Chapter 68 Management of Unruptured Intracranial Aneurysms

Vikram V. Nayar, K.a.i. Frerichs, Arthur L. Day

Subarachnoid hemorrhage, when caused by the rupture of an intracranial aneurysm, has a mortality rate near 50% at 30 days, and approximately half of the survivors sustain irreversible brain damage.¹ To avoid such a catastrophic event, it is important to identify and treat patients who harbor aneurysms that carry a significant risk of rupture. With the increased use of brain imaging in recent medical practice, including noninvasive tests like CT and MR angiography, a growing number of unruptured and usually asymptomatic intracranial aneurysms are being diagnosed. The decision of whether such lesions should be treated, and if so, whether by surgical or endovascular therapy, has been the subject of great controversy. This chapter evaluates the data available for making management decisions for unruptured aneurysms.

Natural History

Intracranial aneurysms are common, and may generally be classified as saccular (hemodynamic or “berry”) and fusiform (dissecting, infectious, arteriosclerotic, or traumatic) types. The saccular type is by far the most common, and is the focus of this chapter.²^–⁷ Autopsy studies have shown that the overall prevalence of intracranial aneurysms in the general population ranges from 0.2% to 9.9%.⁶^–¹⁰ The population-based incidence of aneurysmal subarachnoid hemorrhage varies from 6 to 21.6 cases per 100,000 persons per year.¹¹^–¹⁶

The decision of whether to treat an unruptured aneurysm is based on the likelihood of its rupture during the patient’s lifetime. The natural history study with the longest follow-up comes from Helsinki, Finland, where Juvela and colleagues reviewed a series of 142 patients with unruptured aneurysms followed without treatment from 1956 to 1978 (median follow-up of 19.7 years).¹⁷ An advantage of the study is that it avoids treatment selection bias, because no unruptured aneurysms were treated in Helsinki before 1979. During 2575 person-years, 33 of the 142 patients (23%) had subarachnoid hemorrhage, resulting in an annual rupture rate of 1.3%. The cumulative rates of rupture were 10.5% at 10 years, 23% at 20 years, and 30.3% at 30 years.¹⁷ Twenty-nine of 33 aneurysms that eventually ruptured were smaller than 10 mm in diameter at the time of the original diagnosis (18 were ≤6 mm).^8,^17,¹⁸ Those aneurysms that ruptured were more likely to have increased in size (≥1 mm) compared to those that did not rupture. Notably, the majority of the patients (131 of 142) had previous subarachnoid hemorrhage from another aneurysm, thus comprising a group of patients which may have had a higher rupture rate compared to those with no prior history of subarachnoid hemorrhage. Nevertheless, the study by Juvela and colleagues provided substantial long-term data on the natural history of unruptured aneurysms, and the annual rupture rate of 1.3% was similar to previously published reports (1% to 2.3%).¹⁹^–²¹

Between 1976 and 1997, Tsutsumi and colleagues observed 62 patients who had noncalcified unruptured intradural aneurysms and no prior history of subarachnoid hemorrhage.²² For small aneurysms (<10 mm), the 5- and 10-year rupture risks were 4.5% and 13.9%, respectively, an annual rupture rate similar to that reported by Juvela and colleagues.^8,^17,¹⁸ For large aneurysms (>10 mm), the 5- and 10-year rupture risks were several-fold higher, 33.5% and 55.9%, respectively.

From 2003 to 2006, Ishibashi and colleagues elected to observe unruptured intracranial aneurysms at their institution. Of a total of 419 patients with 529 aneurysms, 19 aneurysms ruptured during the observation period, resulting in a 1.4% annual rupture rate.²³ Eight of the 19 aneurysms that ruptured were under 5 mm in size.

In 1998, a large retrospective international study, Phase I of the International Study of Unruptured Intracranial Aneurysms (ISUIA), evaluated the natural history of 1937 unruptured aneurysms in 1449 patients.²⁴ Patients harboring at least one unruptured aneurysm were divided into two groups: those with no history of subarachnoid hemorrhage (group 1) and those with a history of subarachnoid hemorrhage from another aneurysm (group 2). Patients in these two groups were not selected for surgical repair for various and often unknown reasons. The mean duration of follow-up was 8.3 years. As shown in Table 68-1, for aneurysms smaller than 10 mm, the annual rupture rate was 0.05% in group 1 (727 patients) and 0.5% in group 2 (722 patients). For aneurysms 10 to 24 mm in size, the annual rupture rate was approximately 1% in both groups. For giant aneurysms (25 mm or larger), the rupture rate was 6% in the first year, and declined thereafter. Aneurysms of the posterior circulation (the vertebrobasilar system) were significantly more likely to rupture than aneurysms of the anterior circulation, with basilar apex aneurysms carrying a relative risk of 13.8 compared to other locations. The study concluded that small unruptured aneurysms, particularly those in the anterior circulation with no history of prior aneurysmal subarachnoid hemorrhage, should be left untreated, especially when the morbidity and mortality rates of surgical repair were considered. This study generated significant controversy because its results were substantially different from previously published reports of a 1% to 2% annual rupture risk.^8,¹⁷^–²³

Table 68-1 Annual Rupture Risks as Reported in ISUIA I and II Studies

In 2003, prospective data from the ISUIA was published as Phase II.²⁵ Aneurysms in this study were categorized by size into four groups: under 7 mm, 7 to 12 mm, 13 to 24 mm, and 25 mm or larger. Noncavernous anterior circulation aneurysms had 5-year cumulative rupture rates of 0% to 1.5%, 2.6%, 14.5%, and 40%, by size category, respectively. Posterior circulation aneurysms (including [Pcom] aneurysms, which are usually considered part of the anterior circulation) had cumulative 5-year rupture rates of 2.5%, 14%, 18.4%, and 50%. Although the two ISUIA studies stratified aneurysm size differently, ISUIA II reported rupture rates that were generally higher than ISUIA I (Table 68-1). For aneurysms in the 7-10-mm size range, ISUIA II suggested an annual rupture risk (0.5%–2.9%) that was several-fold greater than the rate of 0.05% to 0.5% purported by ISUIA I, which grouped together all aneurysms under 10 mm in size for the analysis.

Both ISUIA studies have been widely criticized for underestimating the true rupture risk of intracranial aneurysms, and certain limitations may have affected their results:

1. Selection bias: To be included in either study, a patient first had to be recommended for conservative management by a neurosurgeon. It is quite possible that those aneurysms included in the studies were judged to be very low risk, based on their benign morphology or location (i.e., intracavernous). In ISUIA I, internal carotid artery (ICA) aneurysms represented 42% of the total in group 1 and 27% in group 2; of these, cavernous-segment ICA aneurysms represented 16.9% in group 1 and 9.5% in group 2. In addition, a significant number of patients (32.7% in group 1 and 61.2% in group 2) had very small aneurysms (2–5 mm in size) which could also carry a lower rupture risk. Thus, the ISUIA I study population included large numbers of patients who harbored intracranial aneurysms considered to have little risk of rupture, at least during the study duration.

2. Crossover: In both ISUIA phases I and II, some patients who were initially chosen for observation were later advised to have treatment. In phase II, of the 1692 patients in the observation cohort, 534 patients were switched to therapeutic intervention. It is possible that the management strategy changed because of new symptoms or increased aneurysm size, both of which are risk factors for rupture. If those patients were left untreated, the observation cohort rupture rate may have been much higher.

3. Incomplete follow-up: Mean follow-up for the retrospective phase I study was 8.3 years, and for the prospective phase II study was 4.1 years. Of the 193 patients (11%) who died of causes other than subarachnoid hemorrhage, 52 patients died of intracranial hemorrhage. These patients were excluded from the analysis. It is unknown how these hemorrhages were determined to result from causes other than aneurysm rupture.

Given the disparity between the ISUIA data and other published reports, a population analysis based on prevalence of unruptured aneurysms and incidence of subarachnoid hemorrhage would be helpful. To estimate the prevalence of aneurysms, Winn and colleagues reviewed 3684 cerebral angiography studies performed at the University of Virginia between April 1969 and January 1980.¹⁰ During the pre-CT era, the cerebral angiogram was a commonly performed neuroimaging test for a variety of indications. The authors found 24 asymptomatic unruptured aneurysms in 3684 patients, yielding a prevalence rate of 0.65%. Nearly 80% of the aneurysms were smaller than 10 mm. Because only 53% of the patients underwent a complete angiography study, the authors estimated that the true prevalence ranged from 0.65% to 1.3%. That would mean that in a population of 100,000, between 650 and 1300 people would have an unruptured intracranial aneurysm. Given an annual incidence of subarachnoid hemorrhage of 11 per 100,000 people,²⁶ a person with an unruptured aneurysm can be estimated to have a 0.85% to 1.7% yearly risk of rupture.

Risk Factors for Rupture

Aneurysm Size

It is now generally accepted that there is a strong correlation between aneurysm size and risk of rupture. ISUIA I used a historical classification scheme to separate aneurysms into small (<10 mm), large (10–24 mm), and giant (≥25 mm) categories (ISUIA 1998). ISUIA II separated nongiant aneurysms into three groups: <7 mm, 7 to 12 mm, and 13 to 24 mm (ISUIA 2003). Both studies demonstrated increased rupture rates with larger sizes. The exact size beyond which an aneurysm becomes “dangerous,” however, is unclear. In both clinical and autopsy series, aneurysms that present with hemorrhage are most commonly between 7 and 10 mm in size, and many are smaller than 7 mm.²⁷^–²⁹ Aneurysms smaller than 7 mm in size account for 55% of the aneurysms that ruptured in Juvela’s series, 58% of the aneurysms that ruptured in the study by Ishibashi and colleagues, and 22% of the aneurysms that ruptured in ISUIA II.^18,^23,²⁵ Some aneurysms may enlarge prior to rupture, allowing for the premise that unruptured aneurysms can be followed until a change in size is noted.³⁰ Unfortunately, enlargement may more often occur near the time when the aneurysm ruptures, a quite unpredictable (and potentially fatal) moment.

Aneurysm Location

Both phase I and phase II ISUIA studies showed that posterior circulation aneurysms, especially basilar apex aneurysms, have a higher relative rupture rate compared to those at other sites.^24,²⁵ A similar increased risk was also noted for lesions arising from the Pcom artery, a site traditionally considered to be within the anterior circulation. The distribution of aneurysms described in these series differs markedly from that encountered in ruptured aneurysm series. In the ISUIA studies, cavernous and small parasellar ICA aneurysms are highly represented, while in series dealing with ruptured aneurysms, anterior communicating (Acom) and Pcom aneurysm sites predominate.²⁷ In our personal experience, proximal (paraclinoid) ICA aneurysms are much more common than previously described, and may in fact be the most common aneurysm site, particularly in females. In series of ruptured aneurysms, however, lesions at this site are far less frequent, perhaps indicating that their relationship to and reinforcement by the parasellar dura provides some protection against subarachnoid hemorrhage.

Aneurysm Shape

Several reports suggest that aneurysms with irregular morphology, particularly those that are multilobed with daughter domes, have a significantly higher hemorrhage risk compared to smooth-walled, more spherical lesions.^20,^31,³² In recent years, as aneurysm shapes and origins have become more important in determining “coilability,” several quantifiable parameters have been evaluated for their contribution to rupture risk, including aspect ratio, ellipticity index, nonsphericity index, and undulation index. Of these parameters, aspect ratio (aneurysm height/neck width) has correlated best with rupture risk.^27,³³ Several studies have shown that ruptured aneurysms have higher aspect ratios than unruptured aneurysms, but there is no consensus on a threshold value for increased risk.³⁴^–³⁸

Symptoms Other than Rupture

Unruptured aneurysms may present with cranial neuropathy (particularly oculomotor nerve or optic nerve/chiasmal deficits), ischemia, or other symptoms related to mass effect. New nonhemorrhagic symptoms suggest an acute change in the aneurysm (i.e., expansion), indicating a higher risk of imminent rupture compared to asymptomatic lesions. Data supporting this assertion is scant and retrospective, but in general, symptomatic aneurysms are treated with relative urgency, especially small Pcom aneurysms that cause oculomotor deficits.²⁷

Significant Family History

In families that have multiple members with intracranial aneurysms, an aneurysm’s risk of rupture is higher, and rupture may occur at an earlier age, compared to aneurysms that arise in individuals with no known family history.³⁹ Familial intracranial aneurysms are discussed later in this chapter.

Prior History of Aneurysmal Subarachnoid Hemorrhage

Phase I of ISUIA showed that a small unruptured aneurysm had a tenfold increase in rupture risk if it occurred in a patient with a history of subarachnoid hemorrhage from a different aneurysm, rather than in a patient with no history of subarachnoid hemorrhage.²⁴

Risk Factors for Aneurysm Formation

Age and Gender

Female gender seems to be a risk factor affecting both aneurysm formation and growth, with aneurysms 1.6 times more likely to occur in women than in men.^17,⁴⁰ A series of 1230 autopsies showed two peaks in the prevalence of aneurysms in women, ages 40 to 49 and ages 60 to 69, which correlate with a peak incidence of subarachnoid hemorrhage between ages 40 and 60.¹⁵ Interestingly, in this series, the prevalence of aneurysms in men was unchanged across the range of age groups.

Smoking

Cigarette smoking may hasten the growth of a preexisting aneurysm, and may contribute to an increased rupture rate, with hemorrhage occurring at smaller sizes.¹⁷ In smokers, there is an increased ratio of elastase to alpha₁-antitrypsin in the walls of cerebral arteries, which may contribute to aneurysm formation or rupture.^18,⁴¹

Genetic Conditions

In families in which two people have known intracranial aneurysms, first-degree relatives have a 9% to 11% chance of having an aneurysm in adulthood. Autosomal dominant polycystic kidney disease (ADPKD) is associated with a 15% prevalence of intracranial aneurysms.⁴⁰ Genetic conditions with at least some evidence of having an increased incidence of intracranial aneurysms are outlined in Table 68-2.

Table 68-2 Genetic Conditions Associated with Increased Incidence of Aneurysm Formation

Autosomal dominant polycystic kidney disease

Type IV Ehlers-Danlos syndrome

Hereditary hemorrhagic telangiectasia

Neurofibromatosis type 1

Alpha₁-antitrypsin deficiency

Klinefelter’s syndrome

Tuberous sclerosis

Noonan’s syndrome

Alpha-1,4-glucosidase deficiency

Aneurysm Detection

Transfemoral Cerebral Angiography

Digital subtraction angiography (DSA), with selective injections of dye into the intracranial arteries, has long been the gold standard for imaging intracranial aneurysms. Since cerebral angiography is invasive (procedural stroke risk ranges from 0.07% to 0.5%) and requires a 2 to 6 hour hospital stay after the procedure, it is generally not advocated as a screening procedure.^42,⁴³ A more recent innovation in angiography is three-dimensional reconstruction (3D-DSA). This technology permits rotation of the virtual image in any direction, allowing neurosurgeons and endovascular practitioners to evaluate specific anatomic characteristics of the aneurysm, including the neck width and the relationship of the aneurysm to the parent artery and branch vessels. Both conventional and 3D-DSA reveal only the patent lumen of an aneurysm; heavy calcification or intraluminal thrombosis are not visualized, and CT and MRI provide useful complementary information.⁴⁴

Magnetic Resonance Angiography

The quality and the spatial resolution of noninvasive imaging have significantly improved in recent years, approaching that of DSA. Magnetic resonance angiography (MRA) is useful for screening and follow-up, and in some cases, it is sufficient for treatment planning.⁴⁵ Aneurysms with diameters as small as 2 mm and vessels as small as 1 mm can be detected.⁴⁶ Aneurysms 6 mm or more in diameter have been detected with 100% sensitivity. The sensitivity decreased to 87.5%, 68.2%, 60%, and 55.6% for aneurysms with a diameter of 5, 4, 3, and 2 mm, respectively. Three-dimensional reconstructions are valuable for anatomic evaluation; 3D contrast-enhanced MRA may be superior to 3D time-of-flight MRA in the detection of aneurysms.^44,^47,⁴⁸ Imaging of an aneurysm’s morphology and relationship to branch vessels may be improved with 3-Tesla and 7-Tesla time-of-flight MRAs.^49,⁵⁰ MRA does not expose a patient to radiation risks, and thus its utility is quite attractive in those patients with anticipated multiple studies during their follow-up, as long as the area of interest is seen in sufficient detail by this technology.

Computed Tomography Angiography

Multislice CT scanners allow simultaneous acquisition of as many as 64 slices by using multirow detector systems.⁴⁹ The concurrent acquisition of multiple slices results in a dramatic reduction of scan time. The major advantages of multislice CT are a longer scanning range, shorter scanning times, and a higher two-axis resolution.⁵¹ The high scan speed of multislice helical CT permits scanning with a smaller slice thickness than is possible with conventional helical CT.⁵² As a result, volumetric data with superior resolution in the z-axis can be obtained. Acquired data can be reformatted to provide 3D angiographic images, called CT angiography (CTA). Many physicians routinely use CTA in clinical practice. A recent review reported that the sensitivity of the CTA ranged from 53% for 2-mm aneurysms (Fig. 68-1) to 95% for 7-mm aneurysms. The overall specificity was 98.9%, but there was interstudy heterogeneity.⁵³ A meta-analysis comparing CTA with DSA in the diagnosis of cerebral aneurysms revealed that CTA had an overall sensitivity of 93.3% and a specificity of 87.8%.⁵⁴

FIGURE 68-1 Three-dimensional reconstruction allows small intracranial aneurysms to be detected on CT angiography. The arrow points to a 2-mm aneurysm at the right middle cerebral artery bifurcation. An 8-mm right posterior communicating artery aneurysm is also seen.

The advantages of CTA over DSA are the following:

1. Data can be obtained more quickly and less expensively.

2. CTA provides additional anatomic information (information on surrounding bony structures and the presence of calcium or atheromas).

3. CTA can be used for the rapid planning of craniotomies for clip obliteration, and the preliminary determination of whether aneurysms are suitable for coil embolization.

4. CTA subjects patients to virtually no stroke risk and negligible discomfort.

The disadvantages of CTA are as follows:

1. Less sensitive and specific than the standard method, DSA, for the detection of cerebral aneurysms.

2. Difficulty detecting aneurysms at the skull base, such as those of clinoidal or cavernous segments of the ICA, due to the aneurysms’ proximity to bone.

3. Sensitive to bolus timing, and opacified veins may be incorporated into the reconstructed image,⁴⁴ making interpretation of the arterial tree anatomy difficult.

4. May be inadequate in patients with left ventricular failure due to suboptimal opacification of the intracranial vasculature.⁵⁵

5. Like standard CT, it is subject to motion artifact.

6. Provides no information regarding flow in all phases of the bolus transit in the cerebral vessels.

7. Has not been validated to detect vasospasm or other flow-limiting lesions with the same reliability as DSA.

At present, most cerebrovascular centers still use DSA to plan the surgical treatment of aneurysms. The use of DSA will likely diminish as CT and MR technology continues to improve. Endovascular intervention, which still depends on fluoroscopic technology, will remain closely linked to DSA.

Indications for Treatment

Aneurysmal subarachnoid hemorrhage carries a high fatality rate. In the retrospective part of ISUIA, 66% of the patients whose aneurysms ruptured died (83% in group 1 and 55% in group 2).²⁴ In the natural history study by Juvela and colleagues, 52% of patients whose aneurysms ruptured died.⁸ Tsutsumi and colleagues reported a mortality rate of 86% after subarachnoid hemorrhage.²² Thus, there are clear reasons to obliterate an unruptured aneurysm that has a significant risk of rupture.

Before deciding to treat an unruptured aneurysm, the cumulative risk of rupture over the patient’s expected lifetime needs to be estimated and weighed against the risks of treatment.⁵⁶^–⁶⁰ Juvela advocated that all young and middle-aged patients with unruptured aneurysms should be surgically treated, regardless of the size of the aneurysm.¹⁷ White and Wardlaw suggest that in patients under age 50 with no prior history of subarachnoid hemorrhage, all posterior circulation aneurysms and those anterior circulation aneurysms 7 mm or larger should be treated.⁶¹ In patients over age 50 with no prior history of subarachnoid hemorrhage, they favor treatment for posterior circulation aneurysms larger than 7 mm and anterior circulation aneurysms larger than 12 mm.

In 2000, after the publication of phase I of ISUIA, the Stroke Council of the American Heart Association issued the following recommendations for the management of patients with unruptured intracranial aneurysms.⁶²

The existing body of knowledge supports the following recommendations (options) regarding the treatment of UIAs:

1. The treatment of small incidental intracavernous ICA aneurysms is not generally indicated. For large symptomatic intracavernous aneurysms, treatment decisions should be individualized on the basis of patient age, severity, and progression of symptoms, and treatment alternatives. The higher risk of treatment and shorter life expectancy in older individuals must be considered in all patients and favors observation in older patients with asymptomatic aneurysms.

2. Symptomatic intradural aneurysms of all sizes should be considered for treatment, with relative urgency for the treatment of acutely symptomatic aneurysms. Symptomatic large or giant aneurysms carry higher surgical risks that require a careful analysis of individualized patient and aneurysmal risks and surgeon and center expertise.

3. Coexisting or remaining aneurysms of all sizes in patients with SAH due to another treated aneurysm carry a higher risk for future hemorrhage than do similar sized aneurysms without a prior SAH history and warrant consideration for treatment. Aneurysms located at the basilar apex carry a relatively high risk of rupture. Treatment decisions must take into account the patient’s age, existing medical and neurologic condition, and relative risks of repair. If a decision is made for observation, re-evaluation on a periodic basis with CT/MRA or selective contrast angiography should be considered, with changes in aneurysmal size sought, although careful attention to technical factors will be required to optimize the reliability of these measures.

4. In consideration of the apparent low risk of hemorrhage from incidental small (<10 mm) aneurysms in patients without previous SAH, treatment rather than observation cannot be generally advocated. However, special consideration for treatment should be given to young patients in this group. Likewise, small aneurysms approaching the 10-mm diameter size, those with daughter sac formation and other unique hemodynamic features, and patients with a positive family history for aneurysms or aneurysmal SAH deserve special consideration for treatment. In those managed conservatively, periodic follow-up imaging evaluation should be considered and is necessary if a specific symptom should arise. If changes in aneurysm size or configuration are observed, this should lead to special consideration for treatment.

5. Asymptomatic aneurysms of ≥10 mm in diameter warrant strong consideration for treatment, taking into account patient age, existing medical and neurologic conditions, and relative risks for treatment.

After the above recommendations were issued, prospective results from phase II of ISUIA were published, which showed that 7- to 10-mm aneurysms had a higher rupture rate than was suggested by the phase I study. In light of those results, most neurosurgeons and endovascular practitioners give strong consideration to treating aneurysms ≥7 mm in diameter in patients who are not elderly. As stated in recommendation (4) above, treatment may be favored for some smaller aneurysms, with worrisome anatomic features, young patient age, or significant family history (Table 68-3).

Table 68-3 Factors That Influence Management of Unruptured Intracranial Aneurysm

Favoring Treatment	Favoring Observation
Patient Factors
Age <70	Age >70
Prior SAH from another aneurysm	Significant medical comorbidities
Family history of intracranial aneurysms	Patient preference
Symptoms caused by aneurysm
Size
Size approaching ≥7 mm	Size much <7 mm
Location
Within the subarachnoid space	Intracavernous or clinoidal segment
Posterior circulation aneurysm	Small superior hypophyseal aneurysm
Shape
Irregular with bleb	Regular
Daughter dome	Unilobed
High aspect ratio	Low aspect ratio