← All articles
Gloved hands lifting the frosted lid of a cryogenic shipping container as vapour spills out

Approved Gene Therapy: Who Qualifies, How Long It Has Been Followed, What Is Known About Harm

ResearchMedicine and treatment

Six one-time genetic treatments are licensed in the United States for sickle cell disease, beta-thalassemia, hemophilia and SMA. What the labels say about eligibility, how many years anyone has actually been followed, and what the boxed warnings record.

12 min read

Headlines about gene therapy keep using one word: cure. If the diagnosis is yours or your child's, that is not the useful word. Three questions are, and all three are answered in the trial papers and the label text: who is eligible right now, how many years anyone has actually been followed after the infusion, and what has gone wrong in the people already treated.

This is informational material. It does not replace a consultation with a doctor, and nothing below is a recommendation for or against any treatment.

What has been licensed, and for whom?

Five conditions have a licensed one-time genetic treatment in the United States: sickle cell disease, beta-thalassemia, hemophilia A, hemophilia B and SMA. The eligibility wording below is the regulator's, not ours.

ProductUS approvalWho qualifies on the label
Casgevy (exagamglogene autotemcel)8 Dec 2023 (sickle cell); age floor lowered 1 Jul 2026Age 2 and older with sickle cell disease and recurrent vaso-occlusive crises, or transfusion-dependent beta-thalassemia
Lyfgenia8 Dec 2023, same FDA announcementAge 12 and older with sickle cell disease and a history of vaso-occlusive events
Zynteglo (betibeglogene autotemcel)17 Aug 2022Adults and children with beta-thalassemia who need regular red-cell transfusions
Hemgenix (etranacogene dezaparvovec)22 Nov 2022Adults with hemophilia B on factor IX prophylaxis, or with current or past life-threatening hemorrhage, or repeated serious spontaneous bleeds
Roctavian (valoctocogene roxaparvovec)29 Jun 2023Adults with severe hemophilia A and no pre-existing antibodies to the AAV5 vector on an FDA-approved companion test
Zolgensma (onasemnogene abeparvovec)24 May 2019Children under 2 years with bi-allelic SMN1 mutations

Two access claims need a footnote. Casgevy's beta-thalassemia approval is usually dated 16 January 2024, but that date comes from a company timeline rather than an agency release, so it is unconfirmed here. And on 24 November 2025 the agency's own news release approved an intrathecal onasemnogene product (Itvisma, onasemnogene abeparvovec-brve) for SMA from age 2 — one intrathecal dose that does not scale with body weight; the intravenous Zolgensma label still reads under 2 years.

Approval is also not availability. Beqvez (fidanacogene elaparvovec) was FDA-approved for hemophilia B in 2024; in February 2025 its manufacturer said it would stop global development for low demand, and on 15 May 2025 the European Commission withdrew the EU authorization at the company's request, for commercial reasons.

For the blood disorders, the word "infusion" hides the procedure. Casgevy, Lyfgenia and Zynteglo are autologous: your own stem cells are collected, modified outside the body, and returned after myeloablative busulfan conditioning. That is a transplant admission, not a clinic appointment.

How long has anyone been followed?

Between about one and about ten years, depending on the product.

Casgevy, sickle cell (CLIMB SCD-121, open-label single-group phase 3): ages 12–35 with at least two severe crises in each of the two prior years, cells CRISPR-edited at the erythroid enhancer of BCL11A. Forty-four were infused, median follow-up 19.3 months (range 0.8–48.1). The primary endpoint — no severe crises for at least 12 consecutive months — was met by 29 of 30 evaluable patients (97%; 95% CI 83–100).

Casgevy, transfusion-dependent beta-thalassemia: same design and ages, genotypes including β0/β0. Transfusion independence in 32 of 35 (91%; 95% CI 77–98), mean total hemoglobin 13.1 g/dL while independent, no deaths and no cancers in that paper.

Lyfgenia: in the pivotal cohort, 25 of 25 evaluable patients had resolution of severe vaso-occlusive events, from a median 3.5 per year in the preceding 24 months, with follow-up reaching 37.6 months. The manufacturer's trial page reports a larger set — 30 of 32 free of severe vaso-occlusive events, and 28 of 32 free of any vaso-occlusive event, between months 6 and 18 — sponsor material, not a peer-reviewed paper.

Zynteglo has the longest published follow-up here: 63 people (phase 1/2 n=22; phase 3 n=41), median 5.9 years, range 2.9–10.1. Transfusion independence was 15 of 22 (68.2%) in phase 1/2 and 37 of 41 (90.2%) in phase 3; once reached, it was still present at last follow-up, and that report recorded no malignancy, no insertional oncogenesis and no replication-competent lentivirus.

Hemgenix (HOPE-B, open-label phase 3): 54 men with factor IX of 2 IU/dL or less, each with at least six months of prophylaxis as a within-patient lead-in, then one infusion; AAV5 antibodies were not an exclusion. In the final five-year analysis the adjusted annualized bleeding rate over months 7–60 was 1.52 against 4.16 during the lead-in — a 63% reduction (95% CI 24–82) — with mean factor IX 36.1 ± 15.7 IU/dL at five years. Fifty of 54 completed five years. The reasons for the other four — two deaths, one liver transplant, one withdrawal of consent — come from a conference presentation of that same final analysis, not from the journal paper.

Roctavian is the cautionary curve. Factor VIII was high early (mean increase 41.9 IU/dL at weeks 49–52, 95% CI 34.1–49.7, in 132 participants), and at five years treated bleeds in the rollover population (n=112) were down 83.3%. But week-260 factor VIII in the modified intention-to-treat set (n=132) was median 6.2 IU/dL (mean 13.7; IQR 2.4–14.2), and 25 of 134 patients resumed prophylaxis at some point. At the two-year analysis the half-life of factor VIII production was modeled at 123 weeks (95% CI 84–232) — the falling curve, written as a number.

Zolgensma (STR1VE, completed US phase 3): 22 symptomatic infants. The coprimary endpoints were sitting unassisted for at least 30 seconds at the 18-month visit — 13 of 22 (59%; 97.5% CI 36–100) — and survival free of death or permanent ventilation at 14 months — 20 of 22 (91%) against 6 of 23 (26%) in an untreated historical cohort, not a randomized control group.

Zolgensma, treated before symptoms (SPR1NT): in infants with two copies of SMN2 treated at six weeks of age or younger, 14 of 14 sat unassisted for at least 30 seconds by 18 months, and all were alive without permanent ventilation at 14 months. That is a different population from STR1VE — children found before symptoms, not after — and the two are separate single-group trials, so the comparison between them is not a randomized one.

What is known about harm?

Lyfgenia carries a boxed warning for hematologic malignancy. The label documentation records two patients given an earlier manufacturing version of the product (Group A) who developed acute myeloid leukemia, and one Group C patient who developed myelodysplastic syndrome. The 2022 journal paper reported no blood cancers through 37.6 months — an earlier, shorter window; the label is the later safety document. The label also notes that people with α-thalassemia trait (−α3.7/−α3.7) may develop anemia with erythroid dysplasia requiring ongoing transfusions.

Zolgensma carries a boxed warning for acute liver failure with fatal outcomes, alongside serious liver injury and raised transaminases, with higher risk when the liver is already impaired. The label requires a steroid course around the infusion and liver testing for at least three months, and platelet count is part of required monitoring. The label says "fatal outcomes" without printing a count, so no count is given here.

Zynteglo's approval-history labeling carries an insertional-oncogenesis warning; no such cancer appeared in the long-term report above.

Casgevy, Lyfgenia and Zynteglo all require myeloablative busulfan, and the trial papers state the adverse-event pattern matches busulfan and autologous transplantation — neutropenia, thrombocytopenia, mucositis, febrile neutropenia. Conditioning has its own irreversible consequences, which belong in the pre-treatment conversation at the transplant centre.

What none of this shows

  • No randomized comparison against today's best alternative care. The pivotal evidence behind these licences is phase 3 and mostly single-arm: the sickle-cell and thalassemia trials are single-group; the SMA trial used a historical untreated cohort; the hemophilia trials compared each man with his own lead-in — stronger than no comparator, still not randomization.

  • A 12-month crisis-free endpoint is not a lifetime. Casgevy's published median is about 19 months, Zynteglo's 5.9 years with a maximum near 10. "One shot, settled forever" lies outside what has been measured.

  • Durability is not a class property. Factor IX was around 36 IU/dL five years after Hemgenix; factor VIII had a median of 6.2 IU/dL at week 260 after Roctavian — mild-hemophilia territory, not a normal clotting level.

  • Zolgensma's result stops at 14–18 months of age. It says nothing about walking, spine or breathing at school age, and nine of the 22 infants had not reached the sitting endpoint.

  • Eligibility is narrow by design. None of these products was studied as treatment for sickle-cell trait, non-transfusion-dependent thalassemia, mild hemophilia, or SMA past the age on the label.

What to do with this

Three checkable things are more useful at the treating centre than the word cure.

Whether the person meets the label, item by item — diagnosis, severity, age, and for the AAV products the anti-AAV5 antibody result, since Roctavian's label requires that test. "Not eligible" is information about a label, not a verdict about the disease; active trials and their enrolment criteria are public on ClinicalTrials.gov.

Which document your number came from. A trial paper, a regulator's label and a company release say different things about the same product; the Lyfgenia malignancy question is exactly that gap.

What the follow-up protocol is — liver tests and their duration for the AAV products, blood counts for the autologous ones, and the endpoint the specialist will watch: crisis-free intervals, transfusion independence, or factor activity in IU/dL. Those numbers describe the outcome in your own record, so they are worth keeping in one place over the years rather than scattered across discharge letters.

The decision belongs to you and a haematologist or neurologist who can see the whole record. This text is a map of what has been measured and of where the measurements stop.

Sources

Informational material. It does not replace a consultation with a doctor and is not a recommendation for or against treatment.

ShareTelegramWhatsAppLinkedIn

Articles in this section are educational and are not medical advice, a diagnosis, or a prescription. Consult a qualified professional before acting on anything you read here.

Read next