Predicting Long-Term Durability of a Gene Therapy 

AAV gene therapies rely on stable episomes in target cells to sustain therapeutic protein production. However, in self-renewing tissues such as the liver, where cells continually proliferate and die, the long-term durability of these therapies remains uncertain. In the face of limited long-term data and biological uncertainties, CSL Behring called on MetrumRG’s Quantitative Systems Pharmacology (QSP) expertise to address the question of  durability of transgene expression for a hemophilia B gene therapy.

MetrumRG developed a sophisticated ABM to simulate the dynamics of transgene expression in liver tissue following  treatment with etranacogene dezaparvovec, a liver-targeted AAV-based gene therapy. By incorporating detailed physiological parameters—such as cell turnover, patient age, and transduction efficiency—the ABM accounted for the complex interplay between biological variables and therapeutic outcomes. Virtual populations were constructed, capturing population variability and biological uncertainties, and generating predictions of transgene expression durability for up to 20 years post-treatment.

It was projected that long term durability of response would be achieved.  The level of factor IX activity achieved was stable and sufficient to avoid additional therapy over twenty years. Liver cell turnover was identified as a critical factor determining long-term transgene expression. This highlighted the importance of accounting for tissue physiology in gene therapy design. A clear understanding of durability is essential to evaluating the gene therapy's long-term risk-benefit profile, especially given that patients typically receive only a single dose.

This MetrumRG-CSL Behring collaboration provided quantitative insights into the durability of an AAV-based gene therapy. By predicting the long-term stability of transgene expression, the ABM framework empowered stakeholders to understand the implications of biological and product characteristics on therapeutic outcomes. This computational framework is adaptable to other AAV-based gene therapies, broadening its potential value across diverse therapeutic areas.