from gene therapies that deliver a new gene on an episomal vector. The episome can be lost. A CRISPR edit cannot.

The question for CRISPR therapies is never "will the edit fade?" but rather:

Were enough cells edited? A single dose might edit 90% of target cells, which could be sufficient for a clinical benefit.
Are the edited cells long-lived? Editing a stem cell is more durable than editing a mature cell that will eventually die and be replaced.
Will the un-edited cells eventually outnumber the edited ones? In tissues with active turnover, this is a theoretical concern over very long timeframes.

The FDA's 15-Year Follow-Up Mandate

Recognizing that gene therapy is still a young field and that long-term safety and durability questions remain open, the FDA requires manufacturers of approved gene therapies to conduct long-term follow-up studies lasting 15 years after treatment.

This is not optional — it is a condition of approval. These studies track:

Durability of the therapeutic effect — Is the protein still being produced? Are clinical benefits maintained?
Delayed adverse events — Including the risk of insertional mutagenesis (cancer risk from integrating vectors), immune reactions, and organ toxicity
Overall health outcomes — How do treated patients fare compared to those on conventional therapy?

For patients, this 15-year follow-up period means two things. First, you will be asked to participate in long-term monitoring, including periodic blood tests, clinical evaluations, and reporting of any health changes. Second, it means the medical community is being honest that we do not yet have complete answers — we are gathering them in real time, and every patient who participates in follow-up is contributing to knowledge that will help future patients.

The European Medicines Agency (EMA) has similar requirements, and in some cases has mandated post-authorization efficacy studies that track real-world outcomes beyond the controlled clinical trial setting.

The Re-Dosing Problem: Anti-AAV Antibodies

What happens if an AAV-based gene therapy fades? Can you just get another dose?

Unfortunately, in most cases, no — at least not with current technology.

Here is why: when your body encounters an AAV vector for the first time, your immune system mounts an antibody response against the viral shell (capsid). These anti-AAV antibodies can persist for years, possibly for life. If a second dose of the same AAV vector is administered, these antibodies will neutralize it before it can deliver its therapeutic cargo to the target cells.

This is one of the most significant challenges in the gene therapy field. It means that for patients whose AAV-based therapy fades — like some Roctavian recipients — there is currently no straightforward way to "top up" the treatment.

Researchers are working on several strategies to overcome this barrier:

Alternative AAV serotypes. If the first treatment used AAV5, perhaps a second dose could use AAV8 or another serotype that the antibodies do not recognize. This is being tested, but cross-reactivity between serotypes complicates the approach.
Immunosuppression protocols. Temporarily suppressing the immune system before re-dosing could theoretically allow the second dose to evade antibodies. Clinical trials exploring this approach are underway.
Plasmapheresis (IgG depletion). Physically filtering antibodies out of the blood before re-administration is another approach being investigated.
Non-viral delivery systems. Lipid nanoparticles (like those used by Intellia's NTLA-2001) do not trigger the same kind of persistent immune memory as AAV capsids. This means re-dosing with lipid nanoparticle-based therapies may be feasible — a significant potential advantage.
Engineered capsids. Companies are developing novel AAV capsids that are designed to evade pre-existing antibodies while still efficiently delivering genes to target tissues.

The re-dosing challenge is one reason why many researchers and companies are increasingly interested in non-viral delivery methods and in making sure the first dose of any gene therapy is as effective and durable as possible.

What Patients Should Expect: A Practical Guide

If you are a patient or caregiver considering gene therapy, here is a realistic framework for thinking about durability:

Questions to Ask Your Treatment Team

What type of vector or editing tool does this therapy use? Integrating vectors and direct gene editing (CRISPR) are more likely to be permanent. Episomal AAV vectors may fade in some tissues.
What tissue is being targeted, and how quickly does it turn over? Therapies targeting neurons (which rarely divide) have better durability prospects than those targeting liver cells (which turn over slowly but continuously).
How long is the follow-up data? A therapy with 5+ years of stable data gives more confidence than one with only 1-2 years. Ask for the latest published results.
What happens if the effect diminishes? Is re-dosing possible? Are there backup treatment options? What monitoring will detect a decline early?
What does long-term follow-up involve? Understand the schedule of visits, blood tests, and evaluations you will need to commit to over the next 15 years.

Setting Realistic Expectations

No gene therapy yet has 20+ years of follow-up data. The field is young. Even the most promising results come with the caveat that ultra-long-term outcomes are still being studied.
"Permanent" is a biological expectation, not a guarantee. When we say a CRISPR edit or an integrating vector is "permanent," we mean the DNA change itself is permanent. Whether that translates to a permanent clinical benefit depends on many factors — cell survival, immune responses, the disease itself.
Monitoring is not a sign that something is wrong. Regular follow-up after gene therapy is standard of care and a regulatory requirement. It does not mean your doctor expects problems — it means the field is being responsible about tracking long-term outcomes.
Your experience may differ from the average. Clinical trial data reports medians and averages, but individual responses vary. Some patients may have stronger or weaker responses than the published numbers suggest.

The Emotional Reality

It is worth acknowledging that the uncertainty around durability can be emotionally difficult. Many patients undergo gene therapy hoping for a cure — a single treatment that frees them from a lifetime of medication, infusions, or symptoms. For some therapies and some patients, that hope is being realized. For others, the reality is more nuanced.

If you are dealing with a disease like hemophilia A, where the durability of current gene therapy options is uncertain, it is reasonable to feel frustrated. The best thing you can do is stay connected with your treatment team, participate in follow-up studies, and stay informed about the rapid advances happening in this field.

The Road Ahead

Gene therapy durability is one of the most active areas of research in medicine. The lessons learned from first-generation therapies like Roctavian are directly informing the design of next-generation treatments. Key trends include:

Shift toward gene editing over gene addition. Permanent edits avoid the episome durability problem entirely. Expect to see more CRISPR, base editing, and prime editing therapies entering clinical trials.
Improved vector engineering. New AAV capsids with better tissue targeting, higher transduction efficiency, and lower immunogenicity are in development. The goal is to get more therapeutic gene into more cells with a single dose.
Non-viral delivery platforms. Lipid nanoparticles, virus-like particles, and other non-viral systems may eventually replace AAV for many applications, potentially enabling re-dosing and avoiding the antibody barrier.
Combination approaches. Some researchers are exploring combining a transient gene editing treatment with a durable cell therapy — for example, using in vivo CRISPR to edit liver cells while simultaneously engineering stem cells for long-term benefit.

The field is moving fast. Therapies that seemed futuristic five years ago are now treating real patients, and the data on their long-term durability grows richer every year.

The Bottom Line

Gene therapy durability is not one-size-fits-all. Here is the simplest way to think about it:

Therapy Type	Expected Durability	Why
CRISPR editing of stem cells (Casgevy)	Likely permanent	Permanent edit in self-renewing cells
Lentiviral gene addition to stem cells (Lyfgenia)	Likely permanent	Integrated into chromosomes of self-renewing cells
In vivo CRISPR editing of liver (NTLA-2001)	Likely very long-lasting	Permanent edit, but not all cells edited
AAV delivery to neurons (Zolgensma)	Likely very long-lasting	Neurons rarely divide, so episome persists
AAV delivery to liver (Roctavian)	May fade over years	Episomal gene lost as liver cells turn over

The field is learning in real time, and every year of follow-up data brings more clarity. If you are a patient considering gene therapy, the most important thing you can do is have an open conversation with your care team about what is known, what is uncertain, and what the monitoring plan looks like. Gene therapy is not magic — but for many patients, it is the closest thing medicine has ever offered to a genuine cure.

Sources

Frangoul, H., et al. (2024). "Exagamglogene autotemcel for severe sickle cell disease." New England Journal of Medicine. Long-term follow-up data from the CLIMB SCD-121 trial.
Locatelli, F., et al. (2024). "Exagamglogene autotemcel for transfusion-dependent beta-thalassemia." New England Journal of Medicine. CLIMB THAL-111 trial results.
Mahlangu, J., et al. (2023). "Two-year outcomes of valoctocogene roxaparvovec therapy for hemophilia A." New England Journal of Medicine, 388(8), 694-705.
Gillmore, J.D., et al. (2024). "CRISPR-Cas9 in vivo gene editing for transthyretin amyloidosis." Updated long-term follow-up data from the NTLA-2001 Phase 1 study, presented at AHA Scientific Sessions.
Mendell, J.R., et al. (2021). "Five-year extension results of the Phase 1 START trial of onasemnogene abeparvovec in spinal muscular atrophy." JAMA Neurology, 78(7), 834-841.
FDA Guidance for Industry. (2020). "Long Term Follow-Up After Administration of a Gene Therapy Product." U.S. Food and Drug Administration.
Nathwani, A.C., et al. (2014). "Long-term safety and efficacy of factor IX gene therapy in hemophilia B." New England Journal of Medicine, 371(21), 1994-2004.
High, K.A. & Roncarolo, M.G. (2019). "Gene therapy." New England Journal of Medicine, 381, 455-464.
BioMarin Pharmaceutical. (2024). Press release regarding Roctavian commercial strategy update and EU market withdrawal.
Vertex Pharmaceuticals and CRISPR Therapeutics. (2025). Casgevy long-term follow-up data presentations at ASH Annual Meeting.