Endovascular Treatment of Anterior Circulation Aneurysms

REZA JAHAN, GARY DUCKWILER, AND FERNANDO VINUELA

The prevalence of intracranial aneurysms in the general population is not well known. Most autopsy studies estimate the prevalence to be ~2 to 5% .^1–3 The largest study published thus far found intracranial aneurysms in 1.6% of 87,772 autopsies.⁴ Intracranial aneurysms are a major cause of nontraumatic subarachnoid bleeding, with morbidity and mortality exceeding 50%.^5–8 Neurosurgical clipping has been established as the “gold standard” for treatment of intracranial aneurysms.⁹ Due to the high surgical morbidity and mortality in selected patients, much interest has developed in endovascular treatment of intracranial aneurysms. This technique allowed treatment of aneurysms in high-risk patients. The current technique of endovascular coiling of aneurysms was first performed in 1990 and reported by Guglielmi et al^10,11 Since then, endovascular coiling has been shown to have a major role in management of intracranial aneurysms. Recently, it has been shown to be a safer procedure compared with surgical clipping in selected patients with ruptured intracranial aneurysms.¹²

Despite attempts at understanding the natural history of aneurysms, the fate of an aneurysm is still unpredictable in an individual patient. In general, treatment is undertaken when the risk of treatment does not exceed the predicted morbidity of the aneurysm. In this chapter, endovascular treatment of aneurysms in the anterior circulation will be discussed.

Endovascular Treatment Background

Approximately 90% of intracranial aneurysms occur in the anterior circulation with the majority in the anterior communicating and posterior communicating location.¹³ Generally, endovascular treatment of aneurysms in the anterior circulation consists of occlusion of the aneurysm and the parent artery or obliteration of the aneurysm sac with preservation of the parent artery.

The first reports of endovascular occlusion of a parent artery were described in the 1970s.^14–17 Since then, various authors have substantiated the utility of the technique in the treatment of selected intracranial aneurysms.^18–20 In the anterior circulation, for aneurysms below the circle of Willis off of the internal carotid artery (ICA), the procedure involves temporary balloon occlusion (TBO) of the ICA to assess collateral flow through the circle of Willis. This procedure is performed with the patient fully awake. This allows the operator to perform neurological assessment of the patient while the ICA is temporarily occluded. Abrupt occlusion of the ICA is not performed as it results in a 26% stroke rate and 12% mortality.²¹ Several techniques have been developed to increase the sensitivity of the test and ensure that few patients suffer delayed ischemic symptoms from occlusion of the ICA. These include neurological testing for 15 or 30 minutes, electroencephalographic monitoring, and induced systemic hypotension during balloon occlusion.^{18–20,22–25} Physiological studies to assess cerebral blood flow and cerebrovascular reserve during balloon occlusion have also been performed. These include utilization of transcranial doppler, Tc HMPAO SPECT, and Xenonenhanced CT.^24,26–28 Based on these physiological studies it has been proposed to separate patients undergoing TBO into low-, moderate-, and high-risk patients.^24,29 The high-risk group would be those that do not tolerate the TBO with neurological deterioration during the test. The moderate-risk group would be those that neurologically tolerate the TBO but who on blood flow studies show a significant and asymmetric decrease in cerebral blood flow to less than 30 ml/100 g of tissue per minute. The low-risk group would tolerate the TBO but show a mild asymmetric decrease in cerebral blood flow >30 ml/100 g of tissue per minute). In light of this classification into low-, moderate-, and high-risk patients, two approaches have emerged in the management of patients undergoing TBO; a selective approach and a universal approach. In a selective approach those with moderate or high risk would undergo external carotid to internal carotid artery bypass (EC/IC bypass) prior to ICA occlusion.^25,30 This is justified by studies showing a low risk of ischemic complications in patients undergoing TBO and parent vessel occlusion.^30–32 In addition, longterm follow-up hemodynamic studies of patients who are undergoing elective occlusion of the ICA have shown no diminishment in cerebrovascular reserve.³³

In contrast to the selective approach, in the universal approach, all patients undergoing parent vessel occlusion would undergo EC/IC bypass.³⁴ The universal approach is justified by studies that show that even the low-risk patients as identified by TBO incur a significant risk of long-term ischemic complications following parent vessel occlusion without bypass.³⁴ In addition, there are theoretical advantages of EC/IC bypass in these patients. First, the bypass reduces flow and turbulence through the circle of Willis, increasing the potential for aneurysmal thrombosis and shrinkage. Furthermore, it reduces the risk of development of contralateral flow related aneurysms due to increased circulation through the circle of Willis.^25,34,35

The techniques of TBO and parent vessel occlusion have been described extensively. Early reports involved the use of detachable balloons for testing, followed by occlusion of the vessel with the same balloon.^14–17,20 Alternatively, a separate double-lumen balloon can be used for test occlusion followed by a detachable balloon for occlusion.²⁰ However, permanent occlusion is achieved by utilization of detachable coils. This removes the technical and clinical complications related to premature detachment of the balloon or deflation and distal migration of detachable balloons. The balloons used for test occlusion are typically compliant balloons necessary to minimize damage to the ICA during inflation. Systemic heparinization is necessary to prevent thromboembolic complications during balloon occlusion. Following the procedure, the heparin can be reversed with protamine sulfate or digested.

In the anterior circulation, parent vessel occlusion is typically reserved for wide-neck symptomatic cavernous aneurysms or giant wide-neck aneurysms of the supraclinoid ICA. The objectives of treatment are twofold. The first objective is to prevent bleeding from the aneurysm, particularly with supraclinoid ICA aneurysms. Occlusion of the parent artery allows thrombosis of the aneurysm, and transmural stress due to flow in the aneurysm is removed. To accomplish this, one must ensure that substantial collateral flow to the aneurysm is absent. This may require, at times, occlusion above and below the aneurysm with trapping of the aneurysm to exclude it from the circulation. Furthermore, with giant ophthalmic aneurysms, continued flow to the aneurysm can be seen through the ophthalmic artery following occlusion of the ICA below the ophthalmic artery. In this instance, coils would be first placed into the aneurysm and then allowed to fall into the ICA for occlusion of the parent vessel.

The second objective of parent vessel occlusion is to shrink the aneurysm, thus removing its mass effect. This is particularly true with cavernous aneurysms. Cavernous aneurysms are not generally life threatening, as they rarely cause subarachnoid bleeding.³⁶ The treatment of cavernous aneurysms is aimed at alleviation of symptoms related to mass effect, carotid cavernous fistula, decreased visual acuity, or headaches not responding to medical therapy.^22,36 Most cavernous aneurysms are large or giant at the time of diagnosis and thus require parent vessel occlusion. Small asymptomatic cavernous aneurysms do not pose an immediate risk to the patient and thus can be observed conservatively and only treated if showing evidence of growth or becoming symptomatic.³⁷ Aneurysms treated by carotid occlusion have been shown to involute with reduction of mass effect.^27,37,38

In conclusion, the utilization of parent vessel occlusion in the management of anterior circulation aneurysms must be individualized for each patient to allow maximum protection of the patient from acute and chronic ischemic complications and to minimize technical complications from the TBO, parent vessel occlusion, and EC/IC bypass surgery.

Endovascular Occlusion of the Aneurysm Sac

General Considerations

Over the past decade, endovascular occlusion of aneurysms with microcoils has become an acceptable alternative to surgical clipping. Guglielmi et al first reported utilization of detachable coils in 1991.^10,11 The Food and Drug Administration approved the platinum coil in 1995 for treatment of intracranial aneurysms. Since then, there has been much controversy over the selection of patients for endovascular coiling versus clipping. The results of a major trial comparing endovascular coiling to surgical clipping showed that in a population of patients with subarachnoid bleeding with aneurysms suitable for coiling and clipping, endovascular occlusion was safer at 1 year follow-up.¹² It is most certain that coiling will continue to play a major role in the management of patients with aneurysms. This role will likely grow in the future as improvements in technology address the current limitations of the coiling procedure.

As the technology evolves, the selection criteria for patients undergoing coiling continue to be revised. With the introduction of Guglielmi detachable coils (GDC) in early 1990, patient selection was primarily based on the patient’s medical condition and the location of the aneurysm. In short, this included patients that were not good surgical candidates. Initial success in a population of patients with complete occlusion of the aneurysm was reported in two large studies to be 50%³⁹ and 56%.⁴⁰ Based on early technical experience and angiographic outcomes, it was noted that aneurysm anatomy and morphology play a significant role in the determination of successful occlusion.⁴¹ An aneurysm with neck width of 4 mm or greater is considered wide-necked with initial endovascular occlusion possible in a few of these aneurysms, with high rates of recanalization.⁴¹ Zubillaga et al reported the success rate of initial endovascular occlusion in narrow- and wide-neck aneurysms⁴¹ to be 85% and 15%, respectively. In our own center’s experience of 11 years with 916 aneurysms treated in 818 patients, small aneurysms (less than 10 mm) with small necks (less than 4 mm) had an initial success rate with complete occlusion in 75.4% of cases.³⁹ Complete occlusion was achieved in only 41.2% of small aneurysms with wide necks. On follow-up, small-neck aneurysms had a recanalization rate of 5.1%, whereas wide-neck aneurysms had recanalization rates of 20%.

Neck width is not the only morphological determinant of successful occlusion. For instance, an aneurysm with a 3 mm neck and a 3 mm body diameter would have unfavorable morphology for coiling, although the neck diameter is less than 4 mm. Thus, the ratio of the aneurysm sac width to the width of the aneurysm neck should also be considered. Aneurysms with a sac-to-neck ratio less than 2:1 should be considered wide necked. Debrun et al reported their results of embolization of 339 aneurysms in 329 patients,⁴⁰ dividing the aneurysms into those with sac-to-neck ratio greater or less than 2:1. In a group of patients selected for treatment primarily based on aneurysm geometry, a complete occlusion rate of 80% was noted when the sac-to-neck ratio was at least 2, and 58% when the ratio was less than 2.

The reason for the low initial success rate and recanalization in wide-neck aneurysms is multifactorial. Wideneck aneurysms do not allow dense packing of the aneurysm. Studies have shown that there is a direct correlation between aneurysm packing density and recanalization rates.^42,43 Even in narrow-neck aneurysms only ~20 to 30% of the aneurysm volume is filled with platinum coils, with the remaining 70 to 80% being filled with thrombus.⁴⁴ As the majority of the aneurysm is filled with thrombus, natural thrombolytic processes can act on this thrombus and recanalize the aneurysm. This is particularly true in large and giant aneurysms where the thrombus remains unorganized for a long period following embolization.^44–49 There have been mainly two approaches to address this limitation of platinum coil technology. The first is to hasten the conversion of the thrombus to collagen scar tissue, providing a stable firm scaffold that resists coil compaction and recanalization.⁴⁹ The second is filling the aneurysm cavity with complete or near-complete exclusion of thrombus.^50,51

The first approach, enhancing the organization of thrombus, relies on surface modification of the platinum coil to accomplish this goal. Platinum coils are biologically inert and produce little or no inflammatory response. The utilization of biologically active material as an embolic agent to promote scar formation inside the aneurysm has been reported by Murayama et al⁴⁹ The Guglielmi platinum detachable coils were covered with a bioabsorbable polymer that enhances the intra-aneurysmal inflammatory response. Histologic studies in animals showed that this material enhanced clot maturation. The new coils, called Matrix detachable coils, are now being utilized for embolization of aneurysms. Definitive clinical data to show reduced recanalization rates with Matrix coils remain to be seen. There has not yet been a direct trial comparing the platinum coils to the new Matrix coils.

The second approach, enhanced filling of the aneurysm has been reported by Kallmes et al^50,51 As stated above, recanalization of aneurysms is in part related to the density of the packing with embolic material. As the majority of the aneurysm filled with platinum coils is thrombus, recanalization can occur as the clot is lysed. The hydrocoil embolic system (HES) was developed to address this.⁵¹ The HES is a hybrid hydrogel–platinum coil device with the hydrogel expanding in blood to about three times its initial diameter. Thus, the majority of the aneurysm volume is filled with the hydrogel, with small amounts of thrombus within the aneurysm cavity. Initial clinical experience with the HES has been reported in 11 patients.⁵⁰ Volumetric packing was improved with the HES compared with standard platinum coils. Long-term follow-up studies to show reduced recanalization rates with the HES have not been reported.

Morphological factors other than neck size and sac-toneck ratio have been reported to be important determinants of anatomical outcome after coiling. Gonzalez et al reported results of embolization of anterior communicating artery aneurysms.⁵² Over a 12-year period of treating over 1000 intracranial aneurysms, 135 anterior communicating aneurysms were identified. These were divided into those pointing anteriorly versus those pointing posteriorly with respect to the axis of the pericallosal artery, as described by Proust et al⁵³ In addition, aneurysms were classified with respect to orientation of the neck. This included aneurysms with necks that were arising from the anterior communicating artery (pure anterior communicating artery aneurysms) and those with the neck pointing toward an A1 segment of the anterior cerebral artery (ie, arising from the junction of the A1 and A2 segments). Orientation of the dome and neck had a significant impact on anatomical success and recanalization. For instance, aneurysms with the neck oriented toward an A1 have significant less risk of recanalization, while pure anterior communicating artery aneurysms have a significant risk of recanalization. Anterior communicating aneurysms pointing anteriorly have significant better anatomic results with higher rates of complete occlusion. Posteriorly pointing anterior communicating aneurysms have a higher risk of incomplete occlusion and higher recanalization.

Other configurations unfavorable to successful coiling include the branching pattern near the neck of the aneurysm. This is particularly true of middle cerebral and posterior communicating aneurysms. The branching vessels can obscure visualization of the aneurysm neck, increasing the risk of coil protrusion into the parent artery or adjacent branch vessel. This is often the case with middle cerebral artery aneurysms. A branch arising near the neck of the aneurysm, which commonly occurs with posterior communicating artery aneurysms, is an unfavorable configuration for coiling as well. These facts are well reflected in a reported series of aneurysm embolizations. Debrun et al reported their experience of coiling 339 aneurysms, and middle cerebral and posterior communicating artery aneurysms were underrepresented due to this anatomical limitation with branching at the neck region.⁴⁰ In the randomized International Subarachnoid Aneurysm Trial (ISAT),¹² middle cerebral aneurysms constituted only 14% of the aneurysms treated as the branching pattern at the middle cerebral artery trifurcation resulted in unfavorable configuration for coiling. The majority of patients with middle cerebral artery aneurysms were sent to surgery in the ISAT.

Treatment of Wide-Neck Aneurysms

As the above discussion indicates, the endovascular treatment of cerebral aneurysms with coils has significant challenges. This is particularly true with respect to the treatment of wide-neck aneurysms. Several technical advances have improved the successful treatment of wide-neck aneurysms. These include the introduction of three-dimensional coils, which, with their complex threedimensional shape provided some additional flexibility with respect to the ability to treat patients with wide-neck aneurysms.⁵⁴ Moret et al⁵⁵ developed a balloon remodeling technique by which the parent vessel could be protected with the use of a nondetachable balloon catheter. Then, in 1994, the application of endovascular stents for the treatment of experimental carotid sidewall aneurysms in animals was first reported,^56–58 followed by a report by Higashida et al describing the first use of endovascular stents to treat wide-neck aneurysms in humans.⁵⁹ These technical advances have allowed for endovascular treatment of broad-neck aneurysms that otherwise would not have been amenable to safe treatment with coiling.

The application of an endovascular stent to support the coil embolization deserves further mention, as it has significantly broadened the range of aneurysms amenable to the coiling procedure. Stent-supported coiling of cerebral aneurysms provides several important theoretical and technical advantages. First, the stent provides for protection of the parent vessel during coiling, and this, in part, facilitates more complete aneurysm packing. Second, the stent can produce flow redirection with disruption of the aneurysm inflow and outflow zones contributing to stasis and thrombosis in the aneurysm. Third, the stent may provide a scaffold across the neck of the aneurysm for endothelial growth. Balloon remodeling can often facilitate adequate embolization of complex aneurysms; however, it does not provide the durable advantages achieved with the application of an endovascular stent.

Thus, stent-assisted coiling has been utilized to increase the success of acute aneurysm occlusion and reduce recanalization.^60,61 Initially, only balloon-expandable coronary stents were available for use. Although success in selected cases was reported, the inflexibility of these stents made navigation through the cerebrovasculature difficult. Recently, stents specifically designed for use in the cerebrovasculature have become available.^60,61 Initial experience with the first generation of these stents reported an overall complication rate of 10.7%.⁶¹ Of 56 patients treated, there were three ischemic events and one death related to stent placement. Fiorella et al reported stent-assisted coiling in 19 patients with 22 aneurysms.⁶⁰ Two clinically significant events were reported, both of which were secondary to thromboembolic events related to stent placement. One of these patients died after thrombolysis was attempted, while the other made an excellent recovery.

The stent used in these reports was the Neuroform stent from Boston Scientific Corporation (Natick, MA). The flexibility of this stent greatly facilitates navigation of the tortuous cerebral vasculature not accessible with balloon-expandable stents. In addition, even the tortuous arterial segments themselves, such as the carotid siphon, can be stented. The challenge with this technology has been the actual stent deployment rather than navigation to the site of interest. However, a second-generation Neuroform stent is now available with an improved delivery catheter, which allows for better-controlled stent deployment. Long-term follow-up of patients has not been yet reported, and it remains to be seen whether stent-assisted coiling reduces recanalization rates of wide-neck aneurysms.

Clinical Outcomes of Embolization of Anterior Circulation Aneurysms

Published clinical data suggest that endovascular embolization of ruptured and unruptured cerebral aneurysms carries permanent procedural morbidities of 3 to 9%, mortality up to 8%, and rehemorrhage rates up to 4%.^62–75 Although these studies do not report results based on anatomic location of the aneurysm (i.e., anterior vs. posterior circulation), morbidity and mortality with respect to aneurysm location can be inferred from several reported studies. Murayama et al reported the experience at our institution over an 11-year period with 818 patients harboring 916 aneurysms.³⁹ Six hundred forty-eight aneurysms (71%) were anterior circulation aneurysms. In-hospital morbidity and mortality was reported as 91% of patients having no change or improved clinical condition after embolization. Six percent of patients had a new neurological deficit, and mortality was 3%. Dividing this group into those with acute subarachnoid hemorrhage (SAH) and those with incidental aneurysms, in-hospital morbidity and mortality in those with acute SAH were 7.2% and 6.4% and, in those with incidental aneurysms, 4.5% and 0.8%, respectively. Long-term clinical follow-up was obtained in 768 of the 818 patients from 3 months to 8 years postembolization. Compared with postdischarge from the hospital, of 768 patients, 131 (17.1%) had clinical improvement, 531 (69.1%) were unchanged, and 37 (4.8%) had neurological deficits. Mortality was 5.7%, and the incidence of delayed rupture was 1.6% (12 patients). Ten of the 12 delayed ruptures were large and giant aneurysms. The introduction of the balloonassisted technique was assessed in this series. Since 1996, this technique has improved results of aneurysm embolization, with the greatest benefit noted in small aneurysms with wide necks. In this group of aneurysms, a 60% rate of complete occlusion was noted, compared with 41.2% prior to the introduction of the balloonassisted technique. Most notable in this series is the rate of recanalization of giant aneurysms, which was noted to be higher than 50%. This is an unacceptable rate of recurrence, particularly because the majority of late ruptures in this series were large and giant aneurysms. Parent artery occlusion or direct clipping may be the best treatment currently for the cure of giant anterior circulation aneurysms.

Our center has evaluated coil embolization morbidity and mortality with respect to aneurysm location.⁷⁶ Comparing 585 aneurysms in the anterior circulation to 239 aneurysms in the posterior circulation, morbidity was 5.6% (anterior circulation) versus 7.1% (posterior circulation). Mortality was 1.4% in the anterior circulation versus 3.3% in the posterior circulation. Thus, there was no significant difference in morbidity and mortality as a function of aneurysm location. This is in sharp contrast to surgical experience, where aneurysm location plays a major role in outcome. With current catheter technology, catheterization of intracranial vessels, whether in the anterior or posterior circulation, can be safely performed. Morphological characteristics of the aneurysm are the most important factor determining success in anatomic occlusion and morbidity and mortality rates, as discussed in the previous sections.

Questions regarding morbidity and mortality in relation to any medical procedure are best addressed in a randomized clinical trial. The International Subarachnoid Aneurysm Trial evaluated neurosurgical clipping versus coiling in 2143 patients with ruptured intracranial aneurysms.¹² These were patients who had ruptured aneurysms that were suitable for neurosurgical clipping and endovascular coiling. The majority of patients randomized in the study had anterior circulation aneurysms. Of the 2143 aneurysms randomized, 2.7% were of posterior circulation. The remainder was anterior circulation aneurysms with 50.5% anterior cerebral artery, 32.5% internal carotid artery, and 14.1% middle cerebral artery. The majority of the patients were of good clinical grade, and the median age in each arm of the study was 52 years. At 1 year, data were available on 1594 patients with poor outcome defined as Modified Rankin Scale 3 to 6 in 23.7% of endovascular patients and 30.6% of surgical patients. The relative and absolute risk reductions in dependency or death in favor of endovascular treatment were 22.6% and 6.9%, respectively. The risk of rebleeding after 1 year was 2 per 1276 and 0 per 1081 patient-years for patients allocated endovascular and neurosurgical treatment, respectively. It must be emphasized that this risk of rebleeding has been taken into account in the final analysis of outcome at 1 year, with endovascular treatment showing better clinical results. The ISAT study with over 4000 patientyear follow-ups now has assessed the risk of rebleeding in the endovascular versus surgical patients to be 0.15% and 0.07% per year, respectively. Thus, it is calculated that to lose the benefit gained by endovascular treatment due to rebleeding would require at least 70 years.

There has yet been no study to compare endovascular and surgical treatment of unruptured aneurysms. In addition, there has been no randomized trial to compare endovascular and surgical treatment of aneurysms in poor-grade patients. As mentioned above, the majority of patients in the ISAT were of good clinical grade. There have been single-center studies that have addressed this issue. Groden et al⁷⁷ compared operative and endovascular treatment of anterior circulation aneurysms in poor-grade patients (Hunt and Hess grade IV and V). They reported on the treatment of 40 patients, 21 treated surgically and 19 treated by endovascular means. The incidence of cerebral vasospasm did not differ significantly between the two groups. There was one surgical- and three endovascular procedural-related complications with clinical consequences. Clinical outcome was assessed at 6 months. Good outcome was obtained in six (29%) of the surgically treated patients and six (30%) of the endovascularly treated patients. Outcome was similar after surgical and endovascular approaches. In this population, surgery was the treatment of choice in patients with intracerebral bleeding with mass effect for whom hematoma evacuation may be necessary. Alternatively, endovascular treatment is considered in certain poor-grade patients within the first 72 hours of hemorrhage where an edematous brain is a major argument against early surgery. The risks of surgery in the elderly would also indicate a endovascular approach in this population, as would the presence of severe vasospasm requiring an angioplasty, which can be performed in the same session that the embolization is done.

Conclusion

Endovascular treatment of aneurysms was initially introduced to provide a viable option for patients considered a high risk for surgical treatment. The technology has advanced rapidly to where endovascular therapy is now playing a major role in the management of patients with cerebral aneurysms. The selection of patients for endovascular therapy is continuing to be defined. Techniques of balloon-supported and stent-assisted coiling have improved anatomical outcomes following embolization. In addition, coil modifications such as the polymer-coated coils are now introduced to further improve immediate and long-term anatomical results of coiling procedure. To become the first line of treatment for cerebral aneurysms, endovascular embolization must provide a permanent occlusion of the treated aneurysm. A single endovascular tool is not likely to accomplish this but, rather, a combination of techniques such as stents and polymer-coated coils. A randomized trial in a population of patients with ruptured aneurysms has shown improved clinical outcome at 1 year in the endovascularly treated group compared with the surgically treated group. Randomized trials to compare the endovascular and surgical treatment of unruptured aneurysms are in the planning stages.

References

1. Chason JL, Hindman WM. Berry aneurysms of the Circle of Willis. Neurology 1958;8:41–44

2. de la Monte SM, Moore GW, Mong MA, Hutchins GM. Risk factors for the development and rupture of intracranial berry aneurysms. Am J Med 1985;78:957–964

3. Di Bonito L, Giarelli L. Pathogenesis of cerebral hemorrhage considered in the light of objective lesions in the arteries of the brain (in 226 cases studied by autopsy). Minerva Med 1975;66: 4391–4398. Italian