TL;DR
This study uses computational fluid dynamics (CFD) on coronary CT angiography (CCTA) scans to predict how coronary artery plaques change over time. They found that specific hemodynamic markers, like wall shear stress and helicity, can predict whether a plaque will progress or regress, offering a non-invasive way to stratify risk beyond just measuring stenosis severity.
What Problem Does This Paper Address?
Coronary artery disease (CAD) is a major health concern, and understanding how plaques within coronary arteries change (progress or regress) is crucial for predicting cardiovascular events and guiding treatment. Current methods for assessing plaque morphology, like intravascular ultrasound (IVUS) and optical coherence tomography (OCT), are invasive, costly, and not widely accessible. Coronary CT angiography (CCTA) is a non-invasive imaging technique that visualizes plaques, but its ability to predict dynamic changes has been limited. While computational fluid dynamics (CFD) can model blood flow and its interaction with plaques, longitudinal studies linking baseline hemodynamic parameters to subsequent plaque changes are scarce. This paper aims to bridge this gap by using CCTA-derived CFD to identify hemodynamic predictors of plaque progression and regression.
Methods
The researchers employed a retrospective analysis of 22 patients with CAD who underwent two CCTA scans over an average interval of two years.
- CCTA Image Acquisition and Processing: Standard CCTA scans were performed, and images were processed on a Siemens workstation.
- Plaque Measurement: Plaque characteristics such as stenosis degree, lumen area, plaque volume, and remodeling index were quantified using dedicated software. Lesions were categorized into progression (≥5% diameter stenosis increase), stable (-5% to 5% change), or regression (≥5% diameter stenosis decrease).
- CFD Model Creation:
- Anatomical Model Reconstruction: Three-dimensional models of the coronary arteries and left ventricle were reconstructed from CCTA data using algorithms like fast marching and level-set methods.
- Meshing: The reconstructed models were meshed into millions of elements to balance accuracy and computational efficiency.
- Boundary Conditions: Patient-specific inlet flow conditions (based on blood pressure and heart rate) and outlet pressure conditions were applied. Hyperemic flow conditions were simulated for CT-FFR calculations, while resting state conditions were used for other hemodynamic parameters.
- Hemodynamic Calculation: The Navier-Stokes equations were solved using the finite volume method (OpenFOAM) to simulate blood flow.
- Hemodynamic Parameter Extraction and Normalization: Key hemodynamic indices, including wall shear stress (WSS), wall shear stress gradient (WSSG), oscillatory shear index (OSI), relative residence time (RRT), gradient oscillatory number (GON), and vorticity, were extracted. Crucially, these parameters were normalized to values from adjacent healthy reference segments to account for inter-patient variability.
- Statistical Analysis: Logistic regression was used to identify independent hemodynamic predictors of plaque progression and regression. Receiver operating characteristic (ROC) curve analysis was performed to assess the predictive performance of these models.
Datasets
The study utilized retrospective CCTA data from 22 patients with coronary artery disease. These patients underwent two CCTA examinations with a median interval of 22 months. No new datasets were introduced; the study focused on analyzing existing patient data.
Key Insights and Ideas
- Hemodynamics as a Key Predictor: The study strongly suggests that hemodynamic forces, beyond just the degree of arterial narrowing (stenosis), play a significant role in determining plaque fate.
- Normalization is Crucial: Normalizing hemodynamic parameters to healthy segments is vital for reducing inter-patient variability and isolating the localized effects of plaque on blood flow.
- Distinct Predictors for Progression vs. Regression: Different hemodynamic parameters are associated with plaque progression versus regression, highlighting the complex interplay of forces involved.
- Helical Flow and Progression: Higher maximum helicity (indicating more swirling, helical flow) was linked to an increased risk of plaque progression. This suggests that complex flow patterns can stress the vessel wall and promote inflammation.
- Low Shear Stress and Progression: Lower normalized minimum time-averaged wall shear stress (Min.TAWSS) was found to be protective against plaque progression. This aligns with the understanding that low WSS regions are prone to lipid accumulation and inflammation.
- Vorticity and GON for Regression: Lower average gradient oscillatory number (Ave.GON) and average vorticity were associated with plaque regression. This implies that smoother, less disturbed flow conditions may facilitate plaque regression.
- Non-invasive Risk Stratification: The identified hemodynamic markers offer a promising avenue for non-invasive risk stratification, potentially identifying high-risk plaques that might be missed by traditional stenosis measurements alone.
Key Results
- Progression Predictors:
- Normalized minimum time-averaged wall shear stress (Min.TAWSS) was independently protective against progression (OR=0.38, p<0.001).
- Normalized maximum helicity was an independent predictor of progression (OR=1.44, p=0.016).
- Regression Predictors:
- Normalized average gradient oscillatory number (Ave.GON) was an independent predictor of regression (OR=0.10, p=0.001).
- Normalized average vorticity was an independent predictor of regression (OR=0.13, p=0.019).
- Predictive Performance:
- The multivariable model for plaque progression achieved an AUC of 0.78.
- The multivariable model for plaque regression achieved an AUC of 0.83.
- These models demonstrated good discrimination with reasonable sensitivity and specificity.
Limitations and Future Work
The authors acknowledge several limitations:
- Model Simplifications: The study used a Newtonian, incompressible flow model with rigid walls. Transitional turbulence, which can occur in severe stenoses or recirculation zones, was not fully captured and could influence hemodynamic metrics.
- Medication and Lifestyle: The study did not deeply explore the effects of medication adjustments (e.g., statins) or lifestyle changes on plaque progression, nor did it analyze potential adverse cardiac events.
- Small Sample Size: The study involved a relatively small sample size (22 patients, 34 lesions), limiting the generalizability of findings.
- Observational Nature: The study is observational, and its conclusions are considered preliminary hypotheses requiring further validation.
Future work suggested includes:
- Incorporating more complex flow models (e.g., non-Newtonian, transitional turbulence) to better capture real-world hemodynamics.
- Conducting larger-scale studies with stratification by treatment intensity and including adverse cardiac event analysis.
- Investigating the impact of medication adjustments and lifestyle changes on plaque dynamics.
- Performing larger-scale observational studies to validate these preliminary findings.
Summary updated