Scientists create digital twins to advance Pompe disease treatment
Nexviazyme outperforms Lumizyme in simulated clinical trial
Scientists at Sanofi created digital counterparts of people with Pompe disease to support drug development and advance Pompe disease treatment.
In a simulated clinical trial of digital twins with infantile-onset Pompe disease (IOPD), Sanofi’s Nexviazyme (avalglucosidase alfa), an enzyme replacement therapy, outperformed its first-generation Lumizyme (alglucosidase alfa) in a head-to-head efficacy comparison, a study found. The study “highlights the value of applying a digital twin-based approach to support rare disease drug development,” the researchers wrote.
The digital counterparts, representing actual patients in disease history and severity, were developed and validated using treatment and response data from several Sanofi-sponsored clinical trials that enrolled Pompe patients. The study, “Quantitative Systems Pharmacology-Based Digital Twins Approach Supplements Clinical Trial Data for Enzyme Replacement Therapies in Pompe Disease,” was published in Clinical Pharmacology & Therapeutics.
Pompe is a rare genetic disorder caused by the deficiency of acid alpha-glucosidase (GAA), an enzyme that degrades glycogen, a complex sugar molecule. The resulting buildup of glycogen in body tissues, especially muscles, disrupts the function of affected cells and gives rise to a wide range of symptoms. IOPD is the most severe and rapidly progressive form of Pompe, marked by residual GAA activity and symptom onset within the first year of life with signs of heart muscle damage (cardiomyopathy). In the late-onset form (LOPD), symptoms generally emerge after age 1, and patients can have up to 40% of normal GAA activity.
Enzyme replacement therapy, or ERT, is the standard Pompe disease treatment, and provides patients with a lab-made version of the GAA enzyme. Approved ERTs include Lumizyme, sold as Myozyme outside the U.S., and the next-generation Nexviazyme, designed to deliver GAA to muscle cells more effectively than Lumizyme.
Generating digital twins
To generate the digital twins, the Sanofi-led team looked to quantitative systems pharmacology (QSP), a computer-based technique that applies mathematical models to characterize biological systems, disease processes, and drug pharmacology. It’s increasingly being used to help guide the development of new therapies and has been used by the U.S. Food and Drug Administration in a clinical pharmacology review.
The team developed a QSP model representing key elements of Pompe disease biology, including glycogen buildup in tissues and an increase in urinary Hex4, a biomarker for glycogen storage. They used the model to generate digital twins of each Pompe patient enrolled in their clinical program, accounting for demographics, disease burden, and treatment history. In simulations, one digital twin with IOPD was treated with Nexviazyme and the other with Lumizyme, and the levels of tissue glycogen and urine Hex4 were compared.
“The digital twins approach enables the supplementation of clinical trial data in rare indications like IOPD, which are often characterized by substantial [onset age and symptom] variability,” the researchers wrote. “Because clinical trials in these indications are necessarily small, this patient [variability] can impede the interpretation of the clinical response.”
In the first step of model development, the researchers successfully replicated glycogen metabolism in healthy individuals using publicly available data, considering body weight and height associated with age and sex.
To generate digital twins, the model was calibrated to each LOPD patient enrolled in COMET, a Phase 3 clinical study (NCT02782741) that compared the two ERTs. A single digital twin was developed for each patient by optimizing residual GAA activity based on the observed urine Hex4 before the study’s start.
The researchers validated the model using data from three other ERT clinical trials not used in model development: the COMET extension study, a Phase 3 trial (NCT00158600), and real-world data from the Pompe Registry.
The model was then extended to IOPD patients by lowering a GAA residual activity parameter to below 2%. Digital twins were created representing each IOPD patient enrolled in the Phase 2 Mini-COMET trial (NCT03019406), which compared the two ERTs in patients who didn’t fully respond to Lumizyme. This model was validated using data from the Phase 2/3 study called AGLU01602 (NCT00059280), which tested Lumizyme in treatment-naïve IOPD patients younger than 6 months.
The team simulated a head-to-head comparison of Lumizyme and Nexviazyme in Mini-COMET digital twins. Each digital twin was matched in age, sex, and body weight to the corresponding patient, and had an optimized parameter for residual GAA activity and treatment response based on urine Hex4 levels. Three treatment regimens were applied: Lumizyme alone, Lumizyme at a higher dose followed by a switch to Nexviazyme, and Nexviazyme alone.
The analysis predicted that treatment with Nexviazyme, either after switching or alone, resulted in more pronounced reductions in urine Hex4 compared with Lumizyme. It also indicated that the enhanced reduction in urine Hex4 following Nexviazyme was due to better glycogen clearance from skeletal muscles, or those attached to bones.
“The Pompe QSP model can be deployed in multiple stages of the R&D lifecycle,” the scientists wrote. “Other applications of the QSP model include the exploration of alternative drug modalities, combination with other treatments, and first-in-human dose selection for other novel drugs for Pompe disease.”
About the Author
