Extradural Internal Carotid Artery Caliber Dysregulation Is Associated With Cerebral Aneurysms
Background and Purpose—Flow-induced hemodynamic forces are critical in extra- and intracranial arterial caliber regulation and have been proposed to mediate intracranial aneurysm (IA) formation and rupture. We hypothesized that vascular structural control may be impaired in patients harboring brain aneurysms and sought to examine any differences in extradural internal carotid artery (ICA) caliber profiles.
Methods—Ninety-six catheter 2-dimensional angiograms were divided into 3 subgroups: (1) ICA leading to IA (n=38), (2) matched contralateral ICA (n=25), and (3) ICA from nonaneurysmal controls (n=33). ICA diameter was measured proximally beyond the bulb (DProx) and distally at the extradural point of maximal dilation (DMaxDist), yielding maximal distal-to-proximal ratio (RMdp).
Results—Unlike non-IA controls that tapered smoothly, ICAs leading to IA consistently demonstrated focal sites of abnormal dilation in the distal cervical or petrous extradural segments. RMdp was greater in ICAs leading to IA compared with non-IA controls (1.17±0.1 versus 1.0±0.08; P<0.0001). Matched-pair analysis showed RMdp to be higher in ICAs leading to IA than the corresponding contralateral ICAs (1.19±0.1 versus 1.07±0.11; P=0.001); RMdp from contralateral ICAs was greater than non-IA controls (P=0.005). Among ICAs leading to IA, women showed higher RMdp (1.11±0.12 versus 1.05±0.11; P=0.02) with no relationship to intradural IA location.
Conclusions—Measurements of the extradural ICA in patients harboring intradural IA suggest an association with a remote upstream abnormal vascular caliber control consistent with a diffuse flow–mediated structural dysregulation showing laterality and sex dependence.
- extradural internal carotid artery
- flow-mediated regulation
- intracranial aneurysm
- luminal caliber
- optimality principle
Intracranial aneurysms (IA) are localized arterial dilations within the cerebrovasculature that, when ruptured, lead to subarachnoid hemorrhage that remains associated with significant case fatality.1 Although several local and systemic processes have been correlated with the pathogenesis of IAs, the mechanisms leading to this arteriopathy remain uncertain.2 Mechanical forces have been well established as important factors contributing to the formation of aneurysms in the absence of significant comorbidities.3 In particular, wall shear stress has been shown to play a central role in this mechanism because of its importance in controlling vessel caliber in accordance with the vascular optimality principle.2–4 This principle, which was previously demonstrated to be at work in the cerebral circulation,4 posits that vessel geometry has evolved to minimize pumping energy expenditure.3–6
Previous research has primarily focused on arterial branching sites and bifurcations that are designed to optimize workload by preserving physiological wall shear stress.4,6,7 However, the optimality principle also applies to the extradural internal carotid artery (ICA), which under normal physiological circumstances exhibits smooth tapering because of constant blood flow and lack of branches along its extradural course. Pathological geometric alterations of the artery have been implied to occur in patients with aneurysms.4,8 Although these abnormalities have been extensively studied in close proximity to the aneurysm, particularly at intradural bifurcations, few studies evaluated the correlation between geometric alterations of the extradural ICA and IA formation.2,9
During diagnostic cerebral angiography, it was subjectively noted that patients with IAs also presented with unexpected focal dilations of the extra- and intracranial ICA (Figure I in the online-only Data Supplement). Invariably, these dilations were noticed upstream of the location of the aneurysms and at considerable distance of what might be considered the vicinity of the lesions. In the region immediately after this focal expansion, the arterial diameter returned to normal. The purpose of this study was to evaluate these differences retrospectively in the caliber along the cervical and extradural ICA, as seen on cerebral angiography, between patients with aneurysms and control subjects. Potential variations in vessel caliber at sites upstream of the aneurysm location may have important clinical implications for the understanding of IA formation and possibly form the basis for a screening tool.
Patient Selection and Demographics
Two-dimensional (2D) biplane diagnostic cerebral angiograms of the cervical carotid artery were obtained using a calibrated high-resolution flat-panel biplane angiography system (Siemens Artis, Malvern, PA). All angiograms performed by the senior author (A.M.M.) between 2009 and 2012 were considered for the study. To be deemed suitable for inclusion, either the anteroposterior or the lateral x-ray angiographic view captured during acquisition was required to include the entirety of the extradural ICA starting from the common carotid artery bifurcation and to exclude the intracranial ICA to avoid inadvertent aneurysm visualization and consequently measurement bias. In addition, only healthy patients aged >30 years were included to ensure adequate age matching with patients harboring aneurysms.
Where available, bilateral angiograms were also included, resulting in a total of 96 angiograms from 51 patients (66.7% women). Thirty-eight angiograms from 33 patients showed the presence of intradural aneurysms (68.4% women; mean age, 56.2 years). Contralateral data were available for 25 patients (72% women; mean age, 56.7 years). Contralateral data were free of any aneurysmal lesions both in the ICA and in distal vessels.
The control group consisted of 33 angiograms from 18 patients (60.6% women; mean age, 51.3 years) in which the postangiographic diagnosis excluded the presence of aneurysms, dissections, stenoses, or arteriovenous malformations. The indications for angiography performed in the control group included epistaxis (n=3), tumor (n=3), venous sinus thrombosis (n=4), vasculitis (n=2), intracranial hemorrhage (n=3), acute stroke (n=2), and WADA test (n=1).
Morphological Measurements and Validation
Manual 2D measurements of the ICA proximal (DProx) and widest distal (DMaxDist) diameters were performed by 2 trained medical operators blinded to the presence of aneurysms and to each other measurements, and the average was reported. Either the anteroposterior or lateral view was used depending on which provided the best visualization while including the entirety of the extradural ICA starting from the common carotid artery bifurcation and excluding the intradural ICA to avoid inadvertent aneurysm visualization. DProx measurement was made along the cervical ICA closest to the carotid bulb where the vessel walls first become parallel at the end of the bulb taper. DMaxDist measurement was made along the cervical, petrous, or extradural cavernous segments of the ICA (Figure 1). The site of each measurement was recorded based on anatomic segments of the ICA,10 and the corresponding maximal distal-to-proximal ratio was evaluated as RMdp=DMaxDist/DProx.
The interobserver variability (2 observers, 96 samples per feature) was analyzed for the proximal and distal measurements using the intraclass correlation coefficient.11 The intraclass correlation coefficient, taking values between 0 and 1, is a statistical measure used to estimate the methodological reliability and reproducibility, with higher values representing higher agreement between sets of observations (>0.75 implying excellent agreement). The intraclass correlation coefficient for the proximal ICA measurement (DProx) was 0.88, whereas the intraclass correlation coefficient for the maximal distal ICA measurement (DMaxDist) was 0.93.
To evaluate the potential confounding effect of windowing variability, additional measurement validation was performed after application of the Laplacian edge detection filter. Details are available in the online-only Data Supplement.
The data were analyzed using JMP statistical software (version 9.02; SAS Institute, Cary, NC). Data are presented as mean±SD. The Student t test was performed independently for individual parameters. The Tukey honestly significance difference (HSD) pairwise test was used to evaluate the statistical significance among patient cohorts. Statistical significance was assumed for P<0.05. Matched-pair analysis was conducted between aneurysmal ICAs and their corresponding contralateral ICAs.
Receiver operating characteristic curve analysis of RMdp was performed, and the area under the curve index was computed. The optimal discriminating RMdp cutoff, as well as corresponding sensitivity and specificity values, was determined for each patient group.
RMdp Is Elevated in Patients With Aneurysms
Univariate statistical analysis for RMdp was performed on 4 subgroups: (1) 38 aneurysms versus 58 contralateral data plus control group, (2) 25 aneurysms versus 25 corresponding contralateral data, (3) 38 aneurysms versus 33 controls, and (4) 25 contralateral angiograms versus 33 controls (Table). The RMdp in the aneurysmal group was significantly greater (1.17±0.1) compared with contralateral (1.07±0.11; P=0.0008) and control groups (1.0±0.08; P<0.0001). Significance was also observed for RMdp in the contralateral group compared with non-IA controls (P=0.005). Notably, an RMdp of <1 (indicating ICA tapering) was not recorded in any of the ICA leading to IA, whereas 5 of 25 contralateral angiograms showed a continuous taper and decrease in ICA diameter distally.
Statistical significance among the patient groups (aneurysm versus contralateral versus control) was assessed using the Tukey HSD pairwise test for comparing RMdp (Figure 2A). RMdp showed a good discrimination performance between aneurysmal ICA and healthy ICA, with an area under the curve of 0.92 (Figure 2B). Age and sex distribution did not differ significantly between groups (Table I in the online-only Data Supplement).
Dependence of RMdp on Aneurysm Location, Type, and Ruptured Status
Of the 38 aneurysmal angiograms, the site of maximal dilation (DMaxDist) was either in the cervical or petrous segments of the ICA in 31 cases (82%), significantly upstream of the associated aneurysm. In 7 cases (18%), DMaxDist was measured along the extradural cavernous segment.
Regarding the aneurysm locations, 7 aneurysms originated from the paraclinoid ICA, 4 from the ophthalmic segment, 5 from the posterior communicating artery, 7 from the middle cerebral artery, 13 from the anterior communicating artery, 1 from the anterior cerebral artery, and 1 from the anterior choroidal artery. There was no significant difference in RMdp between proximal (ICA, posterior communicating artery, anterior choroidal artery) and distal aneurysms (anterior communicating artery, middle cerebral artery, anterior cerebral artery) at 1.18±0.12 versus 1.15±0.09 (P=0.46), respectively, or between bifurcation (n=21) and sidewall (n=17) aneurysms at 1.14±0.09 versus 1.19±0.11 (P=0.12), respectively. RMdp was not significantly different between ruptured (n=10) and unruptured (n=28) aneurysms (1.18±0.12 versus 1.16±0.10; P=0.78).
Multivariate analysis for RMdp was performed, including patient group (aneurysm versus contralateral versus control), sex, and age as variables. The presence of aneurysm was a highly significant factor affecting RMdp (P<0.0001). Overall, women were associated with a significantly larger RMdp (1.11±0.12) compared with men (1.05±0.11; P=0.02). This was also true among ICAs leading to IA, and in this group, women had higher RMdp compared with men (1.20±0.11 versus 1.10±0.06; P=0.01), with no relationship to intradural IA location. No significant age dependence was observed.
To our best knowledge, this is the first study to demonstrate a clear correlation between abnormal caliber changes of the extradural ICA and presence of intradural cerebral aneurysms. The aneurysmal group displayed a significantly greater RMdp, which exhibited sex dependence but did not vary with age. Importantly, the widest measurement used to compute RMdp was consistently recorded at a location significantly proximal to the IA.
Notably, the extradural ICA dilations, while not observed in controls, were also present to a lesser extent in the ICAs contralateral to the aneurysmal side. These observations are in line with other studies2,9 reporting morphological and structural changes in the common carotid artery and extradural ICA, which were also prominent on the contralateral side. Spallone et al9 reported significant changes in the extradural ICA morphology in patients with aneurysms compared with matched controls. Variations in the vasculature, namely kinks, coils, and stenosis, were also commonly noted along the extradural ICA on the side contralateral to the aneurysm.9 Maltete et al2 related a specific extradural ICA phenotype, characterized by hypertrophic remodeling and altered elasticity, to IA formation. Previously, it has been proposed that failure of the vasculature to adhere to the optimality principle is reflected in locally altered bifurcation geometry.4,6,12,13 Our data suggest that pathological alterations in vasculature geometry also occur along the extradural ICA at locations remote from branching points and bifurcations in patients with aneurysms. Despite the localized appearance of the dilation, its presence in the artery contralateral to the IA implies a possible global defect at a histomorphological level within the vessel walls. This indicates the need for future research to examine hemodynamic forces and biomechanical vascular properties outside the immediate vicinity of the aneurysm at a more systemic level.
Additional research is required to validate the effect of artery caliber dysregulation on hemodynamic forces and aneurysm formation and to examine this particular effect in correlation with other possible risk factors. Given that the inclusion criteria required that patients have separate cervical angiograms, this selection methodology may introduce unknown bias and is a limitation of the study. Furthermore, this study is retrospective in nature, and its conclusions will require validation by prospective randomized studies with a larger sample size; only then can its use be determined with respect to prediction of the presence of aneurysm in a clinical setting.
Although flow-related forces have been previously implicated in the genesis of IAs, these correlations focused on cerebral bifurcation sites in the vicinity of the aneurysm. The current study suggests an association between localized dysregulation of the cervical ICA diameter and presence of downstream distal intradural IAs. The results suggest a focally impaired vasoregulatory response with local deviation from the vascular optimality principle at sites significantly distant and upstream from the cerebral aneurysm.
Sources of Funding
This work was supported by a grant from the National Institutes of Health (NIH-R21HL102685). The first author was financially supported by the German National Academic Foundation and the Royal College of Surgeons in Ireland Alumni Fund.
The senior author has received unrestricted research funding from Codman Neurovascular, Stryker Neurovascular, Microvention Inc, Siemens Inc, Ansys Inc, and CD-Adapco Inc for research that is unrelated to the submitted work. The other authors report no conflicts.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.113.001762/-/DC1.
- Received April 11, 2013.
- Accepted September 4, 2013.
- © 2013 American Heart Association, Inc.
- Hop JW,
- Rinkel GJ,
- Algra A,
- van Gijn J
- Rossitti S,
- Löfgren J
- Rossitti S,
- Löfgren J
- Markl M,
- Wegent F,
- Zech T,
- Bauer S,
- Strecker C,
- Schumacher M,
- et al
- Martys JL,
- Wjasow C,
- Gangi DM,
- Kielian MC,
- McGraw TE,
- Backer JM
- Meng H,
- Wang Z,
- Hoi Y,
- Gao L,
- Metaxa E,
- Swartz DD,
- et al