PNIPAM, a medically graded copolymer, consists of 0.75 mol fraction N-isopropylacrylamide, and 0.25 mol fraction N-n-propylacrylamide. The LCST of PNIPAM is 33.92˚C. The copolymer turns into sedimentation in water at 37˚C about 3.2 seconds. The embolic liquid agent we developed is a mixture of 1 g PNIPAM and 30 g powder iohexol dissolved in water to 100 ml in total volume. After the PNIPAM and iohexol completely dissolved in water, the mixture was sterilized at 120˚C for 20 minutes and stored at 20˚C. The quality of sterilization was bacteriologically confirmed for each bath. The mixture had a viscosity of 24－36 mPa·s at 20˚C. The mixture was delivered through both Prowler-10 (Cordis, Micro Therapeutics, USA) and NO.1.8 French Balt Magic catheter (Balt, Montmorency, France).

Swine rete mirabile is a fine network of arteries with connections across the midline to the contralateral rete situated at the termination of each ascending pharyngeal artery as it perforates the cranial base. This vascular network has some morphological similarities to a human plexiform arteriovenous malformations (AVM) nidus, and it has been previously used for assessment of vascular histological responses of numerous embolic agents.^1,4 Thirty domestic pigs were used in this study. The animals were 3 to 4 months old, weighed 25 to 30 kg, and were of mixed sex. After an overnight fast, each pig was premedicated with intramuscular 20 mg/kg of ketamine and 0.5 mg atropine. Then 0.10 mg/kg of midazolam was intravenous administered in a bolus. General anesthesia was maintained with 1% ketamine intravenous administration continuously.

A 6F vascular sheath was placed into the right common femoral artery using the Seldinger technique. A 5-6 F guiding catheter flushed with cool saline at 20˚C was used to select the common carotid artery, through which a microcatheter was placed coaxially for superselective catheterization of the ascending pharyhgeal artery. The tip of the microcatheter was placed distal to the pharyngeal branch of ascending pharyngeal artery to ensure that infusions were delivered to the rete mirabile only. Selective angiography was performed to ensure proper positioning and to study the anatomic configuration and blood-flow pattern of the rete mirabile.

The goal of our study was to evaluate the feasibility, reproductivity, and safety of the embolization process with PNIPAM. The criteria for evaluating included: ① ease of delivery through the microcatheter, ② controllability (ability to define the starting and ending points of embolization, radiopacity, injection speed and volume), ③ extent and depth of penetration into the rete,④ permanence of occlusion of the rete (long-term follow-up), ⑤ gluing of the microcatheter to the artery.⁸

In order to prevent the mixture precipitation in the microcatheter, the rete embolization was performed using a three-stage technique. The technique consisted of first infusing the mocrocatheter with cool saline at 20˚C to prevent precipitation, followed by injection of the PNIPAM embolic mixture (1 ml/min). After the procedure, the microcatheter was flushed with cool saline again. One experienced neuroradiologist performed all of the embolizations.

A protocol was developed to study certain hemodynamic consequences (assessed by angiography) and histological changes from the PNIPAM mixture embolization. In all animals, 0.5 to 1.0 ml cool saline followed by 2.5 to 3.3 ml PNIPAM was injected under continuous fluoroscopic observation. Controls consisted of either contralateral retia infused with saline or noncatheterized and uninjected specimens. Three intervals for follow-up evaluation of hemodynamic and histopathologic consequences of the infusions were performed as follows: acute (the same day of injection), subacute (2 weeks from embolization) and chronic (1 to 6 months from embolization). All follow-up studies included bilateral common carotid and ascending pharyngeal artery angiography to assess the extent and persistence of occlusion of embolized ascending pharyngeal arteries and retia.

Necropsy was performed immediately after all the animals were sacrified by lethal injection of overdose phentobarbital (100 mg/kg). Each rete was carefully exposed and dissected from the cavernous sinus. The specimen was grossly inspected for various post embolization changes, including texture/consistency, thrombosis, extravasation, inflammatory/granulation response, and fibrosis. Each rete was placed into 10% formalin (37˚C) for fixation. Standard techniques were used for preparing sections of the rete for light microscopy. Sections were stained with hematoxylin and erosin. An experienced neuropathologist evaluated the histopathologic changes in the embolized retia.

RESULTS

Technical and clinical evaluation
During embolization, the mixture was clearly visualizaed fluoroscopically. The PNIPAM mixture was used for rete embolization in 30 swine, 28 swine survived the embolization, 2 swine died. One swine died of cerebrum infarction and the other died of embolic agent reflux into the occipital artery because of fast injection (>2 ml/min) in preliminary study. We then performed in vitro evalution efficacy of the PNIPAM mixture using the apparatus as Kazakawa and his colleagues illustrated.⁹ It was found that the in vitro AVM model could be successful embolized with the PNIPAM injection rate at 0.75 to 1.5 ml/min, while the PNIPAM would be flowed through the model if injection was too fast and precipitation in the microcatheter would occur if injection too slow. The mixture was injected at the rate of 1 mm/min in all other procedures.

Moderate vasospasm of the ascending pharyngeal artery occurred twice during selective micro- catheterization. Papaverine infusion through micro- catheter did not lessen vasospasms, and the PNIPAM mixture injection was performed then. The common carotid angiography after embolization presented that the retia remained partially filled by ramus anastomoticus and arteria anastomotica  (Figs. 1, 2).

view in a new window

Fig. 1. Swine common carotid angiography demon-　strated the ascending pharyngeal artery, the rete mirabile (arrow) and the internal carotid artery. Fig. 2. Common carotid angiography post embolization showed that the rete mirabile was embolized.

Subacute follow-up was performed in 6 swine, in which the ascending pharyngeal and rete artery recanalization did not occur. It was observed that the ramus anastomoticus and/or arteria anastomotica filling the rete surround part in 4 swine. In this group, all animals survived in the interval without prominent symptoms.

The chronic group included 16 swine. Follow-up angiography of 1, 2, 3 and 6 months after embolization was performed in 5, 5, 3 and 3 swine respectively. The ascending pharyngeal and rete artery recanalization presented in 2 swine. It was verified that the two swine were those two in which the ascending pharyngeal artery vasospasm occurred during selective microcatheterization. It suggested that vasospasm lessened blood flow of the ascending pharyngeal artery, thus disturbed penetration of the embolic agent. In 8 of other 14 cases, it was observed that the rostral and/or surround part of the retia were supplied by the ramus anastomoticus and/or arteria anastomotica issued from the external carotid, which suggested that previous injection of the embolic agent did not fill the communicant region between the rete and the external carotid branches, and the external carotid blood supply to the rete enhanced after embolization (Figs. 1, 2). In this group, all animals survived the follow-up period.

Injection of the PNIPAM mixture by using a pump at the rate of 1 ml/min through the microcatheter was readily controlled with fluoroscopy in all emboli- zations because of the contrast of iohexol. The starting and ending points of the embolic procedure could be exactly determined with continuous fluoroscopy. There were no difficulties encountered withdrawing the microcatheter after embolization. The inner wall of the microcatheter was smooth and there was no copolymer adhered to the wall.

Histopathologic evaluation
Necropsy was performed immediately after the animals were sacrified. In two cases died of procedure complications, there were no pronounced gross changes in brain. In acute and subacute groups, gross inspection showed that the embolized retia were spongy and soft. “Gel casting” of the distal ascending pharyngeal and proximal retial artery with gelatinoid material was seen, which was similar to the appearance of latex injection specimens of vasculature. The embolized retia seemed no significant change in comparison with the contralaterals. In chronic group, the caudal portions of embolized retia were notable corrugativus and retractive.

The rete vessel lumen was filled with the pink material (the PNIPAM cast) with several neutrophils scattered in acute specimens, which indicated that there was mild inflammatory reaction. The PNIPAM mixture penetrated into the vessels of 100－150 µm in diameter. No signs of subintimal bleeding and angionecrosis were presented (Fig. 3). There was no copolymer inside the internal carotid artery.

In subacute and chronic groups, the copolymer was still found inside the rete arteries with mononuclear cells and eosinophils scatting inside and around it. The muscular layer was loosened like ischemia, with most muscular nuclei degraded, like presentations after radiosurgery (Figs. 4－6).¹⁰

view in a new window

Fig. 3. Photomicrograph of an rete obtained 2 hours after embolization with the copolymer showed that the rete vessels were filled with precipitated copolymer and some neutrophils (arrow) integrated inside and around the copolymer (HE, Original magnification 10×3.3). Fig. 4. Photomicrograph of an rete obtained 12 days after embolization showed that the copolymer  (arrow) remained inside the rete vessels with several neutrophils and eosinocytes scattered (HE, Original magnification 4×3.3）. Fig. 5. Photomicrograph of a rete obtained 2 months after embolization demonstrated that the copolymer still remained inside the rete vessels (arrow) (the copolymer was stripped off during pathological section dealing) without endothelial denuding and necrosis (HE, Original magnification 4×3.3). Fig. 6. Photomicrograph of a rete obtained 3 months after embolization demonstrated that the copolymer still remained inside the rete vessels (arrow) (HE, Original magnification 10×3.3).

DISCUSSION

With the development of new embolic agents and advances in guidewire and microcatheter technologies, endovascular treatment of cerebral AVMs has become increasingly safe and effective. At present, most researchers inject liquid materials, especially cyanoacrylates and EVAL/EVOH, into the AVM nidus and achieve higher success rates in anatomic cures by embolization alone and in the prevention of AVM recurrence and hemorrhage.^2,11-13 Cyanoacrylates are adhesive agents and difficult to handle; sometimes, they cause the microcatheter to adhere the artery and their neurotoxicity has not been formally evaluated.¹ The EVAL/EVOH, as a nonadhesive agent, has advantages for AVM embolization compared with Cyanoacrylates. On the other hand, the EVAL/EVOH should be dissolved in organic solvents (DMSO or ethanol and organic solvents would cause damage to the artery and the surrounding tissue of the AVM.^2,14 An ideal embolic material should be ① satisfactorily biocompatible, ② effective in blood vessel embolization, ③ nontoxic, non-carcinogenic, non-teratogenic, ④ easily available, ⑤ easy to inject through various catheters, ⑥ effective in causing non-injurious inflammation and inducing thrombosis, and ⑦ radiopaque.^15,16

Since the unique thermal behavior of PNIPAM in aqueous media was previously described, many researchers have been engaging in the properties and applications of PNIPAM. It is soluble in water at low temperature, but becomes insoluble when the temperature exceeds the LCST. The conversion between hydrophilic and hydrophobic states is thermally reversible. The PNIPAM had been examined for various possible applications including mechano-chemical actuators, drug delivery devices, and a recovery system of cultured cells.⁷ In 1996, Matsumaru et al⁶ determined the LCST of the PNIPAM for the embolic material. In their study，the PNIPAM with the LCST of 24－26˚C showed appropriate precipitation. To prove the occlusion of vessels in vivo, they injected the copolymers into a rabbit kidney through a microcatheter. The extent of embolization was evaluated by angiography and histological examination. An acute toxicity test of the PNIPAM was performed in comparison with that of the NIPAM monomer. The copolymer used in their study showed no acute toxicity in mice. Water solubility, non-adhesiveness, and non-toxicity are the advantages of the use of thermosensitive polymers as an embolic material. By changing the LCST, various embolic materials can be designed. Based on their results, it was believed that the application of thermosensitive polymers as a new embolic material was very promising.

In our study, we performed rete embolization using the PNIPAM mixture in 30 domestic swine. Except 2 cases in preliminary study, in which the animals died of too fast injection of embolic agent, the other 28 pigs survived either in the embolizing procedures or in follow-up periods with injecting rate at 1 ml/min using a pump. The follow-up angiography revealed that the agent could embolize the retia permanently. The mixture was radiopaque during injection under continuous fluoroscopic observation. The inner wall of the microcatheter was smooth and there was no copolymer adhered to the wall.

In acute group, the rete vessel lumen was filled with the PNIPAM cast with several neutrophils scattered, which indicated that there was mild inflammatory reaction. The PNIPAM mixture penetrated into the 100－150 µm vessels in diameter. No signs of subintimal bleeding and angionecrosis were present. In subacute and chronic groups, the copolymer was still found inside the rete arteries, scattered mononuclear cells and eosinocytes were found inside and around it. The muscular layer was loosened like ischemia, with most muscular nuclei degraded, like presentations after radiosurgery.¹⁰

In conclusion, as a thermosensitive embolic material, PNIPAM has optimized characteristics including embolic instability, better biocompatibility, radiopacity with iohexol, non-adhesiveness. Because of its properties, PNIPAM has great potential as a therapeutic non-adhesive embolic agent.

REFERENCES

1.Wright KC, Greff RJ, Price RE. Experimental evaluation of cellulose acetate NF and ethylene-vinyl alcohol copolymer for selective arterial embolization. J Vasc Interv Radiol 1999;10:1207-1218.

2.Hamada JI, Kai Y, Morioka M, Kazekawa K, Ishimaru Y, Iwata H, et al. A nonadhesive liquid embolic agent composed pf ethylene vinyl alcohol copolymer and ethanol mixture for the treatment of cerebral arteriovenous malformations: experimental study. J Neurosurg 2002;97:889-895.

3.Chaloupka JC, Huddle DC, Alderman J, Fink S, Hammond R, Vinters HV. A reexamination of the angiotoxicity of superselective injection of DMSO in the swine rete embolization model. AJNR Am J Neuroradiol 1999;20:401-410.

4.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: 1107-1115.

5.Murayama Y, Vinuela F, Ulhoa A, Akiba Y, Duckwiler GR, Gobin YP, et al. Nonadhesive liquid embolic agent for cerebral arteriovenous malformations: preliminary histopathological studies in swine rete mirabile. Neurosurgery 1998;43:1164-1175.

6.Matsumaru Y, Hyodo A, Nose T, Tto S, Hirano T, Ohashi S. Application of thermosensitive polymers as a new embolic material for intravascular neurosurgery. J Biomater Sci Polym Ed 1996;7:795-804.

7.Jeong B, Kim SW, Bae YH. Thermosensitive sol-gel reversible hydrogels. Adv Drug Delivery Rev 2002;54: 37-51.

8.Klisch J, Yin L, Requejo F, Eissner B, Scheufler KM, Kubalek R, et al. Liquid 2-poly-hydroxyethyl-methacrylate embolization of experimental arteriovenous malformations: feasibility study. AJNR Am J Neuroradiol 2002;23:422-429.

9.Kazekawa K, Iwata H, Shimozuru T, Sampei K, Sakaki N, Morikawa N, et al. Nontoxic embolic liquids for treatment of arteriovenous malformations. J Biomed Mater Res 1997;38:79-86.

10.De Salles AA, Solberg TD, Mischel P, Massoud TF, Plasencia A, Goetsch S, et al. Arteriovenous malformation animal model for radiology: the rete mirabile. AJNR Am J Neuroradiol 1996;17:1451-1458.

11.Jahan R, Murayama Y, Gobin YP, Duckwiler GR, Vinters HV, Vinuela F. Embolization of arteriovenous malfor- mations with onyx: clinicophathological experience in 23 patients. Neurosurgery 2001;48: 984-987.

12.Hamada JI, Kai Y, Morioka M, Kazekawa K, Ishimaru Y, Iwata H, et al. A mixture of ethylene vinyl alcohol copolymer and ethanol yielding a nonadhesive liquid embolic agent to treat cerebral arteriovenous malformations: initial clinical experience. J Neurosurg 2002;97:881-888.

13.Debrun GM, Aletich V, Ausman JI, Charbel F, Dujovny M. Embolization of nidus of brain arteriovenous malfor- mations with N-butyl cyanoacrylate. Neurosurgery 1997;40:112-121.

14.Sampei K, Hashimoto N, Kezekawa K, Tsukahara T, Iwata H, Takaichi S. Histological changes in brain and vasculature after intracarotid infusion of organic solvents in rats. Neuroradiology 1996;38:291-294.

15.Shi ZS, Huang ZS, Zhang YW, Qi T, Chu M, Bao D. Experimental study of Eudragit mixture as a new nonadhesive liquid embolic material. Chin Med J 2002; 115:555-558.

16.Turjman F, Massound TF, Vinters HV, Ji C, Tardy M, Guglielmi G, et al. Collagen microbeads experimental evaluation of an embolic agent in the rete mirabile of swine. AJNR Am J Neuroradiol 1995;16:1031-1036.