Patient population
This retrospective study was approved by the Institutional Review Board of Satoh Neurosurgical Hospital, and the protocols used in the study were approved by the Committee of Human Subjects Protection of the Satoh Neurosurgical Hospital, Hiroshima, Japan. All methods were performed in accordance with the relevant guidelines and regulations. Informed consent was obtained from all patients to use their clinical data. Forty-nine UIAs from 49 consecutive patients between September 2017 and August 2019 were analyzed. Four patients were excluded due to poor imaging, resulting in a total of 45 aneurysms enrolled in this study (13 men and 32 women). The mean age of patients was 65.07 ± 9.57 years (mean ± SD), with a range of 39 to 78 years. The locations of the aneurysms were as follows: 19 in the middle cerebral artery (MCA) (42.2%), 15 in the anterior communicating artery (AComA) (33.3%), five in the internal carotid-posterior communicating artery (IC-PC) (11.1%), four in the tip of the basilar artery (BA-tip) (8.9%), and two in the anterior choroidal artery (AChorA) (4.4%). Outcomes for these aneurysms included clipping (7 cases), coiling (2 cases), subarachnoid hemorrhage (1 case), and follow-up (35 cases) (Table 1). Treated aneurysms (10 cases) were analyzed using image datasets taken less than three months before treatment, and untreated aneurysms (35 cases) were analyzed based on the initial diagnosis.
CTA conditions
Three-dimensional computed tomographic angiography (CTA) was performed using a multidetector CT system (Activion-16; Canon Medical Systems). The Real-Prep scan mode was used with 100 ml of contrast agent (Iomeron, iodine concentration 350 mg/mL; Eizai Pharmaceutical) injected at 3 mL/s into the antecubital vein. The imaging parameters included a 16-cm field of view, 512×512 matrix, section thickness of 0.5 mm, scanning time of 8 s, and a total of 201 images. The original volume data were interpolated into a matrix of 1024×1024 pixels and a thickness of 0.3 mm using a workstation (Ziostation-2, Ziosoft/AMIN, Tokyo).
MRI conditions
Magnetic resonance cisternography (MRC), in combination with magnetic resonance angiography (MRA), was performed for each patient on a 3T unit (Signa Pioneer; GE Healthcare, Milwaukee, WI). A heavily T2-weighted 3D fast spin-echo sequence was used for the MRC with the following parameters: TR/TE, 1900/100 ms; field of view, 180×180 mm; matrix, 356×256; section thickness, 1.2 mm; NEX,1; bandwidth, 31.2 kHz; echo-train length, 128; acquisition time, 6 min. Sixty axial images were acquired. The volume data were interpolated into 1024×1024 pixels matrix and a thickness of 0.3 mm using the workstation.
CFD conditions
Computational fluid dynamics (CFD) was performed using a commercial CFD package (Hemoscope v1.4, EBM Corp., Tokyo). Blood was assumed to be an incompressible Newtonian fluid with a density of 𝜌 = 1050 kg/m3 and viscosity 𝜇 = 0.004 Pa·s. Vascular geometries were reconstructed from CTA datasets. The mesh size was 0.25 mm in far-wall regions. The near-wall mesh consisted of three layers, with a width of 0.125 mm and a height of 0.05 mm at the wall-nearest mesh. Unsteady pulsatile simulations were carried out as a time step was adjusted to keep the Courant number less than one. As for boundary conditions, a time-averaged flow rate 𝑄 was first estimated using a formula 𝑄 = (𝜏𝜋⁄32𝜇) 𝐷3, where 𝜏 is a magnitude of WSS (𝜏 =1.5 Pa) 17 and 𝐷 is the diameter of the inlet artery. The distribution of flow rate at vascular bifurcations was set to be a ratio of the cubic diameters of the branches. A physiological pulsatile waveform was imposed on the steady flow rate18.
Data extraction
1. Aneurysmal morphology
Based on the CTA datasets, a surface mesh of the vascular geometries was obtained. Aneurysmal shape indices (neck width, dome depth, aspect ratio, dome surface area, and dome volume) was computed using a surface mesh (Hemoscope v1.4, EBM Corp., Tokyo). The neck width was the diameter of a circle with an area equal to that of the neck. Dome depth was defined as the length from the neck center towards the dome perpendicular to the dome.
2. Identification of blebs and PAC
Blebs were defined as secondary focal bulges distinguishable from an aneurysmal dome based on 3D-CTA datasets. The peri-aneurysmal environment was visualized using 3D multifusion CTA and MRC images9. The presence of PAC was judged based on whether the aneurysms overlapped with the surrounding structures. If aneurysms harbored blebs, the location of the bleb relative to PAC was determined. If they overlapped each other, the bleb location was classified as either central or marginal; a central-type bleb had no overlap with a marginal region of PAC, and a marginal-type bleb had at least a partial overlap with a marginal region of PAC.
3. Aneurysmal hemodynamics
The mean WSS at the parent artery (WSS-p) and aneurysmal dome (WSS-d) were computed, including their ratio (normalized WSS, NWSS). The location of the parent artery was chosen immediately upstream of the aneurysm. In aneurysms harboring a bleb, the mean, minimum, and maximum WSS at the bleb including the bleb neck were computed, and the ratio of the maximum WSS at the bleb to that of the whole dome was computed (normalized maximum WSS, NMWSS).
Statistical analysis
A categorical regression analysis was carried out to identify an independent variable associated with bleb formation among aneurysmal morphology, hemodynamics, and PAC by quantifying nominal, ordinal, and numeric variables using optimal scaling and assigning numerical values to the different categories. The dependent variable was the presence of blebs, and the independent variables were age, sex, shape (dome depth, aspect ratio, dome volume), hemodynamic indices (WSS-p, NWSS), and the presence of PAC. Descriptive analysis of continuous variables was performed, including the calculation of the means, standard deviations, medians, and 99% confidence intervals (99% CI) for all variables. The correlations between the quantitative and qualitative variables were analyzed using Pearson’s or Spearman’s correlation analyses. Accordingly, statistical differences in aneurysmal shape and hemodynamic indices were analyzed between groups using the Kruskal–Wallis test followed by the Steel–Dwass multiple comparison test. Data were analyzed using the IBM Statistical Package for the Social Sciences (SPSS) for Windows version 25 (IBM Corp., Armonk, NY) with the Categories module and R version 4.0.2. Statistical significance was set at p < 0.05.
Table 1. Clinical features.
Category
|
|
|
|
Nr.
|
45
|
|
|
Male/Female
|
13/32
|
|
|
Age (y.o.)
|
65.07±9.57
|
|
|
Location (Nr.)
|
45
|
MCA
AComA
IC-PC
BA-tip
AChorA
|
19
15
5
4
2
|
|
|
Outcome (Nr.)
|
Clipped: 7
Coiled: 2
SAH: 1
Follow up:35
|
AChorA, anterior choroidal artery; AComA, anterior communicating artery; BA-tip, tip of the basilar artery; IC-PC, internal carotid-posterior communicating artery; MCA, middle cerebral artery; SAH, subarachnoid hemorrhage.