0148-396X/79/0403-0210$02.00/0 NEUROSURGERY Copyright © 1979 by the Congress of Neurological Surgeons Vol. 4, No. 3, 1979 Printed in U.S.A. Cyanoacrylate Occlusion of Carotid-Cavernous Fistula with Preservation of Carotid Artery Flow Charles W. Kerber, M.D., William O. Bank, M.D., and Laurence D. Cromwell, M.D. Section of Neuroradiology, Department of Radiology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania (C. W. K.); University of California at San Francisco, San Francisco, California (W. O. B.); and University of Washington School of Medicine, Seatile, Washington (L. D.C.) We report a new treatment for carotid-cavernous fistula. Using a flow-guided, balloon-tipped microcatheter, we catheterize the fistula itself, verify balloon entry into the fistula with fluoroscopy and x-ray films, and then infuse the tissue adhesive isobutyl-2-cyanoacrylate with careful fluoroscopic control. Three patients have had their fistulas occluded, with preser- vation of flow through the internal carotid artery. This balloon microcatheter allows the radiologist to perform a reversible test occlusion. All three patients had neurological changes during or after the procedure, and in one we inadvertently occluded several distal middle cerebral artery branches without permanent neurological deficit. No patient became blind or developed 3rd, 4th, or 6th nerve palsy from the treatment. This technique seems to have promise as another method for the obliteration of carotid-cavernous fistula. Key words: Balloon microcatheter, Carotid-cavernous fistula, Catheter therapy, Cyanoacrylate, Therapeutic embolization INTRODUCTION Carotid-cavernous fistula (CCF) is a difficult problem to treat. Carotid ligation below and above the fistula (the trapping procedure) is not always successful. It is frequently difficult to predict whether the patient will tolerate the loss of a carotid artery, and morbidity has included both cerebral and ocular ischemia. Balloon occlusion of the internal carotid artery (1, 14, 15, 17), obstruction of the cavernous sinus with detachable balloons (3, 16), controlled particulate embolization (2, 5, 6, 11), direct surgical repair (12, 13), and electrothrombosis (4) have all been tried in the treatment of CCF. We have developed another technique based upon the pre- cise endarterial deposition of the tissue adhesive isobutyl-2- cyanoacrylate (IBCA) (Bucrylate; Ethicon, Inc., Somerville, New Jersey), delivered through a balloon-tipped, flexible mi- crocatheter (7-10). Blood flow directs the microcatheter into the fistula, which permits not only precise radiographic locali- zation but also the performance of a reversible test occlusion before permanent deposition of polymer. This paper reports our experience with this new technique in three patients. TECHNIQUE We have found this system of occlusive therapy to be helpful in the management of intracranial and spinal arteriovenous malformations and have now expanded its use to include the treatment of carotid-cavernous fistulas. We begin by premedicating our patients with 24 mg of oral dexamethasone in divided doses the day before treatment and continue this regimen for 3 to 6 additional days. Relaxation is achieved and pain is controlled primarily by hypnotic tech- niques supplemented with oral Valium (diazepam: Roche Lab- oratories, Nutley, New Jersey) (18). We prefer our patients to be as awake and cooperative as possible. With the use of a standard Seldinger percutaneous approach, a catheter is placed in each femoral artery (or another accessible vessel) ai.d both are directed with fluoroscopic control into the appropriate carotid artery. One catheter is used to deliver the contrast agent for precise localization of the microcatheter and the other delivers the balloon microcatheter into the intracra- nial circulation. The patient is not heparinized systemically, but the catheters are perfused with heparnized saline during the procedure (1.5 ml of 1:1000 heparin is added to 500 ml of 210 isotonic saline and pressure infusion is maintained at 20 to 40 drops per minute). The microcatheter (calibrated leak balloon microcatheter: Cook, Inc., Bloomington, Indiana) is squirted through the outer coaxial catheter and is finally manipulated by simultaneously advancing the outer catheter while varying the degree of bal- loon microcatheter inflation. Blood flow is allowed to guide and aid the final placement. Dynamic fluoroscopy and instant electronic subtraction are used frequently during the procedure. After the position of the balloon is verified with either x-ray films or fluoroscopy, the contrast agent is washed out of the balloon with 5% dextrose and water (the cyanoacrylate poly- merizes upon contact with ionic material). Immediately before delivery the IBCA is mixed with tantalum powder (Kennametal Inc., Latrobe, Pennsylvania) so that it is radiopaque, and the mixture is then injected through the microcatheter. The balloon is seen to fill on the fluoroscope, the glue passes distally, and when the radiopaque adhesive has reached the desired position the balloon is deflated, allowing incoming blood to polymerize the glue. The microcatheter is withdrawn immediately. A follow-up angiogram is then performed. If this initial treatment is inadequate, additional balloon microcatheters are repositioned and more adhesive is injected. CASE REPORTS Case | A 64-year-old woman had noted the development of a typical left tic douloureux 20 years before admission in October 1976. Ten years before, she had had an unsuccessful alcohol block of the left infraorbital nerve. Five months before, a left retromastoid craniectomy had been performed but had been discontinued because of air embolization. Five days later she had been re-explored without incident and a microvascular decompression of the left Sth nerve had been performed. The patient’s pain recurred and she underwent a percuta- neous radiofrequency left trigeminal rhizolysis on August 12, 1976. The following morning she noted diplopia, pulsatile tinnitus, and headache. She had a left orbital bruit, a left 6th nerve paresis, and hypesthesia and hypalgesia in the distribu- tion of the left Ist trigeminal division. The patient was readmitted 2 months later with a left 6th nerve paresis, hypesthesia and hypalgesia in the left V, distri- March 1979 bution, left trigeminal motor dysfunction, and bruits over both orbits and both carotid bifurcations (left greater than right) that could be stopped by left cervical carotid compression. Angiography showed a left CCF (Fig. 1A). The patient was readmitted on October 27, 1976, for thera- peutic embolization. A catheter was directed into the fistula and the posterior channels were occluded (compare Fig. 1B with Fig. 14). There was considerable improvement in her symptoms and she was discharged. Five days after the occlusive therapy, the patient developed severe frontal headaches, and at another hospital a right cere- bellar hemorrhage was diagnosed and evacuated. She did well after that operation, with slight cerebellar dysfunction persist- ing. and was discharged on methyldopa therapy for hyperten- sion. The patient noted progressive left proptosis and a decrease in visual acuity during the following weeks. A repeat attempt at therapeutic embolization was unsuccessful. We could not manipulate the microcatheter into the fistula. Five days later we tried again and were able to pass the balloon into the CAROTID-CAVERNOUS FISTULA 211 cavernous sinus (Fig. 1C). We filled the remainder of the cavernous sinus with 0.3 ml of tantalum-impregnated IBCA. Follow-up angiography showed occlusion of much of the cav- ernous sinus and preservation of flow through the internal carotid artery (Fig. 1D). Two and one-half hours after the therapeutic embolization the patient developed a nominal and receptive dysphasia, a slight right hemiparesis, and a mild right 6th nerve paresis. By the next day she had marked improve- ment of the right 6th nerve paresis, and her weakness and dysphasia had resolved. She was discharged and has apparently done well with no bruit and with improvement in her diplopia. She has occasional pain about the left eye, but no chemosis or proptosis. She refuses to return for follow-up angiography. Case 2 A 57-year-old man was struck by a car and sustained frac- tures of the hip and lower leg. After that accident he was unconscious for 4 days, but he awoke neurologically intact. During hospitalization he required a splenectomy and the repair of a liver laceration. He recovered satisfactorily and was Fic. lL. Case |. Lateral subtracted views, selective left internal carotid arteriograms. A, immediate shunting of blood into the enlarged cavernous sinus, superior ophthalmic vein, straight sinus, and jugular vein is visible. B, after the first treatment, the posterior channels that filled the straight sinus no longer are visible. As might be expected, the patient’s visual symptoms deteriorated because of the enlargement of the superior ophthalmic vein. C, the balloon microcatheter is within the cavernous s fistula with the fluoroscope is a more accurate method of Its position was verified by frontal views as well. Watching the balloon enter the ing balloon position. D, after deposition of the IBCA-tantalum mixture, only a portion of the cavernous sinus fills with contrast agent. There is preservation of the internal carotid artery and ophthalmic artery flow. 212 KERBER et al. discharged. One chart note described conjunctival redness, O.D. Seven weeks after the injury, the patient was seen by an ophthalmologist for ocular symptoms and was referred to Presbyterian-University Hospital. On admission 8 weeks after injury, a corrected visual acuity of 20/200 O.D., right chemosis. and slight right proptosis were found. A right 6th nerve paresis was also present and the right optic disc was blurred, with distention and arterialization of the veins. Angiography demonstrated a right CCF (Fig. 24). Attempts to direct a microcatheter into the fistula at that time were unsuccessful. The patient’s vision deteriorated during the following week and a second attempt to place the balloon microcatheter was successful on February 14, 1978 (Fig. 2B). While the balloon was being manipulated at the fistula orifice, the patient expe- rienced two episodes of transient left-sided weakness and dysarthria. After verification that the catheter lay within the fistula, approximately 0.4 ml of tantalum-impregnated isobu- tyl-2-cyanoacrylate was delivered into the fistula (Fig. 2C) and Neurosurgery, Vol. 4, No. 3 the catheter was removed. An angiogram performed 20 minutes after the treatment showed partial closure of the fistula, some filling of the cavernous sinus with contrast agent, and slow flow in two branches of the middle cerebral artery. By the next day the patient’s vision had improved from 20/ 200 to 20/60. The ocular tension had decreased from 25 to 10 mm Hg, and the conjunctival prolapse no longer protruded beyond the lids. The vessels of the iris, which had been thought to be neovascular, had receded markedly. There was no change in the 6th nerve paresis. The patient was discharged 2 days later. He returned on March 4, 1978, for follow-up angiography. Further improvement in visual acuity and complete resolution of the 6th nerve paresis were noted. No other abnormalities were present. Angiography showed normal flow in the middle cerebral artery, with the fistula closed (Fig. 2D). Comment. Transient neurological changes were experienced by the patient when the balloon occluded flow through the carotid artery. This observation suggests that he might not have tolerated the trapping procedure. Fic. 2. Case 2. A, lateral subtracted view, selective right internal carotid arteriogram. The contrast agent fills the entire cavernous sinus and the superior ophthalmic vein in addition to opacifying the carotid artery and its intracranial branches. B, lateral view of the skull. The balloon is within the cavernous sinus. C, lateral view of the skull. The tantalum-impregnated isobutyl-2 yanoacrylate has been deposited. and the balloon has been withdrawn. D, lateral follow-up subtracted arteriogram, selective right internal carotid injection. Preservation of the internal carotid artery is evident All of the intracranial branches fill March 1979 Case 3 A 60-year-old woman had developed postherpetic neuralgia involving all three divisions of the right trigeminal nerve 41 months before this admission. The herpes infection had been complicated by a Staphylococcus superinfection that left con- siderable scarring. She had developed excruciating facial pain that varied in intensity. The pain had remitted but had then recurred. Two months before this admission, the patient had had a percutaneous trigeminal block that had produced satisfactory analgesia and hypesthesia of the right V» distribution, with complete relief of her pain. Unfortunately, she also had devel- oped minimal weakness of the left hand and had noted a noise in the right ear. The weakness had then cleared completely. Angiography had demonstrated a CCF. Catheter occlusive therapy had been requested then, but because she had been relatively asymptomatic and had had no visual deterioration, we had elected not to treat her. Later both of her eyes had become edematous, her right eye had begun to protrude, and her visual acuity had decreased. The patient was admitted to the hospital in April 1978, and angiography showed enlargement of the veins draining the ‘CF. We placed a balloon microcatheter into the fistula, verified its position, and then infused approximately 0.4 ml of tantalum- impregnated IBCA. This mixture filled a proximal portion of the superior ophthalmic vein, the cavernous sinus, and a por- tion of the clivus venous plexus. The follow-up angiogram showed patency of the right internal carotid artery and good filling of all of its intracranial branches. Only a small quantity of the contrast agent entered the cavernous sinus. Although the patient remained asymptomatic throughout and after the procedure, a follow-up computerized tomographic scan revealed tantalum in several branches of the right middle cerebral artery. She was asymptomatic for about | week but then developed vague transient cerebral ischemic symptoms. These cleared spontaneously 3 weeks later, and she has re- mained asymptomatic since. On May 21, 1978, the patient was examined. Her right proptosis had resolved, and her visual acuity was 20/70 O.S. and 20/50 O.D. She has refused further angiography. DISCUSSION The treatment of carotid-cavernous fistula remains difficult and dangerous. The organ distal to the fistula is unforgiving, and great care must be taken not to occlude its blood supply. We try to avoid techniques that sacrifice the carotid artery and do everything possible to preserve its integrity. There are problems with all available treatment plans. Sur- gical attack is a formidable undertaking even with new micro- vascular techniques. With releasable balloon therapy, down- stream embolization is possible, and frequently one must sac- tifice the carotid artery. Downstream embolization may also be a complication of the use of adhesives. There are some advantages to the use of the microcatheter and liquid polymer, however. The microcatheter itself is made of soft silicone and has to date caused no visible irritation of any intracranial vessel that we have explored. Having a balloon on the end of a catheter allows one to perform a preliminary test occlusion to see how the patient will tolerate the permanent treatment. Using the fluoroscope dynamically tells us when deposition of the adhesive has occluded the fistula. Cyanoacrylate polymerizes only upon contact with ionic material such as blood, blood vessel wall, or contrast agent, and its polymerization is extremely rapid, generally within | to CAROTID-CAVERNOUS FISTULA 213 2 seconds. With some experience it is possible to control this polymerization. As the glue solidifies, it fills the small recesses of the cavernous sinus. The main disadvantage of cyanoacrylate is that it is an experimental drug. Approval from the Food and Drug Admin- istration for its investigational use is required. It is difficult to learn its use, and considerable animal experimentation is nec- essary. Its rapid polymerization is, at the same time. an advan- tage and a disadvantage. This precise event—the change from liquid to solid must be visualized carefully with the fluoro- scope. Infusion of more IBCA after polymerization may trap the catheter and cause poiymer to zeflux back along the vessel and into normal areas, as happened with our third patient. Finally, we would like to emphasize that this material may be dangerous to the operator, too. IBCA is a powerful adhesive, and we know of one worker who accidentally glued his eye shut. We strongly suggest that when working with this adhesive both gloves and eye protection be worn. CONCLUSION Precise deposition of the polymer isobutyl-2-cyanoacrylate has allowed us to occlude three carotid-cavernous fistulas while preserving carotid blood flow. Use of the calibrated leak bal- loon microcatheter permits catheterization of the fistula and gives us the opportunity to make a reversible test occlusion before polymer deposition. Using dynamic fluoroscopy and instant electronic subtraction, one may verify exact placement of the balloon during that test. We have produced mild but disturbing neurological changes in our patients, once during manipulation of the catheter and twice after glue deposition. Fortunately, these have been tem- porary, and all patients have improved markedly. No perma- nent cranial nerve deficits were produced despite the presence of this powerful vascular occluder around the intracavernous cranial nerves. There is as yet no long term follow-up, but the short term results are encouraging. Treatment with this technique is not easy. The rapidity with which the cyanoacrylate polymerizes and its permanence make its introduction a moment of truth. Much animal experimen- tation is necessary to achieve the operator competence required to make this undertaking safe and effective for the patient. ACKNOWLEDGMENTS This experimental study was approved by the Food and Drug Administration and by the University of Pittsburgh School of Medicine Committee on the Use of Humans for Experimental Purposes. Received for publication, July 27, 1978; accepted, November 25, 1978. Reprint requests: Charles W. Kerber, M.D., Department of Radiol- ogy. University of Pittsburgh School of Medicine, Presbyterian-Uni- versity Hospital, DeSoto at O'Hara Streets, Pittsburgh, Pennsylvania 15261. REFERENCES 1. Bahuleyan, K., Nelson, L. R., and Peck, F. C., Jr. Occlusion of carotid cavernous fistula with a balloon catheter. Surg. Neurol., 3: 283-287, 1975. . Black. P.. Uematsu, S., Perovic, M., and Walker, A. E. Carotid cavernous fistula: A controlled embolus technique for occlusion of fistula with preservation of carotid blood flow. J. Neurosurg., 38: 113-118, 1973. 3. Debrun, G., Lacour, P.. Caron, J. P.. Hurth, M.. Comoy, J.. Keravel, Y.. and Loisance, D. Traitment de fistules artériovei- nN 214 KERBER et al. neuses et d’anévrysmes par ballon gonflable et largable. Nouv. Presse Med.. 4: 2315-2318, 1975 4. Hosobuchi, Y. Electrothrombosis of carotid cavernous fistula. J Neurosurg., 42: 76-85, 1975. 5. Ishimori, S.. Hattori, M., Shibata, Y., Shizawa, H., and Fujinaga, R. Treatment of carotid cavernous fistula by Gelfoam emboliza- tion. J. Neurosurg., 27: 315-319, 1967. 6. Jaeger, R. Intracranial aneurysms. South. Surgeon, 15: 205-217, 1949. 7. Kerber, C. Intracranial cyanoacrylate: A new catheter therapy for arteriovenous malformations. Invest. Radiol., 10: 530-538, 1975 8. Kerber, C. Balloon catheter with a calibrated leak. Radiology, 120: 547-550, 1976. 9. Kerber, C. W. A system of catheter occlusive therapy. Invest. Radiol., 11: 370, 1976. 10. Kerber, C. W., Bank, W. O., and Cromwell, L. D. The calibrated leak balloon microcatheter: A device for arterial exploration and catheter occlusive therapy. In press. 11. Ohta, T., Nishimura, S., Kikuchi, H., and Toyama, M. Closure of carotid cavernous fistula with polyurethane foam embolus. J. Neurosurg., 38: 107-112, 1973. 12. Parkinson. D. Carotid cavernous fistula: Direct repair with pres- ervation of the carotid artery. Technical note. J. Neurosurg., 38: 99-106, 1973. 13. Parkinson, D.. Downs, A. R., Whytehead, L. L., and Syslak, W. B. Carotid cavernous fistula: Direct repair with preservation of ca- rotid. Surgery, 76: 882-889, 1974. 14. Picard, L.. Lepoire, J., Montaut, J., Hepner, H.. Roland, J.. Gu- yonnaud, J. C., Jacob, F., and Andre, J. M. Endarterial occlusion of carotid cavernous sinus fistulas using a balloon tipped catheter. Neuroradiology, 8: 5-10, 1974. 15. Prolo, D. J. and Hanbery, J. W. Intraluminal occlusion of a carotid cavernous sinus fistula with a balloon catheter: Technical note. J. Neurosurg. 37-242, 1971. 16. Serbinenko, F. A. Balloon catheterization and occlusion of major cerebral vessels. J. Neurosurg., 41: 125-145, 1974. 17. Von Rupp, N., Piger, A., and Ultsch, B. Karotis-Kavernosus-Fistel: Behandlungen mit einem Ballon katheter. Fortschr. Geb. Roent- genstr. Nuklearmed., 122: 215-217, 1975. 18. Zanetti, P. H. and Sherman, F. E. Experimental evaluation of a tissue adhesive as an agent for the treatment of aneurysms and arteriovenous anomalies. J. Neurosurg., 36: 72-79, 1972. COMMENTS The authors are to be congratulated on this imaginative technique. The IBCA method, made possible by an ingenious delivery system for the calibrated leak balloon catheter, may be superior to the parallel method of Debrun, who uses single or multiple detachable balloons from carotid artery catheteri- zation. However, Debrun et al. have had outstanding success with 17 carotid-cavernous fistulas, essentially completely oc- cluding these with the detachable balloon technique (1). The possibility of performing a functional test by inflating the microcatheter balloon is a valuable asset of the Kerber tech- nique, especially when contrasted with surgical trapping. In the unanesthetized patient the ischemia can be reversed before permanent neurological deficit ensues. As the authors indicate, the precision of the team is crucial. Dr. Kerber, with 4 years of experience in polymer work in animals and man, has reported here complications that cleared in two of three patients. This should be an indication of the potential hazards to other workers who have not had experience with the technique. However, in summary, the group has demonstrated a highly creative solution to a most difficult clinical problem. E. Ralph Heinz, M.D. Durham, North Carolina 1. Debrun, G., Lacour, P., Caron, J. P., Hurth, M., Comoy, J., and Keravel, Y. Detachable balloon and calibrated-leak balloon tech- Neurosurgery, Vol. 4, No. 3 niques in the treatment of cerebral vascular lesions. J. Neurosurg., 49: 635-649, 1978. From the earliest efforts at arterial thrombosis to the more recent methods of venous occlusion, the treatment of a carotid- cavernous fistula has varied in complexity and attendant risks to the patient. Cavernous sinus thrombosis with preservation of blood flow through the carotid artery has been advanced as the optimal therapy. Such a method involves the sequential occlusion of the venous side of this arteriovenous shunt either by the operative implantation of a foreign body into the cavernous sinus or by the release of a balloon after maneuver- ing it into the sinus on a catheter. An extension of this latter technique (namely, the deposition of a thrombosing substance within the cavernous sinus through a catheter) is discussed in this paper. Success in treatment of a carotid-cavernous fistula must be measured by the ease of implementation of a method and the avoidance of systemic, neurological, and ocular sequelae. A technique designed to preserve carotid blood flow is desirable only if the short and long term consequences of this goal do not undermine the well-being of the patient. The intent of the authors is to preserve carotid artery flow. However, in only one of their three patients did they demon- strate flow through the carotid artery after the development of neurological abnormalities. Their contention, therefore, that the carotid artery remains open in the other two patients is unproven. Furthermore, the fistula was incompletely obliter- ated in these same two patients (Cases | and 3) with formation of a “pseudoaneurysm,” the consequences of which are unclear at this time. Neurological abnormalities in all three patients may have resulted from the isobutyl-2-cyanoacrylate (IBCA) or, in Cases 1 and 3, from delayed carotid occlusion. The second patient developed left-sided weakness and dysarthria during catheter manipulation and later showed slow flow in two branches of the middle cerebral artery; perhaps embolization from the catheter occurred during manipulation. The third patient had incomplete closure of the fistula, demonstrable IBCA with tantalum in branches of the middle cerebral artery on CT scan, and “ischemic” symptoms. The authors have admitted the difficulties of their technique. They have not properly considered its safety. The long term effects on tissues from therapeutic embolization with IBCA are distressing and include excessive local toxicity as well as a potential carcinogenic effect (3). In a previous study in dogs, Dr. Kerber demonstrated the propensity for pulmonary em- bolization and occlusion of the carotid artery (1). The local toxicity of IBCA as well as its potential carcinogenic effect have properly resulted in its limitation by the Food and Drug Administration to investigational use in humans only in life- threatening situations, in terminal illness, or in those conditions in which alternative therapy is contraindicated due to unac- ceptable risk or lack of success. Such criteria are not satisfied by this relatively benign condition, for which alternative meth- ods of treatment are generally successful and safe. Eighty per cent of patients will tolerate the occlusion of one carotid artery (2). The controlled positioning of a balloon at the site of a fistula with segmental occlusion of one internal carotid artery is considerably simpler and safer for a patient than the method herein described with its attendant need for multiple procedures, the persistence of a portion of the fistula in some, the neurological sequelae, and the potentially unde- sirable short and long term effects of IBCA. Donald J. Prolo, M.D. San Jose, California March 1979 I. Kerber, C. Experimental arteriovenous fistula: Creation and per- cutaneous catheter obstruction with cyanoacrylate. Invest. Radiol., 10: 10-17, 1975. . Miller, J. D., Jawad, K., and Jennett. B. Safety of carotid ligation and its role in the management of intracranial aneurysms. J. Neurol. Neurosurg. Psychiatry, 40: 64-72, 1977. 3. White, R. 1, Strandberg, J. V., Gross, G. S., and Barth, K. H. Therapeutic embolization with long-term occluding agents and their effects on embolized tissues. Radiology, 125: 677-687, 1977. N AUTHORS’ REPLY TO DR. PROLO’S COMMENTS 1. The fistulas were completely obliterated in all patients. See figures and text. 2. There is no study that shows even a potential carcinogenic effect of IBCA. The IBCA we use (Ethicon, Inc.) has about the same tissue reactivity as silk suture (Kerber et al., in prepara- CAROTID-CAVERNOUS FISTULA 215 tion; Cromwell et al., in press; Zanetti and Sherman, Ref. 18 in paper) and is not the same compounding as that used by White. 3. Dr. Prolo quotes my reference out of context. 4. Dr. Prolo’s contention that two patients may have had occluded carotid arteries later is one that concerns me, too. If one looks at the histories of those patients, one understands their steadfast refusal to return to the hospital. 5. If 80% of patients tolerate simple balloon occlusion of the carotid, one assumes that 20% do not. That in itself is a powerful reason to search for alternate treatment modalities for this affliction. All of our publications have stressed that these techniques are difficult. Perhaps the time has come to put the problems into perspective: compared to clipping an anterior cerebral artery aneurysm or excising an arteriovenous malformation, catheter therapy should not even raise a sweat. Charles W. Kerber, M.D Pittsburgh, Pennsylvania