CHAPTER 23 Aneurysms
Aneurysms are focal abnormal dilatations of an artery. In 1997, Schievink proposed that “intracranial arterial aneurysms are acquired lesions that are most commonly located at the branching points of the major cerebral arteries coursing through the subarachnoid space at the base of the brain.” He pointed out that “intracranial arteries have an attenuated tunica media and a lack of external elastica lamina” and that “intracranial arterial aneurysms have a thin tunica media or none, and the internal elastica lamina is either absent or severely fragmented.” These observations are generally correct but do not attempt to distinguish among the diverse diseases that give rise to arterial aneurysms. The concept that the arterial bloodstream first “expands” and then “bursts” an “aneurysmal herniation of the wall” is now considered too simple to explain the complex features of arterial aneurysms. Different classification schemes have been based on aneurysm size (small vs. large vs. giant), location (posterior circulation vs. anterior circulation), clinical presentation (ruptured vs. unruptured), morphology (saccular vs. fusiform), or etiology (false or traumatic aneurysms, dissecting aneurysms, flow-related aneurysms, infectious aneurysms). Each of these classifications has advantages, but we will use the etiologic classification in this chapter because therapeutic decision-making can be improved by a better understanding and recognition of the many different lesions grouped together as aneurysms (Fig. 23-1).
FIGURE 23-1 Same morphologic type of aneurysm (fusiform aneurysm) depicted in four different patients. The term fusiform aneurysm corresponds to an enlargement of the entire vessel circumference over a segment of the artery and is a morphologic description rather than a pathomechanism. Although these aneurysms look similar, their pathomechanisms are completely different, inferring different treatment strategies. The four different underlying etiologies are aneurysm in the presence of atrial myxoma most likely resembling a neoplastic aneurysm (A), acute transmural dissection with subarachnoid hemorrhage (B), aneurysmal dilatation present in familial candidasis (C), and fusiform lumen of a partially thrombosed aneurysm following a healed dissecting process (D).
The prevalence of arterial aneurysms is reported to be 3% to 5% in Western populations, which is approximately 10 times higher than the frequency of arteriovenous malformations (AVMs) in the same group. At autopsy, the overall frequency of aneurysms in the general population ranges from 0.4% to 10%. The recent meta-analysis by Rinkel and colleagues found a prevalence of 2.3% in adults. More than 50% of aneurysms identified at postmortem examination are asymptomatic. In Western countries, the average annual incidence of subarachnoid hemorrhage (SAH) is approximately 10 cases per 100,000 people per year. In both autopsy and clinical series, the incidence of arterial aneurysms and of arterial aneurysm–associated SAH increases with age from the third decade, peaking at the sixth decade. The incidence of arterial aneurysms in children is very low, even in familial arterial aneurysms or associated diseases. Intracranial aneurysms are more common in adult women than adult men but more frequent in boys than girls. The relative frequency of females increases systematically with age, so that the male-to-female ratio changes from 3 : 1 in children up to 8 years of age to 1.2 : 1 in the 10- to 20-year-old age group, reverses in the 5th decade (male to female: 0.9 : 1), and reaches 1 : 3 in the seventh decade of life.
The incidence of multiple intracranial aneurysms appears to be extremely variable, depending in part on the patient population; whether the series included surgical, radiologic, or autopsy findings; and whether the aneurysms were ruptured or unruptured. The prevalence of multiple intracranial aneurysms varies between 25% and 31% in autopsy series and 12% to 26% in larger clinical series. Female patients account for 60% to 81% of patients with multiple aneurysms. The internal carotid artery (ICA) and middle cerebral artery (MCA) are most prone to form multiple aneurysms.
A special subgroup of multiple aneurysms is the “mirror-like” or “twin” aneurysms at symmetric sites on each side. Twin aneurysms occur in 5% to 10% of all aneurysm patients but in as many as 36% of all multiple aneurysm patients. They are found on all intracranial vessels but are most common in the MCA. Twin aneurysms may be discovered in the same clinical context as “classic” saccular aneurysms (e.g., during SAH), but their origin is likely to be different. The occurrence of mirror-like aneurysms, their association with familial disease, and their tendency to rupture earlier in life than “classic” aneurysms suggests that congenital weakness of the vessel wall may be an underlying cause (Fig. 23-2).
FIGURE 23-2 A to D, Multiple mirror aneurysms in a single patient who harbored bilateral symmetric MCA aneurysms, bilateral symmetric PICA aneurysms, and an anterior communicating artery aneurysm. The multiplicity and the mirror or twin occurrence suggest a congenital predisposition to be more likely than degenerative causes alone for the appearance of these aneurysms.
The intracranial vascular system is composed of multiple different vascular segments. Each segment may have a unique “segmental identity” that carries with it selective vulnerability to specific triggers. During cephalic vasculogenesis, a defect in migrating cells might be transmitted to the next generation of cells, resulting in a clonal distribution of the defect. The synchronous occurrence of mirror-like aneurysms suggests that these aneurysms develop in segments that share the same defective characteristics. For those reasons, “twin” aneurysms might be a better term than “mirror-like” aneurysms.
Intracranial aneurysms are found approximately twice as often in first-degree relatives of patients with aneurysmal SAH than in the general population. Within families, the greater the number of members affected, the greater the risk that any one member will have an aneurysm, up to four times the incidence in the general population. Females predominate (54% to 70%) in familial SAH. In males, the SAH occurs 7 to 10 years earlier in familial forms of aneurysms than in nonfamilial forms. Sibling pairs with arterial aneurysms tend to bleed within the same decade. Most familial arterial aneurysms arise in the MCA territory (50%), especially on the right side, and tend to bleed at relatively small arterial aneurysm size (<6 mm at rupture). The rate of multiplicity is not higher in familial arterial aneurysms, but collected data suggest that fewer arterial aneurysms remain asymptomatic in SAH families.
It is not recommended to screen the general population or to screen families with only one affected family member (relative risk, 1.8), because the resulting slight increase in life expectancy does not offset the risk of postoperative sequelae. However, screening programs should be implemented for first-degree relatives of patients with familial forms of SAH because these individuals are four times more likely to experience a SAH than the general population.
Aneurysms that present with similar features, such as SAH, may result from entirely different disease processes, including bacterial or viral infections and inflammations, connective tissue diseases, traumatic or spontaneous dissections, neoplasms, radiation therapy, and high-flow cerebral AVMs. Therefore, in this chapter we will differentiate among “classic saccular,” dissecting, infectious, and traumatic aneurysms and will discuss their individual pathophysiologies in specific subdivisions.
Saccular aneurysms are the most frequently encountered form of aneurysms (70%-80%). They typically arise at arterial bifurcations and resemble berry-like outpouchings of the vessel wall. Their etiology and pathophysiology are complex and poorly understood. Genetic factors contribute to their development as underlined by (1) the familial form of aneurysms described previously and (2) the many genetic diseases that manifest with aneurysmal dilatation. Endogenous factors such as aging, elevated blood pressure, vessel wall shear stresses, and anatomic variations of the circle of Willis as well as exogenous risk factors such as cigarette smoking are thought to contribute to the formation and/or rupture of an aneurysm. In addition to these “offensive” factors, there may also be “defective defense mechanisms,” such as improper spontaneous self-repair of a vessel wall. The roles of these different factors in a given individual are unknown. Therefore, the concept of individual host response must be considered when assessing the etiology of an aneurysm in a given patient. This individual host response is dependent on the individual life span and the regeneration capacity of arteries in normal or in pathologic circumstances, at their branching or nonbranching portions.
Aging, hypertension, and/or smoking are thought to lead to general thickening of the intima of arteries, which is most pronounced distal and proximal to branching sites. These thickened intimal “pads” are inelastic and lead to an increased strain on more elastic parts of the vessel wall adjacent to these pads. Increased strain on the vessel leads to an increased activity of metalloproteinases and other extracellular matrix-degrading proteins, culminating in aneurysmal outpouching of the vessel wall. A vicious circle may then ensue as turbulent flow against the outpouching increases the strain on the vessel wall. Such hemodynamic stress may be especially significant at sites of congenital weakness of the vessel wall. For example, aneurysms often form in the vicinity of anatomic variations such as incomplete fusion of the basilar artery or unilateral absence of the horizontal segment of an anterior cerebral artery (A1) (Fig. 23-3). However, these theories are not able to explain why certain patients develop aneurysms and others with similar “offensive exposures” do not. Therefore, individual defects in host response (e.g., a missing defensive line) may be present that finally lead to the formation of an aneurysm.
FIGURE 23-3 A and B, In the absence of a contralateral A1 segment, aneurysms tend to form more often on the ipsilateral anterior communicating ramus. Whether this is due to increased hemodynamic shear stress forces or to a relative immaturity (and therefore increased fragility) of the vascular system is unclear.
The construction and maintenance of blood vessels are the result of complex biologic factors and events that involve repetitive steps and feedback to the vascular tree. The structural integrity of arteries is maintained and modified according to hemodynamic or metabolic demands (e.g., shear stress forces), which are genetically programmed and controlled. Vascular remodeling is an active and adaptive process of structural alteration and includes changes in cellular processes such as cell growth, cell death, cell migration, and production or degradation of the extracellular matrix. Shear stress forces activate specific (genetically programmed) steps that alter the balance of the mediators of remodeling (e.g., metalloproteinases, nitric oxide synthase, platelet-derived growth factor [PDGF], and transforming growth factor β1). Aging decreases the capacity of the vessel wall to adapt to variations in the hemodynamic parameters or to compensate for stressful events. This progressive aging process is an acquired vulnerability of the vessel wall, which will differ in males and females and with patient age. Congenital alterations in these programs could well result in a variant, defective, or absent reconstruction of the vessel wall. Similarly, we believe that the association of arterial variants with arterial aneurysms points to an incomplete maturation process of the arterial wall. The lack of cell selection that such a pattern implies may preserve “weaker,” that is, less mature endothelial cells that will later develop arterial aneurysms when subject to secondary triggers such as hemodynamic stress.
In the individual patient there is a complex interplay between the aggressive “offensive” factors and the host response. Formation of an aneurysm, therefore, indicates failure of the repair system at a given period in time and/or the persistence of the abnormal triggering factors responsible for the failure of the remodeling processes. Therapy can then be targeted either toward augmenting the host’s defensive system (not yet possible) or toward ameliorating local, aggressive “offensive” factors leading to the aneurysm. The ultimate goal is to stimulate the vascular remodeling system to repair the local injury to the vessel wall triggered by the offensive insult. Spontaneous resolution of berry-like aneurysms, seen after correction of abnormal flow dynamics, holds the promise of such therapy in the future (Fig. 23-4).
FIGURE 23-4 A and B, Spontaneous resolution of an ophthalmic artery aneurysm that was associated with a high-flow shunting lesion of the ophthalmic artery that spontaneously regressed after occlusion of the shunt. This example testifies for the repair system of the vascular system after removing locally aggressive factors.
Most saccular aneurysms of the intracranial circulation remain undetected until they rupture and present as SAH. Arterial aneurysms are the most important cause of primary nontraumatic SAH and account for 65% to 85% of all SAH (Fig. 23-5). After acute SAH, recurrent hemorrhage from the aneurysm poses continued serious threat to the patient, with an estimated risk of re-rupture of more than 20% at 2 weeks and 40% at 6 months. Therefore, treatment of ruptured aneurysms is mandatory.
FIGURE 23-5 Subarachnoid hemorrhage. A, On cranial CT, a subarachnoid hemorrhage is typically hyperdense and fills the subarachnoid spaces. B, On FLAIR-weighted images, blood in the subarachnoid space fails to suppress like usual CSF and is therefore bright as well. The distribution of blood may shed light on the location of the aneurysm, which may be of special importance in the presence of multiple aneurysms. C, Intraoperative photos following opening of the dura, with extensive blood in the subarachnoid space and brain swelling (“angry brain”).
Patients typically complain about a history of abrupt onset of the “worst headache of their life.” Loss of consciousness, meningismus, focal neurologic deficits, and nausea may be present. Many patients report earlier, milder headaches over the days preceding the acute event, most likely representing small “sentinel bleeds” from an unstable aneurysm. In a large series of ruptured intracranial saccular aneurysm, a significantly higher rate of rebleeding was found in patients with poor overall clinical status compared with those with good clinical status.
The clinical consequences of SAH are so severe that emphasis should be placed on the prevention of rupture once the aneurysms are detected incidentally. Therefore, it is useful to identify factors that stabilize or destabilize the aneurysm. Factors known to predict increased risk of future hemorrhage include prior SAH from a different aneurysm, large aneurysm size, and location in the posterior circulation, as shown by the International Study of Unruptured Intracranial Aneurysms (ISUIA) study (Table 23-1). Additional factors suggesting increased risk of rupture include familial SAH, smoking, specific aneurysm morphology (e.g., multilobulated berry aneurysms with “daughter” aneurysms or aneurysmal “blebs”), multiplicity of aneurysms, and the presence of arterial variations in the vicinity of an aneurysm (pointing toward an incomplete maturation that may preserve “weaker” cells that may be more prone to rupture).
Despite the known risk of complications from an intracranial aneurysm, the decision not to intervene should certainly be considered in the appropriate clinical circumstances. Once treatment is considered, the choice of treatment should consider first the patient’s needs and, second, the quality of the treatment available at a given center. One should therefore avoid quoting results reported in the literature that may not reflect the individual physician’s experience. Treatment options include surgical clipping and endovascular therapies (e.g., coiling). Depending on the configuration of the aneurysm, additional endovascular techniques such as the balloon remodeling technique or stenting have been proposed. Over the course of the last years, especially since the results of the International Subarachnoid Aneurysm Trial (ISAT) have been published, the endovascular treatment of cerebral aneurysms has gained more and more importance, not only in those aneurysms that are difficult to access surgically (i.e., those in the posterior circulation) but also in uncomplicated aneurysms of the anterior circulation. It was found that patients harboring ruptured aneurysms that were rated by both the neurosurgeon and the interventional neuroradiologist as possible candidates for their respective therapies had better outcomes concerning morbidity, dependency, and mortality when being treated endovascularly at the 2- and 12-month follow-up evaluations. However, the risk of aneurysm recurrence and even rebleeding is higher in coiled aneurysms than in clipped ones, requiring follow-up imaging in specific intervals. Whether this increased risk is counterbalanced by the better immediate outcome after endovascular treatment remains a matter of debate.
CT is a widely available, noninvasive method to detect aneurysms and their major associated complications such as SAH, intraparenchymal hemorrhage, and acute hydrocephalus. CT angiography (CTA) is extremely fast, relatively free from motion artifact, and particularly suitable for an emergency situation with an unstable patient. It is not painful, requires little to no sedation to obtain a diagnostic study, and typically uses limited numbers of personnel, who are usually already in place within the hospital at the time required.
Multislice CT (MSCT) is a technique in which scanning is performed continuously as the CT table is drawn through the gantry (helical or spiral scanning). At least two rows of detector elements (and, at present, as many as 320 rows of detector elements) are employed in the z-axis to create almost isotropic voxels. Although there is no standard protocol, the technique typically entails injection of high-concentration (300 to 400 mg/mL) iodinated contrast media at a dose of 1 to 2 mL/kg. Injection is made through an antecubital vein at a flow of 3 to 5 mL/s up to a total of 60 to 100 mL. The scanned region extends from the C1 vertebra to the vertex. What slice thickness is possible depends on scanner performance (Fig. 23-6). Generally, slice thickness ranges from 0.5 to 1.25 mm, with a reconstruction interval of up to 0.6 mm. In many centers, a scan is started using a triggering technique to optimize acquisition of early arterial phase images. After real-time CT bolus tracking, the region of interest is placed at the internal carotid artery, and scanning is started automatically when the contrast agent in the ICA reaches 80 Hounsfield units (HU). With the use of triggering and real-time CT the increased attenuation of acute subarachnoid blood has no effect on the ability to define the vessels during 3D reconstruction, because it remains lower than the attenuation of the enhanced vessels in all cases. With this technique even small vascular structures can be imaged. The CT data can then be reconstructed to produce maximum intensity projection (MIP) or 3D representations employing shaded surface displays (SSD) or volume rendering techniques (VRT). The main drawback of the MIP images is vessel superposition, which may impair appreciation of vessel relationships. Multiplanar reconstructions of variable thickness may overcome superposition problems. The VRT allows direct 3D analysis; however, bone or venous superposition in the cavernous sinus region may render carotid siphon or posterior communicating artery aneurysms difficult to appreciate. Compared with single-slice CT, the concurrent acquisition of multiple slices and the superior resolution in z-axis in MSCT allows for a dramatic reduction of scanning time and an improvement of visualization of aneurysms.
FIGURE 23-6 CTA (A) and DSA (B) in the detection of a small MCA aneurysm. The major advantages of CTA over other angiographic methods is its added information about the surrounding bony structures, the possible detection of associated calcifications close to the aneurysm neck, the depiction of extraluminal parts of the aneurysms (such as present in partially thrombosed aneurysms), and the accurate 3D demonstration of the aneurysm and its neck. Because CTA is widely available and CT is the first diagnostic choice to detect subarachnoid hemorrhage, many centers perform CTA immediately after CT detection of a subarachnoid hemorrhage for subsequent treatment planning.
All studies that compare CTA with digital subtraction angiography (DSA)—the current gold standard for aneurysm detection—have found that the sensitivity for detecting aneurysms is strongly dependent on the size and location of the aneurysms. Large and medium-sized aneurysms are detected by MSCT in nearly 100% of cases. Small and even medium-sized aneurysms that arise from the intracavernous or supraclinoid carotid artery may be obscured by bony structures or the cavernous sinus and are detected less often. Detection rates vary from 75% to 98% according to different sources. Although further advances with new generation CTs can be anticipated, some aneurysms smaller than 3 mm and close to the carotid siphon near the clinoid processes may still be missed even with the latest CTA technology. Therefore, the negative predictive value is still inadequate, so patients with clinical suspicion but negative CT angiograms will still go on to DSA. DSA can also disclose uncommon causes of SAH, such as dural arteriovenous fistulas, vasculitis, small and micro AVMs, and transmural dissections, which can be missed by CTA. Positive CTA findings, however, show a high specificity and may provide sufficient information to plan therapy, obviating DSA. In our opinion, CTA is the method of choice for studying acutely ruptured aneurysms since it provides immediate triage in the emergency setting.
MR pulse sequences can exploit blood motion to visualize vascular structures directly and without the use of intravascular contrast material. However, MRI and MR angiography (MRA) are seldom used in the setting of acutely ruptured aneurysms because patient monitoring may be difficult, imaging time is typically more extended compared with CT, the field of view is often limited, and specific MR artifacts (discussed later) are likely to result in a lower sensitivity and specificity of aneurysm detection. Although T2-weighted (T2W) fluid-attenuated inversion recovery (FLAIR) sequences are supposed to detect SAH (see Fig. 23-5), artifacts in the perimesencephalic cistern are often present that make identification of true SAH difficult. Nonetheless, MRA may still be useful for screening and follow-up of previously detected aneurysms, because it is noninvasive, uses no ionizing radiation, and requires no iodinated contrast media.
As with any imaging technique, MRA has its own artifacts and problems, which must be recognized to avoid misdiagnosis. MRA usually displays normal anatomy accurately. In the presence of vascular disease, especially aneurysms, however, technical shortcomings may cause inaccurate depiction of aneurysm size, configuration, and neck morphology, all of which are important characteristics for choosing the appropriate therapy.
TOF MRA is usually performed using a flow-compensated gradient-refocused sequence to ‘saturate’ stationary tissues, causing them to show only low signal intensity. Blood upstream of the imaging volume, however, is ‘unsaturated.’ When the unsaturated blood flows into the imaging volume, it is bright compared with the saturated stationary background tissues.
PC MRA creates images by depicting the shifts induced in the phases of moving spins as the blood flows in the presence of ‘flow-encoding’ gradients. Using phase difference images, the signal intensity of the phase shift is proportional to the vector velocity of flow perpendicular to the image plane and stationary background tissue is suppressed. The flow-encoding gradients can be applied in any or multiple directions depending on the desired flow sensitivity.
Although both PC MRA and TOF MRA can visualize normal vessel anatomy, there are intrinsic problems that may degrade MRA quality, especially for aneurysm detection. These problems are complex flow, slow flow, flow stasis and recirculation, and thrombus visualization. These artifacts are more pronounced in small aneurysms (<5 mm). Complex flow due to turbulence, eddy currents, and nonlaminar or vortex flow result in intravoxel phase dispersion and loss of signal on both TOF MRA and PC MRA. PC sequences are sensitive only to a specified range of velocities. Within the aneurysm dome, turbulence causes a broad spectrum of rapidly changing velocities, which, in turn, causes intravoxel phase dispersion and signal loss. The signal loss can potentially lead to misjudgment of the size of the aneurysm. 3D TOF is less affected by complex flow signal loss because of small voxel size and shorter echo time. Concerning slow flow, it is known that flow near the aneurysm dome is often significantly reduced. Recirculation or even flow stasis may occur. The reduced velocities cause increased saturation effects within the imaging volume and decreased signal intensity within the vessel lumen. Intraluminal signal may even become imperceptible and, therefore, lead to a misjudgment in the size of the aneurysm. Flow stasis and recirculation result in signal loss due to saturation effects and intravoxel dephasing. Slow-flow signal loss is particularly a problem with 3D TOF imaging of vessels deep within the imaging volume. The MIP processing technique contributes to the problem because it ignores signal intensities that fall below a certain threshold. With PC techniques, normally effective arterial velocity encoding factors may fail to image slow flow within an aneurysm. Thrombus and tissues with short T1 relaxation times such as fat or blood degradation products can interfere with vascular imaging. Methemoglobin within thrombus has a short T1 relaxation time and generates high signal on TOF images, simulating intraluminal flow, therefore overestimating the size of the aneurysm lumen. Both TOF MRA and PC MRA are subject to signal loss from the magnetic susceptibility effects of deoxyhemoglobin and hemosiderin/ferritin. The patent lumen of the aneurysm may be obscured or the margins of the residual lumen may appear indistinct. Likewise, other tissues with short T1 relaxation times (i.e., fat) may mimic flow on TOF MRA.
Many of these problems can be overcome with CE MRA, which is performed in a manner analogous to conventional contrast angiography. Instead of relying on blood motion to create intravascular signal, a contrast agent is introduced to change the T1 relaxation of blood. Blood can be directly imaged regardless of flow, which alleviates many of the problems inherent to TOF and PC angiography. Contrast medium can reduce spin saturation due to its T1 shortening effect. Shorter echo time then diminishes phase dispersion. Therefore, sensitivity to turbulence is dramatically reduced and in-plane saturation effects are eliminated. The technique allows a small number of slices oriented in the plane of the vessels of interest to image an extensive region of vascular anatomy (equal to the field of view) in a short period of time. At present the major shortcoming of CE MRA is its limited spatial resolution. Small aneurysms are likely to remain undetected using this technique. With the different MR angiographies, sensitivities and specificities of 50% to 100% have been reported depending on the cited source, the technique used, the reconstruction algorithm used, and the employed field strength.
In our practice, we use TOF techniques for screening and evaluating unruptured aneurysms for neck morphology and aneurysm configuration. We especially rely on the raw TOF MRA data. For the follow-up of treated aneurysms both TOF and contrast-enhanced MRA techniques may be helpful.
Despite recent advances in noninvasive diagnostic vascular neuroimaging by CTA and MRA, diagnostic cerebral angiography combined with 3D rotational angiography remains the “gold standard” for evaluating patients presenting with SAH and suspected intracranial aneurysms. DSA and 3D DSA have the highest temporal and spatial resolution and provide the most precise depictions of intracranial vessel morphology and hemodynamic analysis.
In SAH, the aim of diagnostic angiography is to demonstrate or to confirm the presence of an aneurysm and to provide detailed anatomic and hemodynamic information for planning of endovascular or surgical treatment, including precise morphology and size of the aneurysm neck and dome and the presence and location of any perforating vessels arising in relation to the aneurysm. These techniques also demonstrate any concurrent vasospasm, additional aneurysms, or associated vascular lesions. Should the studies prove negative for aneurysm, they must then address vasculitis, cranial dural arteriovenous fistulas, and sinus thrombosis as possible alternate causes for hemorrhage.
Diagnostic angiography after SAH typically requires injections into each of the four major vessels: both internal carotid arteries and both vertebral arteries. However, if injection of a single vertebral artery opacifies the contralateral vertebral artery sufficiently by retrograde filling, with excellent depiction of the intradural segments of the opposite vertebral and the posteroinferior cerebellar arteries, then three-vessel injection may be sufficient. Conversely, if no aneurysm has been detected by four-vessel angiography, the study must be extended to depict both external carotid arteries (six-vessel study) to rule out a dural arteriovenous shunt as the possible cause of the SAH. The global vessel examination usually begins with biplane anteroposterior and lateral acquisitions, using a field of view large enough to cover the intradural arteries and draining sinuses. For aneurysmal detection and pretherapeutic planning, additional 3D angiography is considered the gold standard. If it is not available, additional oblique views with magnification should be obtained for aneurysm detection and detailed pretherapeutic morphologic analysis. The volume of contrast medium injected should be adapted to the vessel caliber and the hemodynamic conditions and may be 9 mL with a flow of 3 mL/s for the internal carotid artery and 12 mL with a flow of 4 mL/s for a large-caliber vertebral artery, allowing reflux in the opposite intracranial vertebral segment. Films are typically acquired at a rate of three frames per second for the arterial phase and may be decreased for the venous phase. Flat panel technology allows for nondistorted high-quality images and a reduced radiation dose. For 3D rotational angiography, an angiographic C-arm rotation around the patient’s head is necessary. Typically, this rotation covers a total angular range of 200 degrees at a speed of 40 degrees/s. A “mask” rotation is performed first to provide a subtraction mask, after which 15 to 20 mL of contrast medium is injected selectively in the studied artery throughout the second rotation. These subtracted images are transferred for further evaluation to a workstation in which volume renderings, surface renderings, and virtual endoscopic views can be generated using dedicated software. This technique provides a precise delineation of the aneurysm in relation to the parent vessel and possible perforators and an optimal analysis of its neck and dome characteristics. Because each 3D projection is defined in space by two angles that appear on the computer screen, the optimal working view for endovascular treatment can be determined. Some artifacts may degrade 3D rotational angiography. Patient movements during rotational acquisition will degrade 3D reconstruction. Metallic objects such as previously placed aneurysm coils may render appreciation of the truly filled aneurysmal lumen difficult. Large aneurysms may fill only partially during rotational acquisition, resulting in incomplete reconstruction and underestimation of lumen size on 3D angiography. Less often, an anterior communicating artery aneurysm may not be visualized clearly on 3D angiography because of flow competition phenomenon from the contralateral A1. Finally, windowing may artificially enlarge or reduce the size of an aneurysm or its neck.
The high diagnostic accuracy of DSA, however, must be weighed against the risk of permanent or transient neurologic deficits, silent microemboli, or potential non-neurologic risks as a direct result of performing the DSA. Non-neurologic risks are mainly hematomas at the puncture site. Most of these are minor hematomas. However, hematomas necessitating blood transfusion or surgery, peripheral emboli, and arteriovenous fistulas may also happen. Other general risks include risks related to iodinated contrast media, especially allergy and nephrotoxicity, and concerns about the radiation dose. The most important risks are neurologic complications related to the angiography. Rates are reported to be as high as 1.8% for angiographies performed for diagnostic workup of SAH and 0.3% for angiographies performed for the diagnostic workup of unruptured aneurysms. Patients with associated atherosclerotic cerebrovascular disease have a higher risk of stroke or silent microemboli from diagnostic cerebral angiography that may be related to difficulties in probing the vessels (elongation of vessels), the presence of fragile arteriosclerotic plaques that might be scraped off the vessel wall (especially during the inversion maneuver when using a Simmons catheter), and the instability of fresh thrombus within ulcerating plaques, which might embolize during the contrast agent injection. Dissections by the catheter or the guidewire may also account for the just-mentioned neurologic deficits.
Despite these considerable risks, in our practice we use DSA and 3D DSA for the workup of all intracranial aneurysms to get the most precise structural and hemodynamic information before an endovascular or neurosurgical procedure. The best working position and the exact aneurysm dimensions can be evaluated, and potential complications of the SAH (vasospasm) or associated vascular diseases can be evaluated.
Aneurysms typically arise at branch points on the parent artery. The branch point may be formed by the origin of a side branch from the parent artery (e.g., posterior communicating artery) or by subdivision of a main arterial trunk into two trunks (e.g., MCA or terminal ICA bifurcation, tip of the basilar trunk). Sidewall aneurysms are rare and if present usually arise at a turn or curve in the artery and point in the direction that the blood would have gone if the curve was not present. At both the branch points and the curves, local alterations in intravascular hemodynamics are present that exert high shear stress forces on those regions that receive the greatest force of the pulse wave. Therefore, the aneurysm dome typically points in the direction of the maximal hemodynamic thrust in the pre-aneurysmal segment of the parent artery. Aneurysms that are encountered on a straight, nonbranching segment of an intracranial artery should raise the suspicion of a dissecting process, because saccular aneurysms are infrequently encountered at these sites. More than 90% of all saccular intracranial aneurysms are located at one of the following five sites:
In the anterior circulation, other common sites of aneurysms are the origins of the ophthalmic, superior hypophysial, and anterior choroidal arteries. In the posterior circulation the origin of the posterior inferior cerebellar artery (PICA), anterior inferior cerebellar artery (AICA), and the superior cerebellar artery (SCA) and the junction of the basilar and vertebral arteries are likely sites of aneurysm formation.
The size, configuration, and neck morphology of the aneurysm and the location of adjacent perforating arteries are important determinants in the decision whether to treat by neurosurgical approaches (e.g., clipping of the aneurysm) or by endovascular techniques (e.g., coiling) and, if endovascular, what kinds of devices and techniques to employ for the procedure. Therefore, the size, configuration, and neck morphology of the aneurysm and the location of adjacent perforating arteries should be described meticulously in the report.
Aneurysms most commonly form along the supraclinoid segment of the ICA immediately distal to the origins of each large branch. Overall, 35% of all intracranial aneurysms arise at five sites along the supraclinoid segment of the ICA:
Aneurysms arising from the ICA close to the ophthalmic artery constitute about 5% of all intracranial aneurysms and are more frequent in women than men. They typically arise from the superior wall of the carotid artery at the distal edge of the origin of the ophthalmic artery close to the roof of the cavernous sinus (Figs. 23-7 to 23-9). At this point, the ICA changes direction from superior toward posterior, so the maximal hemodynamic force is directed toward the superior wall of the carotid artery just distal to the ophthalmic artery. Therefore, these aneurysms project upward toward the optic nerve. Supraophthalmic aneurysms are easily approached endovascularly but may be complicated to expose surgically, because the ophthalmic artery has a variable origin and course and because multiple folds of the dura enclose the region of the optic foramen and clinoid process. Supraophthalmic aneurysms are often large with complex, multilobulated shape. Many are wide-necked aneurysms that may require remodeling techniques. Unruptured aneurysms may become symptomatic due to headaches or compression of cranial nerves.
FIGURE 23-7 Internal carotid artery aneurysm at the origin of the supraophthalmic artery (“ophthalmic aneurysm”) that was incidentally found in a patient who was evaluated for headaches. A, T2W MR sequences demonstrate a flow void that is directed into the subarachnoid space. B, TOF MRA is able to demonstrate the origin of the aneurysm that was verified by DSA (C).
FIGURE 23-8 A, Supraophthalmic ICA aneurysms are often broad based and typically large. The origin of the ophthalmic artery has to be specially considered to avoid erratic emboli entering the artery during endovascular therapies. B, Most often, 3D rotational angiography is necessary to determine the exact localization of the artery in relation to the neck of the aneurysm.
Superior hypophysial artery aneurysms arise just distal to the origin of the superior hypophysial artery from the medial or posterior wall of the ICA where the curvature of the ICA is convex medially (Fig. 23-10). In this location they lie lateral to the pituitary stalk and point medially under the optic chiasm. Medial expansion of the aneurysm may compromise the perforating arteries to the floor of the third ventricle, the optic nerves, the chiasm, the pituitary stalk, and the hypophysial vascular supply.
FIGURE 23-10 Aneurysm of the internal carotid artery that is directed medially into the superior part of the sella and most closely resembles a superior hypophysial artery aneurysm that in this patient led to a circumscribed subarachnoid hemorrhage, because the superior aspect of the aneurysm is not completely extradural but points toward the dural fold surrounding the cavernous sinus. CTA is especially helpful to detect the close spatial relationship between the aneurysm and the bone, whereas 3D rotational angiography can plan the subsequent endovascular treatment that was carried out with a complete and dense coil packing in this patient.
The posterior communicating and anterior choroidal arteries arise from the posterior wall of the ICA, where the ICA forms a posteriorly convex curve as it ascends to its terminal bifurcation under the anterior perforated substance. The most frequent ICA aneurysm is the posterior communicating artery aneurysm, which constitutes about 25% of all intracranial aneurysms (Figs. 23-11 to 23-14). These aneurysms arise near the apex of the posteriorly convex turn, immediately superior to the distal edge of the origin of the posterior communicating artery. They point downward and posteriorly toward the oculomotor nerve, so the posterior communicating artery is usually found inferomedial to the neck of the aneurysm and the anterior choroidal artery is found superior or superolateral to the neck of the aneurysm. The oculomotor nerve enters the dural roof of the cavernous sinus lateral to the posterior clinoid process and medial to a dural band that runs between the tentorium cerebelli and the anterior clinoid process. Posterior communicating artery aneurysms larger than 4 to 5 mm may compress the oculomotor nerve at its entrance into the dural roof, causing ophthalmoplegia.
FIGURE 23-11 This patient had a sudden oculomotor nerve palsy on the right side. A, Nonenhanced CTA demonstrated a dense structure within the subarachnoid spaces within the pentagon of the basal cisterns (arrow). Because of the clinical presentation and this suspicious CT finding, DSA was performed that demonstrated a typical posterior communicating artery aneurysm (B), that, owing to its downward and posterior direction, had led to irritation of the third cranial nerve.
FIGURE 23-12 A to C, Classic posterior communicating artery aneurysm that originated above the distal edge of the origin of the posterior communicating artery that is found inferomedial to the aneurysm. Endovascular therapy was performed that resulted in complete occlusion of the aneurysm with preservation of the parent artery.
FIGURE 23-13 DSA (A) and corresponding surgical aspect (B) of a posterior communicating artery aneurysm (asterisk) and its close spatial relationship to the oculomotor nerve (III) and the tentorium cerebelli. ICA, internal carotid artery; SV, sylvian vein.
FIGURE 23-14 Unenhanced (A) and enhanced (B) cranial CT images in a patient with a spontaneous oculomotor nerve palsy on the left side demonstrating an “additional” vessel within the pentagon of the basal cisterns (arrow in B) that was found to be a classic posterior communicating artery aneurysm on DSA (C