Pharmacological Research · Nov 6, 2026

Patient-anchored pharmacological modeling prioritizes combined miR-4651/miR-6126 for experimental validation in Hutchinson–Gilford progeria syndrome

By Joo EY, Park JS, Kim SJ et al.

Hutchinson–Gilford progeria syndrome (HGPS) is characterized by accelerated vascular aging, but therapeutic development is constrained by sparse longitudinal clinical data and the limited number of independent biological models. We developed a patient-anchored framework integrating longitudinal cardiovascular phenotyping, endothelial transcriptomics, LMNA cryptic splicing, biological-line–preserving virtual populations, treatment counterfactuals and lipid nanoparticle (LNP) delivery optimization. The index patient showed progressive carotid disease and worsening diastolic indices despite preserved systolic function. In HGPS endothelial cells, disease-associated transcriptional structure persisted after adjustment for cell-cycle and flow-related effects across 20,404 non-cell-cycle genes (Spearman ρ = 0.706), with inflammatory activation remaining directionally positive in all five leave-one-line-out refits. Hierarchical modeling generated 2,000 virtual profiles per condition for uncertainty propagation without increasing biological sample size or statistical power. Under the prespecified treatment operators, patient-informed counterfactual simulations ranked combined miR-4651/miR-6126 above either single agent across 6-, 12- and 24-month horizons; however, a separate uncalibrated structural sensitivity model showed that dual-treatment superiority was not robust to matched total miRNA cargo and hypothetical competitive or antagonistic interactions. A four-dimensional computational LNP optimization screen evaluated 18,072 regimens, retained 2,567 Pareto-nondominated solutions and narrowed these to 30 model-prioritized delivery configurations for experimental evaluation. Together, these analyses connect observed HGPS vascular pathology with biologically constrained therapeutic prioritization, identify combined miR-4651/miR-6126 as a model-prioritized candidate requiring matched-dose experimental validation, rather than an intervention with established superiority, and define a model-prioritized LNP delivery space for HGPS vascular disease.

doi.org/10.1016/j.phrs.2026.109745Open accessPubMed 42994541Interventions