There is no single answer to how long gene therapy lasts. Duration depends on the product, the target cells, the disease, the outcome being measured and the evidence collected so far. A one-time dose describes how treatment is given; it does not establish lifelong benefit. The most useful question is: “For this therapy, in patients like these, which benefits have been maintained, for how long, and with what uncertainty?”
This guide is for readers comparing long-term studies, not for choosing a treatment without specialist advice. It separates clinical benefit, biological activity and safety monitoring, which are often combined in headlines even though they answer different questions.
Define what “still working” means
| Measure | What it may show | What remains unanswered |
|---|---|---|
| Biological activity | A protein, enzyme or other marker remains detectable or improved. | Whether the change produces a meaningful long-term benefit. |
| Clinical function | A motor milestone, symptom measure or organ outcome is maintained. | How the same person would have fared without treatment. |
| Major events | Survival, hospitalisations or another prespecified event over time. | How much later risk remains beyond observed follow-up. |
| Safety follow-up | Observed adverse events and potential delayed harms. | Proof that all rare or very late harms have been excluded. |
Read the number still being followed
Imagine a fictional study that treated 20 people. At year one, 20 have assessments; at year three, 16 do; at year five, 12 do. A statement that “10 of 12 retained a response at year five” is not the same as demonstrating that 10 of all 20 treated participants retained it. The eight missing assessments may reflect recent enrolment, withdrawal, illness, death or something else. Each explanation has different implications.
Do not automatically count missing participants as successes or failures. Read the participant flow diagram, reasons for missing data and the prespecified analysis. Ask whether the result is robust to plausible alternative assumptions.
One real example: the START follow-up in SMA
A 2021 report followed 13 of the original 15 START participants after onasemnogene abeparvovec. Median time since dosing was 5.2 years overall. The ten participants in the therapeutic-dose group maintained previously acquired motor milestones and remained alive without permanent ventilation at the reported cutoff. This is evidence of sustained benefit in that cohort, not proof of lifelong benefit for every patient.
Read the dose groups and additional-treatment histories carefully. Some participants received another SMA therapy during follow-up. The study was a small observational extension, not a randomised comparison of lifelong outcomes. This example explains how to read a follow-up paper; it is not presented as the latest complete evidence review for that product.
Gene replacement versus repeat-dosing approaches
“One time” and “repeated” describe dosing patterns, not a universal ranking of effectiveness. Gene addition, genome editing, RNA-based treatments and protein replacement act through different mechanisms. A repeated treatment can maintain benefit while dosing continues; a single administration can have a durable effect without proving that the effect will never diminish. Some products cannot simply be given again using the same approach.
A fair comparison needs a defined disease, population, outcome and follow-up period. Also compare treatment burden, safety, access and the consequences of stopping or changing treatment. Do not compare a short trial in mildly affected adults with a long-term cohort of severely affected infants and attribute all differences to the dosing schedule.
A practical durability worksheet
- Name the intervention precisely. Product, mechanism, dose and whether other treatments continued.
- Describe the patients. Age, genotype, disease stage, baseline function and prior treatment.
- Choose the outcome. A biomarker, a functional change, an event or a patient-reported measure?
- Record the time window. Median follow-up and range, not only the longest-followed participant.
- Count evaluable participants. How many contribute at each time point, and why are data missing?
- Check the comparison. Randomised control, natural history, within-person baseline or no comparator?
- Check additional treatment. Rescue therapy or a second disease-modifying treatment may affect interpretation.
- State a bounded conclusion. Say what was sustained over the observed period; do not extrapolate to a lifetime.
What monitoring schedule is appropriate?
There is no universal monitoring timetable for all gene therapies. Product-specific guidance, the trial protocol and the treating centre determine the plan. FDA guidance describes risk-based long-term follow-up for delayed adverse events; extended surveillance is not identical to routine care or proof of continued efficacy.
Ask the team which visits assess immediate safety, long-term risks, disease progression and treatment response. Keep a record of the product, treatment date, centre and follow-up contact. New or concerning symptoms should be discussed with the treating team rather than waiting for a research visit.
How to write a conclusion that matches the evidence
Instead of “Gene therapy lasts forever,” try: “In this cohort, the specified outcome was maintained through the reported follow-up; later outcomes and applicability to other patients remain uncertain.” Instead of “The therapy stopped working because one biomarker fell,” ask whether the change was clinically meaningful and whether functional outcomes changed too.
For the rest of a trial paper, use How to read gene-therapy trial results. For SMA-specific interpretation, read SMN2 copy number and prognosis. The research tools directory can help locate the registered study and its original publications.
Sources and further reading
- ClinicalTrials.gov: how to read study results.
- Mendell and colleagues: Five-Year Extension Results of the Phase 1 START Trial (2021).
- FDA: Long Term Follow-up After Administration of Human Gene Therapy Products (2020).
- NHLBI: discussion of gene therapies for sickle cell disease and remaining long-term questions (2023).