J Neurosurg 77:515-524, 1992 Endovascular treatment of posterior circulation aneurysms by electrothrombosis using electrically detachable coils GUIDO GUGLIELMI, M.D., FERNANDO VINUELA, M.D., GARY DUCKWILER, M.D, JACQUES D1ON, M.D., PEDRO LYLYK, M.D., ALEX BERENSTEIN, M.D., CHARLES STROTHER, M.D., VIRGIL GRAVES, M.D., VAN HALBACH, M.D., DOUGLAS NICHOLS, M.D., NICK HOPKINS, M.D., ROBERT FERGUSON, M.D., AND IVAN SEPETKA, M.S. Department of Neurological Sciences (Therapeutic Neuroangiography), University of Rome, Rome, Italy; Departments of Radiology, University of California, Los Angeles, California; New York University, New York, New York; University of Wisconsin, Madison, Wisconsin; University of California, San Francisco, California; Mayo Clinic, Rochester, Minnesota; University' of Virginia, Charlottesville, Virginia. and Baptist Memorial Hospital, Memphis, Tennessee; F.L.E.N.L, Buenos Aires, Argentina; Department of Neurosurgery, State University of New York, Buffalo, New York; and Target Therapeutics, San Jose. California v" In a multicenter study, 120 patients with intracranial aneurysms presenting a high surgical risk were treated using electrolytically detachable coils and electrothrombosis via an endovascular approach. The results of treatment in patients with posterior fossa aneurysms (42 patients with 43 aneurysms) are presented. The most frequent clinical presentation was subarachnoid hemorrhage (24 cases). The clinical follow-up periods ranged from 1 week to 18 months. Complete aneurysm occlusion was obtained in 13 of 16 aneurysms with a small neck and in four of 26 wide-necked aneurysms. A 70% to 98% thrombosis of the aneurysm was achieved in 22 of 26 aneurysms with a wide neck and in three of 16 small-necked aneurysms. One aneurysm could not be treated due to a technical complication. Two cases required postprocedural surgical clipping of a residual aneurysm. One patient (originally in Hunt and Hess Grade V) experienced procedural rupture of the aneurysm requiring an emergency parent artery occlusion. He eventually died 5 days later. Another patient (originally in Grade IV) had coil migration and posterior cerebral artery territo~ ischemia. A third patient developed a permanent neurological deficit (hemianopsia) after complete occlusion of a wide-necked basilar bifurcation aneurysrn. One patient, harboring an inoperable giant basilar bifurcation aneurysm, died from aneurysm bleeding 18 months after partial occlusion. Overall morbidity and mortality rates related to treatment were 4.8% (two cases) and 2.4% (one case), respectively (2.6% and 0% if considering only patients in Hunt and Hess Grades I, II, and III). It is suggested that this technique is a viable alternative in the management of patients with posterior fossa aneurysms associated with high surgical risk. Longer angiographic and clinical follow-up study is necessary to determine the long-term efficacy of this recently developed endovascular occlusion technique. Close postoperative angiographic and clinical monitoring of patients with wide-necked subtotally occluded aneurysms is mandatory to check for potential aneurysmal recanalization, regrowth, and rupture. 9 aneurysm 9 multicenter study 9 posterior circulation 9 endovascular therapy 9 electrothrombosis 9 electrolysis 9 detachable coil KEY WORDS NEURYSMS of the posterior circulation account for 5% to 10% of all intracranial aneurysms. 7 Because of the difficulties in surgical intervention in the posterior fossa, the outcome in these cases is poorer than for aneurysms of the anterior circula- A J. Neurosurg. / Volume 77 / October, 1992 tion. 5~ In an effort to widen the alternatives for therapeutic management of intracranial aneurysms, an endovascular approach utilizing detachable platinum coils and electrothrombosis has been developed? '2 To date, 120 patients have been treated with this technique in a 515 G. Guglielmi, et a/. TABLE 1 Pr+pfileof 38 patiet+t+ with po,+to.w circulation (l ] ~[Jlt~Ii'LS]~+.~pre.sentin~, in good ttew+o/o~ica]condition * AneuCase Age, Clinical No. (yrs), PresenSex tation Aneurysm Size+ rysm Location Diameter Neck lndication for Treatment ('oil Length (cm) Result ( ~ Occlusion) Complications Related to Teehnique Time to Postop Anglogram Result (% Occluzion) Time to FollowUp Anglogram Result (% Occlu- come Outsion) 1 33, F SAH PICA small small MI 16 80 none 6 days 80 2,5 mos 80 2 60, F mass bas bif giant wide inoperable 255 70 none I & 5 wks 70 8 mos 70 good deadw 3 58, M SAH VBjcl large wide MI 180 95 none I wk 95 -- -- good 4 46, M SAH bas bif large wide prey explored 120 70 none 3 wks 80 -- -- good 5 56, F SAH bas bif small wide partially clipped 38 70 none 1 & 3 wks 689 mos 1 yr 70 95 95 2 mos --- 70 --- good good~ good:l: 6 69, F SAH bas bif small wide prey explored 15 95 none 1 wk 95 7 mos 70 good 7 57. F SAH bas bif small small prey explored 55 100 none 4 days/1 mo 100 14 mos 100 good 8 44, F mass bas bif giant wide high risk 155 85 none I wk 95:~ good 9 70, F SAH SCA small wide prey explored 115 90 none 5 mos 80 7 mos 95:1: good 10 56, M SAH bas bif small wide med condit 120 80 thrombus 6 hrs progressing to PCA occlusion 100 -- -- hemianopsia II 29, F incid find has bif small small high risk 23 100 none 1 & 3 wks 100 4 mos 100 good 12 34, M mass has bifP~ (PT) giant wide inoperable 200 95 none 4 days 90 3 mos 85 good 13 42, F SAH VBjct small small partially clipped 14 100 none 3 mos 100 -- -- good incid find VB jct small small prey explored 4 100 none -- -- good 14 70, M mass bas bif (MT) giant wide inoperable 53 98 none 3 wks 95 6 mos 95 good 15 56, F bas bif large wide prev explored 255 98 none I ~- mos 95 7 mos 100:~ good 16 38, M mass VB jcl giant wide inoperable 400 50 none 3 & 8 days 4 mos 95 good 17 56, F incid find bas bif large small endocarditis 195 100 none . 18 52, F SAH AICA small small high risk 41 98 none 7 days 98 -- -- 19 62, M mass VB jct large wide inoperable 60 80 none 1 mo 85 4 mos 98:~ good 20 47, F mass bas bif large wide high risk 275 90 none 3 days 95 -- -- good 21 48. M incid find PICA small small hearl disease 18 100 none 6 days 100 -- -- good 22 43, F mass bas bifP~ large wide high risk 150 90 none 4 days 95 -- -- good 23 55, F mass bas bif large small prey explored 70 95 none . . . . good 24 45. F mass bas bit'P~ (PT) giant wide prey explored 335 80 none . . . . good 25 50, M incid find SCA small small high risk 60 100 none . . . . good mass I 85 . . . deadl[ good * Patients in H u n t and Hess ~ Grade I, I1, or IlL SAH = subarachnoid hemorrhage: PICA = posterior inferior cerebellar artery: MI = myocardial infarction; mass = mass effect; bas bif = basilar bifurcation: VB jet = vertebrobasilar junction; prey explored = previously explored: SCA = superior cerebellar artery: med condit = medical condition; PCA = posterior cerebral artery; incid find = incidental finding; P~, P2, P3 = first, second, a n d third segments of PCA; AICA = anterior inferior cerebellar artery: PT = partially thrombosed; MT = mostly thrombosed. - - = procedure not performed. t Aneurysm size (diameter): small = < 12 ram: large = 12-25 mm; giant = > 25 ram. Aneurysm neck size: small = < 4 mm; wide = > 4 m m . Additional embolization procedure for more coils. wDied because of bleeding from aneurysm 18 months after treatment. II Died 1 m o n t h after treatment because of marantic endocarditis. T A B L E 1 (conIimted) ---> 51 6 J. Neurosurg. / Volume 77 / October, 1992 Endovascular thrombosis of aneurysms TABLE 1 (crmtmlwd)* Compli- Time Coil Result cations to Result Length (% Occlu- Related Postop (% Occlu(cm) sion) to Tech- Angiosion} nique gram Time to F o l l o w - Result Up (% Occlu- OutAngiosion) come gram large wide prey explored bas bif- small wide high risk P~ bas bif giant wide inoperable (MT) bas bif large wide prey ex(MT) plored bas bif small wide high risk 100 95 none . . . . good 28 95 none . . . . good 90 90 none . . . . good 8 85 none 2 days 85 -- -- good 10 90 none 7 days 100~: -- -- good bas bif bas bif basbif P: P3 bas bif bas bif 245 270 230 23 85 170 95 95 90 100 I00 95 none none none none none none . . . 3 mos 90 1 mo 90 2 mos 100 . . . 1~ mos 98 -3 mos -. -- -90 -- Aneurysrn Age, Clinical AneuSize't Case (vrs), Presen- rysm -Sex tation Loca- Diam- Neck No. tion eter 26 53, F SAH 27 46, F SAH 28 74, F SAH 29 51, F SAH 30 48, F 31 32 33 34 35 36 incid find 51, F mass 52, M SAH 57, M SAH 45, F SAH 40, F SAH 65, F mass 37 13, F Indication for Treatmerit bas bif giant giant giant small large large wide high risk wide high risk wide inoperable small inoperable wide inoperable wide high risk . -- good good good good good good SAH P I C A large small prey ex35 100 none 1 mo 100 --good (MT) plored 38 40, F SAH SCA small small high risk 8 100 none . . . . good * Patients in Hunt and Hesr Grade I, II, or II1. SAH = subarachnoid hemorrhage; PICA = posterior inferior eerebellar artery; MI = myocardial infarction; mass = mass effect: bas bif = basilar bifurcation: VB j c t = vertebrobasilar junction; prey explored = previously explored; SCA = superior cerebellar artery: reed condit = medical condition: PCA = posterior cerebral artery; incid find = incidental finding; Px, P2, P3 = first, second, and third segments of PCA: AICA = anterior inferior cerebellar artery,; PT = partially thrombosed: MT = mostly thrombosed. procedure not performed, t Aneurysm size (diameter): small = < 12 ram: large = 12-25 ram: giant = > 25 ram. Aneurysm neck size: small = < 4 ram: wide = > 4 ram. :~Additional embolization procedure for more coils. m u l t i c e n t e r study. T h e results o f t r e a t m e n t in a subset o f 42 p a t i e n t s with 43 difficult posterior circulation a n e u r y s m s are presented. Clinical Material and Methods Case Material This p a p e r presents the operative results in a subg r o u p o f p a t i e n t s enrolled in a n o n g o i n g m u l t i c e n t e r clinical study for t r e a t m e n t o f i n t r a c r a n i a l a n e u r y s m s using the G u g l i e l m i detachable coil (GDC).* Presently, the study is regulated b y a protocol a p p r o v e d by the U n i t e d States F o o d a n d D r u g A d m i n i s t r a t i o n , a n d the device is available only to those i n s t i t u t i o n s a p p r o v e d to participate in the study to d e t e r m i n e the safety a n d efficacy o f the m e t h o d . O f the total 120 cases treated, 42 patients with 43 a n e u r y s m s i n v o l v i n g the posterior circulation were o19erated o n b e t w e e n April, 1990, a n d D e c e m b e r , 1991, via an e n d o v a s c u l a r a p p r o a c h using the G D C (Tables 1 a n d 2). T h e t e c h n i q u e , protocol, a n d p r e l i m i n a r y clinical experience have been described e l s e w h e r e ) In * Guglielmi detachable coil manufactured by Target Therapeutics, San Jose, California. J. Neurosurg. / Volume 77 / October, 1992 this series there were 30 females a n d 12 males, ranging in age f r o m 13 to 74 years ( m e a n 51 years). Selection o f p a t i e n t s for this t r e a t m e n t was based o n six criteria: l) failure o f surgical exploration; 2) p o o r surgical c a n d i d a t e s s e c o n d a r y to h e m o r r h a g e ( H u n t a n d Hess 4 G r a d e IV or V); 3) p o o r surgical candidates s e c o n d a r y to medical c o n d i t i o n s ; 4) inoperability seco n d a r y to a n a t o m i c a l considerations; 5) a n e u r y s m s considered to be difficult a n d to have too high a risk for surgery; a n d 6) regrowth o f the a n e u r y s m after surgery. A total o f 24 p a t i e n t s (57%) presented with intracranial h e m o r r h a g e , a n d 13 (31%) h a d s y m p t o m s seco n d a r y to mass effect. Five a n e u r y s m s were f o u n d incidentally. O f 24 patients p r e s e n t i n g with h e m o r r h a g e , 17 were classified as in H u n t a n d Hess neurological G r a d e I or II, a n d three were in G r a d e III (Table 1). T h e r e m a i n i n g four were treated while in G r a d e IV or V (Table 2). Early t r e a t m e n t (up to 3 days p o s t - h e m orrhage) was p e r f o r m e d in three cases, a n d 10 cases received t r e a t m e n t b e t w e e n 3 a n d 15 days after t h e h e m o r r h a g i c event. T h e origin o f t h e a n e u r y s m s was as follows: t h e basilar artery b i f u r c a t i o n in 23, the vertebrobasilar j u n c t i o n in six, the posterior inferior cerebellar artery (PICA) 517 G. Guglielmi, et al. TABLE 2 Pro[ile ~#[/b~lrl)atiet~t.~ ~ iltz poMeFi(~r('ircttMti(m atl~,z~r)'wz~.~tTre~~v~tin~,in poor nelcrolo~ical co~tdition* . Time to Result lndicaCompli- Time to Result Follow-Up Age, Clinical Aneu- Anearvsm Size+ lion for C o i l Result cations Postop (% Occlu(% OccluCase (vrs). Presen- rysm Length (% Occlu- Relatedto AngioAngloNo. Treat(cm) sion) sion) sion) Sex ration Loca- Diam- Neck menl Technique gram gram lion eter PICA small small grading 39 67, F SAH (Grade V) 8 100 aneurysm rupture, parent vesselocclusion~ none . . . . Outcome dead bas bit" small small grading q3 100 . . . . dead 40 45, F mtrav hem (Grade IV) VBjct small wide grading 8 0 coil migration . . . . unchanged 41 67, M SAH (Grade IV) into It P2 PICA small small grading 8 100 none 2 days 100 --unchanged 42 48. F SAH (Grade V) * Patients in Hunt and Hess4 Grade IV or V. SAH = subarachnoid hemorrhage; PICA = posterior inferior cerebellar arteu; intrav hem = intraventricular hemorrhage; bus bif = basilar bifurcation: VB jct= vertebrobasilarjunction. P2 = second segment of posterior cerebral arteu. -= procedure not performed. f Aneurysm size (diameter): small = < 12 mm" large 12-25 mm; giant = > 25 mm. Aneuusm neck size: small < 4 ram: wide = > 4 mm. Aneuwsm rupture on detachment of first coil; parent vesselocclusion with slraight platinunqcoils. FIG. 1. Diagram of the Guglielmi detachable coil (GDC) and of the modified version of the Tracker microcatheter. A = diameter of the circular memory; B = the distal portion, a platinum coil (0.010 or 0.015 in. in diameter and 4 to 40 cm in length if straightened); C = microsolder connecting the platinum coil to the stainless steel delivery wire; D = detachable area; E = proximal (platinum) marker on the GDC; F = Teflon lamination; G = GDC shaft (diameter 0.010 in.); H = catheter tip marker; I = catheter proximal marker; J = catheter shaft. in five, the P l segment of the posterior cerebral artery (PCA) in four, the superior cerebellar artery (SCA) in three, the anterior inferior cerebellar artery in one, a n d the j u n c t i o n of the P2 and P3 segments of the PCA in one. Twenty aneurysms were small (< 12 mm), 13 were large (12 to 25 mm), and 10 were giant (> 25 mm). Operative Technique Most procedures were performed with the patient u n d e r neuroleptic analgesia. General anesthesia was used in noncooperative patients or where catheterization of difficult small aneurysms required high-quality (motionless) "road m a p p i n g " and subtraction fluoroscopy. A detailed description of the mechanical structure of the electrolytieally detachable coils has already been published; 2 however, modifications have occurred since 518 that description. A 0.5-cm long radiopaque marker (made of platinum) has been added to the G D C (Fig. 1). It is positioned in the stainless steel delivery portion of the coil, 3 cm proximal to the platinum-stainless steel junction. Another radiopaque p l a t i n u m marker, 3 cm proximal to the tip, has been added to a special version of the Tracker microcathetert (Fig. 1). These markers, on both the coil and the microcatheter, have been added with the purpose of increasing the safety of the embolization procedure. After detachment of the first coil, it may be difficult to visualize the platinumstainless steel j u n c t i o n of the second and subsequent coils within the coil mesh. The proximal radiopaque markers (outside the aneurysm) allow precise placement of the j u n c t i o n within the aneurysm even if the actual j u n c t i o n cannot be visualized. It is imperative that the j u n c t i o n emerges from the microcatheter for no more than 2 to 3 ram. Because of its relative stiffness, the stainless steel portion (Fig. l, arrows C-E) could potentially perforate the aneurysm if it is advanced too far. Alignment of the proximal radiopaque markers on the microcatheter and the G D C (in the parent artery) assures that the j u n c t i o n is no more than 2 to 3 m m beyond the microcatheter. A larger-diameter (0.015-in.) version of G D C has also been developed. These coils are used with the special Tracker-18 microcatheter primarily for unruplured large and giant aneurysms. For ruptured aneurysms and for small u n r u p t u r e d aneurysms the combination of the special Tracker- 10 microcatheter and the smaller-diameter (0.010-in.) coils are preferred. The t Modified version of Tracker-10 and Tracker-18 microcatheters manufactured by Target Therapeutics, San Jose, California. J. Neurosurg. / Volume 77 / October, 1992 Endovascular thrombosis of aneurysms Illustrative Cases Case/3: Smal! I/ertebrobasilar Junction Aneuo'sms ~Tth SAH FI(;. 2. Case 13. Lr Vertebral angiogram, lateral view. demonstrating a small vertebrobasilar junction aneurvsm (/on;, arrow). The residual portion of anothcr (partially clipped) vertebrobasilar junction aneurvsm is also visible (short arrow). Rigkt: Artcriogram obta]ncd 3 months after occlusion of the first aneurysm with two Guglielmi detachable coils (GDC's) (14 cm total length), and immediately after occlusion of the second (residual) aneurvsm with one 4-cm long GDC. Both aneupjsms are compleiely occluded while the normal arteries have been preserved. 0.015-in. coils are heavier and less malleable than the 0.010-in. coils. Since the 50% increase in coil diameter results in a 100% increase in volume, they are particularly useful for filling large and giant aneurysms in less time. With the development of the 0.015-in. coils, there are now 10 types of GDC (the first number is the diameter of the circular memory in mm, the second is the length of the coil in cm when straightened): 2 • 4, 3x8,4x 10,5 x 1 5 , 6 x 2 0 , 8 x 2 0 . a n d 8 x 4 0 , all 0.010 in. in diameter; and 5 x 15, 8 x 20, and 8 x 40, all 0.015 in. in diameter. This 42-year-old woman presented with a subarachnoid hemorrhage (SAH) from a vertebrobasilarjunction aneurysm that was subsequently surgically clipped. The follow-up angiogram showed a residual portion of the clipped aneuffsm and the presence of another small aneurysm of the vertebrobasilar junction (Fig. 2 / @ ) . This last aneurysm was initially treated endovascularly. Two GDC's with a total length of 14 cm were used to occlude this aneurysm completely. Three months later, the residual portion of the clipped aneurysm was occluded with one GDC of 2-ram circular memory and 4 cm long. The postembolization angiogram served also as a follow-up angiogram for the first treated aneurysm (Fig. 2 right). Both procedures, performed using a modified version of the Tracker-10 microcatheter, were very well tolerated by the patient. Case i6." Giant Vertebrobasilar Junction Aneurysm ~)th Mass Effect This 38-year-old man presented with left-sided hearing loss, left-sided fifth nerve deficit, and gait ataxia. A cerebral angiogram demonstrated a giant vertebrobasilar junction aneurysm (Fig. 3A and B). This aneurysm was considered inoperable due to its size and location. Via a transfemoral approach, a modified version of a Tracker-18 microcatheter was positioned into the sac of the aneurysm Ten GDC's, 0.015 in. in diameter, 8 mm in circular memory, and 40 cm long, were delivered and electrolytically detached within the sac of the aneurysm. Approximately 50% of the aneurysm was filled with coils. An angiogram performed 3 days postoperatively demonstrated that progressive thrombosis had occurred with occlusion of 85 % to 90% of the aneurysm (Fig. 3C). Another follow-up angiogram performed at FIG. 3. Case 16. A: Right vertebral angiogram, anteroposterior view, demonstrating a giant vertebrobasilar junction aneurysm. B: Left vertebral angiogram demonstrating the giant aneurysm. C: Right vertebral angiogram performed 3 days after deposition of 10 Guglielmi detachable coils (0.015 in. in diameter, total length 4 m). Most of the aneurysm has undergone thrombosis. The distal vascular tree is now visible. D: Left vertebral angiogram obtained 8 days after coil deposition and immediately after balloon occlusion of the right vertebral artery. The right posterior inferior cerebellar arte~ (long arro~) is filling from the left. The (subtracted) meshwork of coils is barely visible (short arrow,s9. Permanent occlusion of the right vertebral artery was performed to lessen the jet effect of the arterial flow in the residual aneurysm J. Neurosurg. / Volume 77 / October, 1992 519 G. Guglielmi, et al. FI(~. 4. Case 17. A: Vertebral angiogram, lateral view, demonstrating a large basilar bifurcation aneuD'sm. B: Intra-aneurysmal angiogram obtained while performing catheterization of the aneurysm with a modified version of a Tracker-18 microcatheter. Angiography was performed to determine proper positioning of the microcatheter tip (long arro;~) and in an attempt to determine the size of the neck by comparing the position of the microcatheter (inflow zone at the neck) with the outflow. Note the contrast material depicting the flow pattern within the aneurysm (two short arrow,~7. C: Plain skull x-ray film showing the network of coils placed within the aneurysm (total coil length 1.95 m). Coils both 0.010 in. and 0.015 in. in diameter were used in this case. D: Immediate postembolization vertebral angiogram showing the aneurysm completely occluded and the parent vessel preserved. 8 days confirmed this result. In the same session, permanent balloon occlusion of the right vertebral artery, below the PICA origin was performed to diminish the arterial flow (jet effect) inside the residual aneurysm (Fig. 3D). A 4-month follow-up angiogram showed further reduction of the aneurysm. By that time, the fifth and eighth cranial nerve deficits had improved and the gait ataxia had almost completely resolved. Case 17: Large Basilar Bifurcation Aneurvsm Found Incidentally This 56-year-old woman suffered multiple acute cerebral, pulmonary, renal, splenic, and lower-extremity emboli from acute endocarditis. A brain computerized tomography (CT) scan performed to assess the effects of cerebral embolism, also showed a basilar artery aneurysm. A cerebral angiogram confirmed the presence of a large basilar bifurcation aneurysm (Fig. 4A). Due to the medical condition of the patient, endovascular treatment was performed while she was under local anesthesia. After introduction of a transfemoral Tracker-18 microcatheter (modified version) into the aneurysmal sac, an intra-aneurysmal angiogram was performed to determine the proper placement of the microcatheter tip as well as the size of the aneurysm neck (Fig. 4B). Two GDC's (0.015 in. in diameter) with a total length of 80 cm, and six GDC's (0.010 in. in diameter) with a total length of 115 cm were deposited and electrolytically detached in the aneurysm (Fig. 4C). The immediate postthrombosis angiogram showed complete aneurysm occlusion, patency of the basilar artery, and normal filling of the distal vascular tree (Fig. 4D). The procedure lasted 3 hours and was well tolerated by the patient. This patient, unfortunately, died 1 month later because of the devastating effects of marantic endocarditis. 52.0 Case 19. Wide-?,k;cked Large I@rtebrobasi[ar Junction Aneuo'sm 14qth Mass Effect This 62-year-old man presented with a 3-month history, of headache and double vision. Neurological examination revealed the presence of a left abducens cranial nerve deficit. Cerebral angiography demonstrated a large vertebrobasilar artery junction aneurysm (Fig. 5 left), which was considered inoperable. A transvascular occlusion of the aneurysm was performed by delivering two GDC's (for a total length of 60 cm) through a transfemoral microcatheter into the sac. At the end of this procedure residual filling of the aneurysm base and portions of the dome was present. A 1-month control angiogram showed progression of intra-aneurysmal thrombosis. At 4 months a second occlusion procedure was performed, with deposition of two GDC's (each 15 cm in length). The postocclusion angiogram showed that, in spite of loose packing, most FIG. 5. Case 19. Left. Vertebral angiogram, anteroposterior view, showing a large inoperable vertebrobasilar junction aneurysm, irregular in shape. Right. Angiogram made after two embolization procedures using four coils with an overall length of 90 cm. In spite of loose packing, most of the aneurvsm is occluded. The posttreatment clinical course was uneventful. (Procedure performed at the University of Wisconsin, Madison, Wisconsin.) J. Neurosurg. / Volume 77/October, 1992 Endovascular thrombosis of aneurysms of the ancurysm had been thrombosed, with preservation of the parent vessel (Fig. 5 r~r Clinically. the headache in this patient had resolved and the sixth nerve deficit improved significantly. ('a.s'e 21. Small PI(\4 .,1iwttrv,~m lI "ilk Small Neck This 48-year-old man suffered an SAH from a posterior communicating artery (PCoA) aneu~,sm. A small aneurysm of" the PICA was also present (Fig. 6 Ir The PCoA aneurysm was successfully clipped surgically, but cardiac disease prevented surgical treatment of the PICA aneurysm; therefore, endovascular endosaccular treatment was performed. After catheterization of the aneurysm with a modified Tracker-10 microcatheter, two GDC's with a total length of 18 cm were used to occlude this aneurysm completely. The parent vessel had been preserved (Fig. 6 ri@t). Results Tables 1 and 2 summarize the clinical presentation, anztomical location, therapeutic results, and complications in this series of 42 patients with 43 posterior circulation aneurysms. Endosaccular aneurysm occlusion ranging from 70% to 100% was achieved in 42 aneurysms. One patient could not be treated due to a technical complication (Case 41 ). A complete aneurysm occlusion was achieved in 13 of 16 aneurysms with a small neck and in four of 26 wide-necked aneurysms. Considerable thrombosis of the aneurysm (70% to 98%) was achieved in 22 aneurysms with a wide neck and in three small-necked aneurysms. Incomplete aneurysm occlusion resulted from partial coil filling of the aneurysm. In wide-necked aneurysms, complete coil filling was diffficult because of the danger of occluding the parent vessel and of distal coil migration. The portions of the aneurysm filled with coils underwent thrombosis in 100% of cases. The time of angiographic follow-up studies ranged from 6 hours to 14 months (mean 3 months) postoperatively (Tables 1 and 2). The clinical follow-up period ranged from 5 days to 18 months. In six cases, more than one procedure was necessary to achieve a satisfactom' aneurysm occlusion (Cases 5, 8, 9, 15, 19, and 30). Several patients had posttherapy magnetic resonance (MR) imaging of the head; GDC's are MR imagingcompatible, and display less local field effect than MR imaging-compatible surgical clips. In this series there were no transient and one permanent neurological deficits related to treatment. This patient (Case 10) developed a hemianopsia after complete occlusion of a wide-necked basilar bifurcation aneurysm. In one patient (Case 41) it has not been possible to assess the clinical impact of an untoward coil migration into the P: segment of the PCA due to pre-existing coma. However, CT scans demonstrated an ischemic zone in the left PCA territory. One patient died in connection with the endovascular therapy. In this patient (Case 39), the aneurysm ruptured after detachment of a first coil, due to penetration of the fresh clot with the microcatheter. The patient, whose condition was originally classified as Grade V, died 5 days later. The overall morbidity and mortality rates related '~o this technique were 4.8% (two cases) and 2.4% (one case), respectively. If considering only patients who presented in Grades I, II, and III, the morbidity rate is 2.6% (one case) and the mortality rate is 0%. One patient (Case 17) died 1 month after treatment because of an unrelated disease. The autopsy showed that marantic endocarditis had produced multiple organ infarctions. One patient (Case 2) harboring a giant inoperable basilar bifurcation aneurysm died of SAH 18 months after partial aneurysm occlusion. The ostia of the PCA's and SCA's were incorporated in the neck of the aneurysm, precluding its surgical clipping or endovascular occlusion. This case has already been reported. Discussion FIG. 6. Case 21. Left: Vertebral angiogram, lateraloblique view, showing a small posterior inferior cerebellar artery, aneurysm (arrow). Right: Vertebral angiogram taken after intra-aneurysmal deposition and detachment of two coils (total length 18 cm) via a modified Tracker-10 microcatheter. The aneurysm is completely occluded. (Procedure performed at New York University, New York, New York.) J. Neurosurg. / Volume 77 / October, 1992 The rationale for treatment of cerebral aneurysm is to avoid bleeding and to alleviate symptoms of mass effect in unruptured aneurysms, or to prevent rebleeding in ruptured ones. Currently, aneurysm treatment with preservation of the parent artery may be performed either by open surgical clipping or through an endovascular endosaccular technique. The advantages of the endovascular procedure are that craniotomy and brain manipulation are unnecessary, that the medical condition of the patient does not affect the timing or performance of the procedure, and that it may be applied in cases where direct clipping efforts have failed. Posterior fossa aneurysms pose difficult surgical problems.6-8 ~o.~2Therefore, an alternative endovaseular technique would be of great benefit in this group of patients. The GDC system was applied to this territory in 42 patients with 43 aneurysms. 521 G . G u g l i e l m i , el al. L#flcacr ~f 7)'ealment It was possible to enter all aneurisms with a microcatheter-microguidewire combination. Immediate complete occlusion could be achieved in 13 aneurisms with a small neck (Cases 7, 11, 13 (two aneurysms), 17, 21, 25, 35, 37-40, and 42). A narrow aneurysm neck holds the coils inside the aneu~,sm and allows dense coil packing without impinging on the parenl vessel. Similar anatomical results (complete aneurysm occlusion) may be more difficult to achieve in aneurysms with a wide neck because of the lack of a natural barrier and the risk of coil impingement on the parent vessel and possible distal coil migration. The majority of cases in this series had intracranial aneurysms considered to be technically difficult and with a high surgical risk. Some aneurysms were considered inoperable (10 cases) and some had already undergone unsuccessful surgical exploration (13 cases). Because of this, a partial treatment (70% to 98 % occlusion of the aneurism) was considered to be acceptable in wide-necked aneurysms. Some of these aneurysms may bleed or rebleed, as happened in Case 2. However, it is known that in 98% of cases the rupture site of a saccular aneurysm is located at its dome and sac and that only 2% of aneurysms rupture from their neck. 9 More than one session was necessary to occlude some of these wide-necked aneurisms (Cases 5, 8, 9, 15, 19, and 30). This was due to the fact that, often, only loose coil packing is achieved in these aneurisms due to the risk of migration. This allows the coils to be pushed by the arterial flow (jet effect) and to become compacted inside the aneurism. A second procedure was then necessary to reocclude the portions of neck and proximal sac re-exposed to the blood flow. When the aneurysm appears to be well packed with coils (as visualized fluoroscopically or on a plain x-ray film), only one- to two-fifths of the volume of the aneurism has actually been filled. This is the reason why the coils are susceptible to be compacted by the blood flow toward the sac and away from the neck of the aneurysm. As already mentioned, it was possible to achieve denser packing (and complete occlusion) in aneurisms with a small neck. In these cases, due to the dense packing and to the smaller cross section of the neck area susceptible to blood flow. the phenomenon of coil remodeling by blood flow is less likely to occur. Furthermore, the smaller the neck, the higher the probability that the mesh of coils will bridge across the neck area, the neck being the location of the elastic lining defect initially responsible for aneurysm formation. The follow-up angiograms in wide-necked aneurisms (Table 1) showed no aneurism regrowth from the arterial wall defect. The phenomenon of coil mesh remodeling, as described above, did occur in many of these aneurysms. Recurrence (defined as recanalization of the aneurism due to lysis of the thrombus within the meshwork of coils) was not observed. Small-necked aneurysms that were completely oc- 522 eluded remained occluded in follow-up angiograms (Table It. These data, of course, must be confirmed with longer-term angiographic follow-up studies. In one case (Case 7, Fig. 7) a follow-up angiogram at 14 months showed persistence of complete aneurism occlusion. This case has already been reported ~ and represents the longest angiographic follow-up interval of the entire series. Thirteen of 43 aneurisms presented with symptoms due to mass effect (Table l). These aneurysms were treated with the goal of preventing an intracranial hemorrhage due to aneurysm rupture. Some of these patients improved clinically after treatment (clinical amelioration was particularly evident in Case 16 (Fig. 3) and in Case 19 (Fig. 5)). However, it is too soon to draw definitive conclusions on the outcome of mass effect with GDC treatment. After aneurysm occlusion, diminished pulsations and the possible retraction of the clot may decrease the compression upon the surrounding central nervous system structures. If technically thasible, surgical removal of the aneurysm may be the only possible treatment when symptoms due to mass effect worsen or do not improve. Postprocedure transient worsening of the symptoms was sometimes observed. This might be due to swelling of the aneurysmal wall. To prevent this occurrence, steroids may be administered before and after the procedure. SaJety of Treatment Some technical and clinical complications were observed in this series (Tables 1 and 2). The entrance into the aneurysm with the microcatheter-microguidewire combination should be performed very carefully. Both the microcatheter and the guidewire are less soft than the platinum coils and there is a risk of aneurism perforation if they are pushed against the aneurysmal wall. The tip of either the microcatheter or the guidewire should never touch the fragile wall of the aneurysm in order to avoid perforation of the wall or of the even more fragile fresh clot at the bleeding site. This complication occurred in one of 43 aneurisms (Case 39). The presence of the proximal markers, on both the microcatheter and the detachable coils, adds to the safety of the procedure. It was always possible to retrieve the GDC from the aneurism before detachment if it was inappropriate in size or position. Failure of electrolytic detachment never occurred. In this series, unlike experience with other forms of endovascular treatment, 3 no early bleeding or rebleeding was observed after partial or incomplete aneurysm occlusion. It may therefore be reasonable to stage the aneurism treatment, as was done in the five cases described above. Analysis of the Complications In Case [0, untoward propagation of the intra-aneurysmal thrombus was observed a few hours after sub- J. Neurosurg. / Volume 77 / October, 1992 Endovascular thrombosis of aneuusms FIG. 7. Case 7. A and B: Vertebral angiograms, anteroposterior (A) and lateral (B) views, showing a small basilar bifurcation aneur)'sm. C and D: The 14-month postocclusion follow-up angiograms showing persistence of complete aneuD,sm occlusion. (Angiogram courtesy of Drs. L. Chi and J. Bello, Montefiore Hospital, New York, New York.) total embolization of a wide-neck basilar bifurcation aneurysm. This led to complete occlusion of the aneurysm but also of the ostia of both PCA's that were originating from the aneurysmal neck. After a stormy postembolization clinical course, the patient was left with a permanent hemianopsia. Case 41, with a ruptured wide-necked vertebrobasilar junction aneurysm, was treated while the patient was neurologically in Grade IV. After several unsuccessful attempts to deliver a 4 • 10 GDC, the decision was made to detach a 2 x 8 GDC. This coil migrated, after detachment, into the P2 portion of the left PCA. The clinical impact of this coil migration could not be assessed because of the persistent comatose state of the patient. A C T scan, however, showed a low-density region in the left occipital lobe consistent with ischemia. The aneurysm was subsequently surgically clipped. Due to this technical and clinical complication, the 2 • 8 GDC was withdrawn from clinical use and a 3 x 8 GDC was substituted. One patient (Case 39), harboring a small ruptured PICA aneurysm, was treated while neurologically in Grade V. Severe vasospasm did not allow the utilization of the road-mapping capability during the procedure. After deposition and detachment of one coil, the microcatheter perforated the aneurysm, likely through the fresh clot. The parent vessel was then immediately occluded with straight platinum coils. The patient died 5 days later. It is important to stress that two of the three complications occurred in patients presenting with severe neurological conditions (Hunt and Hess Grades IV and V). Both the severity of the patient's medical condition and the degree of surgical difficulty led us to use a more aggressive approach in an attempt to save these patients' lives. The limited experience related to this kind of patient precludes our being certain about the indications for the endovascular approach in Grade IV and J. Neurosurg. / Volume 77/October, 1992 V patients. The question arises not from technical limitations but rather from the patient's poor medical condition. In four patients, surgical intervention was performed after the thrombosis. In two patients (Cases 1 and 27), surgical clipping was performed at 1 week and 3 months, respectively, to exclude a neck remnant from the circulation. This allowed the two patients to recover from the effects of SAH. In another patient (Case 5) with a wide-necked basilar bifurcation aneurysm, surgical clipping of the mid-basilar artery was performed after three occlusion procedures, in order to decrease the jet flow into the aneurysm and enhance thrombosis. Temporary balloon occlusion of the basilar artery was performed prior to clipping in order to assess clinical tolerance. In another patient (Case 41) in whom endovascular treatment failed, surgical clipping of a vertebrobasilar junction aneurysm was performed. This technique potentially may be used to occlude the aneurysm in the acute phase of SAH. This endovascular occlusion may eliminate the risk of rebleeding and would allow intensive therapy for vasospasm. In those cases in which an incomplete endovascular occlusion of the aneurysm occurred, an elective surgical clipping of the residual aneurysm may be performed. Conclusions The technique of endovascular endosaccular occlusion of intracranial aneurysms using GDC's has been developed with the aim of providing a therapeutic alternative. The Food and Drug Administration has limited and oriented the use of GDC's only to patients harboring aneurysms with a high risk of surgery. To date, 120 anterior and posterior circulation aneurysms have been treated with this technique; of these, 42 patients with 43 posterior circulation aneurysms are the subject of this paper. 523 G. Guglielmi, et al. With regard lo safety, GDC's adapl to the shape of the aneurysm causing no significant trauma to the aneurysmal wall, due to the fact that they are very soft and pliable. No aneurysm perforation with the platinum coils was noted in the 43 aneurysms treated. The GDC's are easily controllable and retrievable. Softness, retrievability, and atraumatic detachment are the safety features of the technique. In this series, the morbidity and mortality rates related to treatment were 4.8% (two cases) and 2.4% (one case), respectively (2.6% morbidity and 0% mortality if considering only patients who presented in neurological Grades I, II, and Ill, according to Hunt and Hess classification). As to efficacy, a consistent amount ol thrombosis, ranging from 70% to 98%, was achieved in 22 of 26 wide-necked aneurysms and in three of 16 small-necked aneurysms. Complete aneurysm occlusion was achieved in 13 small-necked aneurysms and in four wide-necked aneurysms. One aneurysm could not be occluded due to a technical complication. Electrothrombosis may be an important factor in the formation of a stable thrombus. It may play a role in eliciting celeritous thrombosis that may prevent coil migration. Further studies on the structure of the electrically produced thrombi are necessary. Longer angiographic and clinical follow-up studies are essential to ascertain the long-term efficacy of this recently developed endovascular occlusion technique. Disclosure Part 1: Electrochemical basis, technique, and experimental results. J Neurosurg 75:1-7, 1991 3. Higashida RT, Halbaeh VV, Cahan ED, et al: Detachable balloon embolization therapy of posterior circulation intracranial aneurysms. J Neurosurg 71:512-519, 1989 4. Hunt WE, Hess RM: Surgical risk as related to time of intervention in the repair of intracranial aneurysms. J Neurosurg 28:14-20, 1968 5. Kassell NF, Torner JC, Haley EC Jr, el al: The International Cooperative Study on the Timing of Aneurysm Surgeff. Part 1: Overall management results. J Neurosurg 73:18-36, 1990 6. McMurlry JG lit, Housepian EM, Bowman FO Jr, et al: Surgical treatment of basilar artery aneurysms. Elective circulatory arrest with thoracotomy in 12 cases. J Neurosurg 40:486-494, 1974 7. Peerless SJ, Drake CG: Posterior circulation aneurysms, in Wilkins RH, Rengachary SS (eds): Neurosurgery. New York: McGraw-Hill, 1985, Vol 2, pp 1422-1437 8. Spetzler RF, Hadley MN, Rigamonti D, et al: Aneurysms of the basilar artery treated with circulatory arrest, hypothermia, and barbiturate cerebral protection. J Neurosurg 68:868-879, 1988 9. Weir B: lntracranial aneurysms and subarachnoid hemorrhage: an overview, in Wilkins RH, Rengachary SS (eds): Neurosurgery. New York: McGraw-Hill, 1985, Vol 2, 1308-1329 10. Wilson CB, U HS: Surgical treatment for aneurysms of the upper basilar artery. J Neurosurg 44:537-543, 1976 ll. Ya~argil MG: Microneurosurgery. Clinical Considerations, Surgery of Intracranial Aneurysms and Results. Stuttgart: Georg Thieme, Vol 2, 1984 12. Ya~argil MG, Antic J, Laciga R, el ah Microsurgieal pterional approach to aneurysms of the basilar bifurcation. Surg Neurol 6:83-91, 1976 A patent on the GDC coils is pending in the names of the first and last authors. References 1. Guglielmi G, Vifiuela F, Dion J, et al: Electrothrombosis of saccular aneurysms via endovascular approach. Part 2: Preliminary clinical experience. J Neurosurg 75:8-14, 1991 2. Guglielmi G, Vifiuela F, Sepetka I, et al: Electrothrombosis of saccular aneurysms via endovascular approach. 524 Manuscript received December 18, 1991. Accepted in final form April 8, 1992. This work has been supported by National Institutes of Health Grant HL46286-01. Address reprint requests to: Guido Guglielmi, M.D., Dipartimento di Scienze Neurologicbe, Neuroangiografia Terapeutica, Universita' di Roma, Viale dell'Universita' 30/a, 00185 Roma, Italy. J. Neurosurg, / Volume 77/October, 1992