This study develops a physics-based, multiscale pulmonary arterial growth and remodeling (G&R) framework to connect mechanistic drivers of pulmonary hypertension (PH) with longitudinal changes in clinically relevant metrics. The framework integrates three coupled domains: a morphometric arterial tree for hemodynamics, a constrained mixture theory representation of vessel wall mechanics, and a cellular-scale description of maladaptive remodeling. By combining these elements, the model simulates evolving functional outputs—pulmonary pressure, vessel wall thickness, and wall stiffness—while preserving mechanistic interpretability.
Model parameters were calibrated to previously collected longitudinal measurements from the monocrotaline (MCT) rat model. The calibration targeted three observed quantities: pulmonary pressure, arterial wall thickness, and wall stiffness. A multiobjective optimization approach was applied to fit model parameters to the study-specific MCT dataset, enabling simultaneous reproduction of multiple remodeling features over time.
Disease progression within the G&R framework was driven by three mechanistically interpretable parameters: (1) excess smooth muscle production, (2) remodeling activation, and (3) passive stiffening. Each parameter maps to a biological or structural process that contributes to vascular remodeling. The modular parameterization allows the framework to attribute longitudinal changes in pressure, thickness, and stiffness to specific remodeling mechanisms and to adjust those mechanisms when simulating interventions.
To demonstrate predictive capability, the authors simulated therapeutic intervention within the same disease-specific framework. Rather than fitting separate treatment models empirically, the simulation used functional cell-level responses to therapy to inform changes in the mechanistic parameters. This approach preserved mechanistic consistency between disease evolution and treatment response and allowed direct comparison between predicted and experimentally observed treatment effects on pressure, wall thickness, and stiffness.
Calibration to the MCT dataset reproduced temporal increases in the three target quantities. Goodness-of-fit metrics reported in the source indicate R2 values of 0.81 for pressure, 0.83 for wall thickness, and 0.95 for stiffness. Simulated treatment lowered predicted pressure, wall thickness, and stiffness. The model’s predicted reductions in pressure and wall thickness aligned closely with the experimental treatment effects reported in the source dataset. However, stiffness recovery after simulated therapy was overpredicted relative to experimental observations, indicating a discrepancy between modeled and observed post-intervention mechanics.
Beyond the study-specific MCT data, the framework captured the overall progression of pulmonary pressure increases when applied to aggregated datasets that included both MCT and Sugen–hypoxia animal models. This suggests the platform has utility across different experimental PH phenotypes and can be used to compare remodeling trajectories between models within the same mechanistic modeling paradigm.
The authors note that the model overpredicts recovery of wall stiffness after intervention, implying that additional biological or structural mechanisms—currently not represented by the constitutive forms used—may contribute to persistent vascular stiffening. The source indicates that refinements to constitutive relations are needed to improve predictive capability across phenotypes and to better capture persistent post-treatment stiffness.
The presented G&R framework provides a mechanistically interpretable platform that links cellular and tissue remodeling mechanisms to measurable functional outcomes in PH. By calibrating to longitudinal preclinical data and by simulating treatment effects via parameter adjustments grounded in cell-level responses, the model supports cross-phenotype comparisons and hypothesis testing about drivers of progression and incomplete recovery. The framework highlights where additional mechanistic detail is required to match clinical or experimental post-intervention stiffness trajectories, informing future model development and experimental design.
In summary, the work outlines a multiscale, physics-based model that reproduces longitudinal increases in pulmonary pressure, wall thickness, and stiffness in the monocrotaline rat model and that can simulate therapeutic effects using mechanistic parameter changes. Goodness-of-fit to MCT data was high for the three target metrics (R2 = 0.81, 0.83, and 0.95). While pressure and thickness responses to simulated therapy matched experimental effects closely, stiffness recovery was overpredicted, indicating the need for additional constitutive or mechanistic components to account for persistent stiffening after intervention. The framework is presented as a tool for comparing experimental phenotypes and interventions and for guiding where model refinements are necessary to enhance predictive performance.