Preliminary experience with precipitating hydrophobic injectable liquid in brain arteriovenous malformations

Abstract

Advancement in microcatheter design and emergence of new embolic agents offer better results in endovascular treatment of brain arteriovenous malformations (AVMs). Precipitating hydrophobic injectable liquid (PHIL) (Microvention) is a newly introduced dimethyl sulfoxide-based embolic agent for endovascular use. Herein, we present three patients who underwent endovascular treatment of brain AVMs with PHIL, followed by surgical resection. Endovascular features and same-day surgical handling of the new embolic agent PHIL are presented along with histopathologic changes in the acute stage in brain AVMs are presented, and its major differences from Onyx. In our series, PHIL had moderate inflammatory reaction in the acute stage without any associated angionecrosis that is different than Onyx which cause mild inflammatory reaction with angionecrosis. Smallest vessel containing PHIL was 2.9 μm compared to 5 μm with Onyx, which suggests better penetration.

Diagn Interv Radiol 2016; 22:184189

Turkish Society of Radiology 2015

INTERVENTIONAL RADIOLOGY

TECHNICAL NOTE

Preliminary experience with precipitating hydrophobic injectable liquid in brain arteriovenous malformations

Naci Koer Hakan Hanmolu ebnem BaturSedat Giray Kandemirli Osman KzlklZihni SanusBge zCivan IlakMehmet Yaar Kaynar

ABSTRACT

Advancement in microcatheter design and emergence of new embolic agents oer better results in endovascular treatment of brain arteriovenous malformations (AVMs). Precipitating hydro-phobic injectable liquid (PHIL) (Microvention) is a newly introduced dimethyl sulfoxide-based embolic agent for endovascular use. Herein, we present three patients who underwent endovascular treatment of brain AVMs with PHIL, followed by surgical resection. Endovascular features and same-day surgical handling of the new embolic agent PHIL are presented along with histopathologic changes in the acute stage in brain AVMs are presented, and its major dierences from Onyx. In our series, PHIL had moderate inammatory reaction in the acute stage without any associated angionecrosis that is dierent than Onyx which cause mild inammatory reaction with angionecrosis. Smallest vessel containing PHIL was 2.9 m compared to 5 m with Onyx, which suggests better penetration.

Treatment of brain arteriovenous malformations (AVMs) includes endovascular embolization, surgical resection, and stereotactic radiotherapy, alone or in combination (1). In the past, n-butyl cyanoacrylate (nBCA)was used as an adjunct to surgery;

however, introduction of a dimethyl sulfoxide (DMSO)-based embolic agent, Onyx (ev3, Covidien), and advancement in microcatheter design oered better endovascular results in brain AVM treatment (2). Recently, a new DMSO-based embolic agent, precipitating hydro-phobic injectable liquid (PHIL) (Microvention) has been introduced for endovascular use. PHIL is a nonadhesive co-polymer (polylactide-co-glycolide and polyhydroxyethylmethacrylate)-based liquid embolic material suspended in DMSO. An iodine (triiodophenol) component is bound to co-polymer for radiopacity. Possible advantages of PHIL are its ease of use, faster plug formation, and less computed tomography (CT) and magnetic resonance imaging (MRI) artifact during imaging follow up. PHIL was selected as the embolic agent in view of its potential advantages (3, 4). Three patients with brain AVMs had partial embolization with PHIL followed by same-day surgical resection in our institution. Here, we present our preliminary experience with PHIL in brain AVMs from an endovascular, surgical, and pathologic perspective, and outline its major dierences from Onyx.

Cases

From the Division of Neuroradiology, the Department of Radiology (N.K. nkocer@ istanbul.edu.tr, S.G.K., O.K., C.I.); the Departmentsof Neurosurgery (H.H., Z.S., M.Y.K.) and Pathology (.B., B..), stanbul University Cerrahpaa School of Medicine, stanbul, Turkey.

Received 27 June 2015; revision requested 18 July 2015; revision received 31 July 2015; accepted 3 August 2015.

Published online 19 January 2016. DOI 10.5152/dir.2015.15283

In our institution, nine cases with AVM and arteriovenous stulae underwent PHIL embolization after obtaining informed patient consent, between August 2014 and May 2015. Herein, we focused on three of six brain AVMs who had partial embolization and same-day surgical nidus resection. Clinical, endovascular, surgical, and pathologic data were retrospectively collected and analyzed.

Case 1: A 20-year-old female patient presented with seizure and an unruptured AVM (Spetzler-Martin Grade 3) located in the right temporal lobe. Main arterial feeders were temporal branches of the right middle cerebral artery. AVM had a mixed stulous and plexiform nidus with cortico-subcortical location. Nidus size was 42.5 cm. AVM had both supercial and deep venous drainage, with ectasia in veins and venous aneurysm (Fig. 1).

Case 2: A 33-year-old female patient presented with vertigo and an unruptured AVM (Spetzler-Martin Grade 2) located in the right parieto-occipital sulcus. Main feeder was the parietal

184

branch of the right middle cerebral artery with a partial supply by parietal branches of the right posterior cerebral artery, additionally enlargement in arterial feeders was noted. AVM had a mixed stulous and plexiform nidus located sulcally. Nidus size was 22.5 cm. AVM had a single supercial venous drainage, with no ectasia in veins (Fig. 2).

Case 3: A ve-year-old male patient presented with nausea, vomiting, and temporary loss of consciousness due to a ruptured AVM (Spetzler-Martin Grade 3) located in the right cerebellar hemisphere. Main feeders were the right posterior inferior cerebellar artery and the right superior cerebellar artery, with enlargement in arterial feeders. AVM had a mixed stulous and plexiform nidus with cortico-subcortical location. Nidus size was 3.52.5 cm. AVM had a supercial venous drainage, with partially thrombosed aneurysm in drainage vein (Fig. 3).

a

b

Technique

c

e

d

Embolic agents

Onyx is a nonadhesive liquid embolic agent consisting of 48 mol/L ethylene and 52 mol/L ethylene vinyl alcohol, dissolved in DMSO and mixed with tantalum powder for radiopacity. Onyx is available in ready-touse vials of 1.5 mL at three dierent viscosities, Onyx 18, 20, and 34 (concentrations of ethylene vinyl alcohol: 6%, 6.5%, and 8%; respectively). The vials are shaken for at least 20 minutes for a homogeneous solution (5).

PHIL is a nonadhesive co-polymer (polylactide-co-glycolide and polyhydroxyethylmethacrylate)-based liquid embolic material suspended in DMSO. It has iodine

Main points

Onyx is a widely used nonadhesive liquid DMSO-based embolic agent mixed with tantalum powder for radiopacity.

Precipitating hydrophobic injectable liquid (PHIL) is a newly introduced nonadhesive

DMSO-based liquid embolic agent with an iodine component for radiopacity.

Main advantages of PHIL compared to Onyx are its ease of use, faster plug formation, and less CT and MRI artifact at follow-up imaging.

PHIL is more brittle during vessel dissection and less pliable during surgery in comparison to Onyx.

PHIL causes moderate vascular inammatory eect and no angionecrosis on same-day resected specimens.

Figure 1. ae. A 20-year-old female with an unruptured temporal arteriovenous malformation (AVM). Anteroposterior (AP) subtracted-digital subtraction angiography (DSA) (a) shows the AVM fed by the middle cerebral artery. PHIL and Onyx were used as embolic agents. Postembolization early arterial AP subtracted-DSA (b) shows almost total obliteration; however, minimal residual nidus lling and stagnation of perinidal vascular structures were noted in venous phase (c). Coronal XperCT (d) shows PHIL and Onyx embolization of the AVM with notable artifact on CT due to Onyx. Immediately after surgery, lateral subtracted-DSA (e) shows total nidus removal without stagnated vascular structures.

(triiodophenol) component bound to a co-polymer for radiopacity. It comes as ready to use pre-lled syringes of 1 mL at three dierent concentrations; PHIL 25%, 30%, and 35% (viscosity in centistokes 16, 36, and 72, respectively) and syringes pre-lled

with 1 mL DMSO. Embolic capacity of 25%, 30% and 35% PHIL is 0.85, 0.87, and 0.94 mL, respectively, per 1 mL of PHIL. Higher concentrations are preferred in the presence of a stulous component, whereas lower concentrations are used to achieve distal access.

Preliminary experience with PHIL in brain AVM 185

Endovascular treatment

In our institution, the main strategy for AVM treatment is to achieve complete nidus obliteration by endovascular route alone or in combination with same-day total microsurgical resection in incompletely embolized AVMs, if surgically accessible. Diencephalic and brain stem AVMs are outside of this strategy.

Endovascular treatment was performed under general anesthesia using a biplane at

panel digital subtraction unit (Allura Xper 20/20, Philips Healthcare). During endovascular treatments, feeding arteries were catheterized with DMSO-compatible detachable Apollo (Covidien; in cases 1, 2, and 3) and nondetachable Marathon (Covidien; in c) tip microcatheters. After proper positioning of microcatheter, its dead space was lled with DMSO. PHIL 25% was used as the embolizing agent in all three cases. After plug formation,

the slow injection style used for PHIL was not dierent from injection of Onyx used in AVM treatment (2). In case 1, Onyx was injected into the nidus through the same microcatheter after PHIL injections. Immediately after embolization, routine Xper CT was performed to rule out possible intracranial complications. All three cases underwent same day surgery because of incomplete nidus occlusion by endovascular route.

During endovascular treatment, features of PHIL such as visibility, ease of plug formation, tendency to backow, and behavior in arterial and venous sides were noted.

Surgical resection

After partial embolization, the patients were transferred to neurosurgery operating room under general anesthesia for total nidus resection. During surgical resection, the form PHIL takes within the vessel, its pliability for resection, the degree of blood loss and compressibility of nidus were noted. In case 1, dierences between PHIL and Onyx embolized parts of the nidus were also noted.

Histopathologic examination

Partially embolized and surgically resected AVM nidi were sent to the pathology lab within 47 hours after embolization. Specimens were xed in 10% buered formalin and embedded in paraffin. Serial 23 m sections were obtained and stained with hematoxylin-eosin and elastica van Gieson stain prior to review.

Specimens were evaluated for vascular and perivascular inammation, mural an-

a b

c

Figure 2. ac. A 33-year-old female with an unruptured parieto-occipital AVM. Lateral subtracted-DSA (a) shows the AVM fed mainly by the middle cerebral artery with a partial supply by the posterior cerebral artery. Postembolization AP plain radiography (b) shows partial PHIL embolization of AVM. Postsurgery lateral subtracted-DSA (c) shows total nidus removal.

a b c

Figure 3. ac. A five-year-old male with a ruptured cerebellar AVM. Lateral subtracted-DSA (a) prior to embolization shows the AVM fed by posterior inferior cerebellar artery and superior cerebellar artery (SCA) branches. Postembolization AP plain radiography (b) and coronal XperCT (c) show partial PHIL embolization of AVM via SCA branches with minimal artifact noted on CT.

186 MarchApril 2016 Diagnostic and Interventional Radiology Koer et al.

gionecrosis and parenchymal hemorrhage. In order to evaluate the degree of embolic agent penetration, the smallest vessel that contained the embolic agent was identied and measured. Extravasation of the embolic agent and foreign body giant cells in the embolic cast were investigated.

a b

Results

Partial AVM embolization using PHIL alone or in combination with Onyx was performed successfully in all three cases. Nidus obliteration rates were 90%, 80%, and 70% in cases 1, 2, and 3, respectively. Endovascular embolization was terminated in case 2, because of a small distal artery perforation with resultant subarachnoid hemorrhage during microcatheterization of an arterial feeder.

Microsurgical nidus resections were performed successfully in all three cases without any complications. During surgery, PHIL was noted within the vessels as a white chalky material that was less pliable compared to Onyx (Fig. 4).

On day 1, all cases had control digital subtraction angiography. Case 1 and 2 had total nidus resection. Partial lling was found in an unresected area of the nidus in case 3. On discharge, case 1 and 2 had modied Rankin scores of 0, and case 3 had modied Rankin score of 1.

Histopathologic exam of the surgical specimens from PHIL-embolized cases (cases 2 and 3) revealed moderate amount of vascular and acute perivascular inammation without any evidence of chronic inammatory response (Fig. 5). Angionecrosis was not seen in PHIL-lled vessels.

On microscopic examination, lumina of the embolized vessels appeared as hyaline eosinophilic material admixed with yellowish material that had linear and perpendicular conguration at the edge of the vessels (Fig. 5). Some parts of the embolized vessels were completely lled with the embolic material, whereas in other parts the embolic material lled a small portion of the vessel lumen, with thrombus lling the remainder of the lumen. Blood vessels ranging from 2.9 m to 5 mm in diameter were lled with PHIL (Fig. 6).

Foreign body giant cells were not ob-served and the vessel wall integrity was maintained in all embolized vessels. No evidence of perivascular extravasation of embolic material was observed in any specimen. All specimens showed evidence of

hemorrhage in the surrounding perinidus brain parenchyma.

PHIL and Onyx were both used as embolic agents in case 1 (Fig. 5). Histopatho-

logic exam showed moderate amount of vascular inammation in PHIL embolized vessels and less vascular inammation in Onyx-lled vessels compared with PHIL.

Figure 4. a, b. Intraoperative image of PHIL. Vessel lled with the embolic agent PHIL (a, arrows) forms a rough intravascular column. Extravasation of PHIL (b, arrow) is also seen under arachnoid mater. Dissection of PHIL-lled vessel was easy, similar to Onyx (a). Meticulous handling of the vessel with low voltage bipolar cauterization is required to maintain vessel wall integrity (b).

a b

Figure 5. a, b. Photomicrograph (a) shows endovascular PHIL material with polymorpho nuclear leukocytes (hematoxylin eosin staining [HE], 400). Onyx (b, arrow) is layering over the PHIL inside the vessel (HE, 200).

Figure 6. Photomicrograph showing PHIL as a hyaline eosinophilic material in lumina of embolized vessels (HE, 200). Smallest vessel occluded (arrow) by PHIL was 2.9 m in our specimens (inset, HE, 400).

Preliminary experience with PHIL in brain AVM 187

No sign of angionecrosis was seen in PHIL embolized vessels, whereas signs of angio-necrosis was evident in Onyx-lled vessels. Smallest vessels containing PHIL and Onyx were 50 m and 32 m, respectively.

Discussion

PHIL is a newly introduced DMSO based embolic agent that is gaining wider use in AVM treatment. In this paper, we share our preliminary radiologic and surgical experience in same-day combined treatment of AVMs, acute stage histopathologic ndings of PHIL, and its major dierences from Onyx.

PHIL comes as ready to use pre-lled syringes of 1 mL and does not require shaking prior to injection. Onyx is shaken for 20 minutes for a homogenous solution that requires anticipation of estimated time of usage (6).

Both PHIL and Onyx are DMSO-based liquid embolic agents that require DMSO-compatible microcatheters. As shown in case 1, Onyx can be delivered through the same catheter after PHIL injection, which oers the advantage of switch between agents during endovascular therapy without injecting additional DMSO.

PHIL has limited visibility during passage through the microcatheter (especially PHIL 25%). After passage through the tip of microcatheter, PHIL has a lower visibility than Onyx. PHIL is bound to iodine for radi-opacity and does not lose its homogeneity during prolonged injections, which oers the advantage of consistent visibility. Onyx uses tantalum for radiopacity, which can cause inhomogeneity during prolonged injections and decrease visibility. Onyx also suers from self-hiding eect when used in large amounts due to dense radiopaque saturation (7).

Plug formation with PHIL is faster because of less layering eect compared to Onyx, which may oer decreased duration of embolization and uoroscopy time (2, 5, 6). Backow into arterial side is slower with PHIL, which may limit unintended proximal arterial embolization and decrease the amount of embolic agent used.

PHIL has a column eect when it reaches the venous side, which decreases forward ow in the venous side. Decreased forward ow in venous side limits venous obliteration, which may decrease the rate of post-embolization hemorrhage due to venous outow obstruction (8).

Onyx causes CT beam hardening artifact and intense artifact in gradient recalled-echo (GRE) and susceptibility-weighted imaging (SWI) sequences which limit evaluation of perinidus parenchyma during post-treatment follow up (9, 10). PHIL has minimal CT beam hardening artifact and no artifact in GRE and SWI.

Onyx and PHIL have dierent gross appearances inside the embolized vessels; Onyx is dark gray, which limits its use in vascular malformations in the face due to tattoo eect. PHIL has a white color, which may oer aesthetic advantage for use in vascular malformations in the face (11).

Grossly, Onyx forms a continuous smooth surfaced gray-dark column within vessels, whereas PHIL forms a rough-surfaced whit-ish column interspersed by blood clots. During surgical manipulation, Onyx feels as a rubbery, pliable material, whereas PHIL is sti and brittle (12, 13). When PHIL-lled vessel is cauterized with bipolar cautery, crunching and breaking of the intravascular column can be noted (Fig. 4). This feature of PHIL is important when handling small vessels near the distal end of PHIL column, especially when working under the nidus. However, these dierences do not aect the overall surgery, and material properties are easily handled during the rst few minutes of dissection.

All patients underwent embolization with PHIL followed by same-day surgical resection. PHIL was the sole embolization material in cases 2 and 3, while Onyx and PHIL were both used in case 1.

Review of the literature shows limited experience with same-day embolization and surgical resection in brain AVMs. In a case of Onyx embolization with same-day surgical resection, no evidence of associated inammatory inltrate was noted. In four brain AVMs surgically resected one day after embolization, mild acute inammatory reaction was seen in two cases (14, 15). We observed moderate amount of vascular inammation in cases 2 and 3, who were embolized with PHIL only. In case 1, parts of the nidus embolized with Onyx had less vascular inammatory reaction compared with PHIL-embolized parts.

No evidence of chronic inammatory response and foreign body giant cells were observed in our cases since same-day surgical nidus resection was performed.

Angionecrosis was not observed in embolization with PHIL; however, in case 1, where PHIL injection was followed by Onyx injec-

tion, angionecrosis was seen in Onyx lled vessels. Angionecrosis rate of Onyx was reported as 30.2% after an average duration of 18 days (15). Jahan et al. (14) reported angio-necrosis in two of four specimens resected one day after embolization. The most likely cause of angionecrosis after Onyx embolization is the solvent DMSO (16, 17). Angionecrosis was absent in cases 2 and 3, who were embolized by PHIL only (even though PHIL is also a DMSO-based agent) and was only present in case 1, who was embolized with Onyx in addition to PHIL.

Smallest vessel occluded by PHIL was 2.9 m in our specimens (Fig. 6). Natarjan et al. (15) reported the smallest vessel occluded by Onyx to be 5 m with an average diameter of 81.4103.8 m. Jahan et al. (14) reported blood vessels ranging from 80 m to 1 mm in diameter lled with the embolic material.

In conclusion, we acknowledge the limited patient number in this case series. Main possible advantages of PHIL compared with the widely used Onyx from an endovascular perspective are its ease of use, faster plug formation, and less CT and MRI artifact during follow up. Although PHIL has decreased visibility, it has constant homogeneity. From a surgical and histopathologic perspective, in comparison with Onyx, PHIL is brittle during vessel dissection and less pliable during surgery, and causes moderate vascular inammatory eect and no angionecrosis on same-day resected specimens. Histopathologic analysis showed a higher degree of penetration with PHIL than with Onyx. This may be an advantage during the embolization of complex malformations, but may also be an a disadvantage due to small perforator occlusion. Larger prospective multicenter series are needed to conrm the safety and eectiveness of this new embolic material.

Conict of interest disclosure

N.K. has a proctoring and consulting agreement with MicroVention, Inc. and proctoring agreement Covidien, Inc. C.I. has a proctoring agreement with MicroVention, Inc. and Covidien, Inc. The other authors have no conict to report.

References

1. Barr JC, Ogilvy CS. Selection of treatment modalities or observation of arteriovenous malformations. Neurosurg Clin N Am 2012; 23:6375. http://dx.doi.org/10.1016/j.nec.2011.09.010

Web End =[CrossRef]

2. Saatci I, Geyik S, Yavuz K, Cekirge HS. Endovascular treatment of brain arteriovenous malformations with prolonged intranidal Onyx injection technique: long-term results in 350 consecutive patients with completed endovascular treatment course. J Neurosurg 2011; 115:7888. http://dx.doi.org/10.3171/2011.2.JNS09830

Web End =[CrossRef]

188 MarchApril 2016 Diagnostic and Interventional Radiology Koer et al.

3. Leyon JJ, Chavda S, Thomas A, Lamin S. Preliminary experience with the liquid embolic material agent PHIL (Precipitating Hydrophobic Injectable Liquid) in treating cranial and spinal dural arteriovenous stulas: technical note. J Neurointerv Surg 2015. [Epub ahead of print] http://dx.doi.org/10.1136/neurintsurg-2015-011684

Web End =[CrossRef]

4. Gentric JC, Raymond J, Batista A, Salazkin I, Gevry G, Darsaut TE. Dual-lumen balloon catheters may improve liquid embolization of vascular malformations: an experimental study in Swine. AJNR Am J Neuroradiol 2015; 36:977 981.http://dx.doi.org/10.3174/ajnr.A4211

Web End = [CrossRef]

5. van Rooij WJ, Jacobs S, Sluzewski M, van der Pol B, Beute GN, Sprengers ME. Curative embolization of brain arteriovenous malformations with onyx: patient selection, embolization technique, and results. AJNR Am J Neuroradiol 2012; 33:12991304. http://dx.doi.org/10.3174/ajnr.A2947

Web End =[CrossRef]

6. Panagiotopoulos V, Gizewski E, Asgari S, Regel J, Forsting M, Wanke I. Embolization of intracranial arteriovenous malformations with ethylene-vinyl alcohol copolymer (Onyx). AJNR Am J Neuroradiol 2009; 30:99106. http://dx.doi.org/10.3174/ajnr.A1314

Web End =[CrossRef]

7. Katsaridis V, Papagiannaki C, Aimar E. Curative embolization of cerebral arteriovenous malformations (AVMs) with Onyx in 101 patients. Neuroradiology 2008; 50:589597. http://dx.doi.org/10.1007/s00234-008-0382-x

Web End =[CrossRef]

8. Rangel-Castilla L, Spetzler RF, Nakaji P. Normal perfusion pressure breakthrough theory: a reappraisal after 35 years. Neurosurg Rev 2014; 38:399405. http://dx.doi.org/10.1007/s10143-014-0600-4

Web End =[CrossRef]

9. Pierot L, Kadziolka K, Litre F, Rousseaux P. Combined treatment of brain AVMs with use of Onyx embolization followed by radiosurgery. AJNR Am J Neuroradiol 2013; 34:13951400. http://dx.doi.org/10.3174/ajnr.A3409

Web End =[CrossRef]

10. Saatci I, Cekirge HS, Ciceri EF, Mawad ME, Pamuk AG, Besim A. CT and MR imaging ndings and their implications in the follow-up of patients with intracranial aneurysms treated with endosaccular occlusion with onyx. AJNR Am J Neuroradiol 2003; 24:567578.

11. Arat A, Cil BE, Vargel I, et al. Embolization of high-ow craniofacial vascular malformations with onyx. AJNR Am J Neuroradiol 2007; 28:14091414. http://dx.doi.org/10.3174/ajnr.A0547

Web End =[CrossRef]

12. Akin ED, Perkins E, Ross IB. Surgical handling characteristics of an ethylene vinyl alcohol co-polymer compared with N-butyl cyanoacrylate used for embolization of vessels in an arteriovenous malformation resection model in swine. J Neurosurg 2003; 98:366370. http://dx.doi.org/10.3171/jns.2003.98.2.0366

Web End =[CrossRef]

13. Hamada J, Kai Y, Morioka M, et al. A nonadhesive liquid embolic agent composed of ethylene vinyl alcohol copolymer and ethanol mixture for the treatment of cerebral arteriovenous malformations: experimental study. J Neurosurg 2002; 97:889895. http://dx.doi.org/10.3171/jns.2002.97.4.0889

Web End =[CrossRef]

14. Jahan R, Murayama Y, Gobin YP, Duckwiler GR, Vinters HV, Vinuela F. Embolization of arteriovenous malformations with Onyx: clinicopatho-logical experience in 23 patients. Neurosurgery 2001; 48:984995. http://dx.doi.org/10.1097/00006123-200105000-00003

Web End =[CrossRef]

15. Natarajan SK, Born D, Ghodke B, Britz GW, Sekhar LN. Histopathological changes in brain arteriovenous malformations after embolization using Onyx or N-butyl cyanoacrylate. Laboratory investigation. J Neurosurg 2009; 111:105113. http://dx.doi.org/10.3171/2008.12.JNS08441

Web End =[CrossRef]

16. Chaloupka JC, Vinuela F, Vinters HV, Robert J. Technical feasibility and histopathologic studies of ethylene vinyl copolymer (EVAL) using a swine endovascular embolization model. AJNR Am J Neuroradiol 1994; 15:11071115

17. Murayama Y, Vinuela F, Ulhoa A, et al. Non-adhesive liquid embolic agent for cerebral arteriovenous malformations: preliminary histopathological studies in swine rete mirabile. Neurosurgery 1998; 43:11641175. http://dx.doi.org/10.1097/00006123-199811000-00081

Web End =[CrossRef]

Preliminary experience with PHIL in brain AVM 189