Gene Therapy Transduction Threshold
Definition
The transduction threshold is the minimum proportion of cardiomyocytes that must be corrected by a gene therapy vector before the arrhythmic phenotype is abolished. No cardiac vector transduces every cardiomyocyte; the question that determines whether a therapy works is not how much of the heart is corrected but how much correction is enough.
For arrhythmia — as distinct from structural repair — the answer turns out to be surprisingly low: roughly 30–50% of cells, because the myocardium is an electrical syncytium in which corrected cells actively suppress the misbehaviour of their uncorrected neighbours. The same property that makes this threshold low also makes the therapy's safety profile depend on where, not just how many, cells are corrected. (sources/gene-therapy-arrhythmia-cjca-2026, rating: high)
Key Concepts
- Arrhythmias need a trigger and a substrate, and transduction affects the two differently. A trigger is an ectopic action potential from a delayed or early afterdepolarisation (DAD/EAD); a substrate is slow conduction plus unidirectional block capable of sustaining reentry. Correcting cells suppresses the trigger but, if the correction is patchy and alters action potential duration, it can create the substrate.
- The threshold is a property of the tissue, not the vector. It is set by gap junction coupling, cell geometry, and the presence of fibrosis — which means the same vector dose can be sufficient in a channelopathy and insufficient in a cardiomyopathy.
- It is the strongest argument for treating early. Established fibrosis raises the number of cells that must be corrected and independently favours reentry, so a therapy that works in a young mouse may not work in an adult human with a scarred ventricle.
The source–sink problem
- Each cardiomyocyte is electrically coupled to roughly 11 neighbours through gap junctions. When one cell begins to accumulate depolarising charge, current flows into adjacent cells and dilutes the stimulus — the "sink" absorbs the "source".
- Computational modelling estimates that 70–80 contiguous cells must simultaneously reach a suprathreshold DAD or EAD to initiate a propagating action potential in a 1-dimensional strand. In 2-D tissue the requirement rises to ~7,000–8,000 cells and in 3-D to ~700,000–800,000 — which is why random triggered activity almost never propagates in healthy myocardium. (sources/gene-therapy-arrhythmia-cjca-2026, rating: high)
- In electrically or structurally remodelled tissue the threshold falls sharply, which is exactly what allows a single premature beat to start an arrhythmia in a diseased heart.
The stabiliser-cell threshold
Modelling replaces "corrected cells" with "stabiliser cells" — cells rendered incapable of spontaneous diastolic calcium release, as gene therapy for CPVT would do:
| Tissue geometry | Stabilisers required |
|---|---|
| 1-D cable | 30–50%, randomly distributed |
| 2-D and 3-D | 30–40% |
| 3 × 3 or 5 × 5 clusters (rather than dispersed) | Higher, though still <50% in 3-D |
| Reduced gap junction conductance or simulated fibrosis | Raised further |
- The same threshold held when stabilisers were modelled by enhancing IK1 (Kir2.1) rather than by abolishing calcium leak — suggesting the principle generalises beyond CPVT.
What the animals show
- CPVT: transduction rates of 40–66% were sufficient to completely suppress adrenergically induced DAD-mediated ventricular arrhythmia across several murine models. In CASQ2 p.D307H mice, rates below 33% still prevented sustained VT, though non-sustained episodes persisted. (sources/gene-therapy-arrhythmia-cjca-2026, rating: high)
- LQT3: base editing that reduced mutant SCN5A transcripts by only 20% completely prevented carbachol-induced VT, even though the QT interval did not normalise. A partial biochemical correction produced a complete antiarrhythmic one.
- LQT1: a Sup-Rep construct achieving only ~30% reduction of mutant KCNQ1 mRNA normalised QT interval and QT dispersion in transgenic rabbits at baseline and after isoproterenol.
- SQT1: Sup-Rep corrected IKr behaviour in 47% of ventricular cardiomyocytes and substantially recovered QTc and apicobasal APD₉₀ heterogeneity; modelling suggested 50% correction would drastically reduce reentry susceptibility.
- LQT2: in silico modelling predicted that suppressing EADs would require editing only ~10% of cardiomyocytes in 3-D tissue — the lowest estimate in the literature, and a reminder that these thresholds are model-dependent.
The proarrhythmic counter-effect
- Where the therapy directly modulates action potential duration, non-uniform correction increases APD dispersion, which is itself a reentry substrate. Enhancing IK1 in the simulations produced modest APD dispersion; the effect became pronounced in EAD-prone tissue when stabilisers were clustered in 5 × 5 blocks or when a central region was left untreated.
- Subthreshold DADs and EADs that never trigger a premature beat can still create localised conduction block and regional dispersion of refractoriness — enough to facilitate reentry without producing the trigger everyone was measuring.
- The practical implication is that dispersed transduction is safer than clustered transduction at the same overall efficiency, and that "how many cells" is an incomplete way to report a preclinical result.
Is the arrhythmia threshold different from the structural one?
Yes, and by a wide margin.
| Disease class | Correction required | Why |
|---|---|---|
| Inherited arrhythmia (channelopathy) | ~20–50% | Electrophysiological benefit spreads electronotically to neighbouring uncorrected cells — a population-level emergent property |
| Structural cardiomyopathy (DMD, ARVC) | >70% | Each cardiomyocyte must be repaired individually to restore mechanical integrity; sub-threshold mosaicism has caused myofibre necrosis and fibrosis |
The two figures are not in conflict; they measure different things. But the structural threshold has a further consequence for arrhythmic endpoints, because fibrosis and reduced coupling raise the stabiliser requirement — so in cardiomyopathy the two thresholds move in opposite directions. (sources/gene-editing-cv-tcm-2025, rating: medium; sources/aav-gene-therapy-arrhythmia-hr-2024, rating: medium)
Contradictions / Open Questions
- The thresholds rest on models, not on measured human transduction. Several of the preclinical studies that "confirm" a 30–50% threshold inferred transduction from the degree of mutant transcript silencing rather than counting transduced cells, and in the LQT1 rabbit work the exact fraction of transduced cardiomyocytes was not determined. The numbers may be right, but they are not yet measured. (sources/gene-therapy-arrhythmia-cjca-2026, rating: high)
- Suppressing triggers and avoiding reentry may not be simultaneously achievable. All the favourable evidence concerns the trigger; the models that warn about APD dispersion concern the substrate. No preclinical study has been designed to test both endpoints in the same animal with histology to confirm vector distribution — so "safe at 30% transduction" remains an inference.
- Fibrosis may be reversible, which would change the threshold over time. It is unknown whether a gene therapy that addresses the underlying pathophysiology can regress established fibrosis. If it can, late intervention becomes viable; if it cannot, the early-intervention argument becomes a hard constraint rather than a preference.
- Clustered versus dispersed transduction is not controllable in practice. The recommendation that stabilisers be dispersed is not something a systemic AAV can be programmed to do — capsid tropism, dose and coronary anatomy determine distribution. A threshold derived assuming random distribution may therefore be unreachable in a scarred human ventricle.
- Whether the same numbers apply to the human heart is untested. Human cardiomyocytes are larger and the ventricular wall far thicker than in a mouse; whether source–sink buffering scales with cell number, tissue volume or wall thickness has not been established. The authors call for large-animal models precisely because of this.
- Paediatric growth may erode any threshold achieved. Episomal AAV is diluted as cardiomyocytes divide in infancy, so a child treated at 30% transduction may fall below the threshold over years — and re-dosing is blocked by the immunity the first dose induces.
Connections
-
Part of concepts/Gene-Therapy-in-Arrhythmia — the parent overview
-
The review that assembled it: sources/gene-therapy-arrhythmia-cjca-2026 — source–sink modelling, stabiliser thresholds, per-disease transduction data
-
How the vector gets there: concepts/AAV-Gene-Delivery — cargo limits, routes, and the mosaicism problem this threshold quantifies
-
The silencing strategies that rely on it: concepts/Gene-Silencing-Therapy · concepts/SupRep-Therapy — partial knockdown is sufficient precisely because of this threshold
-
The editing strategies that rely on it: concepts/CRISPR-Cas9-in-Channelopathies · concepts/Gene-Editing-Risk-Benefit-Framework
-
The electrical principle underneath: concepts/Cardiac-Refractoriness · concepts/Cardiac-Action-Potential — dispersion of refractoriness and the source–sink relationship
-
Where the threshold is highest: concepts/Arrhythmogenic-Cardiomyopathy · entities/ARVC — fibrosis and reduced coupling both raise it
-
The diseases tested against it: entities/CPVT · entities/Long-QT-Syndrome · entities/Brugada-Syndrome · entities/PKP2