388MEDICINE & HEALTH/RHODE ISLAND

Intracranial Aneurysms: Perspectives On the Diseaseand Endovascular Therapy

Awais Z. Vance, MD, Mahesh V. Jayaraman, MD, Richard A. Haas, MD, Curtis E. Doberstein, MD�

There are approximately 30,000 casesper year in the United States of aneurys-mal subarachnoid hemorrhage (SAH),leaving 60% of patients dead or disabled.However, improvements in treatmenthave occurred in the past few decades; arecent meta-analysis shows decreasingcase fatality rates from 1973 to 2002.1

This article will discuss the epidemi-ology, natural history, diagnosis and treat-ment of intracranial aneurysms. Wehope that the reader will leave with anunderstanding of the strengths and limi-tations of endovascular treatment for in-tracranial aneurysms.

EPIDEMIOLOGY AND NATURALHISTORY

Intracranial aneurysms are present in1-5% of the general population;2 10 to30% of patients will have multiple aneu-rysms.3 The majority of aneurysms occur inthe anterior circulation near the Circle of

Willis with the most common locations indescending frequency being the anteriorcommunicating artery (Acomm), poste-rior communicating artery (Pcomm), andmiddle cerebral artery (MCA).

Ruptured aneurysms most oftenpresent with subarachnoid hemorrhage.After the initial hemorrhage, the highestrisk for rehemorrhage occurs in the first 2weeks at a rate of 1-2%/ day. If untreated,50% of patients will re-hemorrhage within6 months of initial hemorrhage; up to 72%of these patients will suffer death or severedisability.4,5 While historically, patients weretreated 2 to 3 weeks after hemorrhage, theInternational Cooperative study on timingof Aneurysm surgery6,7 showed that earlytreatment was superior. The current stan-dard of care is to treat ruptured intracra-nial aneurysms as soon as reasonably pos-sible after hemorrhage. Upon presentation,patients are assessed clinically using theHunt and Hess grading system, from I

(minimal headache, no other symptoms) toV (comatose, moribund). As would be ex-pected, the lower the grade at presentation,the better the outcome.

RUPTURE RISK OF UNRUPTUREDANEURYSMS

Unruptured aneurysms may presentwith headache, stroke, neurologic defi-cits related to mass effect or as incidentalfindings on imaging. The InternationalStudy of Unruptured Intracranial An-eurysms (ISUIA)8 showed that the riskof rupture was related to the size, loca-tion and history of prior SAH. The 5-year rupture rates are summarized inTable 1. To summarize their findings,larger aneurysms, those in the posteriorcirculation (or Posterior communicatingartery), and those in patients with priorhistory of SAH are all associated withhigher rupture rates. In general, mostsmall (<7mm) aneurysms in the anterior

Coined by Dr. Alexander Margulis in 1967,interventional radiology is a medical spe-cialty devoted to patients’ clinical care inan image-guided, innovative, and mini-mally invasive manner. Dr. Charles Dot-ter first introduced interventional radiol-ogy to the world in 1964, when he percu-taneously dilated a superficial femoral ar-tery stenosis in an 82 year-old woman andaverted an amputation. Since then,interventional radiologists have pioneeredtreatments in many areas of medicine, in-cluding balloon angioplasty and stentingin peripheral vascular disease, catheter di-rected thrombolysis, embolization for ces-sation of bleeding, needle biopsies anddrainages, etc. Many recent and excitingadvances have come in the areas of periph-eral vascular disease, oncology, uterine fi-broid disease, varicose vein management,

What Is Interventional Radiology?Sun H. Ahn, MD, and Mahesh V. Jayaraman, MD

�and dialysis and venous access. Inneurointerventional radiology, endovas-cular therapies for intracranial aneurysms,vascular malformations and atheroscleroticocclusive disease have expanded treatmentoptions for many patients.

Perhaps more important than thetechnical advances, the field has evolvedfrom a procedure-oriented specialty to aclinical practice with commitment to directpatient care. Interventional radiologistsroutinely perform inpatient and outpatientconsultations, make diagnoses, performtreatments, and longitudinally follow theirpatients. This issue highlights some of thelatest advances in the field of interventionaland neurointerventional radiology. Formore information please refer to Society ofInterventional Radiology (www.sirweb.org)and Society of NeuroInterventional Surgery(www.snisonline.org).

Sun H. Ahn, MD, is the Director ofthe Vascular and Interventional Radiol-ogy Fellowship Program at Rhode IslandHospital and an Assistant Professor (Clini-cal) of Diagnostic Imaging.

Mahesh V. Jayaraman, MD, is Assis-tant Professor of Diagnostic Imaging andNeurosurgery.

Disclosure of Financial InterestsThe authors have no financial inter-

ests to disclose.

CORRESPONDENCESun H. Ahn, MDDepartment of Diagnostic ImagingRhode Island Hospital593 Eddy St.Providence, RI 02903phone: (401) 444-5194e-mail:sahn@lifespan.org

389VOLUME 92 NO. 12 DECEMBER 2009

TABLE 1: Prospective 5-year cumulative rupture rates from theInternational study of unruptured intracranial aneurysms (ISUIA)

Aneurysm Size < 7mm 7-12mm 13-24mm >25mmGroup I Group II(no prior (priorSAH) SAH)

Cavernous ICA 0 0 0 3.0% 6.4%Anterior Circulation 0 1.5% 2.6% 14.5% 40%Posterior Circulation 2.5% 3.4% 14.5% 18.4% 50%(incl. PComm)

TABLE 2: Typical indications, advantages and disadvantages for availableaneurysm imaging modalities:

Modality Indications Strengths Weaknesses

MRA • Screening of high risk • No ionizing radiation • Limited sensitivity forasymptomatic patients <5mm aneurysms

• Follow up of previously • No intravenousdetected aneurysms contrast required • Substantial artifact for

• Follow up of previously • Good for coiled clipped aneurysmscoiled aneurysms aneurysms

CTA • Characterize a • Better spatial resolution • Ionizing radiation requiredquestionable MRA then MRAabnormality • Iodinated contrast required

• Occasionally for follow up • Least time consumingof aneurysms difficult to • Less sensitive then DSAvisualize by MRA for <3mm aneurysms

DSA • Suspected aneurysmal SAH • Highest spatial resolution • Minimally Invasive• Treatment planning when

definitive characterization • Can image and treat in the • Ionizing radiation requiredand anatomic definitions same settingnecessary • Iodinated contrast required

circulation in a patient with no prior per-sonal or family history of SAH can befollowed with serial imaging.

Patients with a family history of an-eurysm are also at higher risk for rupturefor small aneurysms.9 While these pro-spective rupture rates are helpful, theultimate decision to treat or observe anunruptured aneurysm depends on sev-eral other factors, including the patient’sage, medical status, family history andpersonal preferences.

IMAGING EVALUATIONThe most appropriate imaging study

for evaluation of intracranial aneurysmsdepends on the clinical presentation. Forpatients with subarachnoid hemorrhage,the most sensitive examination is requiredbecause even very small aneurysms areimportant to detect. For unrupturedaneurysms, a safe (though perhaps lesssensitive) imaging modality is preferredbecause very small aneurysms (<3mm) areunlikely to make a clinical difference. Theavailable imaging modalities include MRangiography (MRA), CT angiography(CTA) and catheter-based digital sub-traction angiography (DSA).

MRA is the least sensitive modality,especially for aneurysms < 3mm in size, butuses no ionizing radiation or intravenouscontrast. The sensitivity of MRA varies withaneurysm size, with MRA being best foraneuryms 5mm or larger. CTA is more sen-

sitive than MRA, especially for aneurysms<5 mm in size. However, the patient is sub-jected to the risks of ionizing radiation andiodinated contrast. In aneurysms greaterthan 5mm in size, the sensitivity for detec-tion is equal to MRA though CTA is supe-rior in characterization of vascular anatomyof adjacent branches.10,11,12, 13, 14

Angiography is the most sensitive ex-amination for intracranial aneurysms. Itis minimally invasive with a low rate ofpermanent neurologic morbidity. Recentstudies have shown that permanent com-plication rates at busy neurovascular cen-ters are extremely low,15 typically much lessthan 0.1%. Angiography’s spatial resolu-tion exceeds that of CTA and MRA byalmost an order of magnitude. Newer 3D-DSA technology allows for detailed recon-structions that can allow for examining theaneurysm from a variety of angles and fa-cilitates treatment planning.

For patients previously treated withsurgical clips, both MRA and CTA haveartifact that precludes adequate evaluationof the treated aneurysm. In those cases,angiography is often necessary. For patientstreated with endovascular coils, MRA is anideal way to follow treated aneurysms, asthere is little artifact caused from the coils.

A common question is: who should bescreened for aneurysms? The current rec-ommendations are to screen patients inwhom at least two immediate relatives havea documented intracranial aneurysm, andall patients with adult polycystic kidney dis-ease.3,16,17 Familial aneurysms represent ap-proximately 15% of all aneurysm patients.

In modern practice, MRA is typicallyused for screening of asymptomatic highrisk patients and for follow up of previ-ously detected or previously coiled aneu-rysms. CTA is often used to characterizea questionable MRA abnormality and oc-

390MEDICINE & HEALTH/RHODE ISLAND

casionally for follow up of aneurysmswhich are difficult to visualize by MRA.DSA is the gold standard and should beperformed on all SAH patients since it isimportant to detect even the smallest ofaneurysms in this population. DSA is alsoused in treatment planning when defini-tive characterization and anatomic defini-tion is needed. These imaging recommen-dations are summarized in Table 2.

TREATMENT OPTIONSThe goal of aneurysm treatment is to

prevent hemorrhage or re-hemorrhage.Treatment should be safe, effective anddurable. The two main options are cran-iotomy with microsurgical clipping orendovascular coiling. Although neurosur-gical clipping is invasive, it has been thetraditional treatment with a long trackrecord, with the major advantage of du-rability. Re-hemorrhage and recurrencerates of surgically clipped aneurysms areextremely low, and routine imaging fol-low up is typically not necessary.

Endovascular therapy for aneurysmshas been performed for several decades, butthe Guglielmi Detachable Coil (GDC,Boston Scientific, Natick, MA) in the early

1990s revolutionized the field. In this pro-cedure, a very small microcatheter is placedinto the aneurysm sac, and progressivelysmaller coils are placed to induce throm-bosis of the aneurysm, all through a smallfemoral arterial access. Initially, coil embo-lization was limited to aneurysms with a nar-row opening (neck) to the aneurysm to fa-cilitate coil placement. However, newer bal-loon and stent-assisted techniques havegreatly expanded endovascular treatmenthorizons. When using balloon-assisted tech-nique, a balloon is inflated at the base ofthe aneurysm in order to prevent coils fromprotruding out of the aneurysm and intothe parent vessel. For stent-assisted coiling,newer endovascular stents specifically de-signed for this purpose have been designed.(Figures 1 and 2) One drawback of stentsis that the required anti-platelet agents canpose a potential problem in acutely rup-tured aneurysms.

COMPARATIVE STUDIES OFCLIPPING AND COILING

When the GDC coil was first intro-duced, the coiling procedure was typicallyreserved for poor Hunt & Hess grade pa-tients. Gradually, however, greater use of

coiling was applied to patients with betterclinical grades. The landmark InternationalSubarachnoid Hemorrhage Trial (ISAT)enrolled patients from 1995 onward,mostly in Europe. Patients with good clini-cal grades who had aneurysms which couldreasonably undergo either surgical orendovascular therapy were randomized.The initial data, published in 2003 (after2143 patients had been randomized),showed a statistically significant reductionin percentage of patients dead or disabledat one year, from 30.9% (surgery) to 23.5%(coiling). Among the initial criticisms ofthe study were the lack of long term fol-low-up, and the fact that surgical experi-ence among the European centers may notbe comparable to those of major US medi-cal centers. However, more recent long-term data have shown that the coiling treat-ment is indeed durable, with a statisticallylower risk of death at 5 years when com-pared with clipping.18 While the risk of re-hemorrhage from a clipped aneurysm wassignificantly lower than that of a coiled an-eurysm, the rates of both were low and didnot overwhelm the early benefit of the lessinvasive procedure. ISAT also establishedthat the rate of recurrent SAH in these pa-

Figure 1. Angiogram imagesfrom a 62 year-old male withan unruptured aneurysm.Lateral (upper left) and Frontal(upper right) angiogramimages demonstrate a 12mmaneurysm (arrows) arisingfrom the internal carotidartery.

Images after endovasculartherapy show coils (arrow,lower left) in the aneurysm.Angiogram following coilingshows no residual filling of theaneurysm (lower right).

391VOLUME 92 NO. 12 DECEMBER 2009

…the currentstandard of care isto treat ruptured

intracranialaneurysms as soon

as reasonablypossible afterhemorrhage.

tients was both from treated aneurysms butalso from other aneurysms, either thosepresent initially or those which formed inthe interim. This stresses the need for long-term follow-up of patients with prior SAH,irrespective of whether they were treatedwith coiling or clipping. While these re-sults from a landmark study cannot be gen-eralized to every patient, they suggest thatif a patient has subarachnoid hemorrhageand an aneurysm that can be treated witheither modality, then coiling is associatedwith a lower rate of death or dependency.A randomized trial at a high volume USneurosurgical center (Barrow NeuroscienceInstitute, Phoenix, AZ) confirmed thesefindings, showing a statistically significantbenefit to coiling over clipping in patientswho could undergo both treatments.19

Unlike ruptured aneurysms, no ran-domized trial compares surgical andendovascular treatment for unrupturedaneurysms. With unruptured aneurysms,the surgical limitations of operating on an

acutely injured brain are not present asthey are in the setting of SAH, and thegoal should be to attempt to give the pa-tient a durable treatment for their aneu-rysm. The ISUIA study, comparing theoutcomes of surgical clipping withendovascular coiling, found that the 1-year morbidity/mortality for endovascularcoiling (9.8%/7.1%) was lower than thatof surgical clipping (12.6%/10.1%).They study found that poorer surgicaloutcome was associated with age greaterthan 50, larger aneurysms, posterior cir-culation location and in patients with his-tory of ischemic cerebrovascular disease.Poorer prognosis associated withendovascular treatment was less depen-dent on patient age and aneurysm sizethough aneurysms >12mm and those lo-

Figure 2. Stent-assisted endovascular therapy of a large aneurysm arising from the left internal carotid artery. Frontal (upper left) and Lateral(upper right) angiograms show the aneurysm (arrows). Note the relatively wide neck (opening) into the aneurysm, as compared with thenarrow neck in Figure 1. In this case, treatment was performed using a stent-assisted technique. Post-coiling images (lower left, lower right)show the coils within the aneurysm, without obstructing parent vessel flow. Note the markers indicating the location of the stent within theinternal carotid artery (arrows, lower left). Post-treatment angiogram shows patent flow through the parent vessel with only minimal fillingof the aneurysm (lower right).

392MEDICINE & HEALTH/RHODE ISLAND

cated in the posterior circulation por-tended a poorer prognosis.8 Higashidaet al performed a retrospective analysis ofdischarge results in 18 states of 2619 pa-tients treated for unruptured aneurysmsat 429 community and academic hospi-tals. They found that endovascular treat-ment of unruptured aneurysms was asso-ciated with statistically significant reduc-tions in morbidity, mortality and decreasedhospital use at discharge.20 However, boththese studies were not randomized, andthe data should be interpreted as such,acknowledging the substantial limitationof patient selection bias.

ENDOVASCULAR THERAPY:LIMITATIONS AND FUTUREDIRECTIONS

The greatest limitations ofendovascular therapy have been the dura-bility of treatment for large and giant an-eurysms, and therefore the implicit needfor follow-up imaging. ISAT showed that

endovascular therapy was indeed durablewith respect to preventing re-hemorrhage,but that patient population was mostly smallaneurysms.18 Murayama et al reported an-eurysm recurrence rates post coiling fromtheir experience at UCLA from 1991-2002.21 They found that progressivelylarger and wider neck aneurysms had pro-gressively higher recurrence rates, rangingfrom 5.2% for small aneurysms (with a smallneck) to 63.2% for giant aneurysms. Themean time to recurrence following treat-ment was 12 months, with most recurrences

by 36 months in a large series.22 They alsofound that the major risk factors for recur-rence included aneurysm size > 10mm andincomplete initial occlusion. Thus one ofthe limitations of endovascular treatmentremains the durability in large and wideneck aneurysms.

The most exciting future develop-ment in endovascular aneurysm therapyis the flow diverting stent (FDS). Unlikestents used to assist in placing coils into ananeurysm, these devices are designed tobe the only therapy used to treat an aneu-rysm. As such, they have a higher degreeof metal coverage than stents currentlyused. The device is placed in the parentvessel harboring the aneurysm, and bydoing so causes marked disruption of theflow vectors leading into the aneurysm.This induces thrombosis of the aneurysm,and animal studies have shown that even-tually a neo-intima grows over the previ-ous opening to the aneurysm. This con-cept is extremely promising, since it ad-

The most excitingfuture development

in endovascularaneurysm therapy

is the flow divertingstent (FDS).

Figure 3. Treatment of a complex, large aneurysm at the vertebro-basilar junctionusing a flow diverting stent. Initial angiogram images prior to treatment (upperleft) show a large aneurysm (arrows) at the confluence of both vertebral arteries toform the basilar artery. Conventional coil embolization or surgery for this lesionwould be difficult. Native images (upper right) show a flow diverting stent (arrows)placed from the right vertebral artery to the basilar artery, in order to disrupt flowacross the neck of this aneurysm and induce thrombosis. Final post treatmentangiogram (lower left) shows stasis of contrast (arrows) well after contrast haswashed out of the normal arteries and veins. It is expected this portion of theaneurysm will slowly thrombose and shrink down over time.

393VOLUME 92 NO. 12 DECEMBER 2009

dresses the issues of recurrence of large andgiant aneurysm. While none of the de-vices are FDA-approved, studies outsideof the US have shown great promise. ThePipeline Embolization Device (Ev3 neu-rovascular, Irvine, CA) is the best studiedto date, with several hundred human casesperformed. (Figure 3) Single center stud-ies have shown very impressive results intreating the large and giant aneurysmsendovascularly using that device.23 Com-plete thrombosis of the aneurysm can takeweeks to months, and eventually the an-eurysm sac collapses and restores a nor-mal parent vessel contour. Remarkably,the small perforating arteries which arisefrom the vessel segment seem to stay patent.Most of the work using FDS has not in-volved bifurcation aneurysms, so their ap-plicability in that situation is unknown.Much work remains to be done with thesedevices, and their use in acutely rupturedaneurysms would require the use of anti-platelet agents. All these issues would needto be addressed before they could sup-plant coils as the major therapy for intrac-ranial aneurysms. If indeed further stud-ies confirm the promising preliminarydata, the landscape for aneurysm therapywill change greatly in the coming decade.

WHAT’S BEST FOR MY PATIENT?One of the most common questions

is: “what’s better: coiling or clipping?”The answer is “it depends...” Certainlythe patient’s clinical condition and medi-cal co-morbidities should be considered.Depending on their morphology and lo-cation, aneurysms may be better suitedfor one therapy over another. While thetrend has been towards a greater percent-age of aneurysms treated endovascularly,many aneurysms are still better suited forsurgical therapy. For example, MCA an-eurysms are often treated surgically ow-ing to their combination of easier surgi-cal exposure than other locations, andtheir tendency to have wide necks andbranch vessel incorporation, anatomicfactors which make endovascular treat-ment more difficult.24 For patients withan aneurysm that can be treated usingeither modality, a frank discussion withthe patient of the benefits and drawbacksof both options is warranted. In our prac-tice, we discuss the cases jointly wheneverpossible and frankly discuss the pros andcons of both therapies with our patients.

CONCLUSIONIntracranial aneurysms are present

in 1-5% of the population, most oftenpresenting with subarachnoid hemor-rhage. For unruptured aneurysms, largersize (especially >6 mm), posterior circu-lation location, personal or family historyof aneurysm rupture are all associatedwith higher prospective rates of hemor-rhage. Treatment of intracranial aneu-rysms has evolved over the past two de-cades, with endovascular coiling offeredas primary therapy in a greater percent-age of patients. This is primarily basedon the results of a large scale randomizedtrial showing improved outcomes withcoiling over surgical clipping in patientswith ruptured aneurysms suited for boththerapies. Both clipping and coiling areeffective at preventing hemorrhage anddurable over the long term, with slightlylower rates of re-hemorrhage with clip-ping compared with coiling. Futureendovascular directions may involve theuse of flow diverting stents, which induceaneurysm thrombosis over time and showearly promise in the most difficult of an-eurysms. Aneurysms should be treated ata center that can offer both surgical clip-ping and coiling in order to result in thebest possible patient outcomes.

REFERENCES1. Nieuwkamp DJ, Setz LE, et al. Lancet Neurol

2009;8:635-42.2. Wiebers DO, Whisnant JP, et al. Lancet

2003;362:103-10.3. Schievink WI. NEJM 1997;336:28-40.4. Hurst RW, RH Rosenwasser. Interventional

Neuroradiology. New York: Informa; 2008.5. Greenberg MS. Handbook of Neurosurgery. New

York: Thieme, 2006.6. Kassell NF, Torner JC, et al. The International

Cooperative Study on the Timing of AneurysmSurgery Part 1. J Neurosurg 1990;73:1, 18-36.

7. Kassell NF, Torner JC, et al. J Neurosurg 1990;73:37-47.

8. International Study of Unruptured IntracranialAneurysms Investigators. Lancet 2003;362:103-10.

9. Broderick JP, Brown RD, et al. Stroke2009;40:1952

10. White PM, Teadsale E, et al. J Neurol NeurosurgPsychiatry 2001;71:322-8.

11. McKinney AM, Palmer CS, et al. Amer JNeuroradiol 2008;29:594-602.

12. Lubicz B, Levivier M, et al. American J Neuroradiol2007;28:1949-55.

13. Deutschmann HA, Augustin M, et al. Amer JNeuroradiol 2007;28:628-34.

14. Adams WM, Laitt RD, A Jackson. Amer JNeuroradiol 2000;21:1618-28.

15. Fifi JT, Meyers PM, et al. J Vascular InterventionalRadiol 2009;20:442-7.

16. Butler WE, Barker FG II, Crowell RM. Neurosurg1996;38:506-15.

17. The Magnetic Resonance Angiography in Rela-tives of Patients with Subarachnoid HemorrhageStudy Group. NEJM 1999;341:1344-50.

18. Molyneux AJ, Kerr R, et al. Lancet 2009; 8:427-33.

19. McDougall CG. Barrow Ruptured AneurysmTrial. American Association of Neurological Sur-geons Annual Meeting, 20-22 Feb 2008. Chi-cago, IL 2008.

20. Higashida RT, Lahue BJ, et al. American JNeuroradiol 2007;28:146-51.

21. Murayama Y, Nien YL, et al. J Neurosurg2003;98:959-66.

22. Raymond J, Guilbert F, et al. Stroke2003;34:1398-403.

23. Lylyk P, M iranda C, et al. Neurosurg2009;64:632-42

24. Jayaraman MV, Do HM, et al. J Stroke Cerebrovas-cular Dis 2007;16:52-6.

Awais Z. Vance, MD, is in his finalyear of residency in the Department ofDiagnostic Imaging and will be pursuingfur ther training in InterventionalNeuroradiology.

Mahesh V. Jayaraman, MD, is Assis-tant Professor of Diagnostic Imaging andNeurosurgery.

Richard A. Haas, MD, is AssociateProfessor (Clinical) of Diagnostic Imagingand Neurosurgery.

Curtis E. Doberstein, MD, is Associ-ate Professor (Clinical) of Neurosurgery andDirector of Cerebrovascular Neurosurgery.

All are at the Warren Alpert MedicalSchool of Brown University.

Disclosure of Financial InterestsThe authors have no financial inter-

ests to disclose.

CORRESPONDENCE TO:Mahesh V. Jayaraman, MDDepartment of Diagnostic ImagingRhode Island Hospital593 Eddy Street, 3rd Floor MainProvidence, RI 02903Phone: (401) 444-5184E-mail: MJayaraman@lifespan.org