Morphological parameters related to ruptured aneurysm in the patient with multiple cerebral aneurysms (clinical investigation) Hong Jun Jeon, Jae Whan Lee, So Yeon Kim, Keun Young Park, Seung Kon Huh Department of Neurosurgery, Brain Research Institute, Yonsei University College of Medicine, Seoul, Republic of Korea Objectives: We evaluated the rupture risk of multiple cerebral aneurysms in aspects of various morphological parameters, and determined which parameter can be a reliable predictor as one aneurysm ruptured, and the others did not. Methods: Between 2007 and 2012, three-dimensional (3D) angiographic images of 85 patients harboring multiple aneurysms (85 ruptured and 104 unruptured aneurysms) were used to assess the following morphological parameters: geometry of the aneurysm itself, e.g. maximal size, aspect ratio, bottleneck ratio, height/width ratio, undulation, and daughter sac; architecture of the aneurysm and surrounding vessels, e.g. aneurysmal angle, vessel angle, inflow angle, parent–daughter angle, and size ratio type I & II. Univariate analysis was applied to all parameters, and significant parameters were identified in multivariate analysis, yielding the cut-off point from receiver-operating characteristic (ROC) curve analysis. Results: On multivariate logistic regression, the aspect ratio [odds ratio (OR), 1.21; 95% confidence interval (CI), 1.05–1.41] and daughter sac (OR, 3.12; 95% CI, 1.05–9.27) were significant parameters in geometries of the aneurysm itself. The size ratio type I (OR, 1.14; 95% CI, 1.05–1.22) and parent–daughter angle (OR, 1.02; 95% CI, 1.00–1.04) were independent parameters in architecture of the aneurysm and surrounding vessels. From the ROC curve, the aspect ratio and size ratio type I had cut-off values of 1.3 and 1.8, respectively. Conclusion: Several morphological parameters were investigated to predict a rupture in multiple cerebral aneurysms using 3D angiogram. The aspect ratio, size ratio type I, daughter sac, and parent–daughter angle were revealed as competent parameters. Keywords: Multiple cerebral aneurysms, Morphological parameters, Aspect ratio, Daughter sac, Size ratio, Parent–daughter angle
Introduction The natural history of cerebral aneurysms has been well described by previous large cohort studies and the size and location of aneurysm has been a wellknown predictor of intracranial aneurysm rupture.1–3 Nonetheless, decision-making for the treatment of unruptured cerebral aneurysms is complex and could demand careful consideration of treatment-related risk and natural history of cerebral aneurysms. Technical advancements in imaging tools, such as digital subtraction angiography (DSA) with threedimensional (3D) reconstruction and 3D computed tomographic angiography (3D-CTA), enabled the precise measurement of morphological parameters of cerebral aneurysms.4–8 From these modalities, various aspects of aneurysm geometry, their relationship Correspondence to: Keun Young Park, Department of Neurosurgery, Brain Research Institute, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemoon-gu, Seoul 120-752, Republic of Korea. Email: [email protected]
1056
ß W. S. Maney & Son Ltd 2014 DOI 10.1179/1743132814Y.0000000393
with surrounding and parent vessel curvatures, and vessel type such as sidewall or bifurcation have been investigated as morphological parameters. Some studies demonstrated that those morphological parameters such as aspect ratio, size ratio, undulation, and nonsphericity index are better predictors than maximal size.5,8,9 Also, hemodynamics into aneurysm sac was investigated with computational fluid dynamics (CFD).10–12 It demonstrated that pathophysiology and geometry of cerebral aneurysm were related with variations of hemodynamics. The present study is a case–control study of ruptured multiple cerebral aneurysms in which morphological parameters can be directly compared between ruptured and unruptured aneurysm without patient-related bias. Our objective is to determine reliable parameters to predict a rupture among multiple cerebral aneurysms in terms of the geometry, and to evaluate the degree of rupture risk by means of such parameters.
Neurological Research
2014
VOL .
36
NO .
12
Jeon et al.
Parameters to predict rupture among cerebral aneurysms
–120u. Non-ionic, iso-osmolar contrast (270 mg per I/ ml, iodixanol; Visipaque, Nydalen, Oslo, Norway) was injected in a y16- to 30-ml volume through a 5–6 Fr catheter (injector velocity of 2–6 ml/second). Image acquisition started 2–3 seconds after contrast injection. The total acquisition time was 12 seconds. The gradient mode on the 3D workstation was used to service translucent 3D images with 60–80% magnification. The size (height, width, and neck) and related-angle of the aneurysm, as well as the surrounding vessels, were measured in reconstructed 3D images. Figure 1 (a) Morphological parameters associated with the aneurysm itself showing on 3D (three-dimensional) workstation (A) maximal size, (B) neck size, (C) maximal perpendicular height, (D) maximal height, (E) maximal width, the circular line displayed a daughter sac, not seen undulation. Related ratio measured such as: aspect ratio (C and B), bottleneck ratio (E and B), and height/width ratio (D and E). (b) Illustration showing schematic measurements.
Materials and Methods Patients The present study was based on the ethical standards of the Institutional Review Board. At our institute from January 2007 to December 2012, consecutive 85 patients harboring ruptured aneurysm among multiple cerebral aneurysms were evaluated and treated by either clipping (N 5 78, 91.8%) or coiling (N 5 7, 8.2%). In clipping, ruptured lesions could be identified by microscopic visual assessment. In coiling, one neuroradiologist and one neurosurgeon reviewed the image and determined the ruptured lesions by consensus. Patients without 3D-DSA or with fusiform, dissecting, infectious, and arteriovenous malformation-related aneurysms were excluded from the study. At the time of hospitalization, all of aneurysms were evaluated by 3D-DSA, and morphological parameters were calculated and obtained. Aneurysm location was briefly divided into four groups (internal carotid artery (ICA): paraclinoid, supraclinoid, and distal ICA; middle cerebral artery (MCA): M1 trunk, bifurcation, and distal MCA; anterior cerebral artery (ACA): A1, anterior communicating artery, and distal ACA; vertebrobasilar artery (VBA): vertebral artery to basilar artery). According to maximal size, the aneurysms were divided into three groups (small: #5 mm, medium: 6–10 mm, large: .10 mm). They were also dichotomized by sidewall and bifurcation type.
3D workstation Three-dimensional rotational angiographic examination with a 15-cm image intensifier was performed on a biplane system (Allura Xper FD 20/20 system; Philips Medical Systems, Best, The Netherlands). The field’s view was 22 6 22 cm, and the frame rate was 30/second. The rotational ranges were from z120 to
Morphological parameters I We measured fundamental geometry of the aneurysm itself on 3D workstation (Fig. 1). Then, we evaluated and calculated the six morphological parameters that were previously described for cerebral aneurysms. Maximal size According to Raghavan’s method, the maximal size was defined as the largest cross-sectional diameter from the 3D image (Fig. 1; A).8 Aspect ratio This parameter was defined as the ratio of maximal perpendicular height to neck diameter of the aneurysm (Fig. 1; C and B).9 Bottleneck ratio and height/width ratio The bottleneck parameter was measured using the ratio of the maximal width to the size of the aneurysmal neck (Fig. 1; E and B) and the height/ width parameter was determined using the ratio between the maximal height and width of the aneurysm (Fig. 1; C and E).13 Undulation and daughter sac We defined the presence of a prominent irregularity on the aneurysmal surface as the undulation, and the daughter sac was a combined small sac on the dome.
Morphological parameters II Six parameters were determined by considering the architecture of the aneurysm and surrounding vessels (Fig. 2). The aneurysmal angle, vessel angle, and inflow angle were classified by the sidewall type and bifurcation type. The parent–daughter angle was only measured for the bifurcation type. Aneurysmal angle It was an inclination between the plane of the aneurysm neck and the maximal height of the aneurysm (Fig. 2a; A).5 The maximal height was measured by the farthest distance from the center point of cross-sectional area of the neck to the dome of the aneurysm. Vessel angle This angle was measured between the centerline of the incoming parent vessel into the aneurysm and the
Neurological Research
2014
VOL .
36
NO .
12
1057
Jeon et al.
Parameters to predict rupture among cerebral aneurysms
Figure 2 Morphological parameters considering the aneurysm and surrounding vessels showing on three-dimensional (3D) workstation: On (a) figures, (A) aneurysmal angle, (B) vessel angle, and (C) inflow angle measured from sidewall type. On (b) figures, parent–daughter angle (AzB/2) measured from bifurcation type. Parent vessels mean diameter calculated as (D1azD1a9zD2bzD2b9zD3czD3c9/6). Size ratio was determined as type I (maximal height/parent vessels mean diameter) and type II (maximal size/parent vessels mean diameter). D1a9, D2b9, D3c9 represent the vessel diameter 1.5 times away from D1a, D2b, D3c.
neck plane (Fig. 2a; B).11,12 The centerline was displayed by connecting the center points between the two cross-sectional area used to measure the average diameter of the parent vessel. Inflow angle The inflow angle was measured between the vector of the maximum height of the aneurysm and the vector of the centerline of the parent vessel displayed in the 3D image (Fig. 2a; C).14,15 Parent–daughter angle It was defined as the average angle between the flow vector of the parent and daughter vessels (Fig. 2b; AzB/2).6 It considers only the architecture of blood vessels surrounding the aneurysm at the bifurcation. Size ratio type I & II This parameter was classified as two type. The size ratio type I was defined as the ratio between the maximal height of the aneurysm and the parent vessels mean diameter.5 The size ratio type II was applied the maximal size as a standard aneurysm measurement instead of the maximal height.7 The mean diameter of parent vessels was measured by the average diameter of the blood vessels surrounding the aneurysm (Fig. 2b; D1azD1a9zD2bzD2b9zD3czD3c9/6).
Statistical analysis Variables measured among multiple cerebral aneurysms were compared using chi-square test (or Fisher’s exact test) and two-tailed independent Student’s t-test for binary and continuous variables, respectively. The arithmetic mean, standard deviation, odds ratio (OR), and 95% confidence interval (CI) of the results were obtained from the univariate analysis. The parameters with significance in the univariate analysis (P , 0.05) were further analyzed using binary stepwise forward logistic regression. To eliminate confounding effects, the Pearson’s correlation coefficient and linear regression
1058
Neurological Research
2014
VOL .
36
NO .
12
were used to identify the relationship among the morphological parameters. Odds ratio related to the aspect ratio, bottleneck ratio, height/width ratio, and size ratio type I & II were adjusted with each 0.1 unit change. Receiver-operating characteristic analysis was performed on the final model and the optimal sensitivity and specificity of competent parameters were obtained by determining the cut-off point using the Youden index. P value less than 0.05 was deemed to be statistically significant. These statistical analyses were performed using IBM SPSS Statistics 20.0 for Windows (IBM Corp., Armonk, NY, USA).
Results Demographics Of the total 189 aneurysms, there were 85 ruptured aneurysms and 104 unruptured aneurysms, respectively. Age, gender, and aneurysm location could not show any significant differences between ruptured and unruptured lesions. The distribution by aneurysm size showed a significant difference between ruptured and unruptured lesions (Table 1, P , 0.001). There were 21 (11.1%) ruptured aneurysms and 81 (42.8%) unruptured aneurysms in the small group, 55 (29.1%) and 21 (11.1%) in the medium group, nine (4.8%) and two (1.1%) in the large group.
Univariate analysis of morphological parameters between the ruptured and unruptured aneurysms Morphological parameters I The maximal size, aspect ratio, bottleneck ratio, height/width ratio, undulation, and daughter sac were statistically different between ruptured and unruptured aneurysms (P , 0.001, Table 2). Morphological parameters II In the sidewall type, the vessel angle (P 5 0.033) and inflow angle (P 5 0.005) were significantly larger in
Jeon et al.
Parameters to predict rupture among cerebral aneurysms
Table 1 Demographic features between ruptured and unruptured aneurysms in multiple cerebral aneurysms Category Age ,45 45–64 >65 Gender Male Female Size (5 mm 6–10 mm .10 mm Location ICA MCA ACA VBA
Ruptured
Unruptured
Total
7 (8.2%) 44 (51.8%) 34 (40.0%)
8 (7.7%) 53 (51.0%) 43 (41.3%)
15 (7.9%) 97 (51.3%) 77 (40.7%)
19 (22.4%) 66 (77.6%)
20 (19.2%) 84 (80.8%)
39 (20.6%) 150 (79.4%)
21 (11.1%) 55 (29.1%) 9 (4.8%)
81 (42.8%) 21 (11.1%) 2 (1.1%)
102 (54.0%) 76 (40.2%) 11 (5.8%)
30 28 21 6
35 40 19 10
(35.3%) (32.9%) (24.7%) (7.1%)
(33.7%) (38.5%) (18.3%) (9.6%)
ruptured aneurysms. In the bifurcation type, aneurysmal angle and parent–daughter angle were significantly smaller in ruptured aneurysms (P , 0.001). Size ratio type I and II were significantly different between ruptured and unruptured aneurysms (P , 0.001, Table 2).
Multivariate analysis of morphological parameters between the ruptured and unruptured aneurysms The correlations among morphological parameters were identified using the Pearson’s correlation coefficient. Size parameters such as the maximal size (r 5 1), neck size (r 5 0.759), maximal perpendicular height (r 5 0.880), maximal height (r 5 0.934), and maximal width (r 5 0.923) showed a strong correlation with each other, and the maximal size with the largest value of area under the curve (AUC 5 0.837) was included as a representative value in the model. Additionally, the maximal size showed a correlation with the size ratio (r 5 0.777) and multivariate analysis made two models by dividing each parameter. The measured angles (aneurysmal angle, vessel angle, and inflow angle) at the aneurysm were also included
65 68 40 16
X2
P value
0.04
0.978
0.28
0.718
53.40
0.001
1.71
0.635
(34.4%) (36.0%) (21.2%) (8.5%)
in the multivariate model, as the results were significant according to the sidewall and bifurcation type. In multivariate analysis, the aspect ratio [OR, 1.21; 95% CI, 1.05–1.41] and daughter sac [OR, 3.12; 95% CI, 1.05–9.27] were significant parameters in geometries of the aneurysm itself. The size ratio type I [OR, 1.14; 95% CI, 1.05–1.22] and parent–daughter angle [OR, 1.02; 95% CI, 1.00–1.04] were independent parameters in architecture of the aneurysm and surrounding vessels (Table 3).
Comparison of ROC results The AUC value of the aspect ratio was 0.83, at a cutoff value of 1.3, with the sensitivity 75.3%, and the specificity 73.1%. The AUC value of size ratio type I was 0.86, at a cut-off value of 1.8, with the sensitivity 82.4%, and the specificity 81.7% (Fig. 3).
Subgroup analysis of morphological parameters in small aneurysms In the small aneurysm group (N 5 102, 54%), univariate and multivariate analysis revealed that the aspect ratio [OR, 1.57; 95% CI, 1.17–2.11], size ratio type I [OR, 1.04; 95% CI 1.01–1.08], and
Table 2 Univariate analysis for morphological parameters of the aneurysm itself Parameter Morphological parameters I Maximal size (mm) Aspect ratio Bottleneck ratio Height/width ratio Undulation Daughter sac Morphological parameters II Aneurysmal angle Sidewall Bifurcation Vessel angle Sidewall Bifurcation Inflow angle Sidewall Bifurcation Parent–daughter angle Size ratio type I
Ruptured
Unruptured
P value
6.8 ¡ 2.5 1.4 ¡ 0.5 1.5 ¡ 0.4 1.2 ¡ 0.4 51/85 (60.0%) 49/85 (57.6%)
4.0 ¡ 2.0 0.9 ¡ 0.4 1.1 ¡ 0.4 0.9 ¡ 0.3 18/104 (17.3%) 14/104 (13.5%)
0.001 0.001 0.001 0.001 0.001 0.001
1.80 1.31 1.27 1.25 7.18 8.75
(1.49–2.17) (1.20–1.43) (1.16–1.39) (1.13–1.38) (3.67–13.98) (4.31–17.78)
72.2 ¡ 64.0 ¡ 34.9 ¡ 52.7 ¡ 120.8 ¡ 131.4 ¡ 72.8 ¡ 2.6 ¡
77.4 ¡ 74.6 ¡ 14.6 ¡ 55.1 ¡ 97.5 ¡ 139.7 ¡ 91.6 ¡ 1.2 ¡
0.242 0.001 0.033 0.607 0.005 0.087 0.001 0.001
0.98 0.97 1.04 0.97 1.03 0.99 0.97 1.24
(0.95–1.01) (0.95–0.99) (1.01–1.07) (0.98–1.01) (1.01–1.06) (0.98–1.00) (0.96–0.99) (1.17–1.33)
18.5 19.0 39.5 27.3 24.2 29.0 26.6 1.1
15.2 17.7 15.2 22.8 31.2 24.5 27.3 0.6
Neurological Research
OR (95% CI)
2014
VOL .
36
NO .
12
1059
Jeon et al.
Parameters to predict rupture among cerebral aneurysms
daughter sac [OR, 26.49; 95% CI 2.58–271.4] could be independent parameters.
Discussion Natural history and risk factors of cerebral aneurysms have been well studied with prospective and retrospective large-scale data.1–3 These studies have shown that the size and location of the cerebral aneurysm are associated with rupture risk, and family history, hypertension, cigarette smoking, and increasing age have been suggested to increase rupture risk. However, the treatment of unruptured aneurysms remains controversy regarding the decision-making process due to frequent rupture in small-size aneurysms.16,17 With the advancement of imaging modalities, some researches have attempted to predict the rupture risk of cerebral aneurysm through the aneurysmal geometry and hemodynamic effect. Unfortunately, such studies may face challenges, including patient-related confounding factors such as hypertension, smoking, and genetic problems that can affect the results. However, when we investigate ruptured aneurysm among multiple cerebral aneurysms, the issues can be resolved by eliminating confounding factors from the patient’s own internal control. As considering only geometry of the aneurysm itself, Ujiie et al. retrospectively reviewed the twodimensional (2D) angiogram from 129 ruptured and 78 unruptured aneurysms, and introduced the aspect ratio defined as the ratio between the height and neck width of the aneurysm.9 Almost 80% of the ruptured aneurysms showed an aspect ratio greater than 1.6, whereas almost 90% of the unruptured aneurysms showed an aspect ratio less than 1.6. These results suggested that the aspect ratio could be an index for predicting the likelihood of aneurysm rupture at the time the measurements were taken. In our study, the aspect ratio showed similar results, but the threshold (51.3) appeared to be low. The reason is the measurements of 2D angiogram can be more exaggerated due to an overlay effect than 3D reconstructed images. Hoh et al. analyzed 67 multiple cerebral aneurysms to identify the relationship between the aneurysm rupture and morphological features.13 Maximal diameter, height, maximal width, bulge height, aspect ratio, and bottleneck factor were significantly associated with ruptured aneurysms by univariate analysis. Best subsets and stepwise multivariable analysis identified the bottleneck factor and
Figure 3 Comparison ROC results of the most reliable morphological parameters.
height/width ratio as significant parameters to predict a rupture. Our study also confirmed that these parameters were significant on a univariate analysis; however, as compared with other morphological parameters, those parameters were not included in the final model. From biomathematical models of cerebral aneurysms, Chitanvis et al. suggested that undulation in the aneurysm wall, presented by irregular morphology of aneurysm surface, could play an important role in aneurysm rupture.18 Similarly, Beck et al. retrospectively analyzed 2D angiograms from 94 ruptured and 53 unruptured aneurysms and found that ruptured aneurysms were associated with irregular multilobar appearance in aneurysms 5–9 mm in size.4 Both studies suggested that the rupture risk could be related to the geometry of aneurysmal surface. However, overlapping images from the 2D angiogram could induce a rather obscure interpretation between undulation and daughter sac on the aneurysmal surface. On the other hand, it can be easily discriminated from the 3D reconstructed image. In this regard, our result showed that the daughter sac was more associated with the rupture status than the undulation. Several studies have examined the architecture of aneurysms and surrounding vessels for rupture risk at
Table 3 Multivariate analysis for morphological parameters Parameter Morphological parameters I
Aspect ratio Daughter sac Size ratio type I Parent–daughter angle
Morphological parameters II
1060
Neurological Research
2014
VOL .
36
NO .
12
OR
95% CI
P value
1.21 3.12 1.14 1.02
1.05–1.41 1.05–9.27 1.05–1.22 1.00–1.04
0.010 0.040 0.001 0.039
Jeon et al.
the cerebral aneurysm and evaluated how blood flow in the aneurysm is being affected by the surrounding vessel architecture within the main bloodstream. Dhar et al. assumed that the physical vector of the incoming flow can affect the morphological features of the aneurysm and showed that a neck plane of ruptured aneurysms was more tilted from the flow direction of aneurysms.5 Through the CFD analysis, Hoi et al. and Hassan et al. investigated the vessel angle and found that a more curved status caused aneurysm growth and elevations in the wall shear stress by increasing flow impingement in the impact zone of aneurysm neck.11,12 Also, Baharoglu et al. used CFD analysis to study the inflow angle of aneurysms of the sidewall type and showed that, as the inflow angle increased, the inflow zone and dome of the aneurysms occurred at a higher inflow velocity and greater wall shear stress.14,15 On the branched type, Lin et al. reported that a smaller parent– daughter angle at the bifurcation vessel was associated with ruptured aneurysms.6 There seemed that the flow vector could cause a more stressful condition because the smaller composite angle at a branch site can more induce irrotational flow in the aneurysm. In our study, those parameters also demonstrated similar results in ruptured aneurysms. Therefore, it could be considered that geometric change and rupture of aneurysms can be influenced with hemodynamic effect through the surrounding vascular architecture. Of several parameters, the size ratio is known to reflect the relationship between the aneurysmal geometry and surrounding vascular architectures. Initially, Dhar et al. retrospectively examined the 3D image from 20 ruptured and 25 unruptured aneurysms and obtained a significant difference from the ratio of the maximal height and the average size of the vessels around the ruptured aneurysms.5 About 78% of all ruptured aneurysms showed a size ratio greater than 2.05, whereas 83% of all unruptured aneurysms showed a size ratio less than 2.05. Similarly, Tremmel et al. demonstrated a higher size ratio (.2.0) related with multiple vortices and complex flow patterns of the intra-aneurysmal sac from the CFD study.19 In the measurement of size ratio, Rahman et al. adjusted the size ratio to be used in the 2D angiogram by changing the maximal height to maximal size as a more typical method of the aneurysm measurement. They reported that the size ratio type II was correlated with aneurysmal rupture in a blind analysis.7 In this respect, we evaluated which parameter is more significant in the ruptured aneurysms of multiple cerebral aneurysms and size ratio type I was a more predictable parameter at the rupture risk. In order to confirm the most reliable parameter, the forward method of multivariate logistic regression
Parameters to predict rupture among cerebral aneurysms
analysis from various parameters can be the appropriate choice with regard to the statistical analysis. From that result, independent morphological parameters indicate the relative risk of a certain aneurysm among multiple cerebral aneurysms. Thus, if an aneurysm has a larger value than 0.1 units in aspect ratio, the rupture risk increases by approximately 21%, and if the aneurysm only has a daughter sac, the risk increases approximately 3.1 times. A higher value of 0.1 units in size ratio type I can raise the risk by approximately 14%; for one degree smaller in the parent–daughter angle, the relative risk increases 2%. Many large volume studies demonstrated that aneurysm location have related rupture risk.1,2,3,20 While we did not show the difference according to location, it seems that small volume or simplified classification related with major territory may affect the results. Nevertheless, size ratio appeared as a significant parameter. Actually, size ratio takes into account not only the aneurysm size itself, but also vessel geometry by integrating it into a quantifiable parameter. In other words, it indirectly accounts for the influence of aneurysmal location on rupture. Several studies have shown that small sized aneurysms ruptured in a large percentage.4,21,22 Our data also presented that ruptured portion of small aneurysm with 5 mm or less was 11.1% (21/189). It seems that aneurysm size has not yet been completely explained and needs to use additional parameters. In this regard, when we analyzed morphological parameters in the small aneurysms, independent parameters appeared as the aspect ratio, daughter sac, and size ratio type I. There are several limitations in this study. First, it was a retrospective study that the certain cerebral aneurysm has already ruptured condition. So, we do not know exactly whether the aneurysm size has been changed after rupture or undulation of aneurysm surface has been caused by rupture. At least, we have to assume a precondition without change at that time. Second, interpretation of statistical results only showed the relative risk. Third, risk of rupture according to location did not clarify, because of the small volume, and it might induce a bias. However, these results may be meaningful processes in that the several parameters with the risk prediction were compared to their reliability among the multiple cerebral aneurysms. Among several morphological parameters, aspect ratio, daughter sac, size ratio type I, and parent–daughter angle were independent predictors related to the geometry of the cerebral aneurysm and surrounding vessels. Furthermore, if the size ratio type I §1.8, aspect ratio §1.3, the presence of a daughter sac and relatively smaller parent–daughter angle appear, it would be considered
Neurological Research
2014
VOL .
36
NO .
12
1061
Jeon et al.
Parameters to predict rupture among cerebral aneurysms
the first therapeutic target among multiple cerebral aneurysms, if not ruptured.
Conclusion The various morphological parameters were compared under the controlled conditions with ruptured aneurysm among multiple cerebral aneurysms. The aspect ratio, daughter sac, size ratio type I, and parent–daughter angle were reliable predictors of aneurysm rupture.
Disclaimer Statements Contributors Hong Jun Jeon, Jae Whan Lee, So Yeon Kim, Keun Young Park, and Seung Kon Huh are all contributors in this study. Funding None. Conflicts of interest There are no conflicts of interest to declare. Ethics approval The present study was approved on the ethical standards of the Institutional Review Board.
References 1 The International Study of Unruptured Intracranial Aneurysms Investigators. Unruptured intracranial aneurysms – risk of rupture and risks of surgical intervention. N Engl J Med. 1998;339:1725–33. 2 UCAS Japan Investigators, Morita A, Kirino T, Hashi K, Aoki N, Fukuhara S, et al. The natural course of unruptured cerebral aneurysms in a Japanese cohort. N Engl J Med. 2012;366:2474– 82. 3 Juvela S, Porras M, Poussa K. Natural history of unruptured intracranial aneurysms: probability of and risk factors for aneurysm rupture. J Neurosurg. 2008;108:1052–60. 4 Beck J, Rohde S, El Beltagy M, Zimmermann M, Berkefeld J, Seifert V, et al. Difference in configuration of ruptured and unruptured intracranial aneurysms determined by biplanar digital subtraction angiography. Acta Neurochir (Wien). 2003;145:861–5; discussion 5. 5 Dhar S, Tremmel M, Mocco J, Kim M, Yamamoto J, Siddiqui AH, et al. Morphology parameters for intracranial aneurysm rupture risk assessment. Neurosurgery. 2008;63:185–96; discussion 96–7. 6 Lin N, Ho A, Gross BA, Pieper S, Frerichs KU, Day AL, et al. Differences in simple morphological variables in ruptured and unruptured middle cerebral artery aneurysms. J Neurosurg. 2012;117:913–9.
1062
Neurological Research
2014
VOL .
36
NO .
12
7 Rahman M, Smietana J, Hauck E, Hoh B, Hopkins N, Siddiqui A, et al. Size ratio correlates with intracranial aneurysm rupture status: a prospective study. Stroke. 2010;41:916–20. 8 Raghavan ML, Ma B, Harbaugh RE. Quantified aneurysm shape and rupture risk. J Neurosurg. 2005;102:355–62. 9 Ujiie H, Tamano Y, Sasaki K, Hori T. Is the aspect ratio a reliable index for predicting the rupture of a saccular aneurysm? Neurosurgery. 2001;48:495–502; discussion 502-3. 10 Cebral JR, Castro MA, Appanaboyina S, Putman CM, Millan D, Frangi AF. Efficient pipeline for image-based patientspecific analysis of cerebral aneurysm hemodynamics: technique and sensitivity. IEEE Trans Med Imaging. 2005;24:457–67. 11 Hoi Y, Meng H, Woodward SH, Bendok BR, Hanel RA, Guterman LR, et al. Effects of arterial geometry on aneurysm growth: three-dimensional computational fluid dynamics study. J Neurosurg. 2004;101:676–81. 12 Hassan T, Timofeev EV, Saito T, Shimizu H, Ezura M, Matsumoto Y, et al. A proposed parent vessel geometry-based categorization of saccular intracranial aneurysms: computational flow dynamics analysis of the risk factors for lesion rupture. J Neurosurg. 2005;103:662–80. 13 Hoh BL, Sistrom CL, Firment CS, Fautheree GL, Velat GJ, Whiting JH, et al. Bottleneck factor and height-width ratio: association with ruptured aneurysms in patients with multiple cerebral aneurysms. Neurosurgery. 2007;61:716–22; discussion 22–3. 14 Baharoglu MI, Schirmer CM, Hoit DA, Gao BL, Malek AM. Aneurysm inflow-angle as a discriminant for rupture in sidewall cerebral aneurysms: morphometric and computational fluid dynamic analysis. Stroke. 2010;41:1423–30. 15 Baharoglu MI, Lauric A, Gao BL, Malek AM. Identification of a dichotomy in morphological predictors of rupture status between sidewall- and bifurcation-type intracranial aneurysms. J Neurosurg. 2012;116:871–81. 16 Lu HT, Tan HQ, Gu BX, Wu-Wang, Li MH. Risk factors for multiple intracranial aneurysms rupture: a retrospective study. Clin Neurol Neurosurg. 2013;115:690–4. 17 Schievink WI, Piepgras DG, Wirth FP. Rupture of previously documented small asymptomatic saccular intracranial aneurysms. Report of three cases. J Neurosurg. 1992;76:1019–24. 18 Chitanvis SM, Dewey M, Hademenos G, Powers WJ, Massoud TF. A nonlinear quasi-static model of intracranial aneurysms. Neurol Res. 1997;19:489–96. 19 Tremmel M, Dhar S, Levy EI, Mocco J, Meng H. Influence of intracranial aneurysm-to-parent vessel size ratio on hemodynamics and implication for rupture: results from a virtual experimental study. Neurosurgery. 2009;64:622–30; discussion 30–1. 20 Wiebers DO, Whisnant JP, Huston J 3rd, Meissner I, Brown RD Jr, Piepgras DG, et al. Unruptured intracranial aneurysms: natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet. 2003;362:103–10. 21 Rinkel GJ, Djibuti M, Algra A, van Gijn J. Prevalence and risk of rupture of intracranial aneurysms: a systematic review. Stroke. 1998;29:251–6. 22 Forget TRJr, Benitez R, Veznedaroglu E, Sharan A, Mitchell W, Silva M, et al. A review of size and location of ruptured intracranial aneurysms. Neurosurgery. 2001;49:1322–5; discussion 5–6.
Introduction The natural history of cerebral aneurysms has been well described by previous large cohort studies and the size and location of aneurysm has been a wellknown predictor of intracranial aneurysm rupture.1–3 Nonetheless, decision-making for the treatment of unruptured cerebral aneurysms is complex and could demand careful consideration of treatment-related risk and natural history of cerebral aneurysms. Technical advancements in imaging tools, such as digital subtraction angiography (DSA) with threedimensional (3D) reconstruction and 3D computed tomographic angiography (3D-CTA), enabled the precise measurement of morphological parameters of cerebral aneurysms.4–8 From these modalities, various aspects of aneurysm geometry, their relationship Correspondence to: Keun Young Park, Department of Neurosurgery, Brain Research Institute, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemoon-gu, Seoul 120-752, Republic of Korea. Email: [email protected]
1056
ß W. S. Maney & Son Ltd 2014 DOI 10.1179/1743132814Y.0000000393
with surrounding and parent vessel curvatures, and vessel type such as sidewall or bifurcation have been investigated as morphological parameters. Some studies demonstrated that those morphological parameters such as aspect ratio, size ratio, undulation, and nonsphericity index are better predictors than maximal size.5,8,9 Also, hemodynamics into aneurysm sac was investigated with computational fluid dynamics (CFD).10–12 It demonstrated that pathophysiology and geometry of cerebral aneurysm were related with variations of hemodynamics. The present study is a case–control study of ruptured multiple cerebral aneurysms in which morphological parameters can be directly compared between ruptured and unruptured aneurysm without patient-related bias. Our objective is to determine reliable parameters to predict a rupture among multiple cerebral aneurysms in terms of the geometry, and to evaluate the degree of rupture risk by means of such parameters.
Neurological Research
2014
VOL .
36
NO .
12
Jeon et al.
Parameters to predict rupture among cerebral aneurysms
–120u. Non-ionic, iso-osmolar contrast (270 mg per I/ ml, iodixanol; Visipaque, Nydalen, Oslo, Norway) was injected in a y16- to 30-ml volume through a 5–6 Fr catheter (injector velocity of 2–6 ml/second). Image acquisition started 2–3 seconds after contrast injection. The total acquisition time was 12 seconds. The gradient mode on the 3D workstation was used to service translucent 3D images with 60–80% magnification. The size (height, width, and neck) and related-angle of the aneurysm, as well as the surrounding vessels, were measured in reconstructed 3D images. Figure 1 (a) Morphological parameters associated with the aneurysm itself showing on 3D (three-dimensional) workstation (A) maximal size, (B) neck size, (C) maximal perpendicular height, (D) maximal height, (E) maximal width, the circular line displayed a daughter sac, not seen undulation. Related ratio measured such as: aspect ratio (C and B), bottleneck ratio (E and B), and height/width ratio (D and E). (b) Illustration showing schematic measurements.
Materials and Methods Patients The present study was based on the ethical standards of the Institutional Review Board. At our institute from January 2007 to December 2012, consecutive 85 patients harboring ruptured aneurysm among multiple cerebral aneurysms were evaluated and treated by either clipping (N 5 78, 91.8%) or coiling (N 5 7, 8.2%). In clipping, ruptured lesions could be identified by microscopic visual assessment. In coiling, one neuroradiologist and one neurosurgeon reviewed the image and determined the ruptured lesions by consensus. Patients without 3D-DSA or with fusiform, dissecting, infectious, and arteriovenous malformation-related aneurysms were excluded from the study. At the time of hospitalization, all of aneurysms were evaluated by 3D-DSA, and morphological parameters were calculated and obtained. Aneurysm location was briefly divided into four groups (internal carotid artery (ICA): paraclinoid, supraclinoid, and distal ICA; middle cerebral artery (MCA): M1 trunk, bifurcation, and distal MCA; anterior cerebral artery (ACA): A1, anterior communicating artery, and distal ACA; vertebrobasilar artery (VBA): vertebral artery to basilar artery). According to maximal size, the aneurysms were divided into three groups (small: #5 mm, medium: 6–10 mm, large: .10 mm). They were also dichotomized by sidewall and bifurcation type.
3D workstation Three-dimensional rotational angiographic examination with a 15-cm image intensifier was performed on a biplane system (Allura Xper FD 20/20 system; Philips Medical Systems, Best, The Netherlands). The field’s view was 22 6 22 cm, and the frame rate was 30/second. The rotational ranges were from z120 to
Morphological parameters I We measured fundamental geometry of the aneurysm itself on 3D workstation (Fig. 1). Then, we evaluated and calculated the six morphological parameters that were previously described for cerebral aneurysms. Maximal size According to Raghavan’s method, the maximal size was defined as the largest cross-sectional diameter from the 3D image (Fig. 1; A).8 Aspect ratio This parameter was defined as the ratio of maximal perpendicular height to neck diameter of the aneurysm (Fig. 1; C and B).9 Bottleneck ratio and height/width ratio The bottleneck parameter was measured using the ratio of the maximal width to the size of the aneurysmal neck (Fig. 1; E and B) and the height/ width parameter was determined using the ratio between the maximal height and width of the aneurysm (Fig. 1; C and E).13 Undulation and daughter sac We defined the presence of a prominent irregularity on the aneurysmal surface as the undulation, and the daughter sac was a combined small sac on the dome.
Morphological parameters II Six parameters were determined by considering the architecture of the aneurysm and surrounding vessels (Fig. 2). The aneurysmal angle, vessel angle, and inflow angle were classified by the sidewall type and bifurcation type. The parent–daughter angle was only measured for the bifurcation type. Aneurysmal angle It was an inclination between the plane of the aneurysm neck and the maximal height of the aneurysm (Fig. 2a; A).5 The maximal height was measured by the farthest distance from the center point of cross-sectional area of the neck to the dome of the aneurysm. Vessel angle This angle was measured between the centerline of the incoming parent vessel into the aneurysm and the
Neurological Research
2014
VOL .
36
NO .
12
1057
Jeon et al.
Parameters to predict rupture among cerebral aneurysms
Figure 2 Morphological parameters considering the aneurysm and surrounding vessels showing on three-dimensional (3D) workstation: On (a) figures, (A) aneurysmal angle, (B) vessel angle, and (C) inflow angle measured from sidewall type. On (b) figures, parent–daughter angle (AzB/2) measured from bifurcation type. Parent vessels mean diameter calculated as (D1azD1a9zD2bzD2b9zD3czD3c9/6). Size ratio was determined as type I (maximal height/parent vessels mean diameter) and type II (maximal size/parent vessels mean diameter). D1a9, D2b9, D3c9 represent the vessel diameter 1.5 times away from D1a, D2b, D3c.
neck plane (Fig. 2a; B).11,12 The centerline was displayed by connecting the center points between the two cross-sectional area used to measure the average diameter of the parent vessel. Inflow angle The inflow angle was measured between the vector of the maximum height of the aneurysm and the vector of the centerline of the parent vessel displayed in the 3D image (Fig. 2a; C).14,15 Parent–daughter angle It was defined as the average angle between the flow vector of the parent and daughter vessels (Fig. 2b; AzB/2).6 It considers only the architecture of blood vessels surrounding the aneurysm at the bifurcation. Size ratio type I & II This parameter was classified as two type. The size ratio type I was defined as the ratio between the maximal height of the aneurysm and the parent vessels mean diameter.5 The size ratio type II was applied the maximal size as a standard aneurysm measurement instead of the maximal height.7 The mean diameter of parent vessels was measured by the average diameter of the blood vessels surrounding the aneurysm (Fig. 2b; D1azD1a9zD2bzD2b9zD3czD3c9/6).
Statistical analysis Variables measured among multiple cerebral aneurysms were compared using chi-square test (or Fisher’s exact test) and two-tailed independent Student’s t-test for binary and continuous variables, respectively. The arithmetic mean, standard deviation, odds ratio (OR), and 95% confidence interval (CI) of the results were obtained from the univariate analysis. The parameters with significance in the univariate analysis (P , 0.05) were further analyzed using binary stepwise forward logistic regression. To eliminate confounding effects, the Pearson’s correlation coefficient and linear regression
1058
Neurological Research
2014
VOL .
36
NO .
12
were used to identify the relationship among the morphological parameters. Odds ratio related to the aspect ratio, bottleneck ratio, height/width ratio, and size ratio type I & II were adjusted with each 0.1 unit change. Receiver-operating characteristic analysis was performed on the final model and the optimal sensitivity and specificity of competent parameters were obtained by determining the cut-off point using the Youden index. P value less than 0.05 was deemed to be statistically significant. These statistical analyses were performed using IBM SPSS Statistics 20.0 for Windows (IBM Corp., Armonk, NY, USA).
Results Demographics Of the total 189 aneurysms, there were 85 ruptured aneurysms and 104 unruptured aneurysms, respectively. Age, gender, and aneurysm location could not show any significant differences between ruptured and unruptured lesions. The distribution by aneurysm size showed a significant difference between ruptured and unruptured lesions (Table 1, P , 0.001). There were 21 (11.1%) ruptured aneurysms and 81 (42.8%) unruptured aneurysms in the small group, 55 (29.1%) and 21 (11.1%) in the medium group, nine (4.8%) and two (1.1%) in the large group.
Univariate analysis of morphological parameters between the ruptured and unruptured aneurysms Morphological parameters I The maximal size, aspect ratio, bottleneck ratio, height/width ratio, undulation, and daughter sac were statistically different between ruptured and unruptured aneurysms (P , 0.001, Table 2). Morphological parameters II In the sidewall type, the vessel angle (P 5 0.033) and inflow angle (P 5 0.005) were significantly larger in
Jeon et al.
Parameters to predict rupture among cerebral aneurysms
Table 1 Demographic features between ruptured and unruptured aneurysms in multiple cerebral aneurysms Category Age ,45 45–64 >65 Gender Male Female Size (5 mm 6–10 mm .10 mm Location ICA MCA ACA VBA
Ruptured
Unruptured
Total
7 (8.2%) 44 (51.8%) 34 (40.0%)
8 (7.7%) 53 (51.0%) 43 (41.3%)
15 (7.9%) 97 (51.3%) 77 (40.7%)
19 (22.4%) 66 (77.6%)
20 (19.2%) 84 (80.8%)
39 (20.6%) 150 (79.4%)
21 (11.1%) 55 (29.1%) 9 (4.8%)
81 (42.8%) 21 (11.1%) 2 (1.1%)
102 (54.0%) 76 (40.2%) 11 (5.8%)
30 28 21 6
35 40 19 10
(35.3%) (32.9%) (24.7%) (7.1%)
(33.7%) (38.5%) (18.3%) (9.6%)
ruptured aneurysms. In the bifurcation type, aneurysmal angle and parent–daughter angle were significantly smaller in ruptured aneurysms (P , 0.001). Size ratio type I and II were significantly different between ruptured and unruptured aneurysms (P , 0.001, Table 2).
Multivariate analysis of morphological parameters between the ruptured and unruptured aneurysms The correlations among morphological parameters were identified using the Pearson’s correlation coefficient. Size parameters such as the maximal size (r 5 1), neck size (r 5 0.759), maximal perpendicular height (r 5 0.880), maximal height (r 5 0.934), and maximal width (r 5 0.923) showed a strong correlation with each other, and the maximal size with the largest value of area under the curve (AUC 5 0.837) was included as a representative value in the model. Additionally, the maximal size showed a correlation with the size ratio (r 5 0.777) and multivariate analysis made two models by dividing each parameter. The measured angles (aneurysmal angle, vessel angle, and inflow angle) at the aneurysm were also included
65 68 40 16
X2
P value
0.04
0.978
0.28
0.718
53.40
0.001
1.71
0.635
(34.4%) (36.0%) (21.2%) (8.5%)
in the multivariate model, as the results were significant according to the sidewall and bifurcation type. In multivariate analysis, the aspect ratio [OR, 1.21; 95% CI, 1.05–1.41] and daughter sac [OR, 3.12; 95% CI, 1.05–9.27] were significant parameters in geometries of the aneurysm itself. The size ratio type I [OR, 1.14; 95% CI, 1.05–1.22] and parent–daughter angle [OR, 1.02; 95% CI, 1.00–1.04] were independent parameters in architecture of the aneurysm and surrounding vessels (Table 3).
Comparison of ROC results The AUC value of the aspect ratio was 0.83, at a cutoff value of 1.3, with the sensitivity 75.3%, and the specificity 73.1%. The AUC value of size ratio type I was 0.86, at a cut-off value of 1.8, with the sensitivity 82.4%, and the specificity 81.7% (Fig. 3).
Subgroup analysis of morphological parameters in small aneurysms In the small aneurysm group (N 5 102, 54%), univariate and multivariate analysis revealed that the aspect ratio [OR, 1.57; 95% CI, 1.17–2.11], size ratio type I [OR, 1.04; 95% CI 1.01–1.08], and
Table 2 Univariate analysis for morphological parameters of the aneurysm itself Parameter Morphological parameters I Maximal size (mm) Aspect ratio Bottleneck ratio Height/width ratio Undulation Daughter sac Morphological parameters II Aneurysmal angle Sidewall Bifurcation Vessel angle Sidewall Bifurcation Inflow angle Sidewall Bifurcation Parent–daughter angle Size ratio type I
Ruptured
Unruptured
P value
6.8 ¡ 2.5 1.4 ¡ 0.5 1.5 ¡ 0.4 1.2 ¡ 0.4 51/85 (60.0%) 49/85 (57.6%)
4.0 ¡ 2.0 0.9 ¡ 0.4 1.1 ¡ 0.4 0.9 ¡ 0.3 18/104 (17.3%) 14/104 (13.5%)
0.001 0.001 0.001 0.001 0.001 0.001
1.80 1.31 1.27 1.25 7.18 8.75
(1.49–2.17) (1.20–1.43) (1.16–1.39) (1.13–1.38) (3.67–13.98) (4.31–17.78)
72.2 ¡ 64.0 ¡ 34.9 ¡ 52.7 ¡ 120.8 ¡ 131.4 ¡ 72.8 ¡ 2.6 ¡
77.4 ¡ 74.6 ¡ 14.6 ¡ 55.1 ¡ 97.5 ¡ 139.7 ¡ 91.6 ¡ 1.2 ¡
0.242 0.001 0.033 0.607 0.005 0.087 0.001 0.001
0.98 0.97 1.04 0.97 1.03 0.99 0.97 1.24
(0.95–1.01) (0.95–0.99) (1.01–1.07) (0.98–1.01) (1.01–1.06) (0.98–1.00) (0.96–0.99) (1.17–1.33)
18.5 19.0 39.5 27.3 24.2 29.0 26.6 1.1
15.2 17.7 15.2 22.8 31.2 24.5 27.3 0.6
Neurological Research
OR (95% CI)
2014
VOL .
36
NO .
12
1059
Jeon et al.
Parameters to predict rupture among cerebral aneurysms
daughter sac [OR, 26.49; 95% CI 2.58–271.4] could be independent parameters.
Discussion Natural history and risk factors of cerebral aneurysms have been well studied with prospective and retrospective large-scale data.1–3 These studies have shown that the size and location of the cerebral aneurysm are associated with rupture risk, and family history, hypertension, cigarette smoking, and increasing age have been suggested to increase rupture risk. However, the treatment of unruptured aneurysms remains controversy regarding the decision-making process due to frequent rupture in small-size aneurysms.16,17 With the advancement of imaging modalities, some researches have attempted to predict the rupture risk of cerebral aneurysm through the aneurysmal geometry and hemodynamic effect. Unfortunately, such studies may face challenges, including patient-related confounding factors such as hypertension, smoking, and genetic problems that can affect the results. However, when we investigate ruptured aneurysm among multiple cerebral aneurysms, the issues can be resolved by eliminating confounding factors from the patient’s own internal control. As considering only geometry of the aneurysm itself, Ujiie et al. retrospectively reviewed the twodimensional (2D) angiogram from 129 ruptured and 78 unruptured aneurysms, and introduced the aspect ratio defined as the ratio between the height and neck width of the aneurysm.9 Almost 80% of the ruptured aneurysms showed an aspect ratio greater than 1.6, whereas almost 90% of the unruptured aneurysms showed an aspect ratio less than 1.6. These results suggested that the aspect ratio could be an index for predicting the likelihood of aneurysm rupture at the time the measurements were taken. In our study, the aspect ratio showed similar results, but the threshold (51.3) appeared to be low. The reason is the measurements of 2D angiogram can be more exaggerated due to an overlay effect than 3D reconstructed images. Hoh et al. analyzed 67 multiple cerebral aneurysms to identify the relationship between the aneurysm rupture and morphological features.13 Maximal diameter, height, maximal width, bulge height, aspect ratio, and bottleneck factor were significantly associated with ruptured aneurysms by univariate analysis. Best subsets and stepwise multivariable analysis identified the bottleneck factor and
Figure 3 Comparison ROC results of the most reliable morphological parameters.
height/width ratio as significant parameters to predict a rupture. Our study also confirmed that these parameters were significant on a univariate analysis; however, as compared with other morphological parameters, those parameters were not included in the final model. From biomathematical models of cerebral aneurysms, Chitanvis et al. suggested that undulation in the aneurysm wall, presented by irregular morphology of aneurysm surface, could play an important role in aneurysm rupture.18 Similarly, Beck et al. retrospectively analyzed 2D angiograms from 94 ruptured and 53 unruptured aneurysms and found that ruptured aneurysms were associated with irregular multilobar appearance in aneurysms 5–9 mm in size.4 Both studies suggested that the rupture risk could be related to the geometry of aneurysmal surface. However, overlapping images from the 2D angiogram could induce a rather obscure interpretation between undulation and daughter sac on the aneurysmal surface. On the other hand, it can be easily discriminated from the 3D reconstructed image. In this regard, our result showed that the daughter sac was more associated with the rupture status than the undulation. Several studies have examined the architecture of aneurysms and surrounding vessels for rupture risk at
Table 3 Multivariate analysis for morphological parameters Parameter Morphological parameters I
Aspect ratio Daughter sac Size ratio type I Parent–daughter angle
Morphological parameters II
1060
Neurological Research
2014
VOL .
36
NO .
12
OR
95% CI
P value
1.21 3.12 1.14 1.02
1.05–1.41 1.05–9.27 1.05–1.22 1.00–1.04
0.010 0.040 0.001 0.039
Jeon et al.
the cerebral aneurysm and evaluated how blood flow in the aneurysm is being affected by the surrounding vessel architecture within the main bloodstream. Dhar et al. assumed that the physical vector of the incoming flow can affect the morphological features of the aneurysm and showed that a neck plane of ruptured aneurysms was more tilted from the flow direction of aneurysms.5 Through the CFD analysis, Hoi et al. and Hassan et al. investigated the vessel angle and found that a more curved status caused aneurysm growth and elevations in the wall shear stress by increasing flow impingement in the impact zone of aneurysm neck.11,12 Also, Baharoglu et al. used CFD analysis to study the inflow angle of aneurysms of the sidewall type and showed that, as the inflow angle increased, the inflow zone and dome of the aneurysms occurred at a higher inflow velocity and greater wall shear stress.14,15 On the branched type, Lin et al. reported that a smaller parent– daughter angle at the bifurcation vessel was associated with ruptured aneurysms.6 There seemed that the flow vector could cause a more stressful condition because the smaller composite angle at a branch site can more induce irrotational flow in the aneurysm. In our study, those parameters also demonstrated similar results in ruptured aneurysms. Therefore, it could be considered that geometric change and rupture of aneurysms can be influenced with hemodynamic effect through the surrounding vascular architecture. Of several parameters, the size ratio is known to reflect the relationship between the aneurysmal geometry and surrounding vascular architectures. Initially, Dhar et al. retrospectively examined the 3D image from 20 ruptured and 25 unruptured aneurysms and obtained a significant difference from the ratio of the maximal height and the average size of the vessels around the ruptured aneurysms.5 About 78% of all ruptured aneurysms showed a size ratio greater than 2.05, whereas 83% of all unruptured aneurysms showed a size ratio less than 2.05. Similarly, Tremmel et al. demonstrated a higher size ratio (.2.0) related with multiple vortices and complex flow patterns of the intra-aneurysmal sac from the CFD study.19 In the measurement of size ratio, Rahman et al. adjusted the size ratio to be used in the 2D angiogram by changing the maximal height to maximal size as a more typical method of the aneurysm measurement. They reported that the size ratio type II was correlated with aneurysmal rupture in a blind analysis.7 In this respect, we evaluated which parameter is more significant in the ruptured aneurysms of multiple cerebral aneurysms and size ratio type I was a more predictable parameter at the rupture risk. In order to confirm the most reliable parameter, the forward method of multivariate logistic regression
Parameters to predict rupture among cerebral aneurysms
analysis from various parameters can be the appropriate choice with regard to the statistical analysis. From that result, independent morphological parameters indicate the relative risk of a certain aneurysm among multiple cerebral aneurysms. Thus, if an aneurysm has a larger value than 0.1 units in aspect ratio, the rupture risk increases by approximately 21%, and if the aneurysm only has a daughter sac, the risk increases approximately 3.1 times. A higher value of 0.1 units in size ratio type I can raise the risk by approximately 14%; for one degree smaller in the parent–daughter angle, the relative risk increases 2%. Many large volume studies demonstrated that aneurysm location have related rupture risk.1,2,3,20 While we did not show the difference according to location, it seems that small volume or simplified classification related with major territory may affect the results. Nevertheless, size ratio appeared as a significant parameter. Actually, size ratio takes into account not only the aneurysm size itself, but also vessel geometry by integrating it into a quantifiable parameter. In other words, it indirectly accounts for the influence of aneurysmal location on rupture. Several studies have shown that small sized aneurysms ruptured in a large percentage.4,21,22 Our data also presented that ruptured portion of small aneurysm with 5 mm or less was 11.1% (21/189). It seems that aneurysm size has not yet been completely explained and needs to use additional parameters. In this regard, when we analyzed morphological parameters in the small aneurysms, independent parameters appeared as the aspect ratio, daughter sac, and size ratio type I. There are several limitations in this study. First, it was a retrospective study that the certain cerebral aneurysm has already ruptured condition. So, we do not know exactly whether the aneurysm size has been changed after rupture or undulation of aneurysm surface has been caused by rupture. At least, we have to assume a precondition without change at that time. Second, interpretation of statistical results only showed the relative risk. Third, risk of rupture according to location did not clarify, because of the small volume, and it might induce a bias. However, these results may be meaningful processes in that the several parameters with the risk prediction were compared to their reliability among the multiple cerebral aneurysms. Among several morphological parameters, aspect ratio, daughter sac, size ratio type I, and parent–daughter angle were independent predictors related to the geometry of the cerebral aneurysm and surrounding vessels. Furthermore, if the size ratio type I §1.8, aspect ratio §1.3, the presence of a daughter sac and relatively smaller parent–daughter angle appear, it would be considered
Neurological Research
2014
VOL .
36
NO .
12
1061
Jeon et al.
Parameters to predict rupture among cerebral aneurysms
the first therapeutic target among multiple cerebral aneurysms, if not ruptured.
Conclusion The various morphological parameters were compared under the controlled conditions with ruptured aneurysm among multiple cerebral aneurysms. The aspect ratio, daughter sac, size ratio type I, and parent–daughter angle were reliable predictors of aneurysm rupture.
Disclaimer Statements Contributors Hong Jun Jeon, Jae Whan Lee, So Yeon Kim, Keun Young Park, and Seung Kon Huh are all contributors in this study. Funding None. Conflicts of interest There are no conflicts of interest to declare. Ethics approval The present study was approved on the ethical standards of the Institutional Review Board.
References 1 The International Study of Unruptured Intracranial Aneurysms Investigators. Unruptured intracranial aneurysms – risk of rupture and risks of surgical intervention. N Engl J Med. 1998;339:1725–33. 2 UCAS Japan Investigators, Morita A, Kirino T, Hashi K, Aoki N, Fukuhara S, et al. The natural course of unruptured cerebral aneurysms in a Japanese cohort. N Engl J Med. 2012;366:2474– 82. 3 Juvela S, Porras M, Poussa K. Natural history of unruptured intracranial aneurysms: probability of and risk factors for aneurysm rupture. J Neurosurg. 2008;108:1052–60. 4 Beck J, Rohde S, El Beltagy M, Zimmermann M, Berkefeld J, Seifert V, et al. Difference in configuration of ruptured and unruptured intracranial aneurysms determined by biplanar digital subtraction angiography. Acta Neurochir (Wien). 2003;145:861–5; discussion 5. 5 Dhar S, Tremmel M, Mocco J, Kim M, Yamamoto J, Siddiqui AH, et al. Morphology parameters for intracranial aneurysm rupture risk assessment. Neurosurgery. 2008;63:185–96; discussion 96–7. 6 Lin N, Ho A, Gross BA, Pieper S, Frerichs KU, Day AL, et al. Differences in simple morphological variables in ruptured and unruptured middle cerebral artery aneurysms. J Neurosurg. 2012;117:913–9.
1062
Neurological Research
2014
VOL .
36
NO .
12
7 Rahman M, Smietana J, Hauck E, Hoh B, Hopkins N, Siddiqui A, et al. Size ratio correlates with intracranial aneurysm rupture status: a prospective study. Stroke. 2010;41:916–20. 8 Raghavan ML, Ma B, Harbaugh RE. Quantified aneurysm shape and rupture risk. J Neurosurg. 2005;102:355–62. 9 Ujiie H, Tamano Y, Sasaki K, Hori T. Is the aspect ratio a reliable index for predicting the rupture of a saccular aneurysm? Neurosurgery. 2001;48:495–502; discussion 502-3. 10 Cebral JR, Castro MA, Appanaboyina S, Putman CM, Millan D, Frangi AF. Efficient pipeline for image-based patientspecific analysis of cerebral aneurysm hemodynamics: technique and sensitivity. IEEE Trans Med Imaging. 2005;24:457–67. 11 Hoi Y, Meng H, Woodward SH, Bendok BR, Hanel RA, Guterman LR, et al. Effects of arterial geometry on aneurysm growth: three-dimensional computational fluid dynamics study. J Neurosurg. 2004;101:676–81. 12 Hassan T, Timofeev EV, Saito T, Shimizu H, Ezura M, Matsumoto Y, et al. A proposed parent vessel geometry-based categorization of saccular intracranial aneurysms: computational flow dynamics analysis of the risk factors for lesion rupture. J Neurosurg. 2005;103:662–80. 13 Hoh BL, Sistrom CL, Firment CS, Fautheree GL, Velat GJ, Whiting JH, et al. Bottleneck factor and height-width ratio: association with ruptured aneurysms in patients with multiple cerebral aneurysms. Neurosurgery. 2007;61:716–22; discussion 22–3. 14 Baharoglu MI, Schirmer CM, Hoit DA, Gao BL, Malek AM. Aneurysm inflow-angle as a discriminant for rupture in sidewall cerebral aneurysms: morphometric and computational fluid dynamic analysis. Stroke. 2010;41:1423–30. 15 Baharoglu MI, Lauric A, Gao BL, Malek AM. Identification of a dichotomy in morphological predictors of rupture status between sidewall- and bifurcation-type intracranial aneurysms. J Neurosurg. 2012;116:871–81. 16 Lu HT, Tan HQ, Gu BX, Wu-Wang, Li MH. Risk factors for multiple intracranial aneurysms rupture: a retrospective study. Clin Neurol Neurosurg. 2013;115:690–4. 17 Schievink WI, Piepgras DG, Wirth FP. Rupture of previously documented small asymptomatic saccular intracranial aneurysms. Report of three cases. J Neurosurg. 1992;76:1019–24. 18 Chitanvis SM, Dewey M, Hademenos G, Powers WJ, Massoud TF. A nonlinear quasi-static model of intracranial aneurysms. Neurol Res. 1997;19:489–96. 19 Tremmel M, Dhar S, Levy EI, Mocco J, Meng H. Influence of intracranial aneurysm-to-parent vessel size ratio on hemodynamics and implication for rupture: results from a virtual experimental study. Neurosurgery. 2009;64:622–30; discussion 30–1. 20 Wiebers DO, Whisnant JP, Huston J 3rd, Meissner I, Brown RD Jr, Piepgras DG, et al. Unruptured intracranial aneurysms: natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet. 2003;362:103–10. 21 Rinkel GJ, Djibuti M, Algra A, van Gijn J. Prevalence and risk of rupture of intracranial aneurysms: a systematic review. Stroke. 1998;29:251–6. 22 Forget TRJr, Benitez R, Veznedaroglu E, Sharan A, Mitchell W, Silva M, et al. A review of size and location of ruptured intracranial aneurysms. Neurosurgery. 2001;49:1322–5; discussion 5–6.
