Surgical Treatment of Childhood Moyamoya Disease -Comparison of Reconstructive Surgery Centered on the Frontal Region and the Parietal Region-

Transcription

1 Surgical Treatment of Childhood Moyamoya Disease -Comparison of Reconstructive Surgery Centered on the Frontal Region and the Parietal Region- Akihiro TAKAHASHI, Hiroyasu KAMIYAMA, Kiyohiro HOUKIN, and Hiroshi ABE Department of Neurosurgery, University of Hokkaido School of Medicine, Sapporo Abstract Indirect revascularization procedures centered on the parietal region, such as encephalo-myo-arterio synangiosis (parietal synangiosis) and direct procedures centered on the frontal - region using both the anterior and the posterior branches of the superficial temporal artery (STA), such as STA to middle cerebral artery anastomosis combined with encephalo-duro-arterio-myo-synangiosis (frontal anastomosis) were compared in childhood moyamoya disease patients. The parietal synangiosis group consisted of 10 sides in five patients, and the frontal anastomosis group consisted of 30 sides in 15 pa tients. The development of postoperative collateral circulation was assessed by external carotid angiography, the neurological outcome was monitored for 2 years after surgery, and the intelligence quotient (IQ) was measured at least 6 months after surgery. Frontal anastomosis achieved superior results compared to the parietal synangiosis assessed by development of collateral circulation, in par ticular to the orbitofrontal artery, the prefrontal artery, and the precentral artery (p < 0.01), and reduction in the incidence of ischemic attacks, such as transient ischemic attacks (p < 0.05). The mean IQ in the frontal anastomosis group was higher than that in the parietal synangiosis group. Vascular reconstruction centered on the frontal region utilizing both the anterior and posterior branches of the STA is more efficacious than only synangiosis centered on the parietal region. Key words: moyamoya disease, pediatric surgery, superficial temporal artery-middle cerebral artery anastomosis, synangiosis, revascularization Introduction Childhood moyamoya disease is characterized by progressive arterial stenosis or occlusion of the ma jor intracranial cerebral arteries and by development of extensive collateral circulation, including a fine network of vessels at the base of the brain referred to as the "moyamoya vessels."","' The major symp tom is cerebral ischemia. The pathogenesis of this disease is unknown, so no basic treatment has been established. Therefore, the symptom of cerebral ischemia is treated by increasing the blood supply to the ischemic brain from the external carotid system using several types of vascular reconstructive surgery: superficial temporal artery-middle cerebral artery (STA-MCA) anastomosis,6) encephalo-myo synangiosis (EMS),') encephalo-duro-arterio-synan giosis (EDAS),11,16) encephalo-myo-arterio-synan giosis (EMAS),18-20) encephalo-duro-arterio-myo synangiosis (EDAMS),10) omentum implantation, 8) and combinations of these procedures. In childhood moyamoya disease cerebral cir culatory insufficiency occurs in the frontal lobe. The regional cerebral blood flow (rcbf) of the frontal lobe is reduced") and the frontal rcbf decreases as a result of hyperventilation,') because the MCA and anterior cerebral artery are affected at practical ly the same time. Therefore, resolution of this fron tal lobe ischemia is a very important target of sur gical revascularization.4,24) However, most indirect Received May 20, 1994; Accepted August 9, 1994 Author's present address: A. Takahashi, M.D., Department of Neurological Surgery, Wayne State University School of Medicine, Detroit, Michigan, U.S.A.

2 bypass procedures such as EDAS, EMS, and EMAS achieve vascular reconstruction centered on the parietal region. The anterior branch of the STA is not used because it travels through the forehead area. Only the posterior branch is used, and the craniotomy is centered on the parietal region. Fol low-up angiography studies",",", 18,19) show that the collateral vascular channels linked to the ex ternal carotid system are localized in the operative field. The collateral circulation flows through the cortical branches of the MCA peripherally, but proximal flow is retrograde, because moyamoya disease is characterized by multiple stenotic and/or occlusive lesions of the cortical arteries.') This con trasts with the findings in atherosclerotic steno-oc clusive disease. Therefore, circulation to the central parietal-temporal area improves angiographically, but collateral formation to the prefrontal area is in evitably inadequate. Another problem is that some cases are refractory to indirect procedures.',',",") To solve these problems, we have created anastomoses between the anterior branch of the STA and cortical branch of the MCA in the frontal region since Here, we compare the postoperative angiographic and clinical findings of conventional synangiosis centered on the parietal region and anastomosis with synangiosis centered on the frontal region. Materials and Methods Twenty children with moyamoya disease were surgically treated in the Department of Neurosur gery, Hokkaido University Hospital. A total of 40 cerebral hemisphere operations were performed. From 1982 to 1985, only synangiosis was per formed centered on the parietal region. Only the posterior branch of the STA was used, since the anterior branch often naturally forms a collateral channel. A horseshoe-shaped scalp incision was made around the posterior branch of the STA above the ear, and a wide craniotomy performed in the parietal region. The middle meningeal artery (MMA) was preserved as extensively as possible. The posterior branch of the STA was laid on the exposed cortex and then covered with temporal muscle. When the MMA was preserved, an EDAMS procedure was used (Fig. 1 left), and when damaged, EMAS.20) When the posterior branch of the STA could not be used, the cortex was covered with muscle alone (EMS). From 1986 to 1991, anastomosis combined with synangiosis centered on the frontal region was per formed to "use all vessels capable of providing future collateral circulation." ",'9) Both the anterior and posterior branches of the STA were utilized. The scalp incision was made directly above the posterior branch of the STA, and from the midway point the scalp incision was extended anteriorly away from the posterior branch of the STA, to reduce ischemia in the scalp. Both the anterior and posterior branches of the STA were adequately dissected, but not cut. Any lateral branch of the STA was also sufficiently dissected and then cut. A burr hole was made over or near the arterial sulcus of the MMA, and after con firming the MMA, a wide craniotomy centered on the frontal lobe was made, preserving the vessel. A window-like incision was made in the dura mater without cutting the MMA. If STA-MCA end-to-side anastomosis in the frontal lobe was possible with the lateral branch of the anterior branch of the STA, the anterior branch of the STA was laid on the exposed cortex. When the lateral branch of the anterior branch could not be used, the anterior branch was cut peripherally, and after adequate drawing over the exposed cortex, the cut end of the anterior branch was anastomosed to one of the frontal branches of the MCA. If a preoperative CBF study revealed another low-perfusion area other than the frontal lobe, the posterior branch of the STA or its lateral branch was also anastomosed to the cor tical branch of the MCA in the low-perfusion area. The largest possible area of the exposed cortex was covered with temporal muscle (STA-MCA anas tomosis with EDAMS) (Fig. 1 right). If the MMA was damaged during the craniotomy, EMAS was performed for indirect revascularization. When both the anterior and the posterior branches of the STA were anastomosed, encephalo-duro-myo-synangio sis (EDMS) or EMS was performed for indirect re vascularization. EDAMS was performed on two sides in one pa tient, EMAS on six sides in three patients, and EMS on two sides in one patient. STA-MCA anastomosis was performed on 25 sides in 15 patients with EDAMS, on three sides in three patients with EMAS, on one side in one patient with EDMS, and on one side in one patient with EMS. The 40 sides were divided into two groups accord ing to the surgical procedure; a "parietal synangiosis group" of 10 sides in five patients (EDAMS, EMAS, and EMS), and a "frontal anastomosis group" of 30 sides in 15 patients (STA-MCA anastomosis com bined with either EDAMS, EMAS, EDMS, or EMS). Table 1 shows the age on admission, age at onset, type of moyamoya disease, and preoperative angiographic stage according to Suzuki and Takaku22> for each group. Differences in distribution

3 Table 1 Clinical features of patients *Values are means } SD. **According to Suzuki and Takaku.22> NS: not significant, transient is chemic attack. Fig. 1 Illustrations showing surgical procedures, left: EDAMS and right: STA-MCA anastomosis with EDAMS. A: Skin incision and crani otomy. B (left): The MMA and posterior branch of the STA are dissected and laid on t he exposed cortex. B (right): The MMA and both anterior and posterior branches of the STA are dissected and laid over the cortex, with the lateral branches of the STA end-to side anastomosed to the cortical branch - of the MCA. C: The exposed cortex is covered with temporal muscle (TM). D: Cranioplasty. The arrow indicates the end-to-side anastomosis be tween the STA and the cortical branch of the MCA. BC: bone chip, BF: bone flap, DM: dura mater, SSV: superficial Sylvian vein. were measured by the X2 test, and differences between the mean values were assessed using a Student's t test. There were no significant differences between the two groups. Collateral formation was reviewed using external carotid angiograms taken 6-12 months after surgery. A transparent plastic template 21) was superimposed on the lateral view of the postoperative external carotid angiograms. The degree of opacification in each cortical branch of the MCA was evaluated for each patient using a four-step scale according to the area of the MCA cortical branch distribution sup plied: 3 points, when two-thirds or more of the area was covered; 2 points, when between two-thirds and one-third was covered; 1 point, when less than one third was covered; and 0 points, when no perfusion was observed. Complications occurring within 1 month of surgery were assessed on the basis of medical records. Postoperative follow-up examinations were con tinued for 2 years after bypass surgery. The neurological outcome was evaluated according to the frequency of ischemic attacks during the year follow ing surgery: excellent when cerebral ischemic attacks had stopped, and good when the frequency of cerebral ischemic attacks had diminished. The surgical effects were evaluated on the 40 operated sides based on the symptoms. The intelligence quotients (IQs) of five patients in the parietal synangiosis group and 11 patients in the frontal anastomosis group were studied using Wechsler Intelligence Scale for Children-Revised 21) at least 6 months after surgery, to investigate the relationship between the operative method and the IQ. The classification of the IQ was: normal (IQ >_

4 80), borderline (IQ 70-79), and mental retardation (IQ < 69). Since the IQs of patients who underwent surgery before 1987 and that of patients under 5 years old were studied using other tests, pre and postoperative IQs could not be compared. Results I. Angiographic results The parietal synangiosis group developed significantly less collateral formation to the or bitofrontal artery, the prefrontal artery, and the Fig. 2 Postoperative external carotid angiograms in patients with only synangiosis centered on the parietal region, the best result with EDAMS (left), and worst with EMS (right). precentral artery (p < 0.01; Student's t-test), which was predominantly from the central artery to cortical arteries of the MCA in the parietal and the temporal lobes (Fig. 2, Table 2). In contrast, the frontal anastomosis group had very good collateral forma tion to the orbitofrontal, prefrontal, and precentral arteries, and was better than in the parietal synangiosis group even in the parietal and temporal lobes (Fig. 3, Table 2). However, even though the cortex was included in the operative fields in both groups, there were also areas where neovasculariza tion did not occur as a result of indirect revasculariza tion (Figs. 2 and 3). Overall examination of the angiograms in both groups suggested that a wider craniotomy generally resulted in better neovascu larization from the MMA and the deep temporal artery. II. Complications In the parietal synangiosis group, two sides (20%) in two patients exhibited ischemic symptoms either immediately postoperatively or in the early postoperative period. All symptoms were transient. In the frontal group, no patient showed deteriorating postoperative symptoms either immediately post operatively or in the early postoperative period, al though two sides (7%) in two patients developed in farcts after crying, both in the latter half of the sec ond week. The incidence of complications in the fron Table 2 Angiographic comparison of collateral formation Opacification of territories supplied by MCA branches was assessed on external carotid angiograms. Opacification score shows area of branch distribution visualized: 3, 2/3 or more; 2, 2/3-1/3; 1, less than 1/3; 0, no perfusion. A: angular artery, AP: anterior parietal artery, AT: anterior temporal artery, C: central artery, MT: middle temporal artery, NS: not significant, OF: orbitofrontal artery, PC: precentral artery, PF: prefrontal artery, PP: posterior parietal artery, PT: posterior temporal artery, TO: temporo-occipital artery, TP: temporo-polar artery.

5 tal anastomosis group was significantly better than in the parietal synangiosis group (p < 0.05; x2 test). IV. IQ In the parietal synangiosis group, two patients (40%) had a normal IQ, one patient (20%) was borderline, and two patients (40%) were mentally retarded. In the frontal anastomosis group, seven pa tients (64%) had a normal IQ, two patients (18%) were borderline, and two patients were mentally retarded. The parietal synangiosis group had a mean IQ of 77.4 } 19.8, while the frontal anastomosis group had a mean IQ of 87.1 } 20.1 (Fig. 4). There was no significant difference. Discussion Fig. 4 Postoperative IQ classified by the Wechsler In telligence Scale for Children-Revised.") No significant difference between groups was ob served. IQ classified into normal, >_ 80; bor derline, 70-79; mental retardation, :569. The vertical bar indicates the mean IQ score and SD. tal anastomosis group was lower than in the parietal synangiosis group, but there was no significant difference.fig. 3 Postoperative patients with (right). STA-MCA with external STA-MCA anastomosis carotid anastomosis with EDAMS angiograms (left), with and EDAMS worst combined with synangiosis centered on the frontal cannot region, the best provide result with STA-MCA immediate revascularization, and are III. Clinical results In the parietal synangiosis group, six sides (60%) had excellent outcome, while in the frontal anastomosis group, 28 sides (93%) were excellent. The postoperative neurological outcome in the fron The present study has shown that development of col lateral circulation, especially in the prefrontal area, is better in patients treated by frontal anastomosis than by parietal synangiosis. Subsequent intellectual development apparently reflects this superior col lateral circulation. Three-dimensional hemodynamic changes after vascular reconstruction in childhood moyamoya disease have been investigated using xenon-133 ('33Xe)-inhalation single photon emission computed tomography (SPECT).24) Inadequate rcbf was ob served in the frontal lobe after parietal synangiosis, but the low-perfusion area disappeared after frontal anastomosis. Isobe et al.4) also examined rcbf after hyperventilation using 133Xe-inhalation SPECT, and reported that rcbf in the frontal lobe after hyperven tilation in patients treated by parietal synangiosis was markedly reduced compared with rcbf in other regions and rcbf in patients treated by frontal anastomosis. These findings show that frontal lobes which are not revascularized remain in a state of cir culatory insufficiency. Therefore, the revasculariza tion procedure, i. e. craniotomy, should be extended to the frontal region. In addition, since the STA is the best blood supply source in the external carotid system,') the anterior branch of the STA should also be utilized whenever possible. Another question is whether STA-MCA anastomosis should be performed. Miyamoto et al." suggested that non-anastomotic bypass procedures not suitable for patients with frequent transient ischemic attacks. In addition, the occurrence of moyamoya disease refractory to indirect procedures has been emphasized. 2,4,'3.") In our study, immediate postoperative aggravation was less in the frontal anastomosis group in which STA-MCA anastomosis

6 was performed, and we have also experienced a case refractory to the indirect procedure which needed another anastomosis procedure." Matsushima et al. "I compared surgical results arteries. after indirect and direct revascularization pro cedures, and found that STA-MCA anastomosis with EMS in the parietal region was superior to EDAS in the parietal region in both development of collateral circulation and postoperative clinical im provement, as in the present study. They concluded that direct anastomosis should be the treatment of first choice. We would like to emphasize that the MCA cortical branch of the anastomosis should be selected on the basis of the preoperative CBF study. However, Karasawa et al.,') based on follow-up caroti,d angiography, found that the STA-MCA anastomosis became increasingly obstructed even though the cortical branches of the MCA were opacified through the fine network produced in the region of the anastomosis. Therefore, both the anterior and posterior branches of the STA should be utilized as synangioses to extend the potential area for neovascularization as widely as possible over the ischemic brain surface. If possible, the side branches of the STA should be anastomosed to frontal cor tical branches of the MCA and other cortical branches in the low-perfusion area. When the lateral branch of the STA cannot be used, the an terior or the posterior branch of the STA should be cut peripherally, and after drawing over the ex posed cortex, the cut end of the STA should be anastomosed to the MCA. In addition, all vessels that might serve as future collateral channels, in cluding the MMA and the deep temporal artery, should be utilized. Some authors have claimed that STA-MCA anastomosis procedures have some considerable disadvantages: the anastomosis is difficult to per form, and some patients deteriorate after anastomosis because of the temporary stasis of blood needed in the cortical artery for direct anastomosis.11,20,261 An STA-MCA anastomosis is difficult to perform in patients with childhood moyamoya disease because the arteries anastomosed are very small, with an outer diameter ranging from 0.5 to 1.0mm, 6) and ar e thin and fragile like veins.") In addition, the ostia of arteriotomies become invisi ble because of their transparency after the blood in the lumen flows out. Staining of the ostium of an 1978 arteriotomy with a pyoktanine blue solution can pro vide visualization,') allowing rapid identification of the ostium and performance of the correct anastomosis. In addition, we select the lateral branch 1977 of the anterior or the posterior branch of the STA as the donor arterial vessel instead of the anterior or the posterior branch itself, because the lateral branch has a suitable caliber matching the small cortical We consider that the combination of a large craniotomy centered on the frontal lobe, utilizing all vessels capable of providing future collateral circula tion, and creating STA-MCA anastomoses in the low-perfusion area whenever possible is the most effective treatment for the childhood moyamoya disease. However, even this procedure cannot pre vent ischemia in the posterior or anterior cerebral artery territories in some patients, so further studies must investigate the need for more extensive vascular reconstruction, such as omentum implantation.) Acknowledgments This work was partly supported by a research grant on spontaneous occlusion of the circle of Willis (moyamoya disease) from the Ministry of Health and Welfare of Japan. References 1) Abe H, Kamiyama H: Surgical treatment of the occlu sion of the circle of Willis, in: Annual Report of the Research Committee on Spontaneous Occlusion of the Circle of Willis of the Ministry of Health and Welfare Japan. 1987, pp (in Japanese) 2) Cahan LD: Failure of encephalo-duro-arterio synangiosis. Pediatr Neurosci 12: 58-62, ) Gotoh F, Fukuuchi Y, Takashima S, Kawamura J, Terayama Y: Local cerebral blood flow and CO2 responsiveness in patients with gmoyamoya disease, hin: Annual Report of the Research Com mittee on Spontaneous Occlusion of the Circle of Willis of the Ministry of Health and Welfare Japan. 1987, pp (in Japanese) 4) Isobe M, Kuroda S, Kamiyama H, Abe H, Mitumori K: Cerebral blood flow reactivity to hyperventilation in children with spontaneous occlusion of the circle of Willis (moyamoya disease). No Shinkei Geka 20: , 1992 (in Japanese) 5) Kamiyama H, Takahashi A, Houkin K, Mabuchi S, Abe H: Visualization of the ostium of an arteriotomy in bypass surgery. Neurosurgery 33: , ) Karasawa J, Kikuchi H, Furuse S, Kawamura J, Sakaki T: Treatment of moyamoya disease with STA-MCA anastomosis. J Neurosurg 49: , 7) Karasawa J, Kikuchi H, Furuse S, Sasaki T, Yoshida Y, Ohnishi H, Taki W: A surgical treatment of g moyamoya hdisease gencephalo-myo synangio sis. hneurol Med Chir (Tokyo) 17 [Part I]: 29-37, 8) Karasawa J, Kikuchi H, Kawamura J, Sakaki T: In

7 t racranial transplantation of the omentum for cerebrovascular moyamoya disease: A two-year follow-up study. Surg Neurol 14: , ) Karasawa J, Kikuchi H, Kobayashi K, Takahashi N, Mitsugi T: Evaluation of angiographical changes after ST-MC anastomosis in gmoyamoya hdisease, in: Proceedings of the 10th Japanese Conference on Surgery of Cerebral Stroke. Tokyo, Nyuron-sha, 1981, pp (in Japanese) 10) Kinugasa K, Mandai S, Kamata I, Sugiu K, Ohmoto T: Surgical treatment of moyamoya disease: Operative technique for encephalo-duro-arterio myo-synangiosis: Its follow-up, - clinical results, and angiograms. Neurosurgery 32: , ) Kobayashi K, Takeuchi S, Tsuchida T, Ito J: Encephalo-myo-synangiosis (EMS) in moyamoya disease: With special reference to postoperative angiography. Neurol Med Chir (Tokyo) 21: , 1981 (in Japanese) - 12) Kudo T: Spontaneous occlusion of the circle of Willis. A disease apparently confined to Japanese. Neurology (Minneap) 18: , ) Matsushima T, Fujiwara S, Nagata S, Fujii K, Fukui M, Hasuo K: Reoperation for moyamoya disease refractory to encephalo-duro-arterio-synangiosis. Acta Neurochir (Wien) 107: , ) Matsushima T, Inoue T, Suzuki SO, Fujii K, Fukui M, Hasuo K: Surgical treatment of moyamoya disease in pediatric patients: Comparison between the results of indirect and direct revascularization pro cedures. Neurosurgery 31: , ) Matsushima Y, Fukai N, Tanaka K, Tsuruoka K, Inaba Y, Aoyagi M, Ohno K: A new surgical treat ment of moyamoya disease in children. A prelimi nary report. Surg Neurol 15: , ) Matsushima Y, Inaba Y: Moyamoya disease in children and its surgical treatment. Introduction of a new surgical procedure and its follow-up angiograms. Childs Brain 11: , ) Miyamoto S, Kikuchi H, Karasawa J, Nagai I, Yamazoe N: Pitfalls in the surgical treatment of moyamoya disease. Operative techniques for refrac tory cases. J Neurosurg 68: , ) Nakagawa Y, Gotoh S, Shimoyama M, Ohtsuka K, Address reprint requests to: K. Houkin, M.D., Depart Mabuchi S, Sawamura Y, Abe H, Tsuru M: ment of Neurosurgery, University of Hokkaido Reconstructive operation for moyamoya disease: School of Medicine, North-15, West-7, Kita-ku, Sap Surgical indication for the hemorrhagic type, and poro 060, Japan. preferable operative methods. Neurol Med Chir (Tokyo) Takei Suzuki moyamoya Proceedings Surgery Y, Ohtsuka Anatomy pp base Fukumoto Tomographic Xenon-enhanced Med reconstructive disease tomography. (trans-eds): Kagaku-sha, transplanted cerebral 1981, ) ) Children-Revised, artery, ) Suzuki Wechsler Gotoh pediatric of Chir pp Tsuru Suzuki H, Y: in J, R, brain. ischemia. 23: using of K, M: of Takaku D; Salamon Indication Shinkei Tsuruoka (Tokyo) Sugimoto T, disease Japanese S, Abe Geka Hemodynamic , Cerebral the of Neurol 1982, Kodama 9: Inaba omentum. cerebral Arch patients Miyasaka surgery Brain. single the , H, computerized (in A: Neurosurgery 25: and pp G, ed Y, Tsuru S, Neurol Med 10th of Japanese)20) Cerebrovascular Wechsler 1983 H, Stroke. Berlin, Disease ) , Ohno blood photon Huang Hiratsuka H, ideal reconstructive showing Nakagawa 7. for with Yonekawa abnormal A Goto Chir Y, 1981 Japanese K, M: Shinagawa (in Y, Shimoyama evaluation Tokyo, Yasargil (Chicago) possible operative K: childhood flow MG: M, Springer-Verlag, gmoyamoya hdisease. Reconstructive Japanese)19) Takei Tokyo, YP Brain (Tokyo) Intelligence Kashiwaba (in tomography. emission Cerebral S, : Nakagawa H, vascularization Japanese)21) (eds): map net-like Y, , Shimoyama Ikota Conference Matsushima Nihon (in H, Scialfa T, operation F, treatment methods, 20: Ohtsuka Nyuron-sha, 30: G: obtained of Japanese)24) Radiologic moyamoya K, circulation Gotoh Anatomic Motegi T, computed Abe vessels , Scale , study Takikawa vascular H, disease: Neurol 1977 Suzuki Tsuru Bunka operation of S, 1976, Kamiyama of the for M, H, in: on Y, by of S, in M moyamoya middle H, Abe cerebral diseases H, and Mitsumori K, problems. No