0148-396 X/89/2503-0424$02.00/0 NEUROSURGERY Vol. 25. No. 3, 1989 Copyright © 1989 by the Congress of Neurological Surgeons Printed in U.S.A. Case reports Transluminal Angioplasty of Intracerebral Vessels for Cerebral Arterial Spasm: Reversal of Neurological Deficits after Delayed Treatment Stanley L. Barnwell, M.D., Ph.D., Randall T. Higashida, M.D., Van V. Halbach, M.D., Christopher F. Dowd, M.D., Charles B. Wilson, M.D., and Grant B. Hieshima, M.D. Departments of Neurological Surgery (SLB, RTH, VVH, CBW, GBH) and Radiology (SLB, RTH, VVH, CFD, GBH), School of Medicine, University of California San Francisco, San Francisco, California We used intracerebral transluminal angioplasty to treat two episodes of symptomatic vasospasm in a patient recovering from an aneurysmal subarachnoid hemorrhage. The procedures were performed after medical therapies, intravascular volume expansion, and induced arterial hypertension failed to alleviate the patient’s neurological condition. The first angioplasty, confined to the right middle cerebral and distal internal carotid arteries, took place more than 30 hours after the onset of left hemiplegia. Despite the subsequent discovery of a small parietal lobe infarct, it brought about a marked improvement in left motor function and may have also limited the spread of necrotic damage. The second angioplasty was necessitated when stenotic segments of the basilar and posterior cerebral arteries caused a 24-hour decline in the patient’s mental status. Although delayed in relation to the onset of symptoms, it successfully reversed the patient’s comatose state. The use of transluminal angioplasty for vasospasm is generally limited to cases where it can be performed shortly after the onset of neurological symptoms; delaying the procedure increases the risk of hemorrhage from reperfused areas of infarction. Our experience with this patient demonstrates that delayed angioplasty can improve vascular flow to ischemic territory, even after infarction, without complications and with resultant improvement in neurological function. (Neurosurgery 25:424-429, 1989) Key words: Angioplasty, Subarachnoid hemorrhage, Vasospasm Intracranial arterial spasm occurs in approximately 30% of patients with aneurysmal subarachnoid hemorrhage, usually within 5 to 12 days of the hemorrhagic event (1, 10, 11). Recognized as a serious complication, it increases the risk of cerebral infarction and contributes to the high mortality and morbidity rates associated with aneurysmal bleeding: less than 50% of patients survive and only 36% of patients return to full employment (9, 12). Many pharmacological agents have been evaluated to treat this condition, but none are known that consistently reverse its clinical manifestations (2-6, 13-15). Medical therapies such as induced arterial hypertension or expanded intravas- cular volume have also been advocated. However, these are associated with complications, and their effectiveness remains to be adequately demonstrated (8). Recently, transluminal angioplasty has been shown to re- verse arterial spasm and restore cerebral perfusion in patients with aneurysmal subarachnoid bleeding. In one report, 33 patients with symptomatic vasospasm benefited from this procedure: they demonstrated a marked improvement in their neurological condition accompanied by angiographic evi- dence of arterial widening (16). In another report, angioplasty reversed the comatose state of a patient that had resulted from vasospasm subsequent to a ruptured basilar artery aneurysm (7). The beneficial effect of transluminal angioplasty has largely been restricted to those cases where the procedure is per- formed shortly after the neurological deficit develops, usually FiG. 1. Noncontrast CT scan obtained on the day the patient within hours. In this case, transluminal angioplasty was effec- presented with a severe headache. The image shows moderate hydro- tive in reversing a fixed neurological deficit that had been cephalus and extensive subarachnoid blood in the sylvian fissures and present for more than 24 hours. The angioplasty produced interpeduncular cistern. There is no evidence of infarction. 424 September 1989 marked improvement in the patient’s condition, despite the occurrence of an infarct. CASE REPORT A 60-year-old woman was taken to the hospital after the sudden onset of headache. A computed tomographic (CT) scan, obtained at admission, showed extensive subarachnoid DELAYED TRANSLUMINAL ANGIOPLASTY 425 blood in the interpeduncular fossa and in both sylvian fissures (Fig. 1). The source of bleeding, an aneurysm of the right posterior communicating artery, was identified 24 hours later by cerebral angiogram (Fig. 2). The angiogram did not show any evidence of vasospasm. Five days after the subarachnoid hemorrhage, a craniotomy was performed to clip the aneurysm. There were no compli- cations during surgery, and a neurological examination when Cc Fic. 2. Cerebral angiograms taken | day after the aneurysmal subarachnoid hemorrhage. A, right internal carotid angiogram (right anterior oblique projection) shows an aneurysm in the right posterior communicating artery. No other vascular abnormalities appear. B, left internal carotid angiogram (anteroposterior projection) shows no abnormalities. C, perfusion in the left vertebral artery (Towne’s projection) is normal. 426 BARNWELL etal. the patient awoke showed no abnormalities. On postoperative Day |, the patient developed left hemiplegia. A CT scan showed no significant edema, hemorrhage. or infarction, and the presumptive diagnosis of cerebral arterial spasm after aneurysmal subarachnoid hemorrhage was made. She was treated vigorously by intravascular volume expansion and induced arterial hypertension, without improvement. Thirty- two hours after the onset of hemiplegia, an angiogram re- vealed severe vasospasm involving the distal supraclinoid segment of the right internal carotid artery and the M1 segment of the right middle cerebral artery (Fig. 3). There was also delayed filling of the distal branches of the middle cerebral artery. With the patient receiving anticoagulation therapy (5,000 units of heparin), an angioplasty was performed by passing a 7.3 French nontapered catheter from the right femoral artery into the proximal right internal carotid artery. A coaxial 2 French-4 French catheter with a custom-designed, nonde- tachable, silicone microballoon (Interventional Therapeutics Corporation, South San Francisco, California) was then nav- igated into the supraclinoid internal carotid artery. Angio- plastic widening of the distal supraclinoid internal carotid artery and the M1 segment of the middle cerebral artery was achieved by repetitive inflation and deflation of the balloon; balloon inflation was performed at low pressures and for short intervals (1 second). A cerebral angiogram, obtained imme- diately after the angioplasty, confirmed that the treated vessels were enlarged and that there was increased filling of the distal branches of the middle cerebral artery (Fig. 4). Eight hours later, the patient’s neurological condition improved: she could flex her left arm and raise her left leg off the bed. Three days after the craniotomy, the patient’s left arm and leg motor function remained unchanged; however, in con- junction with her maintained motor recovery and despite continued intravascular volume expansion therapy, her men- y Fic. 3. Right internal carotid angiogram (anteroposterior projec- tion) obtained 32 hours after the onset of hemiplegia. There are severe narrowings in the internal carotid and middle ‘bral arteries (ar- rows). The distal branches of the middle cerebral artery fill in a delayed fashion, Neurosurgery, Vol. 25, No. 3 Fic. 4. Right internal carotid angiogram (left anterior oblique projection) obtained immediately after the first angioplasty. The lumina of the distal supraclinoid internal carotid and middle cerebral arteries are widened (arrows). The branches of the distal middle cerebral artery fill quickly; there are no branch occlusions. Fic. 5. Noncontrast CT scan, obtained 3 days after the craniot- omy, shows an infarct in the posterior right parietal lobe. tal status started to decline: she became less responsive, would not follow commands, and ultimately was difficult to arouse. A CT scan revealed an infarct in the right posterior parietal lobe (Fig. 5). A cerebral angiogram. performed to evaluate recurrent or new sites of vasospasm, showed no recurrence of the original vasospasm, but new abnormalities in the verte- September 1989 B Fic. 6. Cerebral angiograms taken 3 days after the craniotomy. A, left vertebral angiogram (Towne’s projection) shows severe vaso- spasm of the right posterior cerebral artery, involving both P| and P2 segments (arrows), and the distal basilar artery (curved arrow’). The left posterior cerebral artery is normal. B. left internal carotid angiogram (anteroposterior projection) indicates moderate vasospasm of the middle cerebral artery (arrow). The distal branches fill normally with no branch occlusions. DELAYED TRANSLUMINAL ANGIOPLASTY 427 bral-basilar circulation: there was severe stenosis in the basilar artery and the Pl and P2 segments of the right posterior cerebral artery (Fig. 6). There was also an asymptomatic marked vasospasm of the left middle cerebral artery. A second angioplasty was performed 4 days after the cra- niotomy (24 hours after the onset of coma) to widen the affected segments of the basilar and posterior cerebral arteries (Fig. 7). The patient showed improvement 18 hours after the procedure: she began to follow commands, had spontaneous awakening and speech, and was oriented to her name, the year, and the place. Her left hemiparesis also continued to recover. She was discharged 10 days after the second angio- plasty in stable neurological condition and was transferred to a convalescent hospital. After 3 months, she is resuming an independent lifestyle and has returned to work. Her left hand function continues to improve. DISCUSSION The use of transluminal angioplasty to treat vasospasm associated with aneurysmal subarachnoid hemorrhage was first described in 1984 by Zubkov et al. (16). From their experience with 33 patients, they concluded that cerebral angioplasty was a useful method of reversing arterial spasm and restoring neurological function. Importantly, their pa- tients’ recovery was sustained at follow-up, 3 months later. Intracranial angioplasty is not indicated in all cases of vasospasm. It is most effective when performed soon after the onset of symptoms secondary to vasoconstriction, as can occur in cases related to an aneurysmal subarachnoid hem- orrhage or balloon embolization procedure. In such situa- tions, the clinical response to angioplasty is usually immediate Fic. 7. Left vertebral angiogram (Towne’s projection) after angio- plasty of the basilar artery and the P| and P2 segments of the posterior cerebral artery. Arrows point to the widened segment of the posterior cerebral artery and basilar artery (curved arrow). 428 BARNWELL et al. (7). In contrast, angioplasty may not be warranted for patients with longstanding neurological deficits attributed to infarc- tion; after severe ischemia, reperfusion of the infarcted area could result in hemorrhage. We performed two intracerebral angioplasties in our pa- tient, 3 days apart: the first, to treat hemiplegia precipitated by severe narrowing in the right internal carotid and middle cerebral arteries, and the second, in response to a coma associated with stenoses in the basilar and right posterior cerebral arteries. In both instances, the angioplasty was de- layed for more than 24 hours, while other forms of therapy, intravascular volume expansion and induced arterial hyper- tension, were attempted. Despite these time lags, both angio- plasties successfully restored the patient’s neurological func- tion, although the beneficial effects were not immediately visible. Conceivably, the time course of her recovery was directly related to the duration of her ischemic episode. Our patient’s positive response to her first angioplasty was especially interesting because of the parietal lobe infarct that we subsequently discovered. The stroke, which we were una- ware of at the time of the procedure, testifies to the degree of ischemia she experienced. Yet, despite the developing infarc- tion, the angioplasty produced a dramatic improvement in clinical function. Presumably, it reduced the size of the infarct by increasing blood flow to adjacent areas of the brain, thereby containing the spread of necrosis. Another interesting obser- vation is that we did not precipitate a hemorrhage when we reperfused the infarcted area. In contrast, we did encounter this complication in another patient, 24 hours after an angio- plasty of the middle cerebral artery (7). The delayed angioplasties in this patient were beneficial because they took place before there was more widespread tissue damage. Clearly, if the stroke had progressed further, it is unlikely that we would have seen such a favorable clinical recovery. This report suggests that transluminal angioplasty may be a useful way of reversing fixed neurological deficits, even if it is delayed more than 24 hours. ACKNOWLEDGMENTS The authors thank Cindy Huff for manuscript preparation and Ellyn Cohen for editorial assistance. Received for publication, March 3, 1989; accepted, March 29, 1989. Reprint requests: Stanley L. Barnwell, M.D., Ph.D., c/o Depart- ment of Neurological Surgery, The Editorial Office, 1360 Ninth Avenue, Suite 210, San Francisco, CA 94122. REFERENCES 1. Allcock JM, Drake CG: Ruptured intracranial aneurysms-the role of arterial spasm. J Neurosurg 22:21-29, 1965. 2. Chapleau CE, White RP, Robertson JT: Cerebral vasodilation and prostacyclin. The effects of aspirin and meclofenamate in vitro. J Neurosurg 53:188-192, 1980. 3. Chyatte D, Rush N, Sundt TM: Prevention of chronic experi- mental cerebral vasospasm with ibuprofen and high dose meth- ylprednisolone. J Neurosurg 59:925-932, 1981. 4. Ellis EF, Nies AS, Dates JA: Cerebral arterial smooth muscle contraction by thromboxane A2. Stroke 8:480-486, 1977. 5. Endos S, Suzuki J: Experimental cerebral vasospasm after sub- arachnoid hemorrhage. The participation of adrenergic nerves of cerebral vessel walls. Stroke 10:703-711, 1979. 6. Flamm ES, Yasargil MG, Ransohoff J: Alteration of experimen- tal cerebral vasospasm by adrenergic blockade. J Neurosurg 37:294-301, 1972. 7. Higashida RT, Halbach VV, Cahan LD, Brant-Zawadzki M. Neurosurgery, Vol. 25, No. 3 Hieshima GB:Transluminal angioplasty of intracerebral vessels for treatment of intracranial arterial spasm. J Neurosurg (in press). 8. Kassell NF. Peerless SJ. Durward QJ, Beck DW. Drake CG. Adams HP: Treatment of ischemic deficits from vasospasm with intravascular volume expansion and induced arterial hyperten- sion. Neurosurgery 1 1:337-343, 1982. 9. McKissock W, Paine KW, Walsh L: An analysis of the results of treatment of ruptured intracranial aneurysms: A report of 722 consecutive cases. J Neurosurg 17:762-776, 1960. 10, Mullan S: Conservative management of the recently ruptured aneurysm, Surg Neurol 3:27-32, 1975. 11. Sahs AL, Nishioka H, Torner JC, Graf CJ, Kassell NF, Goettler LC: Cooperative study of intracranial aneurysms and subarach- noid hemorrhage: A long term prognostic study. Arch Neurol 41:1140-1147, 1984. 12. Smith RR, Yoshioka J: Intracranial arterial spasm, in Wilkins RH, Rengachary SS (eds): Neurosurgery. New York, McGraw Hill, 1985, pp 1355-1362. 13. Sundt TM: Chemical management of cerebral vasospasm, in Whisnant JP, Sandok BA (eds): Cerebral Vascular Disease. Pro- ceedings of the 9th Princeton Conference. New York, Grune & Stratton, 1974, pp 77-81. 14. Tani E, Maeda Y, Fukumori T: Effect of selective inhibitor of thromboxane A2 synthetase on cerebral vasospasm after early surgery. J Neurosurg 61:24-29, 1984. 15, Wilkins RH: Attempted prevention or treatment of intracranial arterial spasm: A survey, in Wilkins RH (ed): Cerebral Arterial Spasm. Baltimore, Williams & Wilkins, 1979, pp 542-555. 16. Zubkov YN, Nikiforov BM, Shustin VA: Balloon catheter tech- nique for dilatation of constricted cerebral arteries after aneurys- mal SAH. Acta Neurochir (Wien) 70:65-79, 1984. COMMENT This is a well-written and well-documented case report of transluminal angioplasty for vasospasm. The authors have nicely demonstrated, with the angiograms provided, that their angioplasty resulted in a dramatic immediate reversal of the previously documented segmental angiographic arterial nar- rowing. They have not, however, demonstrated that “angio- plasty can improve vascular flow to ischemic territory.” They also have not demonstrated that the improvement that this patient experienced after the two angioplastic procedures was directly related to such procedures. In each case there was significant delay between the procedure and the beginning of clinical improvement (8 hours after the first procedure and 18 hours after the second procedure). Although it is tempting to postulate that the procedures led to the clinical improve- ment, one should keep in mind the well-known fact that patients with clinical vasospasm do improve spontaneously with considerable frequency. In addition, this patient was also simultaneously treated with the usual measures known to have an effect in vasospasm. It is interesting, however, that the patient had a rather profound deficit for a significant period of time before the procedures, which makes it more likely that, indeed, the procedures were at least partially responsible for the subsequent improvement. Dramatic reports from such a respected group about a possible new treatment for vasospasm are most welcome. We all know the long list of initially encouraging “therapies” for vasospasm that, on wider use, proved to be disappointing. This is no reason to be nihilistic, but it is reason to be cautious. Caution is particularly in order when contemplating a therapy that, at least theoretically, results in increased perfusion to infarcted tissue or to tissue where ischemia has been of long standing. In addition to the well-known risk of hemorrhage and exacerbation of edema, there has been much recent discussion of the potential dangers of reperfusion to the September 1989 ischemic cell per se. In addition, this particular form of therapy requires exquisite expertise. These authors are well recognized as leaders in the area of interventional neurora- diology. It would be wisest if such innovative forms of therapy could be utilized initially only in a few selected centers such as this, where one could be sure that they would be done as safely as possible and that the results would be evaluated with honesty and scientific thoroughness. It would be a shame if reports such as this led to widespread use of such an invasive technique by individuals with less experience. This is a likely 0148-396X/89/2503-0429$02.00/0 NEUROSURGERY Copyright © 1989 by the Congress of Neurological Surgeons DELAYED TRANSLUMINAL ANGIOPLASTY 429 scenario, in view of the magnitude of the problem (vasospasm) and our relatively ineffective therapies for this condition at present. Unfortunately, anecdotal verbal reports of catastro- phes with the use of this technique are becoming more prev- alent. The authors are certainly to be congratulated for their excellent technical results and their courage in pursuing this new frontier. Roberto C. Heros Boston, Massachusetts Vol. 25, No. 3, 1989 Printed in U.S.A. The Case Against Staged Operative Resection of Cerebral Arteriovenous Malformations Michael K. Morgan, M.B., B.S., F.R.A.C.S., and Thoralf M. Sundt, Jr., M.D. Cerebrovascular Research Laboratory, Department of Neurologic Surgery, Mayo Clinic, Rochester, Minnesota Three cases of large cerebral arteriovenous fistulae are presented in which surgical ablation was complicated by brain swelling from hyperperfusion breakthrough believed to be caused by acute intraoperative hypoperfusion superimposed on chronic preoperative hypoperfusion. On the basis of these cases, experimental data, and theoretical considerations, we seriously question the wisdom of using staged surgical resection of cerebral arteriovenous malformation to prevent complications related to alterations in cerebral hemodynamics. The reasons for this concern are: the repeated occurrence of acute-on-chronic hypoperfusion during staged resection; a lack of understanding of the time course for the correction of a disordered autoregulation; risk of hemorrhage between the initial and final resection; difficulty in assessing and substantiating flow reduction after subtotal resection; the rapidity of collateralization; the divergence of flow from large, readily accessible feeding arteries to deep penetrating vessels; and attenuation of the wall thickness in collateral vessels as a consequence of increased flow. (Neurosurgery 25:429-436, 1989) Key words: Autoregulation, Cerebral arteriovenous malformation, Complication, Ischemia, Surgery INTRODUCTION Gowers introduced the concept of cerebral hyperemia caus- ing neurological morbidity on the occlusion of a cerebral artery in 1888 (10). After the introduction of cerebral angiog- raphy and prior to 1950, Elvidge (8), Olivecrona and Riives (26), and Norlen (23) all had contributed to the understanding of cerebral hypoperfusion induced by a cerebral arteriovenous malformation and improvement in perfusion after the extir- pation of the lesion. This angiographic hypoperfusion came to be known as “steal” (22). Spetzler et al. in 1978 (30) married these two concepts (i.e., steal and hyperemia) in the “normal perfusion pressure breakthrough” theory to account for some cases of cerebral hemorrhage and edema arising as a result of cerebral arteriovenous malformation ablation. Central to this theory is the concept that chronic hypoperfusion within the parenchyma affected by cerebral arteriovenous malforma- tions results in impaired cerebral autoregulation and that it remains impaired on restoration of a normal perfusion pres- sure after arteriovenous shunt flow ablation. This combina- tion of impaired autoregulation and normal perfusion pres- sure may result in hyperemia and disruption of the capillary bed. A preferable term rather than “normal perfusion pressure breakthrough” might be “normal pressure hyperperfusion breakthrough” as hyperemia is an important factor in this phenomenon. One suggested method for avoiding this complication of normal pressure hyperperfusion breakthrough is to stage the resection of cerebral arteriovenous malformations (1, 21, 23, 29, 30, 34); however, the clinical significance of the hemody- namic perturbations on extirpation of cerebral arteriovenous malformations has been questioned and alternate explana- tions for cases alleged to be examples of normal pressure hyperperfusion breakthrough are difficult to exclude (e.g., parenchymal damage from retraction and dissection, venous infarction from occlusion of draining veins, and incomplete removal with subsequent rupture) (5, 7, 13, 14, 27, 28, 35, 36). Furthermore, the values for cerebral blood flow changes (CBF) in several cases of alleged normal pressure hyperper- fusion breakthrough have been found only to be hyperemic with reference to the regional CBF (rCBF) before excision of arteriovenous malformations and not hyperemic compared with normal brain (3, 37), suggesting either error in the technique of blood flow evaluation, error in assigning the complication to this phenomenon, or error in the basic theory that autoregulation is impaired. Nonetheless, the theory of hypoperfusion-induced loss of autoregulation is sound and experimental evidence supporting this phenomenon has been reported (17, 30). Three cases of cerebral arteriovenous fistula are presented in which surgical morbidity was thought to be related to acute intraoperative hypoperfusion superimposed on chronic pre-