Journal of Clinical Neuroscience (2000) 7(Supplement 1), 88–91 © 2000 Harcourt Publishers Ltd doi: 10.1054/ jocn.2000.0720, available online at http://www.idealibrary.com on Surgical resection of cerebral arteriovenous malformation combined with pre-operative embolisation Kazuhiro Hongo MD, George Koike MD, Masanori Isobe MD, Takeya Watabe MD, Nobuhito Morota MD, Hiroshi Nakagawa MD Department of Neurological Surgery, Aichi Medical University, Nagakute, Aichi, Japan Summary To assess the importance of pre-operative embolisation, 27 cases of cerebral artriovenous malformation (AVM) treated in this institute between July 1994 and October 1998 were analysed. The patients’ ages ranged from 3 to 70 years (average 36.9) with a follow-up period of 1–41 months (average 19.2). The patient presented with haemorrhage in 21 cases and seizure in five. In 21 of 27 cases, surgical resection of a nidus was performed, gamma knife therapy was applied in three and conservative therapy was chosen in three. Of 21 cases treated surgically, total removal was achieved in 19 cases and a residual nidus was seen in one (a large basal ganglia AVM). In the remaining case, postoperative angiography was not available. Pre-operative embolisation followed by surgical resection of the nidus was performed in seven cases in which there was a large AVM. A volume index was calculated to indicate the size of the nidus using X × Y × Z, where X is the maximum diameter (cm) of the nidus on the lateral angiogram, Y is the diameter (cm) perpendicular to X and Z is the maximum diameter (cm) on the anteroposter or angiogram. The index averaged 45.9 for the cases in which pre-operative embolisation was performed, while it was 5.6 in the cases without embolisation. Pre-operative embolisation was performed to reduce the nidus flow as much as possible, to prevent overload to the surrounding structures. At surgery, the nidus was resected from the surrounding tissue and care was taken not to enter the nidus. Postoperatively, the systolic blood pressure was maintained at 90–100 mmHg for several days in the intensive care unit. The results were excellent in 15 cases, good in three (hemiparesis due to the initial haemorrhage remained in all three), fair in one (a patient with a severe subarachnoid haemorrhage). Two patients died (acute pulmonary oedema and severe meningitis). Minor postoperative bleeding or oozing was seen in three cases. In conclusion, reducing the shunt flow through a nidus in a step-wise fashion with pre-operative embolisation of a large AVM seems to be quite helpful in preventing postoperative haemodynamic overload to the surrounding brain. It is also important not to enter the nidus when it is removed at surgery. This helps to prevent intraoperative and/or postoperative bleeding, and led to successful total removal of the nidus with a good postoperative course. © 2000 Harcourt Publishers Ltd Keywords: arteriovenous malformation, pre-operative embolisation, microsurgery, overload phenomenon INTRODUCTION Surgical removal of an arteriovenous malformation (AVM) is still considered to be one of the most difficult neurosurgical procedures. In recent years, with the introduction of interventional techniques and gamma knife surgery, the indications for surgery in AVM have changed significantly and multidisciplinary treatment has been instituted.1–5 Recently, the authors have utilised pre-operative embolisation in large AVMs to reduce the bleeding during surgery and decrease the risk of postoperative overload to the surrounding brain.6–11 This paper describes the surgical results obtained in treating AVMs and analyses the significance of pre-operative embolisation. CLINICAL MATERIALS AND METHODS Between July 1994 and October 1998, 27 patients (20 males and 7 females) aged 3–70 years (mean 36.9) with angiographically confirmed cerebral AVMs were treated in this institute. The patient presented with haemorrhage in 21 cases and seizures in five; the remaining case was found incidentally. Surgical resection was performed in 21 of 27 cases, a gamma knife was used in three and Correspondence to: Kazuhiro Hongo MD, Department of Neurosurgery, Shinshu University School of Medicine, Asahi 3–1–1, Matsumoto 390-8621, Japan. Tel: +(81) 263 37 2690; Fax: +(81) 263 37 0480; E-mail: khongo@hsp.md.shinshu-u.ac.jp 88 conservative therapy was chosen in three. Before surgical resection of the nidus, embolisation of larger AVMs was performed. As an index of nidus size, the volume index (VI) was calculated using X × Y × Z, where X is the maximum diameter (cm) of the nidus on the lateral angiogram, Y is the diameter (cm) perpendicular to X and Z is the maximum diameter (cm) on the anteroposterior angiogram.8, 11 Pre-operative embolisation was performed 1 week before surgery using Aviten® powder as a particle material. When the nidus was large, the embolisation procedure to reduce the size of the nidus was performed two or three times at 1 week intervals. Surgical resection of the nidus was then carried out. During and after the surgical procedure, the systolic blood pressure was kept between 90 and 100 mmHg for several days, to prevent bleeding from the nidus or the surrounding brain. During this period, the patient was kept in the intensive care unit. The postoperative neurological and angiographic findings were evaluated. The follow-up period was 2–41 months (average 19.2). RESULTS The clinical findings are summarised in Table 1. The SpetzlerMartin grading2 was I in 5 cases, II in 8, III in 10, IV in 3 and V in 1. In 15 of 27 cases, direct surgical removal was performed, in three cases gamma knife therapy was used and in the remaining three cases conservative therapy was chosen in accordance with the patients’ requests (Table 2). The VI of the nidus of all 27 patients ranged from 1 to 112 (average 28.4). Pre-operative embolisation was carried out in 8 of 27 cases. Of 8 patients, embolisation was performed in one patient before gamma knife Pre-operative embolisation and resection of AVM 89 Fig. 1 Volume index and pre-operative embolisation; •, cases with postoperative bleeding or oozing. therapy. One patient underwent pre-operative embolisation three times and another two patients twice. The VI ranged from 3.4 to 101 (average 45.9) in the cases with pre-operative embolisation, and 1–26.3 (average 5.6) in those without embolisation (Fig. 1). The results were excellent in 15 cases, good in 3 (hemiparesis due to the initial haemorrhage remained in all three) and fair in one (a patient with a severe subarachnoid haemorrhage). Two patients died (acute pulmonary oedema and severe meningitis) (Table 3). In three cases, intracerebral bleeding occurred after surgery (Fig. 1). Bleeding occurred in two cases with pre-operative embolisation the VI in these cases was 50.6 and 101) and in one case without pre-operative embolisation (the VI was 15.6). In all three cases, the bleeding was minimal and no fatal haemorrhage occurred. Fig. 2 CT scan showing the subarachnoid and intraventricular haemorrhages. Total removal of the nidus was achieved in 19 of 20 surgical cases (Table 4). A postoperative angiogram was not available for one surgical patient (case 8). This case is described later. There were no complications related to the embolisation. Table 1 Profile of 27 patients with arteriovernous malformations Case No. Age (yrs). Sex 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 56, M 53, M 24, M 27, M 22, F 35, F 53, M 21, M 18, M 31, M 19, M 58, M 3, M 55, F 21, F 46, F 17, M 18, M 70, M 26, M 47, M 49, F 24, M 58, M 62, M 55, F 27, M Location Initial Presentation parietal seizure frontal ICH thalamus ICH temporal seizure occipital seizure occipital ICH parletal ICH basal ganglia ICH parietal ICH occipital ICH temporal ICH frontal ICH frontal ICH thalamus ICH corpus callosum ICH frontal ICH parietal ICH temporal ICH parietal ICH frontal ICH parietal seizure parietal ICH frontal seizure CM junction SAH parietal ICH occipital ICH cerebellum incidental Size (mm) V.I. S-M grade Treatment Preop. Emb. Residual Nidus Outcome Follow-up Period (mos) 50*30*40 10*10*10 10*10*10 70*40*40 40*30*40 20*20*20 10*10*10 50*45*45 40*30*30 15*15*15 20*15*15 25*25*25 10*10*10 15*15*15 70*45*25 10*10*10 25*10*10 50*30*30 20*10*10 20*20*10 45*45*25 40*25*25 50*50*50 10*10*10 35*30*25 10*10*10 20*20*20 60 1 1 112 48 8 1 101 36 3.4 4.5 15.6 1 3.4 78.8 1 2.5 45 2 4 50.6 25 125 1 26.3 1 8 3 1 3 4 4 2 2 4 3 3 1 1 1 3 5 2 2 3 2 2 3 3 3 3 2 2 1 op op gamma cons gamma op op op op op op op op gamma cons op op op op op op op cons op op op op + − − − + − − ++ + + − − − − − − − +++ − − ++ + − − − − − − − * * * − − N.A. − − − − * * − − − − − − − − * − − − − excellent excellent excellent excellent excellent good good dead dead excellent excellent excellent excellent excellent excellent excellent excellent excellent excellent excellent excellent excellent excellent fair good excellent excellent 40 41 27 39 33 37 29 23 26 23 20 20 20 19 20 14 13 11 11 11 9 9 8 6 5 2 2 ICH: Intracerebral hemorrhage, IVH: intraventricular hemorrhage, SAH: subarachnoid hemorrhage, V.I.: volume index, op: direct surgery, cons: conservative therapy, Preop. Emb.: preoperative embolization, N.A.: not applicable © 2000 Harcourt Publishers Ltd Journal of Clinical Neuroscience (2000) 7(Supplement 1), 88–91 90 Hongo et al. B A Fig. 3 Left carotid angiograms showing a nidus in the left parietal area fed by the middle cerebral artery. A: lateral view; B: anteroposterior view. Table 2 Treatment modality S-M grade Table 4 Removal of nidus No. of cases Treatment Modality S-M grade No. of cases direct surgery gamma knife conservative I II III IV V total 5 8 10 3 1 27 5 8 7 1 0 21 0 0 2 1 0 3 0 0 1 1 1 3 I II III IV V total 5 8 6 1 0 20* nidus total removal residual 5 8 6 0 0 19 0 0 0 1 0 1 *: Postoperative angiogram is not performed in one case. Table 3 Operative result S-M grade I II III IV V total No. of cases 5 8 10 3 1 27 Operative Result Excellent Good Fair Poor Dead 5 5 5 0 0 15 0 3 0 0 0 3 0 0 1 0 0 1 0 0 0 0 0 0 0 0 1 1 0 2 CASE STUDY This 47-year-old man had a 10-year history of seizures because of a large left parietal AVM, and was followed-up in the outpatient clinic. His seizures were well controlled until 23 December 1997, when he developed a severe headache. A CT scan showed a mild hemorrhage in the left lateral ventricle (Fig. 2). An angiogram showed a large (4.5 × 4.5 × 2.5 cm) AVM in the left parietal lobe, with a VI of 50.6 and a Spetzler and Martin grade of III (Fig. 3). The nidus was fed mainly from the middle cerebral artery, with some flow from the anterior and posterior cerebral arteries. Two preoperative embolisations were performed, with a 1-week interval between, to occlude the part of the nidus fed by the middle cerebral artery. This reduced the nidus volume by nearly 50%. Direct surgery was then performed. Since the nidus was located near the motor cortex, neuroelectrophysiological monitoring was performed and the motor area was confirmed during surgery, the the nidus was then removed. Postoperatively, the systolic blood pressure was kept between 90 and 100 Journal of Clinical Neuroscience (2000) 7(Supplement 1), 88–91 mmHg for several days, and the patient showed no neurological deficits. The postoperative angiogram showed complete disappearance of the nidus (Fig. 4). However, the postoperative CT scans showed minor bleeding in the cavity after nidus removal. The patient was discharged 1 month after surgery without neurological deficits. DISCUSSION The strategy for treating patients with cerebral AVMs has changed dramatically since the recent advent of intravascular and radiosurgery techniques; 1–3,5 however, direct surgical treatment of an AVM is still an important option.1,13 The goal of AVM surgery is to remove the nidus to prevent future haemorrhage and to improve the neurological status caused by the presence and effects of the AVM. After adequate embolisation of the feeding arteries by an intravascular technique, the nidus is removed surgically. It is very important that the surgical exposure provides access to the major feeding arteries and draining veins. Therefore, pre-operative embolisation primarily targets the feeding arteries that are difficult to approach surgically or the major feeding arteries, to minimise intraoperative bleeding. Another important role of intraoperative embolisation is to reduce the postoperative vascular overload of the surrounding brain,4, 6–9, 11, 14 although a case against staged operative resection of a cerebral AVM has been reported.15 The overload phenomenon may occur after total resection of the nidus. The rationale for step-wise therapy of the AVM is to reduce the vascular overload of the surrounding brain by gradually reducing © 2000 Harcourt Publishers Ltd Pre-operative embolisation and resection of AVM 91 B A Fig. 4 Left carotid angiogram showing complete removal of the nidus. A: lateral view; B: anteroposteior view. the shunt flow through the nidus, especially when the AVM is large. In the present study a large AVM was defined as one with a maximum diameter >5 cm or a VI > 60.11 The authors used pre-operative embolisation mainly for large AVMs and ultimately the postoperative overload phenomenon did not occur. There were three cases with postoperative bleeding. In all three cases, the bleeding was minimal and occurred in the cavity left by removal of the nidus. Collagen particle was used as an embolised material in this series. Particles cannot pass through vessels with a small diameter and are thought to be easily washed away by blood flow as compared with liquid materials, which plug a nidus immediately after injection.16 Recently, however, liquid material has been mainly used in this institute. Postoperatively, it is essential to control the systolic blood pressure, maintaining it between 90 and 100 mmHg, for about 1 week after surgery.8,9,11,17 This is required to prevent both postoperative bleeding and the postoperative vascular overload phenomenon. Antihypertensive drugs or barbiturates can be used during the first few days after surgery to reduce the systolic blood pressure. Postoperative angiograms should be taken 1 week after surgery even in cases where complete resection of the nidus is confirmed by intraoperative angiography or immediate postoperative angiography. The induced hypotension should be maintained until there is no postoperative overload. At surgery, it is important not to enter the nidus while dissecting it. Following this principle helps to prevent intraoperative and postoperative bleeding, and leads to successful total removal of the nidus with a good postoperative course. In conclusion, reducing the shunt flow through a nidus in a stepwise fashion with pre-operative embolisation of a large AVM seems to be quite helpful in preventing the postoperative haemodynamic overload to the surrounding brain. REFERENCES 1. Kobayashi S, Hongo K, Bhagwati SN, Rappaport ZH, Lawton MT, Spetzler RF. Management of arteriovenous malformations: Part II. Surg Neurol 1997; 48: 2–6. © 2000 Harcourt Publishers Ltd 2. Tew J, Levis AI, Reichert KW. Management strategies and surgical techniques for deep-seated supratentorial arteriovenous malformations. Neurosurgery 1995; 36: 1065–1072. 3. Deruty R, Pelissou-Guyotat I, Mottolese AC, Bascoulergue, Turjman F, Gerard JP. Multidisciplinary treatment of cerebral arteriovenous malformations. Neurol Res 1995; 17: 169–177. 4. Sasaki T, Kurita H, Saito I, Kawamoto S, Nemoto S, Terahara A, Kirino T, Takakura K. Arteriovenous malformations in the basal ganglia and thalamus: management and results in 101 cases. J Neurosurg 1998; 88: 285–292. 5. Rosenblatt S, Lewis AI, Tew JM. Combined interventional and surgical treatment of arteriovenous malformations. Neuroimaging Clin North Am 1998; 8: 469–482. 6. Young VL, Kader A, Ornstein E, Baker K, Ostapkovich N, Pile-spellman J, Fogarty-Mack P, Stein B. Cerebral hyperemia after arteriovenous malformation resection is related to “breakthrough” complications but not to feeding artery pressure. Neurosurgery 1996; 38: 1085–1095. 7. Deruty R, Pelissou-Guyotat, Amat D, Mottolese C, Bascoulergue Y, Turjman F, Gerard JP. Complications after multidisciplinary treatment of cerebral arteriovenous malformations. Acta Neurochir (Wien) 1996; 138: 119–131. 8. Hongo K, Kobayashi S, Takemae T, Nakagawa H, Mabuchi S. Staged surgery for a large cerebral arteriovenous malformation to prevent the overload phenomenon. J Clin Neurosci 1998; 5 (Suppl 1): 84–86. 9. Takemae T, Kobayashi S, Sugita K. Perinidal hypervascular network on immediate postoperative angiogram after removal of large arteriovenous malformations located distant from the arterial circle of Willis. Neurosurgery 1993; 33: 400–406. 10. Spetzler RF, Wilson CB, Weinstein P, Hehdorn H, Townsend J. Normal perfusion pressure breakthrough theory. Clin Neurosurg 1978; 25: 651–672. 11. Hongo K, Kobayashi S, Takemae T. Arteriovenous malformation. In: Kobayashi S, Goel A, Hongo K (eds). Neurosurgery of complex tumors and vascular lesions. Churchill Livingstone, Edinburgh, 1997; 127–138. 12. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg 1986; 65: 476–483. 13. de Oliveira E, Tedeschi H, Siqueira MG, Ono M, Rhoton AL. Arteriovenous malformations of the basal ganglia region: rationale for surgical management. Acta Neurochir (Wien) 1997; 139: 487–506. 14. Batjer HH, Devous MD Sr, Meyer YJ, Purdy PD, Samson DS. Cerebrovascular hemodynamics in arteriovenous malformation complicated by normal perfusion pressure breakthrough. Neurosurgery 1988; 22: 503–509. 15. Morgan MK, Sundt TM Jr. The case against staged operative resection of cerebral arteriovenous malformations. Neurosurgery 1989; 25: 429–435. 16. Sorimachi T, Koike T, Takeuchi S, Minakawa T, Abe H, Nishimaki K, Ito Y, Tanaka R. Embolisation of cerebral arteriovenous malformations achieved with polyvinyl alcohol particles: angiographic reappearance and complications. Am J Neuroradiol 1999; 20: 1323–1328. 17. Day AL, Friedman WA, Sypert GW, Mickle JP. Successful treatment of the normal perfusion pressure breakthrough syndrome. Neurosurgery 1982; 11: 625–630. Journal of Clinical Neuroscience (2000) 7(Supplement 1), 88–91