Surgical Neurology 66 (2006) 50 – 55 www.surgicalneurology-online.com Vascular-AVM Retrospective analysis of the surgically treated temporal lobe arteriovenous malformations with focus on the visual field defects and epilepsy Shinji Nagata, MD, PhDa,4, Takato Morioka, MD, PhDa, Koichiro Matsukado, MD, PhDa, Yoshihoro Natori, MD, PhDb, Tomio Sasaki, MD, PhDa a Department of Neurosurgery, Kyushu University Graduate School of Medical Sciences, Fukuoka 812-8582, Japan b Department of Neurosurgery, Aso Iizuka Hospital, Iizuka 820-8505, Japan Received 5 May 2005; accepted 9 December 2005 Abstract Background: Authors investigate the surgical outcomes of the temporal lobe arteriovenous malformation (AVM) with focus on the visual field deficit and seizure. Methods: Between 1981 and 2004, we experienced 294 cases of intracranial AVMs. Among the 294 cases, 45 (15.3%) were located in the temporal lobe. Twenty-six of the 45 cases underwent microsurgical excisions of the AVMs. Results: The male-female ratio of 26 surgically treated temporal lobe AVMs was 15:11. The mean age was 34.2 years, ranging from 7 to 63 years. The sites of lesion were classified as polar in 1, dorsal in 2, laterobasal in 15, and mediobasal in 8. The initial symptoms were hemorrhage in 22 and epilepsy in 4 cases. Arteriovenous malformations were totally removed in all 26 patients and there was no surgical mortality. The visual field deficits were identified in 17 of 22 patients with hemorrhage. Massive hematoma cases that needed emergency operation were 5. Visual field deficits improved in only 2 of the 5 patients after surgery. Among the 7 quadrantanopia patients, 3 resulted in hemianopia after surgery. Seven of 22 hemorrhage patients had history of epilepsy. Although one patient had new postoperative epilepsy, the medical controls of the seizure were good in all 8 patients. Four patients underwent AVM excision for epilepsy without hemorrhage. In two patients, seizures disappeared after surgery. The other two patients had typical psychomotor seizures after the total excision of AVMs. Conclusions: Improvement of visual field deficit due to hematoma was difficult in most cases. Emergency craniotomy for global neurological deterioration due to massive hematoma had improved the visual field deficit in two cases. Although the outcome of seizure associated with hemorrhage was acceptable, the postoperative intractable seizures would remain in cases with epilepsy without hemorrhage. Intraoperative electrocorticography might be requisite for nonruptured temporal lobe AVM cases with epilepsy. D 2006 Elsevier Inc. All rights reserved. Keywords: Arteriovenous malformation; Temporal lobe; Surgery; Visual field defect; Epilepsy 1. Introduction Temporal lobe AVMs account for 12% to 16% of intracranial AVMs [1,2,8,13]. Common clinical symptoms of the temporal lobe AVM are visual field deficits due to the Abbreviations: AVM, arteriovenous malformation; CT, computed tomography; ECoG, electrocorticography. 4 Corresponding author. Tel.: +81 92 642 5524; fax: +81 92 642 5526. E-mail address: snagata@ns.med.kyushu-u.ac.jp (S. Nagata). 0090-3019/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.surneu.2005.12.017 hemorrhage and seizure. From the retrospective analysis of surgically treated temporal lobe AVMs, we will discuss the surgical effects on the temporal lobe AVM with focus on the visual field deficits and seizure. 2. Materials and methods Between 1981 and 2004, we experienced 294 cases of intracranial AVMs at the Department of Neurosurgery, S. Nagata et al. / Surgical Neurology 66 (2006) 50 – 55 Kyushu University Graduate School of Medical Sciences and the Department of Neurosurgery, Aso Iizuka Hospital. Among the 294 cases, 45 were located in the temporal lobe. The microsurgical excisions of the AVMs were performed in 26 cases: 22 cases with ruptured AVMs and 4 nonruptured cases presented by epilepsy. Surgical indication was limited to symptomatic patients with Spetzler and Martin grade I to III. Seventeen patients with non–life-threatening hemorrhage underwent microsurgery in chronic stages. The locations of the temporal lobe AVMs were classified according to Yasargil’s classification [13] as polar, dorsal, laterobasal, and mediobasal. The postoperative outcome of epilepsy was evaluated by Engel Seizure Outcome Scale [4] as class I = free of seizure or residual auras; class II = intermittent, infrequent seizures or relapse after a significant seizure-free period; class III = worthwhile improvement, more than 75% reduction in seizure frequency; and class IV = less than 75% reduction in seizure frequency. 3. Results The incidence of the temporal lobe AVM in our series was 15.3% (45/294) of all intracranial AVMs. Twenty-six patients underwent surgical excision of the temporal lobe AVMs (Table 1). The male-female ratio was 15:11. The mean age was 34.2 years, ranging from 7 to 63 years. The AVM locations were classified as polar in 1, dorsal in 2, laterobasal in 15 (2 anterior and 13 posterior), and 51 mediobasal in 8 (4 anterior and 4 posterior). The initial symptoms for the diagnosis were hemorrhage in 22 cases and epilepsy in 4. Arteriovenous malformations were totally removed in all 26 patients. The complete surgical excisions were confirmed with cerebral angiography. There was no surgical mortality and the permanent surgical morbidity was limited to the visual field defects and seizures. The visual field deficits were identified in 17 of 22 hemorrhage patients: 10 hemianopias and 7 quadrantanopias. Massive hematoma cases that needed emergency operation were 5. All the 5 patients showed hemianopias. Visual field deficits improved in 2 of the 5 patients who underwent emergency operation. The rest of the 8 hemianopia patients did not show improvement of the visual field. Among the 7 quadrantanopia patients, 3 revealed hemianopia after surgery. All these 3 patients had AVMs in the posterior mediobasal locations. Seven of 22 hemorrhage patients had history of epilepsy: 4 laterobasal, 2 mediobasal, and 1 dorsal locations. Intraoperative ECoG monitoring was not performed for these patients. One patient (case 1) had new postoperative epilepsy. The medical controls of the seizure were good in all 8 patients. They could return to their original life in Engel class I or II outcome. Four patients underwent AVM excision for epilepsy without hemorrhage. In two patients, seizures disappeared after surgery. Each of these AVMs was located in the laterobasal and polar sites. Intraoperative ECoG monitoring proved the disappearance of paroxysms Table 1 Clinical features of surgically treated cases with temporal lobe AVM (N = 26) Case Age Sex Site Side S-M B 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 39 12 13 50 26 42 7 23 34 10 28 32 26 25 38 25 34 56 60 54 63 46 33 45 33 36 F M F M M M M F M F M M M M F F F F M F M M M M F F LBp LBa LBp MBp LBp MBa LBp MBa D MBp LBp MBp LBp LBp MBa LBp MBa LBp D LBp LBp MBp LBp MBa LBp P L L R L L L R R R R L L L L L L L R L R L L L R L R 2 3 1 3 2 2 1 2 2 3 3 3 1 2 3 2 3 3 3 2 3 3 2 2 1 1 + + + + + + + + + + + + + + + + + + + + + + VFD PO VFD Hemianopia Hemianopia Quadrantanopia Hemianopia Hemianopia Quadrantanopia Hemianopia Hemianopia Quadrantanopia Hemianopia Quadrantanopia Hemianopia Quadrantanopia Hemianopia Hemianopia Quadrantanopia Improved Hemianopia Hemianopia Hemianopia Hemianopia Hemianopia Hemianopia Improved Quadrantanopia Hemianopia Hemianopia Quadrantanopia Quadrantanopia Quadrantanopia Hemianopia Hemianopia Quadrantanopia Hemianopia Hemianopia S PO S E scale + II Emboli + + + + + + + + + + + + + + + + I I I I I I II IV IV I I + + + + a indicates anterior; B, hemorrhage; D, dorsal; Emboli, preoperative embolization; E scale, Engel Seizure Outcome Scale; F, female; L, left; LB, laterobasal; M, male; MB, mediobasal; P, polar; p, posterior; PO, postoperative; R, right; S, seizure; S-M, Spetzler-Martin grade; VFD, visual field defect. 52 S. Nagata et al. / Surgical Neurology 66 (2006) 50 – 55 Fig. 1. Case 16: A, B: Preoperative left carotid angiography showed a temporal lobe AVM. C: Brain CT just after the embolization showed a massive hematoma in the left temporal lobe. after the total nidal excision that had been recorded from the mediobasal temporal lobe. In the other two patients, each AVM was located in mediobasal and laterobasal sites; psychomotor seizures remained after the total excision of AVM. Intraoperative ECoG monitoring was not performed in these two patients. They refused to undergo a second operation to remove the remote seizure focus. Because of the bizarre behaviors during psychomotor seizures, they were obliged to quit their original jobs. 4. Representative cases 4.1. Case 16 A 25-year-old woman suddenly had seizure with loss of consciousness for the first time. A couple of days later, she visited a nearby hospital. Brain CT showed a small hematoma in the left temporal lobe. She did not show any neurological deficits. Cerebral angiography revealed a left temporal lobe AVM of less than 3 cm in diameter (Fig. 1A and B). She planed to undergo radiosurgery after the embolization. She was hospitalized in the Department of Interventional Neuroradiology for the embolization of the AVM. During the embolization, she suddenly complained of headache and gradually lost consciousness. Brain CT showed a massive temporal lobe hematoma (Fig. 1C). She became comatose after CT with dilated fixed pupil on the left. She was immediately transferred to the operating room as an ultra-emergency case. The AVM was totally removed as well as the hematoma (Fig. 2). She recovered well from anesthesia without any neurological deficits except a complete right homonymous hemianopia. She showed a gradual improvement of the hemianopia during the hospital stay (Fig. 3) and finally got full visual field. Fig. 2. Case 16: The postoperative left carotid angiograms prove the complete total removal of the AVM. S. Nagata et al. / Surgical Neurology 66 (2006) 50 – 55 53 Fig. 3. Case 16: The scotometry 1 week after surgery showed an improvement of the visual field from the right hemianopia to the upper quadrantanopia. The patient finally had a full visual field in 1 month. Fig. 4. Case 23: A, B: Preoperative left carotid angiograms show a laterobasal temporal lobe AVM. Main feeding arteries arise from the temporal branches of the middle cerebral artery. Association of the feeders from the posterior cerebral artery is also observed. C, D: Postoperative angiograms prove the complete total removal of AVM. The anterior cerebral artery is well visualized after surgery. 54 S. Nagata et al. / Surgical Neurology 66 (2006) 50 – 55 4.2. Case 23 A 33-year-old man had seizure with loss of consciousness for the first time when he was working as chief of a gas station. He consulted a nearby hospital and magnetic resonance imaging showed an AVM in the left temporal lobe (Fig. 4A and B). He was referred to us for surgery. He did not show any neurological deficits. Electroencephalography before surgery was normal. After the partial embolization, the AVM was totally removed (Fig. 4C and D). The postoperative course was uneventful and he was discharged from the hospital without any neurological deficits and seizures. He immediately returned to his original job. After a while, he began to have intractable psychomotor seizures. Because of bizarre behaviors such as dashing water on customers, he was dismissed from the job. 5. Discussion Temporal lobe AVM accounts for 12% to 16% of intracranial AVMs [1,2,8,13]. In our series, it was 15.3%. Seizures have been reported to be more common presenting symptoms of temporal lobe AVM patients, occurring in 46% of patients, compared with the incidence of 24% in nontemporal AVM patients [8]. In our series, history of epilepsy was obtained in 14 (31.1%) of 45 cases. The incidence of epilepsy as an initial symptom at diagnosis was 7 (15.6%). About half of the nonruptured temporal lobe AVM patients with epilepsy were passed over for the diagnosis until bleedings occurred. The visual field deficits were observed in 17 cases and were improved in 2, unchanged in 12, and worsened in 3 after surgery. Korosue and Heros [7] have reported similar results as improved in 4.4%, unchanged in 77.8%, and worsened in 17.8%. It must be difficult to expect surgery to improved the visual field deficit. In two cases where the visual field deficit improved after surgery, the common factor was the ultra-emergency evacuation of hematoma and removal of AVM. One case had massive intraparenchymal hematoma with transtentorial herniation sign during the embolization procedure. Ultra-emergency craniotomy had been necessary to save the patient’s life. Iwata et al [6] have described that the interval between the hemorrhage and emergency craniotomy was the significant factor for the favorable outcome in such situations. This might be also true for the outcome of the visual field deficit. However, emergency craniotomy for intracranial AVM only for improvement of visual field could be dangerous without enough information of AVM. Visual improvement might be a fortunate by-product of appropriate neurosurgical management for global neurological deterioration. Three cases with quadrantanopia became hemianopia after surgery. These AVMs were all located at the posterior mediobasal sites. Strong retraction of the temporal lobe to expose the AVM at the parahippocampal gyrus through the subtemporal approach might be the cause of hemi- anopia. The posterior mediobasal location must be a risk factor for the homonymous hemianopia. Although the orbitozygomatic [3] or transinsular [11] approach may reduce the risk of the visual field deficit, gamma knife radiosurgery is the other choice of the treatment for the posterior mediobasal AVMs. Four nonruptured temporal lobe AVM patients and 7 ruptured AVM patients had history of epilepsy and underwent surgery for AVM. Although the laterobasal locations were the most common site of ruptured AVMs with seizure, they were distributed to all subclassified sites. A history of a preoperative seizure has been reported to associate with an increased likelihood of postoperative seizure [10,12]. The risk of persistent epilepsy after surgery is small in those patients who did not have seizure preoperatively. Korosue and Heros [7] had reported that epilepsy would improve in 54.6%, unchanged in 32.7%, and worse in 12.7%. Hoh et al [5] have reported that the association of seizures with intracranial hemorrhage was statistically associated with Engel class I outcomes. Preoperative epilepsy has a strong tendency to improve after excision of the ruptured AVM. None of the 7 patients with ruptured AVM who had history of epilepsy have intractable seizures. Intraoperative ECoG monitoring is not always necessary for the patients with history of epilepsy, particularly for patients with hemorrhages. Among 11 cases with epilepsy in our series, 7 with hemorrhage did not have postoperative intractable seizures. Surgical excision of the epileptic focus should be considered for patients with intractable epilepsy, unless the patient will drop out from social activity because of the bizarre behaviors due to the psychomotor seizures. In our two nonruptured cases, paroxysms recorded from the medial surface of the temporal lobe disappeared after the nidus resection in the laterobasal and polar locations [9]. In these cases, epilepsy disappeared after surgery without anticonvulsant. However, persistent intractable psychomotor seizures remained in the other two patients. Yeh et al [14,15] have described that the remote epileptic activity is particularly prominent in AVMs in the temporal lobe and usually involved medial temporal structures [14,16]. The intraoperative ECoG monitoring should have shown the necessity of the additional excisions of the epileptic focus. When the paroxysms are recorded from the medial temporal ECoG after the complete removal of AVM, the secondary epileptogenetic focus probably due to the focal brain ischemia should be removed. 6. Conclusion Improvement of visual field deficit due to hematoma is difficult in most cases with ruptured temporal lobe AVM. Deterioration due to the operative procedure should be avoided. Although the outcome of seizure associated with hemorrhage is acceptable, the postoperative intractable seizures can remain after the complete total excision of AVMs particularly in cases with seizure without S. Nagata et al. / Surgical Neurology 66 (2006) 50 – 55 hemorrhage. Intraoperative ECoG monitoring might be requisite for nonruptured AVM patients with seizure. References [1] Brown RD, Weibers DO, Forbes G, et al. The natural history of unruptured intracranial arteriovenous malformations. J Neurosurg 1988;68:352 - 7. [2] Drake CG. Cerebral arteriovenous malformations: considerations for and experience with surgical treatment in 166 cases. Clin Neurosurg 1979;26:145 - 208. [3] Du R, Young WL, Lawton MT. bTangentialQ resection of medial temporal lobe arteriovenous malformations with the orbitozygomatic approach. Neurosurgery 2004;54:645 - 52. [4] Engel J, Van Ness PC, Rasmussen TB, et al. Outcome with respect to epileptic seizures. In: Engel J, editor. Surgical treatment of the epilepsies. 2nd ed. New York, NY: Raven Press; 1993. p. 609 - 21. [5] Hoh BL, Chapman PH, Loeffler JS, et al. Results of multimodality treatment for 141 patients with brain arteriovenous malformations and seizures: factors associated with seizure incidence and seizure outcomes. Neurosurgery 2002;51:303 - 11. [6] Iwata T, Yoshimura K, Keller E, et al. Emergency craniotomy for intraparenchymal massive hematoma after embolization of supratentorial arteriovenous malformations. Neurosurgery 2003;53: 1251 - 60. [7] Korosue K, Heros RC. Complications of complete surgical resection of AVMs of the brain. In: Barrow DL, editor. Intracranial vascular malformations. Park Ridge, IL; American Association of Neurological Surgeons; 1990. p. 157 - 68. [8] Malik GM, Seyfried DM, Morgan JK. Temporal lobe arteriovenous malformations: surgical management and outcome. Surg Neurol 1996;46:106 - 15. [9] Morioka T, Nishio S, Hisada K, et al. Disappearance of paroxysm in the medial temporal lobe following blesionectomyQ of an arteriovenous malformation in the lateral temporal lobe. Rinsho noha 1996;38:196 - 200. [10] Piepgras DG, Sundt Jr TM, Ragoowansi AT, et al. Seizure outcome in patients with surgically treated cerebral arteriovenous malformations. J Neurosurg 1993;78:5 - 11. [11] Taniguchi T, Aoki N, Mizutani H. bTransinsular approachQ for the treatment of a medial temporal arteriovenous malformation. Neurosurgery 1993;32:863 - 6. [12] Thorpe ML, Cordato DJ, Morgan MK, et al. Postoperative seizure outcome in a series of 114 patients with supratentorial arteriovenous malformations. J Clin Neurosci 2002;7:107 - 11. 55 [13] Yasargil MG. Microneurosurgery, vol. IIIB. Stuttgart New York7 Georg Thieme Verlag; 1998. p. 112 - 51. [14] Yeh HS, Kashiwagi S, Tew Jr JM, et al. Surgical management of epilepsy associated with cerebral arteriovenous malformations. J Neurosurg 1990;72:216 - 23. [15] Yeh HS, Privitera MD. Secondary epileptogenesis in cerebral arteriovenous malformations. Arch Neurol 1991;48:1122 - 4. [16] Yeh HS, Tew Jr JM, Gartner M. Seizure control after surgery on cerebral arteriovenous malformations. J Neurosurg 1993;78:12 - 8. Commentary The authors’ well-documented clinical series makes two excellent points regarding the management of these common cerebral AVMs. The first is that visual field deficits are common in both hemorrhagic and nonhemorrhagic presentations, and that such deficits are often worsened by operative resection; furthermore, even in patients without preexisting field deficits, surgical resection carries a significant risk to the visual pathways. Although the authors do not specifically comment on the use of embolization in the management of these lesions, my own experience suggests that preliminary embolization significantly improves operative morbidity as regards visual field deficits. The second important observation is that when dealing with patients whose presentation is one of nonhemorrhagic seizures, resection of the AVM alone may yield disappointing results in terms of seizure control. Those patients’ optimal treatment paradigm combines a lesionectomy with resection of the EEG-identified epileptic focus. This appears to be especially true when the mesial temporal structures are involved in seizure generation. Duke S. Samson, MD Department of Neurosurgery UT Southwestern Medical Center Dallas, TX 75235, USA