P-019 Correlation between blood flow and recurrences in coiled patients

Introduction Endovascular treatment with detachable coils has become an established technique for patients with intracranial aneurysms, the drawback being recurrence due to coil compaction or enlargement of a remnant. The contribution of hemodynamics to recurrence is at present unknown. The aim of this study was to explore the relationship between blood flow into the aneurysm and area of recurrence. We hypothesized that the location of the recurrence at first follow-up overlaps with the location of the inflow jet into the original untreated aneurysm. Materials and methods Out of 270 patients who received endovascular treatment for an intracranial aneurysm, 19 patients with a recurrence at first follow-up were included. Five patients with a recurrence were excluded because imaging was unsuitable for processing. Virtual surface models of the original untreated aneurysms and surrounding vessels were segmented with the Vascular Modeling ToolKit (VMTK) using three-dimensional imaging obtained before treatment. Recurrence areas were identified on 6 month follow-up three-dimensional time of flight MR angiography. Virtual surface models were also segmented of the lumen of the recurrences. Both surface models were matched using the Iterative Closest Point registration already available in VMTK (Abstract P-019 figure 1). After post-processing of the flow, it was possible to visualize the exact location of blood flowing into the aneurysm. This flow profile was matched with the location of the recurrence, with the untreated aneurysm neck as a reference frame. This match allowed for calculation of the overlap between area of recurrence and area of inflow (Abstract P-019 figure 2). Results The area of recurrence did not show a strong correlation with the initial inflow jet area in the aneurysm. The mean area of overlap of the inflow area and recurrence area as a percentage of the total recurrence area was only 35.45% (CI 22.83 to 48.07). Two cases did not show any overlap and none of the cases had a complete overlap. The range of overlap area was 0.00–96.35%. Conclusion Our model, construction using high quality geometry and CFD simulations using Fluent, were accurate. The results in this small sample do not support a relation between location of inflow in an untreated aneurysm and location of recurrence after treatment with coils.