P080 Numerical simulation of individual coil placement – A proof–of–concept study for the prediction of recurrence after aneurysm coiling

Julian Schwarting, Fabian Holzberger, Markus Muhr, Martin Renz, Tobias Greve, Tobias Boeckh‐Behrens, Barbara Wohlmuth, Jan S. KirschkeAbstracts 2024 · DOI

Introduction

Brain aneurysms—pathological bulges in the cerebral vasculature—pose a severe risk of rupture and life-threatening subarachnoid hemorrhage. The current clinical standard for treatment is endovascular coiling, a minimally invasive procedure where a neuroradiologist fills the aneurysm sac with platinum coils to induce thrombosis and exclude the lesion from blood circulation.

Despite its efficacy, the success of coiling is often dependent on the clinician's intuition and the specific geometry of the aneurysm. A significant clinical challenge remains: recurrence. If the coils shift or the initial packing is insufficient, the aneurysm can regrow, necessitating complex follow-up interventions. This study presents a high-fidelity numerical simulation framework designed to model individual coil behavior, offering a potential tool for predicting long-term procedural success.

The Limitations of Current Modeling

Traditional computational models often treat the coil mass as a homogeneous "porous blob." While useful for general fluid dynamics, these models fail to capture the mechanical reality of the procedure. In practice, the specific mechanical properties of the wire—its diameter, primary helix radius, and macroscopic shape—dictate how it interacts with the vessel wall and subsequent coils. Furthermore, existing clinical software like AngioSuite often estimates packing density based on 2D imaging, which may not accurately reflect the 3D distribution of the coil mass.

Methodology: High-Fidelity Mechanical Simulation

To bridge this gap, researchers developed a patient-specific simulation pipeline using 3D rotational angiography images from six patients (treated between 2013 and 2016). The study focused on aneurysms in the anterior communicating artery (ACOM), middle cerebral artery (MCA), and internal carotid artery (ICA).

1. Patient-Specific Modeling

High-resolution 3D digital models were generated using ITK SNAP. These models served as the environment for the virtual coiling procedure, accounting for the unique anatomical constraints of each patient.

2. Mechanical Coil Properties

Rather than a simplified volume, the simulation modeled the specific properties of the Stryker Target Helical Ultra coil. This included the wire's physical dimensions and its tendency to seek its pre-formed shape upon deployment.

3. The "Catheter Wobble" Factor

Recognizing that sub-millimeter precision is impossible in a clinical setting, the researchers introduced a sensitivity analysis. They simulated 100 coiling attempts for each aneurysm, randomly shifting the microcatheter tip within a 0.5 mm radius. This allowed the team to measure how sensitive the final coil configuration was to minor variations in placement—a metric they termed "procedural difficulty."

Key Insights and Results

Sensitivity and Stability

The most significant finding was the high variability in coil shapes produced by tiny changes in catheter position. In some patients, the coil successfully framed the aneurysm in 100% of the simulations; in others, success dropped to as low as 16%. This suggests that certain aneurysm geometries are inherently "unstable," making them more difficult to treat effectively in a real-world environment.

Packing Density Discrepancies

The simulated packing densities ranged from 15.82% to 24.02%. Notably, these figures did not correlate with the densities calculated by standard clinical software (AngioSuite). Because the simulation accounts for the 3D physics of the wire rather than 2D projections, the authors suggest their model may provide a more accurate representation of the actual physical packing within the sac.

Predicting Recurrence

While the study utilized a small cohort (n=6), a clear trend emerged between the simulation data and 6-month clinical outcomes. The three patients who experienced recurrence showed:

  • Lower simulated packing densities.
  • Higher "extraaneurysmal volume" (instances where the coil protruded into the parent artery).
  • Lower success rates across the 100 simulated attempts.

This implies that the "ease" of achieving a stable coil frame in simulation could serve as a biomarker for the risk of long-term recurrence.

Limitations and Future Directions

As a proof-of-concept, this study is limited by its small sample size, which prevented the results from reaching statistical significance ($p > 0.05$). Additionally, the model currently assumes rigid vessel walls and only simulates the first "framing" coil, whereas clinical practice involves multiple coils.

Future research will focus on:

  • Elastic Vessel Walls: Incorporating the natural pulsatile flow and flexibility of brain arteries.
  • Multi-Coil Simulations: Modeling the entire sequential coiling process.
  • Large-Scale Validation: Testing the framework on larger patient cohorts to refine its predictive accuracy.

By quantifying the mechanical stability of a planned procedure, this numerical approach moves endovascular planning away from subjective intuition toward a data-driven, patient-specific science.

Summary updated

Abstract

<h3>Introduction</h3> Coiling is the most used method for interventional aneurysm occlusion. The choice of the specific coil has an important influence on secondary regrowth, requiring retreatment, and is made based on expert opinions. Individual simulations of the coiling procedure could provide additional information and save costs for rejected coils.. <h3>Aim of Study</h3> To simulate coiling in different aneurysm geometries and to compare simulation results with clinical recurrence. <h3>Methods</h3> In a proof-of-principle study, we simulated coiling in 6 anterior circulation cases treated with a single coil, out of which 50% had a regrowth, defined as Raymond &amp; Roy class 2 or 3, observed in a follow–up angiography after 6 months using a newly developed numerical approach to simulate coil shapes. We then correlated simulation results with aneurysm recurrence. <h3>Results</h3> The simulation showed various coil shapes depending on the variability in possible microcatheter positions. Aneurysms with a later recurrence showed a tendency for more frequent successful coiling attempts. Further trends suggested lower simulated packing densities in aneurysms with reoccurrence. Simulated packing densities did not correlate with those calculated by conventional software, indicating the potential for our approach to offer additional predictive value. <h3>Conclusion</h3> Our study pioneers a comprehensive numerical model for the simulation of aneurysm coiling, providing insights into individualized treatment strategies and outcome prediction. This simulation framework holds promise for enhancing clinical decision-making and optimizing patient outcomes in endovascular aneurysm treatment after refinement in the future. <h3>Disclosure of Interest</h3> no.
BibTeX
@inproceedings{schwarting2024p,
  author = {Julian Schwarting and Fabian Holzberger and Markus Muhr and Martin Renz and Tobias Greve and Tobias Boeckh‐Behrens and Barbara Wohlmuth and Jan S. Kirschke},
  title = {{P080 Numerical simulation of individual coil placement – A proof–of–concept study for the prediction of recurrence after aneurysm coiling}},
  booktitle = {Abstracts},
  year = {2024},
  doi = {10.1136/jnis-2024-esmint.116},
}
Figure