QSP-based virtual twin approach provides mechanistic insight into enzyme replacement therapies in Pompe Disease

Objectives: A QSP model was developed to better understand the systemic response to avalglucosidase alfa in patients with Pompe disease, a rare disease caused by deficient lysosomal glycogen degradation. This model describes both late-onset (LOPD) and severe infantile-onset (IOPD) phenotypes within a single mechanistic framework. To supplement data from the Mini-COMET (NCT03019406) study comparing avalglucosidase alfa vs. standard-of-care treatment in IOPD patients, a QSP-based virtual twin analysis was performed to account for the low patient number and high phenotypic heterogeneity.Methods: Virtual twins, calibrations of a QSP model to the data of individual patients in clinical studies, consider the demographics, disease burden, and treatment history of each patient to recapitulate their observed disease presentation at baseline and treatment response. By developing virtual twins to all patients in Mini-COMET while leveraging tissue drug exposures predicted by a PBPK model [1], the efficacy of different drug regimens could be directly compared in terms of both the biomarker urine Hex4, which is measured in clinical studies, and also the glycogen burden in skeletal muscle, which was not measured in Mini-COMET. To evaluate its suitability for this application, the model was validated using independent clinical observations of skeletal muscle glycogen and urine Hex4 in IOPD patients following standard-of-care treatment. Validation employed an expanded virtual population of IOPD patients developed from the Mini-COMET virtual twins. Additional virtual twin development and another model validation were performed using data recorded in the ongoing Pompe Registry (NCT00231400) [2].Results: Virtual twins of IOPD patients enrolled in Mini-COMET adequately described observed individual biomarker data. Simulating different drug regimens using the same virtual twin cohort enabled direct comparison of efficacy and compensated for the observed baseline patient heterogeneity in Mini-COMET. This analysis supplements clinical observations by providing mechanistic insight, indicating that the observed enhanced biomarker reduction following avalglucosidase alfa treatment is driven by greater glycogen clearance, and suggesting further glycogen debulking following avalglucosidase alfa treatment even in standard-of-care-treated patients with stable biomarker levels. Conclusions: Drug development efforts in rare indications like Pompe disease often face challenges related to small patient numbers in clinical studies and large heterogeneity within disease phenotypes. The virtual twins approach leverages individual patient data to understand disease heterogeneity and treatment response in IOPD patients, and to supplement clinical observations. Citations: [1] Yau et al. (2024) Manuscript under preparation[2] Mistry et al. (2022) Orphanet J Rare Dis 17, 362