Modified ethoxyethyl cyanoacrylate for therapeutic embolization of arteriovenous malformation Yin-Chao Tseng, Suong-Hyu Hyon, and Yoshito Ikada Research Center for Medical Polymers and Biomaterials, Kyoto University, 53, Kawahara-cho, Shogoin, Sakyo-ku, Kyoto, Japan, 606 War0 Taki, Akira Uno, and Yasuhiro Yonekawa Department of Neurosurgery, National Cardiovascular Center, 5-7-1 Fujishiro-dai, Suita, Osaka, Japan, 565
Aiming at alleviating the problems of using 2-cyanoacrylates as a material for therapeutic embolization, this experiment made some physical modifications by mixing contrast media. It was found that the physicochemical properties of 2-cyanoacrylates can be altered by changing the concentration and the composition of the contrast media added. A 50 wt% cyanoacrylate-50 wt% contrast medium mixture has enough radiopacity for the practical requirement for embolization. A mixture of 50 wt% (ethoxyethyl cyanoacrylate-5 wt% lac-
tide/&-caprolactone copolymer), 25 wt% lipiodol and 25 wt% tetrafluorodibromoethane provides a viscosity of 13.8 cP, a bonding strength of 14.9 kg/cm2, a set time of 6 s, and a spreading in canine blood of 33 mm. It was concluded that the mixture is much more satisfactory than the conventional cyanoacrylates as an embolus material in vitro. The results obtained by in vivo experiments and clinical trials so far suggest that the mixture is very promising as a material for embolization.
INTRODUCTION
2-Cyanoacrylates (CA), which are liquid monomers, polymerize very rapidly on contact with an ionic solution or endothelium. Because of this ability, they have been widely used in many surgical fields as hemostatic agents and tissue adhesives.l*’They also have been reported to reinforce cerebral aneurysm^^,^ and to repair dural cerebrospinal fluid leak^.^-^ Recently, they have attracted considerable attention in therapeutic embolization of cerebra1 arteriovenous malformation (AVM).B-llAlthough ethyl cyanoacrylate (ECA) and isobutyl cyanoacrylate (IBCA) are commonly used as material for therapeutic embolization, some of their physicochemical properties should be improved to obtain higher reliability. First of all, the CA monomers should be mixed with a contrast medium in order to be monitored through angiography, because they are not radiopaque. In addition, they should have appropriate viscosity and set time in blood. A low bonding strength is Address correspondence to Dr. Ikada at Research Center for Medical Polymers and Biomaterials, Kyoto University, 53, Kawahara-cho, Shogoin, Sakyo-ku, Kyoto, Japan, 606. Journal of Biomedical Materials Research, Vol. 24, 65-77 (1990) CCC 0021-9304/90/01OO65- 13$O4.00 0 1990 John Wiley & Sons, Inc.
TSENG ET AL.
66
desirable for removing a microcatheter from the blood vessels easily. Besides, low stiffness of the embolized part is also required as the part needs to be removed with surgical resection. Cromwell et al. modified IBCA by mixing iophendylate to make it more practical and reported that its radiopacity and polymerization time were satisfactory." However, it has been found in our recent report that ECA and IBCA polymers have very high stiffness.13Therefore, we modified the homologous series of CA to develop soft adhesives and found that the homologs with longer ester side groups, especially ethoxyethyl cyanoacrylate (EECA), generate more flexible polymers than the homologs with shorter Furthermore, a lactide (LA)/&-caprolactone(CL) copolymer [P(LAco-CL)] was found to plasticize the resulting CA polymer^.'^ Very recently, we have also reported that tetrafluorodibromoethane (FEB,) and hexafluorodibromopropane are good as contrast medium. l4 In the present work, we will attempt to find a soft embolus material which has the desired radiopacity, viscosity, set time, and spreading in blood as well as a low bonding strength.
MATERIALS AND METHODS
Materials EECA was synthesized with the conventional method from ethoxyethyl cyanoacetate and paraformaldehyde (Y. -C. Tseng, unpublished data). The P(LA-co-CL) was prepared by ring-opening copolymerization of D,L-lactide and &-caprolactonein bulk.15 The copolymer used in this work has a molar chemical composition of 40/60 as D,L-lactide/&-caprolactone ratio and a weight-average molecular weight of 1 X lo5. The mixtures [EECA-P(LA-coCL)] of EECA and P(LA-co-CL) were prepared by dissolving the copolymer in EECA completely at 37°C. Various ratios of FEB, (Daikin Industries LTD., Japan) and lipiodol ultra-fluide (lipiodol) (Kodama Co., Japan) were added into the mixtures at room temperature. The mixtures were stored at 4°C until tests were performed.
Stiffness The stiffness of the polymerized EECA-P(LA-co-CL)mixtures was measured using the apparatus designed and proposed by Hayashi et al.'6f'7for the measurement of pressure-diameter re1 tions of blood vessels. The ex7 by a television system conternal diameter of the specimen was measured sisting of a Vidicon camera and a width analyzer (Hamamatsu Photonics, Hamamatsu, C1000-16 and HTV-C1170). The intraluminal pressure was measured with a pressure transducer (Toyo Baldwin, Tokyo, MPU-0.5-290111), which was attached to the vessel segment via a short catheter. Briefly,
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
67
carotid artery of dogs was resected, stored in Krebs-Ringer solution at 4"C, and transferred to testing within 12 h. In order to remove the initial stressrelaxation effect, the resected test carotid artery, which was stored at 37°C in Krebs-Ringer solution under aeration with a 95% 0, and 5% CO, gas, was kept at 100 mm Hg with Krebs-Ringer solution for 15 min. The intraluminal pressure was then gradually elevated to 200 mm Hg and lowered to 0. This specific inflation and deflation procedure was repeated until a stable pressure-diameter hysteresis loop was formed. The carotid artery was then coated to about 0.3-mm thickness with EECA-P(LA-co-CL)mixtures and subjected to the distension to measure the pressure-diameter relation of the coated carotid. The relation can be expressed by the following equation:
P(D/D, - 1) where D , is the carotid wall diameter when the standard pressure P, (100 mm Hg) is applied and D is the external diameter when the intraluminal pressure is P between the physiological pressure range such as 60160 mm Hg.I7p is a parameter identifying the stiffness of the tube wall. The average p values were obtained by reading on six specimens. In P/P,
=
Viscosity and bonding strength The viscosity of the mixtures of various concentrations of EECA and EECA-5%P(LA-co-CL) with FEB, and lipiodol was measured with a rotational viscometer (Tokyo Keiki Co., LTD., Japan) at 25°C. The bonding shear strengths of bonded stainless strips were measured as references on the change of the bonding strength after mixing contrast media to CA for applying the materials in vivo later on. The measurement was performed in accordance with JISK 6850-76, which is a Japanese standard, using an Instron tensile machine (Shimadzu Seisakusho LTD., Kyoto, Japan) at a separation rate of 10 mm/min. Briefly, the surface (25 X 12.5 mm') of stainless steel test strips (16 x 25 x 100 mm3)was polished with a 240 number sandpaper, rinsed rigorously with acetone, dried at 50°C for 4 h, and kept at 60% RH and 25°C overnight. A 10-pL sample of each mixture was applied to one test strip, another one was lapped on it for adhesion for 24 h. The shear strengths were then determined by averaging at least five readings. Set time and spreading The set time of the mixtures was determined by dropping 10 p L of the mixtures from a height of 9 mm on fresh canine blood, which was placed in 41 mm Petri dish to 7 mm depth, followed by measuring the period of time just from the contacting moment of the mixtures on the blood surface to that of the complete opacification." Fifteen measurements were made to obtain the average value. The spreading of the mixtures was estimated by measuring the diameter of the film formed on the blood surface when 10 pL
TSENG ET AL.
68
of the mixtures was dropped on it as described above. All of these measurements were done at room temperature unless otherwise noted. RESULTS
The solubility limit of lipiodol is 60 wt% both for EECA and EECA5%P(LA-co-CL)at room temperature. On the other hand, FEB, is soluble unlimitedly in these liquids to form stable and completely miscible solution. The EECA polymer is more flexible than the methyl cyanoacrylate, the ECA and the IBCA polymers. To quantitate the polymer flexibility, we determined the @ values of canine carotid arteries coated with the EECA-P(LAco-CL) mixtures. The monomer liquid almost instantly sets to a gel upon coating on the wet artery. The observed p values are given in Table I. The stiffness parameter is seen to decrease as the concentration of P(LA-co-CL) in the mixture increases. This is the same tendency as that reported in our previous work on canine femoral arteries coated with the same rnixt~res.'~ The viscosity of the EECA-lipiodol mixtures increased with an increase of the added amount of lipiodol, whereas the EECA-5%P(LA-co-CL)-lipiodol mixtures had a minimum in viscosity at the lipiodol concentration of approximately 50 wt%. In contrast, the viscosity of EECA as well as EECA5%P(LA-co-CL)became lower upon addition of FEB,. The result is tabulated in Table 11. Table I11 gives the shear strengths of the stainless strips TABLE I of Canine Arteries Coated with Stiffness Parameter Values EECA-P(LA-co-CL) Mixtures
P
Coating ~
None EECA EECA-~%P(LA-CO-CL) EECA-~~%P(LA-CO-CL)
~~
8.35 2 0.71 34.87 f 3.23 23.16 f 2.50 14.22 f 2.58
TABLE I1 Viscosities of EECA and EECA-5%P(LA-co-CL) Mixed with Lipiodol and FEB2. Ratio of Cyanoacrylate to Contrast Medium W/WI
100/0 70/30 60/40 50/50 40/60 0/100
Viscosity (cP) EECA
EECA-5%P (LA-CO-CL)
Lipiodol
FEB,
Lipiodol
FEB,
4.5 8.3 9.5
4.5 4.1 3.8 3.4 3.2 1.7
41.3 32.8 29.3 25.0 28.5 34.9
41.3 15.0 9.8 6.5 4.5 1.7
12.0
14.8 34.9
69
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
TABLE I11 Bonding Strengths between Two Steel Strips Bonded with EECA and EECA-5%P(LA-co-CL)Containing Lipiodol and FEB2 Shear Strength (Kg/cm2)
Ratio of Cyanoacrylate to Contrast Medium
EECA
EECAS%P(LA-co-CL)
W/W)
Lipiodol
FEBz
Lipiodol
FEBz
100/0 70130 60140
130.7 2 2.5 40.5 f 1.5 26.7 f 1.8 21.3 f 2.1 10.6 f 0.5
130.7 f 2.5 107.8 f 3.0 91.9 f 8.6 83.4 k 5.8 80.3 f 4.8
101.5 f 2.7 39.3 f 1.3 27.4 f 2.1 18.4 f 1.6 11.4 f 1.2
101.5 f 2.7 109.6 f 2.5 111.0 k 2.4 106.9 k 4.3 100.0 5 3.6
so/m 40160
bonded with the EECA-lipiodol and EECA-5%P(LA-co-CL)-lipiodol mixtures. As can be seen, the bonding strengths decrease drastically with the increasing amount of lipiodol. On the other hand, no distinguishable change is observed for the EECA-5%P(LA-co-CL)-FEB2 mixtures, while the EECA-FEB, mixtures show a slight decrease in bonding strength on addition of FEB,. The direction of the set time change induced by addition of the contrast medium was different, depending on the nature of the contrast medium used, irrespective of the kinds of CA. It is seen in Table IV that the set time increases with the increasing amount of lipiodol but decreases with an increase of the FEB, amount. As shown in Figure 1, the radiopacity of a 50% (EECA-5%P(LA-co-CL))50% FEB, mixture is good enough. The viscosity and bonding strength of the mixtures from 50 wt% of EECA-5%P(LA-co-CL)and 50 wt% contrast medium containing different ratios of lipiodol to FEB, were determined and plotted against the composition of the contrast media in Figure 2. The viscosity of the mixtures decreases with the increasing amount of FEB,, while the bonding strength decreases with the increasing amount of lipiodol. In the same way, the measured spreading and set times were plotted in Figure 3, where it is clearly seen that as the ratio of FEB, increases, the spreading of the mixture increases while the set time decreases. TABLE IV Set Times of EECA and EECA-S%P(LA-co-CL)Mixed with Lipiodol and FEBz Set Time
Ratio of Cyanoacrylate to Contrast Medium
EECA
EECA-5%~'(LA-co-CL)
(WIW
Lipiodol
FEBz
Lipiodol
FEB2
lOO/O 70130 60140 50150 40160
6.0 f 0.4 8.2 f 0.3 10.0 % 0.4 12.0 2 0.7 18.7 f 0.9
6.0 f 0.4 3.5 f 0.4 2.4 % 0.1 1.8 ? 0.2 1.5 f 0.0
21.8 f 0.7 25.4 f 1.9 27.2 f 1.9 31.8 +- 1.8 36.4 f 2.2
21.8 2 0.7 12.8 f 0.7 5.8 f 0.2 3.6 f 0.2 2.4 ? 0.1
70
TSENG ET AL.
Figure 1. Plain x-ray film after injection of a 50% (ethoxyethyl cyanoacrylate-5% lactide/&-caprolactone copolymer)-50% tetrafluorodibromoethane mixture into the canine carotid artery. The mixture occluded the carotid artery and its branches. The mixture was clearly radiopaque (arrow).
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
n,,
N
*Ot
50 40 0
10
10 40
20 30
30 20
FEB2 and lipiodol conc.
O
FEE2 (ipiodol
50
(wf%)
Figure 2. Bonding strengths a n d viscosities of liquids from-50% (ethoxyethyl cyanoacrylate-5% lactide/&-caprolactone copolymer) with different mixing ratios of tetrafluorodibromoethane (FEBJ to lipiodol.
-
7-7
136
J
32
28
48
1
o
24
l 50 4 0 0
10
30 20
d 20 30
o 10 0 FEE2 40 50 lipiodol
FEE2 and lipiodol conc.
(wt%)
Figure 3. Spreading and set times of liquids from 50% (ethoxyethyl cyanoacrylate-5% lactide/&-caprolactone copolymer) with different mixing ratios of tetrafluorodibromoethane (FEB?)to lipiodol.
71
72
TSENG ET AL.
(a) Figure 4. Intraoperative embolization of the cerebral arteriovenous malformation of the left basal ganglia. (A) Right internal carotid arteriogram showed the cerebral arteriovenous malformation (small arrow), which was fed by the artery of Heubner (large arrow). (B) The catheter and the needle were placed into the artery of Heubner (arrow) by craniotomy. The contrast material was injected and the intraoperative angiogram was performed, revealing the arteriovenous malformation on the left basal ganglia. (C) Via this catheter, a mixture of 50% (ethoxyethyl cyanoacrylate-5% lactide/&-caprolactone copolymer), 25% lipiodol and 25% tetrafluorodibromoethane was injected. Malformation was obliterated.
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
73
0-9 Figure 4.
(continued)
DISCUSSION
The effectiveness of CA monomers as material for therapeutic embolization is dependent upon their properties, which include radiopacity, viscosity, set time, bonding strength, stiffness, etc. If the viscosity is very high,
TSENG ET AL.
74
(4 Figure 4. (continued)
the embolization liquid hardly passes through microcatheters. On the other hand, an extremely low viscosity of the liquid is liable to cause the liquid
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
75
material to spread over passing by the place to be embolized. The optimal viscosity seems to be around 10 cP according to our experience. If it takes a long time to polymerize, namely, to set to gel, the m a t e d would have flown over the site of the malformation. If the set time is too short, surgeons might not have enough time to be able to inject the material to the proper site or it would occlude the microcatheter before penetrating into the central part of the malformation. Polymerization time of 5 to 10 s is most desirable. A high bonding strength easily causes the microcatheter to adhere to blood vessels, resulting in difficulty removing the microcatheter from the blood vessels. Embolization is usually performed either as a primary therapy or in conjunction with surgical resection of defective sites, and may be carried out via the femoral approach or at craniotomy, respectively. In the case of simultaneous use of embolization and surgical resection, low stiffness of the embolized part is required for the easy removal. ECA and IBCA have been reported to be useful for embolization of arteriovenous malformation (AVM), but it is also pointed out that the stiffness of ECA and IBCA polymers are too high to resect the embolized AVM. Recently we have found that the stiffness of the EECA polymers is much lower than that of the ECA and the IBCA p01ymers.l~Furthermore, the present work reveals that the stiffness of the EECA polymer can be reduced by mixing with a large amount of P(LA-co-CL)(Table I). Taking the viscosity into consideration, we selected EECA and an EECA-5%P(LA-co-CL) mixture as the candidate material to be mixed with lipiodol and FEB,. As is apparent from Figure 1, radiopacity of the 50 wt% monomer-50 wt% contrast medium mixture was satisfactory. Even though the 50 wt% (EECA-5%P(LA-co-CL))50 wt% lipiodol has the minimum viscosity among the various mixtures, it is still too high. On the contrary, the EECA-FEB, mixtures are too low in viscosity to be applied. As is shown in Table 11, the 40-50 wt% of lipiodol in EECAlipiodol and the 40 wt% of the FEB, in the EECA-5%P(LA-co-CL)-FEB2 seem to be the most appropriate concentration. They may have a similar tendency also in set time, because it is closely related to its viscosity. An addition of lipiodol to the EECAd%P(LA-co-CL)produced too long set time and an addition of FEB, to the EECA resulted in too short set time for polymerization of the embolus liquids in AVM. The 40 wt% of FEB, in EECA-5%P(LA-coCL)-FEB, may be the best concentration with regard to the set time. In addition, there is another property of the embolus mixture to be adjusted. That is the bonding strength. Some mixtures have too high strength to remove the microcatheter from the blood vessels. Lipiodol added to the EECA and the EECA-5%P(LA-co-CL)is able to reduce the bonding strength. Interestingly, lipiodol and FEB, in the mixture change the viscosity, set time, and bonding strength in the opposite fashion. Therefore, we tried to take advantage of using both lipiodol and FEB, to find an ideal embolus liquid. The mixtures selected contain 50 wt% of the EECA-5%P(LA-co-CL)and 50 wt% of the contrast medium with various compositions of lipiodol and FEB,. According to the experimental results shown in Figures 2 and 3, the
TSENG ET AL.
76
mixture containing 50 wt% of the EECA-5%P(LA-co-CL),25 wt% of lipiodol and 25 wt% of FEB, satisfies all the requirements for the viscosity, the bonding strength, and the set time as well as the spreading in blood. Lipiodol and FEB, are freely miscible with EECA-5%P(LA-co-CL)and can provide a wide range of viscosity, set time, and bonding strength. In order to find the optimal condition of the embolus materials in vivo, we injected the mixtures containing 50 wt% of the EECA-5%P(LA-co-CL)and 50 wt% of the contrast medium with various compositions of lipiodol and FEB, into canine carotid arteries. It was found that the mixture containing 50 wt% of EECA-5%P(LA-co-CL)and 25 wt% of lipiodol and 25 wt% of FEB, could satisfy all the required conditions and gave successful resutts. ClinicaI trials have just started. Figure 4 shows one of the examples of intraoperative treatments of AVM of a human. It is concluded that taking advantage of using the contrast media of lipiodol and FEB, with various concentration and compositions can alter the physicochemical properties of CA. The plasticized EECA with P(LA-to-CL) can satisfy the requirement as an embolus material for AVM by adding appropriate amount of the contrast media, and the material seems to be very promising for clinical application. References 1. J. A. O’Leary, “Tissue adhesives and pelvic hemostasis: An evaluation of isobutyl 2-cyanoacrylate,” J. Surg. Oncol., 3, 117-120 (1971). 2. M. C. Harper and M. Ralston, “Isobutyl2-cyanoacrylate as osseous adhesive in the repair of osteochondral fractures,“ J. Biomed. Mater. Xes., 17, 167-177 (1983). 3. J. 8. Mazur and J. L. Salazar, ”Late thrombosis of middle cerebral artery following clipping and coating of aneurysms,” Surg. Neurol., 10, 131133 (1978). 4. S. N. Chou, H. J. Ortiz-Suarer, and W. E. Brown, “Technique and material for coating aneurysms,” Clin. Neurosurg., 21, 182-193 (1974). 5. J. A. Maxwell, “Use of tissue adhesive in the surgical treatment of cerebrospinal fluid leaks. Experience with isobutyl 2-cyanoacrylate in 12 cases,” J. Neurosurg., 39, 322-336 (1973). 6. R. A. W. Lehman, G. J. Hayes, and A. N. Martins, ”The use of adhesive and lyophilized dura in the treatment of cerebrospinal rhinorrhea. Technical note,” J. Neurosurg., 26, 92-95 (1967). 7. G.D. VanderArk, D.T. Pitkethly, T. B. Ducker, and L. G. Kempe, “Repair of cerebrospinal fluid fistulas using a tissue adhesive,” ]. NeuroStlrg., 33, 151-155 (1970). 8. G. Debrun, F. Vinuela, A. Fox, and C. G. Drake, ”Embolization of cerebral arteriovenous malformations with bucrylate. Experience in 46 cases,” 1. Neurosurg., 56, 615-627 (1982). 9. B. M. Stein and S. M. Wolpert, “Arteriovenous malformations of the brain. Current concepts and treatment,” Arch. Neurol., 37, 1-5, 69-75 (1980). 10. I.R. Whittle, I.H. Johnston, M. Besser, T.S. Lamond, and M. deSilva, ”Experience with bucrylate (isobutyl-2-cyanoacrylate) embolization of cerebral arteriovenous malformations during surgery,” Surg. Neurol., 19, 442-449 (1983).
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
77
11. L. D. Cromwell and A. B. Harris, “Treatment of arteriovenous malformations. A combined neurosurgical and neuroradiological approach,” J. Neurosurg., 33, 151-155 (1970). 12. L. D. Cromwell and C. W. Kerber, ”Modification of cyanoacrylate for therapeutic embolization: Preliminary experience,” Am. J. Xoentgenol., 132, 799-601 (1979). 13. Y. -C. Tseng, S. -H. Hyon, and Y. Ikada, ”Physical modification of 2cyanoacrylate for applications as surgical adhesives,” presented at American Chemical Society Symposium Series, Los Angeles, CA, September 26-29, 1988. 14. Y. -C. Tseng, S. -H. Hyon, Y. Ikada, W. Taki, A . Uno, and Y. Yonekawa, “Radiopaque materials for embolization,” J. Jp.SOC.Biomater., 5(4), 175-180 (1987). 15. T. Nakamura, S. Hitomi, ‘r. Shimamoto, ’3.-H. Hyon, Y. Ikada, S. Watanabe, and Y.Shimizu, “Surgical application of biodegradable films prepared from lactide-6-caprolactone copolymers,” presented at Proceedings of the Sixth European Conference on Biomaterials, Bologna, Italy, September 14-17, 1987. 16. K. Hayashi and T. Nakamura, ”Material test system for the evaluation of mechanical properties of biomaterials,” J. Biomed. Muter. Res., 19, 133-144 (1985). 17. K. Hayashi, H. Handa, S. Nagasawa, A. Okumura, and K. Moritake, “Stiffness and elastic behavior of human intracranial and extracranial arteries,” J. Biomech., 13, 175-184 (1980). 18. J. A. Collins, K. C. Pani, R. A. Lehman, and F. Leonard, ”Biological substrates and cure rates of cyanoacrylate tissue adhesives,” Arch. Surg., 93, 428-432 (1966). Received September 13, 1988 Accepted July 13, 1989
Aiming at alleviating the problems of using 2-cyanoacrylates as a material for therapeutic embolization, this experiment made some physical modifications by mixing contrast media. It was found that the physicochemical properties of 2-cyanoacrylates can be altered by changing the concentration and the composition of the contrast media added. A 50 wt% cyanoacrylate-50 wt% contrast medium mixture has enough radiopacity for the practical requirement for embolization. A mixture of 50 wt% (ethoxyethyl cyanoacrylate-5 wt% lac-
tide/&-caprolactone copolymer), 25 wt% lipiodol and 25 wt% tetrafluorodibromoethane provides a viscosity of 13.8 cP, a bonding strength of 14.9 kg/cm2, a set time of 6 s, and a spreading in canine blood of 33 mm. It was concluded that the mixture is much more satisfactory than the conventional cyanoacrylates as an embolus material in vitro. The results obtained by in vivo experiments and clinical trials so far suggest that the mixture is very promising as a material for embolization.
INTRODUCTION
2-Cyanoacrylates (CA), which are liquid monomers, polymerize very rapidly on contact with an ionic solution or endothelium. Because of this ability, they have been widely used in many surgical fields as hemostatic agents and tissue adhesives.l*’They also have been reported to reinforce cerebral aneurysm^^,^ and to repair dural cerebrospinal fluid leak^.^-^ Recently, they have attracted considerable attention in therapeutic embolization of cerebra1 arteriovenous malformation (AVM).B-llAlthough ethyl cyanoacrylate (ECA) and isobutyl cyanoacrylate (IBCA) are commonly used as material for therapeutic embolization, some of their physicochemical properties should be improved to obtain higher reliability. First of all, the CA monomers should be mixed with a contrast medium in order to be monitored through angiography, because they are not radiopaque. In addition, they should have appropriate viscosity and set time in blood. A low bonding strength is Address correspondence to Dr. Ikada at Research Center for Medical Polymers and Biomaterials, Kyoto University, 53, Kawahara-cho, Shogoin, Sakyo-ku, Kyoto, Japan, 606. Journal of Biomedical Materials Research, Vol. 24, 65-77 (1990) CCC 0021-9304/90/01OO65- 13$O4.00 0 1990 John Wiley & Sons, Inc.
TSENG ET AL.
66
desirable for removing a microcatheter from the blood vessels easily. Besides, low stiffness of the embolized part is also required as the part needs to be removed with surgical resection. Cromwell et al. modified IBCA by mixing iophendylate to make it more practical and reported that its radiopacity and polymerization time were satisfactory." However, it has been found in our recent report that ECA and IBCA polymers have very high stiffness.13Therefore, we modified the homologous series of CA to develop soft adhesives and found that the homologs with longer ester side groups, especially ethoxyethyl cyanoacrylate (EECA), generate more flexible polymers than the homologs with shorter Furthermore, a lactide (LA)/&-caprolactone(CL) copolymer [P(LAco-CL)] was found to plasticize the resulting CA polymer^.'^ Very recently, we have also reported that tetrafluorodibromoethane (FEB,) and hexafluorodibromopropane are good as contrast medium. l4 In the present work, we will attempt to find a soft embolus material which has the desired radiopacity, viscosity, set time, and spreading in blood as well as a low bonding strength.
MATERIALS AND METHODS
Materials EECA was synthesized with the conventional method from ethoxyethyl cyanoacetate and paraformaldehyde (Y. -C. Tseng, unpublished data). The P(LA-co-CL) was prepared by ring-opening copolymerization of D,L-lactide and &-caprolactonein bulk.15 The copolymer used in this work has a molar chemical composition of 40/60 as D,L-lactide/&-caprolactone ratio and a weight-average molecular weight of 1 X lo5. The mixtures [EECA-P(LA-coCL)] of EECA and P(LA-co-CL) were prepared by dissolving the copolymer in EECA completely at 37°C. Various ratios of FEB, (Daikin Industries LTD., Japan) and lipiodol ultra-fluide (lipiodol) (Kodama Co., Japan) were added into the mixtures at room temperature. The mixtures were stored at 4°C until tests were performed.
Stiffness The stiffness of the polymerized EECA-P(LA-co-CL)mixtures was measured using the apparatus designed and proposed by Hayashi et al.'6f'7for the measurement of pressure-diameter re1 tions of blood vessels. The ex7 by a television system conternal diameter of the specimen was measured sisting of a Vidicon camera and a width analyzer (Hamamatsu Photonics, Hamamatsu, C1000-16 and HTV-C1170). The intraluminal pressure was measured with a pressure transducer (Toyo Baldwin, Tokyo, MPU-0.5-290111), which was attached to the vessel segment via a short catheter. Briefly,
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
67
carotid artery of dogs was resected, stored in Krebs-Ringer solution at 4"C, and transferred to testing within 12 h. In order to remove the initial stressrelaxation effect, the resected test carotid artery, which was stored at 37°C in Krebs-Ringer solution under aeration with a 95% 0, and 5% CO, gas, was kept at 100 mm Hg with Krebs-Ringer solution for 15 min. The intraluminal pressure was then gradually elevated to 200 mm Hg and lowered to 0. This specific inflation and deflation procedure was repeated until a stable pressure-diameter hysteresis loop was formed. The carotid artery was then coated to about 0.3-mm thickness with EECA-P(LA-co-CL)mixtures and subjected to the distension to measure the pressure-diameter relation of the coated carotid. The relation can be expressed by the following equation:
P(D/D, - 1) where D , is the carotid wall diameter when the standard pressure P, (100 mm Hg) is applied and D is the external diameter when the intraluminal pressure is P between the physiological pressure range such as 60160 mm Hg.I7p is a parameter identifying the stiffness of the tube wall. The average p values were obtained by reading on six specimens. In P/P,
=
Viscosity and bonding strength The viscosity of the mixtures of various concentrations of EECA and EECA-5%P(LA-co-CL) with FEB, and lipiodol was measured with a rotational viscometer (Tokyo Keiki Co., LTD., Japan) at 25°C. The bonding shear strengths of bonded stainless strips were measured as references on the change of the bonding strength after mixing contrast media to CA for applying the materials in vivo later on. The measurement was performed in accordance with JISK 6850-76, which is a Japanese standard, using an Instron tensile machine (Shimadzu Seisakusho LTD., Kyoto, Japan) at a separation rate of 10 mm/min. Briefly, the surface (25 X 12.5 mm') of stainless steel test strips (16 x 25 x 100 mm3)was polished with a 240 number sandpaper, rinsed rigorously with acetone, dried at 50°C for 4 h, and kept at 60% RH and 25°C overnight. A 10-pL sample of each mixture was applied to one test strip, another one was lapped on it for adhesion for 24 h. The shear strengths were then determined by averaging at least five readings. Set time and spreading The set time of the mixtures was determined by dropping 10 p L of the mixtures from a height of 9 mm on fresh canine blood, which was placed in 41 mm Petri dish to 7 mm depth, followed by measuring the period of time just from the contacting moment of the mixtures on the blood surface to that of the complete opacification." Fifteen measurements were made to obtain the average value. The spreading of the mixtures was estimated by measuring the diameter of the film formed on the blood surface when 10 pL
TSENG ET AL.
68
of the mixtures was dropped on it as described above. All of these measurements were done at room temperature unless otherwise noted. RESULTS
The solubility limit of lipiodol is 60 wt% both for EECA and EECA5%P(LA-co-CL)at room temperature. On the other hand, FEB, is soluble unlimitedly in these liquids to form stable and completely miscible solution. The EECA polymer is more flexible than the methyl cyanoacrylate, the ECA and the IBCA polymers. To quantitate the polymer flexibility, we determined the @ values of canine carotid arteries coated with the EECA-P(LAco-CL) mixtures. The monomer liquid almost instantly sets to a gel upon coating on the wet artery. The observed p values are given in Table I. The stiffness parameter is seen to decrease as the concentration of P(LA-co-CL) in the mixture increases. This is the same tendency as that reported in our previous work on canine femoral arteries coated with the same rnixt~res.'~ The viscosity of the EECA-lipiodol mixtures increased with an increase of the added amount of lipiodol, whereas the EECA-5%P(LA-co-CL)-lipiodol mixtures had a minimum in viscosity at the lipiodol concentration of approximately 50 wt%. In contrast, the viscosity of EECA as well as EECA5%P(LA-co-CL)became lower upon addition of FEB,. The result is tabulated in Table 11. Table I11 gives the shear strengths of the stainless strips TABLE I of Canine Arteries Coated with Stiffness Parameter Values EECA-P(LA-co-CL) Mixtures
P
Coating ~
None EECA EECA-~%P(LA-CO-CL) EECA-~~%P(LA-CO-CL)
~~
8.35 2 0.71 34.87 f 3.23 23.16 f 2.50 14.22 f 2.58
TABLE I1 Viscosities of EECA and EECA-5%P(LA-co-CL) Mixed with Lipiodol and FEB2. Ratio of Cyanoacrylate to Contrast Medium W/WI
100/0 70/30 60/40 50/50 40/60 0/100
Viscosity (cP) EECA
EECA-5%P (LA-CO-CL)
Lipiodol
FEB,
Lipiodol
FEB,
4.5 8.3 9.5
4.5 4.1 3.8 3.4 3.2 1.7
41.3 32.8 29.3 25.0 28.5 34.9
41.3 15.0 9.8 6.5 4.5 1.7
12.0
14.8 34.9
69
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
TABLE I11 Bonding Strengths between Two Steel Strips Bonded with EECA and EECA-5%P(LA-co-CL)Containing Lipiodol and FEB2 Shear Strength (Kg/cm2)
Ratio of Cyanoacrylate to Contrast Medium
EECA
EECAS%P(LA-co-CL)
W/W)
Lipiodol
FEBz
Lipiodol
FEBz
100/0 70130 60140
130.7 2 2.5 40.5 f 1.5 26.7 f 1.8 21.3 f 2.1 10.6 f 0.5
130.7 f 2.5 107.8 f 3.0 91.9 f 8.6 83.4 k 5.8 80.3 f 4.8
101.5 f 2.7 39.3 f 1.3 27.4 f 2.1 18.4 f 1.6 11.4 f 1.2
101.5 f 2.7 109.6 f 2.5 111.0 k 2.4 106.9 k 4.3 100.0 5 3.6
so/m 40160
bonded with the EECA-lipiodol and EECA-5%P(LA-co-CL)-lipiodol mixtures. As can be seen, the bonding strengths decrease drastically with the increasing amount of lipiodol. On the other hand, no distinguishable change is observed for the EECA-5%P(LA-co-CL)-FEB2 mixtures, while the EECA-FEB, mixtures show a slight decrease in bonding strength on addition of FEB,. The direction of the set time change induced by addition of the contrast medium was different, depending on the nature of the contrast medium used, irrespective of the kinds of CA. It is seen in Table IV that the set time increases with the increasing amount of lipiodol but decreases with an increase of the FEB, amount. As shown in Figure 1, the radiopacity of a 50% (EECA-5%P(LA-co-CL))50% FEB, mixture is good enough. The viscosity and bonding strength of the mixtures from 50 wt% of EECA-5%P(LA-co-CL)and 50 wt% contrast medium containing different ratios of lipiodol to FEB, were determined and plotted against the composition of the contrast media in Figure 2. The viscosity of the mixtures decreases with the increasing amount of FEB,, while the bonding strength decreases with the increasing amount of lipiodol. In the same way, the measured spreading and set times were plotted in Figure 3, where it is clearly seen that as the ratio of FEB, increases, the spreading of the mixture increases while the set time decreases. TABLE IV Set Times of EECA and EECA-S%P(LA-co-CL)Mixed with Lipiodol and FEBz Set Time
Ratio of Cyanoacrylate to Contrast Medium
EECA
EECA-5%~'(LA-co-CL)
(WIW
Lipiodol
FEBz
Lipiodol
FEB2
lOO/O 70130 60140 50150 40160
6.0 f 0.4 8.2 f 0.3 10.0 % 0.4 12.0 2 0.7 18.7 f 0.9
6.0 f 0.4 3.5 f 0.4 2.4 % 0.1 1.8 ? 0.2 1.5 f 0.0
21.8 f 0.7 25.4 f 1.9 27.2 f 1.9 31.8 +- 1.8 36.4 f 2.2
21.8 2 0.7 12.8 f 0.7 5.8 f 0.2 3.6 f 0.2 2.4 ? 0.1
70
TSENG ET AL.
Figure 1. Plain x-ray film after injection of a 50% (ethoxyethyl cyanoacrylate-5% lactide/&-caprolactone copolymer)-50% tetrafluorodibromoethane mixture into the canine carotid artery. The mixture occluded the carotid artery and its branches. The mixture was clearly radiopaque (arrow).
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
n,,
N
*Ot
50 40 0
10
10 40
20 30
30 20
FEB2 and lipiodol conc.
O
FEE2 (ipiodol
50
(wf%)
Figure 2. Bonding strengths a n d viscosities of liquids from-50% (ethoxyethyl cyanoacrylate-5% lactide/&-caprolactone copolymer) with different mixing ratios of tetrafluorodibromoethane (FEBJ to lipiodol.
-
7-7
136
J
32
28
48
1
o
24
l 50 4 0 0
10
30 20
d 20 30
o 10 0 FEE2 40 50 lipiodol
FEE2 and lipiodol conc.
(wt%)
Figure 3. Spreading and set times of liquids from 50% (ethoxyethyl cyanoacrylate-5% lactide/&-caprolactone copolymer) with different mixing ratios of tetrafluorodibromoethane (FEB?)to lipiodol.
71
72
TSENG ET AL.
(a) Figure 4. Intraoperative embolization of the cerebral arteriovenous malformation of the left basal ganglia. (A) Right internal carotid arteriogram showed the cerebral arteriovenous malformation (small arrow), which was fed by the artery of Heubner (large arrow). (B) The catheter and the needle were placed into the artery of Heubner (arrow) by craniotomy. The contrast material was injected and the intraoperative angiogram was performed, revealing the arteriovenous malformation on the left basal ganglia. (C) Via this catheter, a mixture of 50% (ethoxyethyl cyanoacrylate-5% lactide/&-caprolactone copolymer), 25% lipiodol and 25% tetrafluorodibromoethane was injected. Malformation was obliterated.
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
73
0-9 Figure 4.
(continued)
DISCUSSION
The effectiveness of CA monomers as material for therapeutic embolization is dependent upon their properties, which include radiopacity, viscosity, set time, bonding strength, stiffness, etc. If the viscosity is very high,
TSENG ET AL.
74
(4 Figure 4. (continued)
the embolization liquid hardly passes through microcatheters. On the other hand, an extremely low viscosity of the liquid is liable to cause the liquid
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
75
material to spread over passing by the place to be embolized. The optimal viscosity seems to be around 10 cP according to our experience. If it takes a long time to polymerize, namely, to set to gel, the m a t e d would have flown over the site of the malformation. If the set time is too short, surgeons might not have enough time to be able to inject the material to the proper site or it would occlude the microcatheter before penetrating into the central part of the malformation. Polymerization time of 5 to 10 s is most desirable. A high bonding strength easily causes the microcatheter to adhere to blood vessels, resulting in difficulty removing the microcatheter from the blood vessels. Embolization is usually performed either as a primary therapy or in conjunction with surgical resection of defective sites, and may be carried out via the femoral approach or at craniotomy, respectively. In the case of simultaneous use of embolization and surgical resection, low stiffness of the embolized part is required for the easy removal. ECA and IBCA have been reported to be useful for embolization of arteriovenous malformation (AVM), but it is also pointed out that the stiffness of ECA and IBCA polymers are too high to resect the embolized AVM. Recently we have found that the stiffness of the EECA polymers is much lower than that of the ECA and the IBCA p01ymers.l~Furthermore, the present work reveals that the stiffness of the EECA polymer can be reduced by mixing with a large amount of P(LA-co-CL)(Table I). Taking the viscosity into consideration, we selected EECA and an EECA-5%P(LA-co-CL) mixture as the candidate material to be mixed with lipiodol and FEB,. As is apparent from Figure 1, radiopacity of the 50 wt% monomer-50 wt% contrast medium mixture was satisfactory. Even though the 50 wt% (EECA-5%P(LA-co-CL))50 wt% lipiodol has the minimum viscosity among the various mixtures, it is still too high. On the contrary, the EECA-FEB, mixtures are too low in viscosity to be applied. As is shown in Table 11, the 40-50 wt% of lipiodol in EECAlipiodol and the 40 wt% of the FEB, in the EECA-5%P(LA-co-CL)-FEB2 seem to be the most appropriate concentration. They may have a similar tendency also in set time, because it is closely related to its viscosity. An addition of lipiodol to the EECAd%P(LA-co-CL)produced too long set time and an addition of FEB, to the EECA resulted in too short set time for polymerization of the embolus liquids in AVM. The 40 wt% of FEB, in EECA-5%P(LA-coCL)-FEB, may be the best concentration with regard to the set time. In addition, there is another property of the embolus mixture to be adjusted. That is the bonding strength. Some mixtures have too high strength to remove the microcatheter from the blood vessels. Lipiodol added to the EECA and the EECA-5%P(LA-co-CL)is able to reduce the bonding strength. Interestingly, lipiodol and FEB, in the mixture change the viscosity, set time, and bonding strength in the opposite fashion. Therefore, we tried to take advantage of using both lipiodol and FEB, to find an ideal embolus liquid. The mixtures selected contain 50 wt% of the EECA-5%P(LA-co-CL)and 50 wt% of the contrast medium with various compositions of lipiodol and FEB,. According to the experimental results shown in Figures 2 and 3, the
TSENG ET AL.
76
mixture containing 50 wt% of the EECA-5%P(LA-co-CL),25 wt% of lipiodol and 25 wt% of FEB, satisfies all the requirements for the viscosity, the bonding strength, and the set time as well as the spreading in blood. Lipiodol and FEB, are freely miscible with EECA-5%P(LA-co-CL)and can provide a wide range of viscosity, set time, and bonding strength. In order to find the optimal condition of the embolus materials in vivo, we injected the mixtures containing 50 wt% of the EECA-5%P(LA-co-CL)and 50 wt% of the contrast medium with various compositions of lipiodol and FEB, into canine carotid arteries. It was found that the mixture containing 50 wt% of EECA-5%P(LA-co-CL)and 25 wt% of lipiodol and 25 wt% of FEB, could satisfy all the required conditions and gave successful resutts. ClinicaI trials have just started. Figure 4 shows one of the examples of intraoperative treatments of AVM of a human. It is concluded that taking advantage of using the contrast media of lipiodol and FEB, with various concentration and compositions can alter the physicochemical properties of CA. The plasticized EECA with P(LA-to-CL) can satisfy the requirement as an embolus material for AVM by adding appropriate amount of the contrast media, and the material seems to be very promising for clinical application. References 1. J. A. O’Leary, “Tissue adhesives and pelvic hemostasis: An evaluation of isobutyl 2-cyanoacrylate,” J. Surg. Oncol., 3, 117-120 (1971). 2. M. C. Harper and M. Ralston, “Isobutyl2-cyanoacrylate as osseous adhesive in the repair of osteochondral fractures,“ J. Biomed. Mater. Xes., 17, 167-177 (1983). 3. J. 8. Mazur and J. L. Salazar, ”Late thrombosis of middle cerebral artery following clipping and coating of aneurysms,” Surg. Neurol., 10, 131133 (1978). 4. S. N. Chou, H. J. Ortiz-Suarer, and W. E. Brown, “Technique and material for coating aneurysms,” Clin. Neurosurg., 21, 182-193 (1974). 5. J. A. Maxwell, “Use of tissue adhesive in the surgical treatment of cerebrospinal fluid leaks. Experience with isobutyl 2-cyanoacrylate in 12 cases,” J. Neurosurg., 39, 322-336 (1973). 6. R. A. W. Lehman, G. J. Hayes, and A. N. Martins, ”The use of adhesive and lyophilized dura in the treatment of cerebrospinal rhinorrhea. Technical note,” J. Neurosurg., 26, 92-95 (1967). 7. G.D. VanderArk, D.T. Pitkethly, T. B. Ducker, and L. G. Kempe, “Repair of cerebrospinal fluid fistulas using a tissue adhesive,” ]. NeuroStlrg., 33, 151-155 (1970). 8. G. Debrun, F. Vinuela, A. Fox, and C. G. Drake, ”Embolization of cerebral arteriovenous malformations with bucrylate. Experience in 46 cases,” 1. Neurosurg., 56, 615-627 (1982). 9. B. M. Stein and S. M. Wolpert, “Arteriovenous malformations of the brain. Current concepts and treatment,” Arch. Neurol., 37, 1-5, 69-75 (1980). 10. I.R. Whittle, I.H. Johnston, M. Besser, T.S. Lamond, and M. deSilva, ”Experience with bucrylate (isobutyl-2-cyanoacrylate) embolization of cerebral arteriovenous malformations during surgery,” Surg. Neurol., 19, 442-449 (1983).
THERAPEUTIC EMBOLIZATION OF A-V MALFORMATION
77
11. L. D. Cromwell and A. B. Harris, “Treatment of arteriovenous malformations. A combined neurosurgical and neuroradiological approach,” J. Neurosurg., 33, 151-155 (1970). 12. L. D. Cromwell and C. W. Kerber, ”Modification of cyanoacrylate for therapeutic embolization: Preliminary experience,” Am. J. Xoentgenol., 132, 799-601 (1979). 13. Y. -C. Tseng, S. -H. Hyon, and Y. Ikada, ”Physical modification of 2cyanoacrylate for applications as surgical adhesives,” presented at American Chemical Society Symposium Series, Los Angeles, CA, September 26-29, 1988. 14. Y. -C. Tseng, S. -H. Hyon, Y. Ikada, W. Taki, A . Uno, and Y. Yonekawa, “Radiopaque materials for embolization,” J. Jp.SOC.Biomater., 5(4), 175-180 (1987). 15. T. Nakamura, S. Hitomi, ‘r. Shimamoto, ’3.-H. Hyon, Y. Ikada, S. Watanabe, and Y.Shimizu, “Surgical application of biodegradable films prepared from lactide-6-caprolactone copolymers,” presented at Proceedings of the Sixth European Conference on Biomaterials, Bologna, Italy, September 14-17, 1987. 16. K. Hayashi and T. Nakamura, ”Material test system for the evaluation of mechanical properties of biomaterials,” J. Biomed. Muter. Res., 19, 133-144 (1985). 17. K. Hayashi, H. Handa, S. Nagasawa, A. Okumura, and K. Moritake, “Stiffness and elastic behavior of human intracranial and extracranial arteries,” J. Biomech., 13, 175-184 (1980). 18. J. A. Collins, K. C. Pani, R. A. Lehman, and F. Leonard, ”Biological substrates and cure rates of cyanoacrylate tissue adhesives,” Arch. Surg., 93, 428-432 (1966). Received September 13, 1988 Accepted July 13, 1989
