Quantitative H2[15O]-PET in Pediatric Moyamoya Disease: Evaluating Perfusion before and after Cerebral Revascularization Felix P. Kuhn, MD,* Geoff Warnock, PhD,* Thomas Schweingruber, MD,† Michael Sommerauer, MD,* Alfred Buck, MD,* and Nadia Khan, MD†
Background: Moyamoya disease (MMD) is an idiopathic intracranial angiopathy with a progressive spontaneous occlusion of the circle of Willis resulting in repeated ischemia if not diagnosed and treated early, especially in children. Prevention of stroke is achieved by revascularization of the affected cerebral regions. Functional imaging techniques such as H2[15O]-Positron emission tomography (PET) allow quantification of cerebral perfusion/blood flow (CBF) and in particular cerebrovascular response after acetazolamide (AZA) challenge. The cerebrovascular reserve (CVR) can then be calculated and used to identify regions at risk of infarct, hence allowing surgery to be specifically targeted and personalized. Methods: Pediatric patients with diagnosed MMD underwent initial H2[15O]-PET scans at baseline and after stimulation with AZA. Indication for surgery was then based collectively on the extent of disease observed clinically and on magnetic resonance imaging, on the arterial territories involved, as seen in angiography and the respective regional CVR observed in PET. Cerebral revascularization surgeries were subsequently performed, tailored to the individual patient. Postoperative assessment of clinical outcome was augmented with follow-up PET (median duration after surgery, 10.4 months). CBF at baseline, after AZA and CVR were compared between presurgery and postsurgery scans in the areas supplied by the major cerebral arteries. Results: Parametric images reflecting CBF, response to AZA and CVR clearly showed deficits in cortical but not subcortical regions or cerebellum. AZA-CBF and CVR deficits were most clear in middle cerebral artery and anterior cerebral artery (ACA) regions. In addition to the clinical symptomatology, angiography, AZA-CBF, and CVR images allowed the laterality of deficits to be clearly visualized for tailored surgery and the indication for targeted ACA or posterior cerebral artery revascularization to be assessed. Comparison of baseline CBF, AZA-CBF, and CVR between presurgery and postsurgery scans in revascularized areas revealed a significant improvement in baseline and AZA-CBF after surgery. Although no significant differences in CVR after revascularization surgery were found, a clear improvement of the deficits apparent in AZA-CBF in revascularized regions was found. Conclusions: We demonstrate that quantitative H2[15O]-PET is a highly useful tool to direct surgical intervention in MMD. Detailed quantitative analysis of CBF changes and CVR after
From the *Department of Nuclear Medicine, University Hospital Zurich; and †Division of Pediatric Neurosurgery, Department of Surgery, Moyamoya Center, University Children’s Hospital Zurich, Zurich, Switzerland. Received December 1, 2014; accepted December 13, 2014. F.P.K. and G.W. contributed equally to this work. Research funds of the Nuclear Medicine Department of the University Hospital Zurich and of the Moyamoya Center, Division of Pediatric Neurosurgery, University Children’s Hospital Zurich.
The authors declare no other conflict of interest. Address correspondence to Felix P. Kuhn, MD, Department of Nuclear Medicine, University Hospital Zurich, CH-8091 Zurich, Switzerland. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.12.017
Journal of Stroke and Cerebrovascular Diseases, Vol. 24, No. 5 (May), 2015: pp 965-971
965
F.P. KUHN ET AL.
966 surgery supports a targeted surgical approach. Key Words: Moyamoya disease— nuclear medicine—positron emission tomography—brain perfusion. Ó 2015 by National Stroke Association
Introduction Moyamoya disease (MMD) is an idiopathic intracranial angiopathy with progressive spontaneous occlusion of the circle of Willis1 resulting in repeated ischemia if not diagnosed and treated early, especially in children. Young patients often suffer from symptoms of hemodynamic insufficiency, such as repetitive transient ischemic attacks (TIAs) and ischemic strokes. Prevention of stroke is achieved by revascularization of the affected vascular territories.2,3 Children with MMD are referred to our MMD center internationally for neurosurgical management. Before cerebral revascularization, a detailed presurgical evaluation protocol is followed. This consists of an initial clinical–neurologic evaluation, magnetic resonance imaging (MRI) to delineate cortical or subcortical ischemia and/ or infarcts, 6-vessel cerebral angiography to demonstrate stenosis, the presence of MMD collaterals, and the involvement of particular arterial territories, (middle cerebral artery [MCA], anterior cerebral artery [ACA], and posterior cerebral artery [PCA]) and a hemodynamic evaluation using H2[15O]-Positron emission tomography (PET) at baseline and after acetazolamide (AZA) challenge, which is invaluable in assessing the regional cerebral baseline blood flow (CBF), cerebrovascular response to vasodilatory challenge, and the cerebrovascular reserve (CVR). Reliable quantification of CBF after vasodilation provoked by AZA is crucial to allow for targeted and personalized surgery, rather than routine revascularization of the bilateral MCA territories alone. In patients with impaired CBF, the response to AZA (AZA-CBF) is suppressed in poorly perfused regions. Typically, CBF at rest appears normal because of homeostatic regulation, but only unaffected regions have the capacity to further increase CBF on challenge. Parametric CVR images highlight this deficit and quantitative analysis of AZA-CBF in the major flow territories in comparison with an unaffected region (eg, cerebellum) can be used to assess the scale of such deficits and measure the improvement after revascularization. Patients found to have a poor CVR have been shown to be good candidates for revascularization surgery.4 A tailored revascularization approach is favored in our MMD center, making particular use of the H2[15O]-PET data to prioritize revascularization in those regions at greatest risk of ischemic stroke (poor CVR/limited response to AZA), including not only MCA but also ACA and PCA territories. Depending on the symptomatology, number of arterial territories
involved, the CBF, CBF response to AZA, and the CVR, cerebral revascularization is performed in 1 or more regions, unilaterally or bilaterally. Combinations of direct and indirect revascularization approaches are used (described in more detail below). The first clinical and neuroradiologic postoperative follow-up is then carried out at 6 months-1 year. In addition to its value in presurgical evaluation of hemodynamics, H2[15O]-PET can provide valuable complementary information for follow-up. The purpose of this study was to use the H2[15O]-PET data from a sample of pediatric patients to directly quantify the changes in CBF soon after tailored revascularization surgery. Generally, the success of revascularization surgery is determined only by the reduction of TIAs and ischemic strokes. We hypothesized that in addition to identifying those regions in greatest need of revascularization, quantitative H2[15O]-PET would provide a measurable indicator of improved CBF in response to revascularization.
Methods Patient Selection For this retrospective study, 20 children with angiographically proven MMD who underwent cerebral revascularization surgery at our institution were considered. MMD was diagnosed based on the clinical presentation, MRI findings, and typical angiographical hallmarks. To identify those regions at greatest risk of ischemic stroke and most likely to respond favorably to revascularization surgery, CBF was measured in each patient using H2[15O]-PET scans at baseline (baseline CBF) and after stimulation with AZA (AZA-CBF). Parametric images of CVR were then calculated as described in the following. Revascularization surgery was subsequently tailored to those cerebral vessel territories with CVR deficits, that is, cerebral revascularization was performed in all symptomatic areas showing limited CVR (unilateral or bilateral).
Tailor-Made Cerebral Revascularization (Types of Surgeries) Revascularization can be performed either by direct bypass surgery from the superficial temporal artery to the ACA (superficial temporal artery [SCA] - anterior cerebral artery [ACA] bypass) for revascularization of the frontal regions or MCA (STA-MCA bypass) for revascularization of the frontolateral, parietal, and temporal regions. In cases where this is technically not
15 QUANTITATIVE H2[ O]-PET IN PEDIATRIC MMD
967
possible, indirect procedures such as encephaloduroarteriosynangiosis (EDAS), encephalomyosynangiosis, and encephalogaleaperiostsynangiosis are performed. The occipital artery can also be used for a direct anastomosis to the PCA or an indirect EDAS for revascularization of the occipital region.
Follow-up Standard postoperative follow-up including clinical, neurologic evaluation along with MRI scanning, and 6-vessel cerebral angiography was performed after revascularization surgery. This was augmented with follow-up PET (median duration after surgery, 10.4 months). Baseline CBF, AZA-CBF, and CVR were compared for presurgery and postsurgery scans. This study focused on CBF and CVR changes before and after targeted regional cerebral revascularization.
Exclusion Criteria Patients who suffered additional cerebral infarcts in the period between initial PET and surgery and patients with posterior circulation involvement and hence undergoing posterior revascularization were excluded.
Image Acquisition PET data were acquired on a full ring PET/CT scanner in 3D mode (PET/CT Discovery STE, GE Healthcare, Waukesha, MI) and corrected for attenuation and scatter by using the corresponding computed tomography (120 kV/80 mA) and manufacturer’s algorithms. The scanner’s axial field of view covered 15.3 cm. Images were reconstructed using a 3-dimensional Fourier rebinning filtered back projection algorithm resulting in a 128 3 128 3 47 matrix with 2.34 3 2.34 3 3.27 mm voxel spacing. In all, 300-800 MBq H2[15O] was administered intravenously using an automatic injection device. The tracer was delivered over 20 seconds. The emission data were acquired as a series of 18, 10-second frames. AZA was administrated 13 minutes before the second PET scan, with dose adjusted according to weight and injection over 2 minutes.
Image Processing Parametric images of CBF were generated using a previously reported method5 in which a standardized arterial input function and image scaling based on the washout rate k2 of H2[15O] are used to derive the CBF. This procedure, based on method by Alpert,6 exploits the fact that k2 is related to the shape and not the scale of the arterial input function and proportional to CBF (k2 5 K1/p, where p is the partition coefficient).7 After application of a 6-mm full width at half maximum Gaussian filter, parametric CVR images were calculated as the difference between AZA-CBF and base-
Figure 1. Cerebral angiography (A, anteroposterior and lateral ICA injections) and PET data (B) from a single patient. Angiography clearly revealed tight stenosis of the ICA-A1-M1 segments and MMD collateral vessels. At baseline (B, baseline), CBF appears close to normal, but after AZA challenge increases in CBF are only apparent in subcortical, PCA, and cerebellar regions. Calculation of CVR images highlights the lack of CVR in bilateral MCA and ACA territories. CVR T2 illustrates the overlay of the CVR image on T2-weighted MRI. The flow territory volume-of-interest definitions for quantitative analysis (after transformation to the individual anatomy) are illustrated on the bottom row, overlaid on the CVR image. These regions were based on blood supply from the cerebral arteries. Vascular territories: light/dark blue–ACA, pink/red–MCA, light/dark green–PCA, yellow– cerebellum. Watershed areas were divided equally between flow territories. Abbreviations: ACA, anterior cerebral artery, AZA, acetazolamide; CBF, cerebral blood flow; CVR, cerebrovascular reserve; MCA, middle cerebral artery; MRI, magnetic resonance imaging; PCA, posterior cerebral artery. Color version of the figure is available online.
line CBF (mL/minute/mL tissue). These images were used for the visual identification of those flow territories with impaired CVR before tailored revascularization surgery. To facilitate the comparison of flow territory CBF with that in the cerebellum, all images were scaled to their individual cerebellar mean CBF. Normalized values below 1 are indicative of impaired CBF. Cerebellar normalized CBF was calculated using an atlas-based
F.P. KUHN ET AL.
968
Table 1. Summary of preoperative clinical and neuroradiologic findings, AZA response and CVR deficits Baseline CBF (normalized)
AZA-CBF (normalized)
CVR (mL/min/mL)
Region
Presurgery
Postsurgery
Presurgery
Postsurgery
Presurgery
Postsurgery
ACA left ACA right MCA left MCA right PCA left PCA right
.785 6 .318 .917 6 .195 .791 6 .246 .892 6 .169 .974 6 .248 1.007 6 .139
.980 6 .251 (*) .990 6 .215 .936 6 .211* .972 6 .197 1.022 6 .204 1.083 6 .160
.604 6 .218 .618 6 .136 .624 6 .222 .633 6 .135 .855 6 .181 .948 6 .105
.707 6 .174* .727 6 .144*** .752 6 .178*** .757 6 .145* .890 6 .178 .993 6 .032
,.001 6 .032 2.015 6 .046 .005 6 .042 2.005 6 .044 .058 6 .058 .058 6 .074
.001 6 .026 2.007 6 .042 .039 6 .040 .022 6 .033 .080 6 .084 .090 6 .068
Abbreviations: ACA, anterior cerebral artery; AZA, acetazolamide; CBF, cerebral blood flow; CVR, cerebrovascular reserve; MCA, middle cerebral artery; PCA, posterior cerebral artery. (*) indicates trend (P approaching .05); *P , .05; ***P , .01.
volume-of-interest method (Fig 1). Cerebellar normalization of CBF images is also valuable for the comparison of presurgery and postsurgery PET in young patients, where global changes in CBF are likely. Regional assessment of CBF was performed after normalization of the PET data to the stereotaxic spatial array of the Montreal Neurological Institute brain template. Normalization to Montreal Neurological Institute space was performed using automated algorithms available in the PMOD software package (version 3.3; PMOD Technologies Ltd, Zurich, Switzerland), except in cases of extreme CBF pathology where manual approximation of the transformation was necessary. Quantitative CBF values were extracted using an atlas outlining the vascular territories. This vascular territory atlas was generated by attributing the predefined brain region volume-of-interests from the previously validated N30R83 atlas by Hammer et al8 to the major vascular territories (illustrated in Fig 1). Watershed regions were divided equally between territory definitions.
Statistical Analysis Differences in baseline CBF, AZA-CBF, and CVR were assessed in each flow territory in which surgery was performed. Changes in CBF and CVR between presurgey and postsurgery PET were assessed using the paired Student t test with Bonferroni correction for multiple comparisons (SAS, SAS software, version 9.2, SAS Institute Inc., Cary, NC).
Results Fourteen patients were assessed in the present study. Six patients were excluded from analysis because of infarcts between initial PET and surgery. Demographic and clinical data for the 14 patients included are shown in Table 1. In all patients, the parametric images reflecting CVR clearly highlighted deficits in cortical rather than subcortical regions or cerebellum. Perfusion deficits were most
evident in MCA and ACA regions. The PCA, superior cerebellar artery, posterior inferior cerebellar artery, and anterior inferior cerebellar artery territories showed the highest response to AZA and the highest CVR. Examples of AZA-CBF and CVR images for a representative patient are shown in Figure 1. Revascularization surgery was tailored to those flow territories with poor response to AZA, in this example, the bilateral MCA and bilateral ACA territories. Normalized baseline CBF, AZA-CBF, and CVR before and after revascularization surgery are summarized for ACA, MCA, and PCA territories in Table 2. Baseline CBF was significantly increased in the left MCA territory (presurgery CBF, .79 6 .25; postsurgery CBF, .94 6 .21; P 5 .014) and showed a trend toward increase in the left ACA territory (presurgery CBF, .79 6 .32; postsurgery CBF, .98 6 .25; P 5 .089). CBF in the PCA territories was not significantly altered. CBF after AZA challenge was significantly increased in all ACA and MCA territories (ACA left, presurgery CBF, .60 6 .22; postsurgery CBF, .71 6 .17; P 5 .034; ACA right, presurgery CBF, .62 6 .14; postsurgery CBF, .73 6 .14; P , .001; MCA left, presurgery CBF, .62 6 .22; postsurgery CBF, .75 6 .18; P 5 .001; MCA right, presurgery CBF, .63 6 .14; postsurgery CBF, .76 6 .15; P 5 .021) but not in the PCA territories. A generalized improvement in CBF was apparent after revascularization surgery, with a more prominent effect on AZA-CBF. Before revascularization, mean CBF in ACA and MCA territories was approximately 40% below that in the cerebellum (Table 1). At follow-up PET, this deficit was reduced to approximately 25%-30 %. No significant improvement in CVR after revascularization was found. Postoperatively, all patients were clinically symptomfree (ie, had no TIAs, strokes, or other neurologic symptoms or new deficits, Table 1) for the duration of follow-up. This was in good agreement with postoperative 6-vessel cerebral angiography, which indicated good distal filling in the respective revascularized recipient arterial territories.
Age, y
Gender
Preoperative symptomatology
Arterial territories involved on cerebral angiogram
1 2 3 4
11 4.5 1 1.7
F F F M
Headaches, bilateral sensorimotor TIA Headaches, left-sided sensorimotor TIA Seizure Left sensorimotor TIA
Bilateral ICA, ACA, MCA Bilateral ICA, ACA, and MCA Right ICA, ACA, MCA, PCA Right ICA, ACA, MCA
5 6 7 8
7 9.5 7 1
F M F F
Bilateral sensorimotor TIA Seizure, developmental delay Left sensorimotor TIA Bilateral sensorimotor TIA
9 10 11 12 13
2 1.4 1.8 10.8 5.7
M F F F M
14
4.5
F
Right-sided sensorimotor TIA Bilateral sensorimotor TIA Seizure, left sensorimotor TIA Bilateral sensorimotor TIA Left sensorimotor TIAs, developmental delay Bilateral sensorimotor TIA
Bilateral ICA, ACA, and MCA Left ICA, ACA, MCA Bilateral ICA, ACA, and MCA Bilateral ICA, ACA, MCA, and PCA Right ICA, ACA, MCA Bilateral ICA, ACA, and MCA Bilateral ICA, ACA, and MCA Bilateral ICA, ACA, and MCA Bilateral ICA and MCA
Subject
Bilateral ICA, ACA, MCA, and PCA
Preoperative distribution of previous ischemia or infarcts (cortical, watershed, or subcortical) depending on ACA, MCA, PCA territories
Preoperative poor AZA response and CVR deficits
Right ACA and MCA watershed None ACA, MCA, MCA-PCA watershed Cortical and watershed right ACA-MCA None Left MCA None MCA right . left, left PCA, right ACA-MCA watershed Right basal ganglia Left MCA Left MCA Left basal ganglia Left MCA
Bilateral MCA and right ACA Left ACA Right ACA, MCA, and PCA Right ACA and MCA
None
Bilateral ACA, MCA, and PCA
15 QUANTITATIVE H2[ O]-PET IN PEDIATRIC MMD
Table 2. Normalized baseline CBF, AZA-CBF, and CVR before and after revascularization surgery
Bilateral ACA and MCA Left ACA and MCA Bilateral ACA and MCA Bilateral ACA, MCA, and PCA Left ACA and MCA Bilateral ACA and MCA Bilateral ACA and MCA Bilateral MCA Bilateral MCA
Abbreviations: ACA, anterior cerebral artery; AZA, acetazolamide; CBF, cerebral blood flow; CVR, cerebrovascular reserve; ICA, internal carotid artery; MCA, middle cerebral artery; PCA, posterior cerebral artery; TIA, transient ischemic attack. Paired t tests with Bonferroni correction for multiple comparisons.
969
F.P. KUHN ET AL.
970
Detailed correlation of the preoperative and postoperative clinical presentation, MRI, and angiogram findings with the PET results will be addressed in a follow-up article.
Discussion MMD is a progressive angiopathy affecting all major supratentorial vascular territories of the brain, and the pediatric form of the disease is responsible for TIAs and ischemic stroke, even before 5 years of age.2,9 To prevent strokes and achieve long-term improvement in quality of life, early identification of this angiopathy is essential.3,10-12 Thus, correct diagnosis and early cerebral revascularization of all symptomatic arterial territories involved is essential.2,13,14 H2[15O]-PET with AZA challenge has been our gold standard to identify the cerebral territories at risk of stroke by providing a regional map of circulatory function in each individual cerebral territory involved. This measurement of function is particularly valuable in the visualization of impaired function before infarct. Hence the technique is invaluable in targeting and planning of cerebral revascularization surgery. In this study, we tested the hypotheses that H2[15O]PET data acquired preoperatively can identify territories in need of revascularization and thus guide the neurosurgeon in a tailored, combined, multiple revascularization approach and that improvements in CBF can be observed after revascularization surgery, supporting such an approach and providing a measurable indicator. Most publications15,16 to date address hemodynamic deficits in adults, are semi-quantitative, and focus mainly on the MCA territory. Hence, surgery is usually targeted to the MCA territory with the STA-MCA bypass in adults and/or the indirect revascularization with EDAS being frequently described in children. A limited number of older publications have actually addressed the territorial perfusion deficits observed in children.4,17-20 Our quantitative PET protocol as part of the presurgical workup, particularly the parametric maps of CVR, is used to clearly demonstrate the involvement of the ACA, MCA, and PCA territories preoperatively. The noninvasive quantitative H2[15O]-PET CBF technique used in our MMD center5 has been demonstrated to correlate very well with the classical fully quantitative, invasive (arterial catheter) method. However, in addition to its use in comparing cortical CBF with an unaffected region, we still considered normalization to cerebellar CBF necessary to compare scans made many months apart. Global changes in CBF due to the effects of growth in these young patients have the potential to influence the data. At the diagnostic stage, quantitative values allow the calculation of the parametric CVR image, which is of great value in visualizing regions with impaired CBF. Quantitative analysis of AZA-CBF images with compari-
son to cerebellar CBF is a valuable adjunct to visual examination of the CVR images. Indeed, improvement of CBF after tailored revascularization surgery was most clear in the quantitative AZA-CBF analysis. Although the response to AZA in cortical regions was not restored to cerebellar levels, the deficit was reduced from approximately 40% before surgery to 25%-30 % after surgery (Table 2). The lack of a significant improvement in CVR in our patient group may be because of limited statistical power with only 14 patients but also because of a bias toward impaired baseline CBF in our pediatric group. Indeed, 6 of the 14 children included in the analysis had suffered from cortical infarcts before their initial presentation. As such, the baseline CBF was much lower than normal and negative CVR values in ACA/MCA territories (Table 2) indicating that CVR is not only limited, but that a steal phenomenon is responsible for reduced CBF in these regions during increased demand in adjacent territories. In addition, the relatively short interval between initial presurgery PET and postsurgery follow-up PET (median duration, 10.4 months) could have been an additional factor because brain tissue chronically hypoperfused would require longer to recover (in terms of CVR). Improved baseline- and AZA-CBF after surgery may thus indicate the early phase of recovery. Continued follow-up PET in the years after surgery could confirm further recovery of CBF and improvements in CVR.
Conclusions In our group of pediatric moyamoya patients, H2[15O]PET not only guided us in MMD as a progressive tailoring surgery to the cerebral areas at risk of further strokes but also allowed the efficacy of revascularization to be quantified, mainly by improvement in the response to AZA challenge. This coincided well with the children being stroke-free postoperatively. Acknowledgments: The study team thanks all their technicians for the excellent technical support.
References 1. Suzuki J, Takaku A. Cerebrovascular ‘‘moyamoya’’ disease. Disease showing abnormal net-like vessels in base of brain. Arch Neurol 1969;20:288-299. 2. Khan N, Schuknecht B, Boltshauser E, et al. Moyamoya disease and moyamoya syndrome: experience in Europe; choice of revascularisation procedures. Acta Neurochir 2003;145:1061-1071. discussion 1071. 3. Scott RM, Smith ER. Moyamoya disease and moyamoya syndrome. N Engl J Med 2009;360:1226-1237. 4. Ikezaki K, Matsushima T, Kuwabara Y, et al. Cerebral circulation and oxygen metabolism in childhood moyamoya disease: a perioperative positron emission tomography study. J Neurosurg 1994;81:843-850. 5. Treyer V, Jobin M, Burger C, et al. Quantitative cerebral H2(15)O perfusion PET without arterial blood sampling,
15 QUANTITATIVE H2[ O]-PET IN PEDIATRIC MMD
6.
7.
8.
9.
10. 11.
12.
13.
a method based on washout rate. Eur J Nucl Med Mol Imaging 2003;30:572-580. Alpert NM, Eriksson L, Chang JY, et al. Strategy for the measurement of regional cerebral blood flow using short-lived tracers and emission tomography. J Cereb Blood Flow Metab 1984;4:28-34. Arigoni M, Kneifel S, Fandino J, et al. Simplified quantitative determination of cerebral perfusion reserve with H2(15)O PET and acetazolamide. Eur J Nucl Med 2000; 27:1557-1563. Hammers A, Allom R, Koepp MJ, et al. Three-dimensional maximum probability atlas of the human brain, with particular reference to the temporal lobe. Hum Brain Mapp 2003;19:224-247. Kuroda S, Ishikawa T, Houkin K, et al. [Clinical significance of posterior cerebral artery stenosis/occlusion in moyamoya disease]. No Shinkei Geka 2002;30:1295-1300. Kuroda S, Houkin K. Moyamoya disease: current concepts and future perspectives. Lancet Neurol 2008;7:1056-1066. Research Committee on the P, Treatment of Spontaneous Occlusion of the Circle of W, Health Labour Sciences Research Grant for Research on Measures for Infractable D. Guidelines for diagnosis and treatment of moyamoya disease (spontaneous occlusion of the circle of Willis). Neurol Med Chir 2012;52:245-266. Smith ER, Scott RM. Spontaneous occlusion of the circle of Willis in children: pediatric moyamoya summary with proposed evidence-based practice guidelines. A review. J Neurosurg Pediatr 2012;9:353-360. Hayashi T, Shirane R, Tominaga T. Additional surgery for postoperative ischemic symptoms in patients with moyamoya disease: the effectiveness of occipital arteryposterior cerebral artery bypass with an indirect procedure: technical case report. Neurosurgery 2009;64: E195-E196. discussion E196.
971 14. Ishikawa T, Kamiyama H, Kuroda S, et al. Simultaneous superficial temporal artery to middle cerebral or anterior cerebral artery bypass with pan-synangiosis for moyamoya disease covering both anterior and middle cerebral artery territories. Neurol Med Chir 2006;46: 462-468. 15. Honda M, Ezaki Y, Kitagawa N, et al. Quantification of the regional cerebral blood flow and vascular reserve in moyamoya disease using split-dose iodoamphetamine I 123 single-photon emission computed tomography. Surg Neurol 2006;66:155-159. discussion 159. 16. Marushima A, Tsurushima H, Suzuki K, et al. Timecourse analysis of brain perfusion single photon emission computed tomography using a three-dimensional stereotactic region-of-interest template in patients with moyamoya disease. World Neurosurg 2011; 76:304-310. 17. Kuroda S, Houkin K, Kamiyama H, et al. Regional cerebral hemodynamics in childhood moyamoya disease. Childs Nerv Syst 1995;11:584-590. 18. Kuwabara Y, Ichiya Y, Sasaki M, et al. Cerebral hemodynamics and metabolism in moyamoya disease–a positron emission tomography study. Clin Neurol Neurosurg 1997;99(Suppl 2):S74-S78. 19. Saito N, Nakagawara J, Nakamura H, Teramoto A. Assessment of cerebral hemodynamics in childhood moyamoya disease using a quantitative and a semiquantitative IMP-SPECT study. Ann Nucl Med 2004;18: 323-331. 20. Shirane R, Yoshida Y, Takahashi T, et al. Assessment of encephalo-galeo-myo-synangiosis with dural pedicle insertion in childhood moyamoya disease: characteristics of cerebral blood flow and oxygen metabolism. Clin Neurol Neurosurg 1997;99(Suppl 2):S79-85.
Background: Moyamoya disease (MMD) is an idiopathic intracranial angiopathy with a progressive spontaneous occlusion of the circle of Willis resulting in repeated ischemia if not diagnosed and treated early, especially in children. Prevention of stroke is achieved by revascularization of the affected cerebral regions. Functional imaging techniques such as H2[15O]-Positron emission tomography (PET) allow quantification of cerebral perfusion/blood flow (CBF) and in particular cerebrovascular response after acetazolamide (AZA) challenge. The cerebrovascular reserve (CVR) can then be calculated and used to identify regions at risk of infarct, hence allowing surgery to be specifically targeted and personalized. Methods: Pediatric patients with diagnosed MMD underwent initial H2[15O]-PET scans at baseline and after stimulation with AZA. Indication for surgery was then based collectively on the extent of disease observed clinically and on magnetic resonance imaging, on the arterial territories involved, as seen in angiography and the respective regional CVR observed in PET. Cerebral revascularization surgeries were subsequently performed, tailored to the individual patient. Postoperative assessment of clinical outcome was augmented with follow-up PET (median duration after surgery, 10.4 months). CBF at baseline, after AZA and CVR were compared between presurgery and postsurgery scans in the areas supplied by the major cerebral arteries. Results: Parametric images reflecting CBF, response to AZA and CVR clearly showed deficits in cortical but not subcortical regions or cerebellum. AZA-CBF and CVR deficits were most clear in middle cerebral artery and anterior cerebral artery (ACA) regions. In addition to the clinical symptomatology, angiography, AZA-CBF, and CVR images allowed the laterality of deficits to be clearly visualized for tailored surgery and the indication for targeted ACA or posterior cerebral artery revascularization to be assessed. Comparison of baseline CBF, AZA-CBF, and CVR between presurgery and postsurgery scans in revascularized areas revealed a significant improvement in baseline and AZA-CBF after surgery. Although no significant differences in CVR after revascularization surgery were found, a clear improvement of the deficits apparent in AZA-CBF in revascularized regions was found. Conclusions: We demonstrate that quantitative H2[15O]-PET is a highly useful tool to direct surgical intervention in MMD. Detailed quantitative analysis of CBF changes and CVR after
From the *Department of Nuclear Medicine, University Hospital Zurich; and †Division of Pediatric Neurosurgery, Department of Surgery, Moyamoya Center, University Children’s Hospital Zurich, Zurich, Switzerland. Received December 1, 2014; accepted December 13, 2014. F.P.K. and G.W. contributed equally to this work. Research funds of the Nuclear Medicine Department of the University Hospital Zurich and of the Moyamoya Center, Division of Pediatric Neurosurgery, University Children’s Hospital Zurich.
The authors declare no other conflict of interest. Address correspondence to Felix P. Kuhn, MD, Department of Nuclear Medicine, University Hospital Zurich, CH-8091 Zurich, Switzerland. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.12.017
Journal of Stroke and Cerebrovascular Diseases, Vol. 24, No. 5 (May), 2015: pp 965-971
965
F.P. KUHN ET AL.
966 surgery supports a targeted surgical approach. Key Words: Moyamoya disease— nuclear medicine—positron emission tomography—brain perfusion. Ó 2015 by National Stroke Association
Introduction Moyamoya disease (MMD) is an idiopathic intracranial angiopathy with progressive spontaneous occlusion of the circle of Willis1 resulting in repeated ischemia if not diagnosed and treated early, especially in children. Young patients often suffer from symptoms of hemodynamic insufficiency, such as repetitive transient ischemic attacks (TIAs) and ischemic strokes. Prevention of stroke is achieved by revascularization of the affected vascular territories.2,3 Children with MMD are referred to our MMD center internationally for neurosurgical management. Before cerebral revascularization, a detailed presurgical evaluation protocol is followed. This consists of an initial clinical–neurologic evaluation, magnetic resonance imaging (MRI) to delineate cortical or subcortical ischemia and/ or infarcts, 6-vessel cerebral angiography to demonstrate stenosis, the presence of MMD collaterals, and the involvement of particular arterial territories, (middle cerebral artery [MCA], anterior cerebral artery [ACA], and posterior cerebral artery [PCA]) and a hemodynamic evaluation using H2[15O]-Positron emission tomography (PET) at baseline and after acetazolamide (AZA) challenge, which is invaluable in assessing the regional cerebral baseline blood flow (CBF), cerebrovascular response to vasodilatory challenge, and the cerebrovascular reserve (CVR). Reliable quantification of CBF after vasodilation provoked by AZA is crucial to allow for targeted and personalized surgery, rather than routine revascularization of the bilateral MCA territories alone. In patients with impaired CBF, the response to AZA (AZA-CBF) is suppressed in poorly perfused regions. Typically, CBF at rest appears normal because of homeostatic regulation, but only unaffected regions have the capacity to further increase CBF on challenge. Parametric CVR images highlight this deficit and quantitative analysis of AZA-CBF in the major flow territories in comparison with an unaffected region (eg, cerebellum) can be used to assess the scale of such deficits and measure the improvement after revascularization. Patients found to have a poor CVR have been shown to be good candidates for revascularization surgery.4 A tailored revascularization approach is favored in our MMD center, making particular use of the H2[15O]-PET data to prioritize revascularization in those regions at greatest risk of ischemic stroke (poor CVR/limited response to AZA), including not only MCA but also ACA and PCA territories. Depending on the symptomatology, number of arterial territories
involved, the CBF, CBF response to AZA, and the CVR, cerebral revascularization is performed in 1 or more regions, unilaterally or bilaterally. Combinations of direct and indirect revascularization approaches are used (described in more detail below). The first clinical and neuroradiologic postoperative follow-up is then carried out at 6 months-1 year. In addition to its value in presurgical evaluation of hemodynamics, H2[15O]-PET can provide valuable complementary information for follow-up. The purpose of this study was to use the H2[15O]-PET data from a sample of pediatric patients to directly quantify the changes in CBF soon after tailored revascularization surgery. Generally, the success of revascularization surgery is determined only by the reduction of TIAs and ischemic strokes. We hypothesized that in addition to identifying those regions in greatest need of revascularization, quantitative H2[15O]-PET would provide a measurable indicator of improved CBF in response to revascularization.
Methods Patient Selection For this retrospective study, 20 children with angiographically proven MMD who underwent cerebral revascularization surgery at our institution were considered. MMD was diagnosed based on the clinical presentation, MRI findings, and typical angiographical hallmarks. To identify those regions at greatest risk of ischemic stroke and most likely to respond favorably to revascularization surgery, CBF was measured in each patient using H2[15O]-PET scans at baseline (baseline CBF) and after stimulation with AZA (AZA-CBF). Parametric images of CVR were then calculated as described in the following. Revascularization surgery was subsequently tailored to those cerebral vessel territories with CVR deficits, that is, cerebral revascularization was performed in all symptomatic areas showing limited CVR (unilateral or bilateral).
Tailor-Made Cerebral Revascularization (Types of Surgeries) Revascularization can be performed either by direct bypass surgery from the superficial temporal artery to the ACA (superficial temporal artery [SCA] - anterior cerebral artery [ACA] bypass) for revascularization of the frontal regions or MCA (STA-MCA bypass) for revascularization of the frontolateral, parietal, and temporal regions. In cases where this is technically not
15 QUANTITATIVE H2[ O]-PET IN PEDIATRIC MMD
967
possible, indirect procedures such as encephaloduroarteriosynangiosis (EDAS), encephalomyosynangiosis, and encephalogaleaperiostsynangiosis are performed. The occipital artery can also be used for a direct anastomosis to the PCA or an indirect EDAS for revascularization of the occipital region.
Follow-up Standard postoperative follow-up including clinical, neurologic evaluation along with MRI scanning, and 6-vessel cerebral angiography was performed after revascularization surgery. This was augmented with follow-up PET (median duration after surgery, 10.4 months). Baseline CBF, AZA-CBF, and CVR were compared for presurgery and postsurgery scans. This study focused on CBF and CVR changes before and after targeted regional cerebral revascularization.
Exclusion Criteria Patients who suffered additional cerebral infarcts in the period between initial PET and surgery and patients with posterior circulation involvement and hence undergoing posterior revascularization were excluded.
Image Acquisition PET data were acquired on a full ring PET/CT scanner in 3D mode (PET/CT Discovery STE, GE Healthcare, Waukesha, MI) and corrected for attenuation and scatter by using the corresponding computed tomography (120 kV/80 mA) and manufacturer’s algorithms. The scanner’s axial field of view covered 15.3 cm. Images were reconstructed using a 3-dimensional Fourier rebinning filtered back projection algorithm resulting in a 128 3 128 3 47 matrix with 2.34 3 2.34 3 3.27 mm voxel spacing. In all, 300-800 MBq H2[15O] was administered intravenously using an automatic injection device. The tracer was delivered over 20 seconds. The emission data were acquired as a series of 18, 10-second frames. AZA was administrated 13 minutes before the second PET scan, with dose adjusted according to weight and injection over 2 minutes.
Image Processing Parametric images of CBF were generated using a previously reported method5 in which a standardized arterial input function and image scaling based on the washout rate k2 of H2[15O] are used to derive the CBF. This procedure, based on method by Alpert,6 exploits the fact that k2 is related to the shape and not the scale of the arterial input function and proportional to CBF (k2 5 K1/p, where p is the partition coefficient).7 After application of a 6-mm full width at half maximum Gaussian filter, parametric CVR images were calculated as the difference between AZA-CBF and base-
Figure 1. Cerebral angiography (A, anteroposterior and lateral ICA injections) and PET data (B) from a single patient. Angiography clearly revealed tight stenosis of the ICA-A1-M1 segments and MMD collateral vessels. At baseline (B, baseline), CBF appears close to normal, but after AZA challenge increases in CBF are only apparent in subcortical, PCA, and cerebellar regions. Calculation of CVR images highlights the lack of CVR in bilateral MCA and ACA territories. CVR T2 illustrates the overlay of the CVR image on T2-weighted MRI. The flow territory volume-of-interest definitions for quantitative analysis (after transformation to the individual anatomy) are illustrated on the bottom row, overlaid on the CVR image. These regions were based on blood supply from the cerebral arteries. Vascular territories: light/dark blue–ACA, pink/red–MCA, light/dark green–PCA, yellow– cerebellum. Watershed areas were divided equally between flow territories. Abbreviations: ACA, anterior cerebral artery, AZA, acetazolamide; CBF, cerebral blood flow; CVR, cerebrovascular reserve; MCA, middle cerebral artery; MRI, magnetic resonance imaging; PCA, posterior cerebral artery. Color version of the figure is available online.
line CBF (mL/minute/mL tissue). These images were used for the visual identification of those flow territories with impaired CVR before tailored revascularization surgery. To facilitate the comparison of flow territory CBF with that in the cerebellum, all images were scaled to their individual cerebellar mean CBF. Normalized values below 1 are indicative of impaired CBF. Cerebellar normalized CBF was calculated using an atlas-based
F.P. KUHN ET AL.
968
Table 1. Summary of preoperative clinical and neuroradiologic findings, AZA response and CVR deficits Baseline CBF (normalized)
AZA-CBF (normalized)
CVR (mL/min/mL)
Region
Presurgery
Postsurgery
Presurgery
Postsurgery
Presurgery
Postsurgery
ACA left ACA right MCA left MCA right PCA left PCA right
.785 6 .318 .917 6 .195 .791 6 .246 .892 6 .169 .974 6 .248 1.007 6 .139
.980 6 .251 (*) .990 6 .215 .936 6 .211* .972 6 .197 1.022 6 .204 1.083 6 .160
.604 6 .218 .618 6 .136 .624 6 .222 .633 6 .135 .855 6 .181 .948 6 .105
.707 6 .174* .727 6 .144*** .752 6 .178*** .757 6 .145* .890 6 .178 .993 6 .032
,.001 6 .032 2.015 6 .046 .005 6 .042 2.005 6 .044 .058 6 .058 .058 6 .074
.001 6 .026 2.007 6 .042 .039 6 .040 .022 6 .033 .080 6 .084 .090 6 .068
Abbreviations: ACA, anterior cerebral artery; AZA, acetazolamide; CBF, cerebral blood flow; CVR, cerebrovascular reserve; MCA, middle cerebral artery; PCA, posterior cerebral artery. (*) indicates trend (P approaching .05); *P , .05; ***P , .01.
volume-of-interest method (Fig 1). Cerebellar normalization of CBF images is also valuable for the comparison of presurgery and postsurgery PET in young patients, where global changes in CBF are likely. Regional assessment of CBF was performed after normalization of the PET data to the stereotaxic spatial array of the Montreal Neurological Institute brain template. Normalization to Montreal Neurological Institute space was performed using automated algorithms available in the PMOD software package (version 3.3; PMOD Technologies Ltd, Zurich, Switzerland), except in cases of extreme CBF pathology where manual approximation of the transformation was necessary. Quantitative CBF values were extracted using an atlas outlining the vascular territories. This vascular territory atlas was generated by attributing the predefined brain region volume-of-interests from the previously validated N30R83 atlas by Hammer et al8 to the major vascular territories (illustrated in Fig 1). Watershed regions were divided equally between territory definitions.
Statistical Analysis Differences in baseline CBF, AZA-CBF, and CVR were assessed in each flow territory in which surgery was performed. Changes in CBF and CVR between presurgey and postsurgery PET were assessed using the paired Student t test with Bonferroni correction for multiple comparisons (SAS, SAS software, version 9.2, SAS Institute Inc., Cary, NC).
Results Fourteen patients were assessed in the present study. Six patients were excluded from analysis because of infarcts between initial PET and surgery. Demographic and clinical data for the 14 patients included are shown in Table 1. In all patients, the parametric images reflecting CVR clearly highlighted deficits in cortical rather than subcortical regions or cerebellum. Perfusion deficits were most
evident in MCA and ACA regions. The PCA, superior cerebellar artery, posterior inferior cerebellar artery, and anterior inferior cerebellar artery territories showed the highest response to AZA and the highest CVR. Examples of AZA-CBF and CVR images for a representative patient are shown in Figure 1. Revascularization surgery was tailored to those flow territories with poor response to AZA, in this example, the bilateral MCA and bilateral ACA territories. Normalized baseline CBF, AZA-CBF, and CVR before and after revascularization surgery are summarized for ACA, MCA, and PCA territories in Table 2. Baseline CBF was significantly increased in the left MCA territory (presurgery CBF, .79 6 .25; postsurgery CBF, .94 6 .21; P 5 .014) and showed a trend toward increase in the left ACA territory (presurgery CBF, .79 6 .32; postsurgery CBF, .98 6 .25; P 5 .089). CBF in the PCA territories was not significantly altered. CBF after AZA challenge was significantly increased in all ACA and MCA territories (ACA left, presurgery CBF, .60 6 .22; postsurgery CBF, .71 6 .17; P 5 .034; ACA right, presurgery CBF, .62 6 .14; postsurgery CBF, .73 6 .14; P , .001; MCA left, presurgery CBF, .62 6 .22; postsurgery CBF, .75 6 .18; P 5 .001; MCA right, presurgery CBF, .63 6 .14; postsurgery CBF, .76 6 .15; P 5 .021) but not in the PCA territories. A generalized improvement in CBF was apparent after revascularization surgery, with a more prominent effect on AZA-CBF. Before revascularization, mean CBF in ACA and MCA territories was approximately 40% below that in the cerebellum (Table 1). At follow-up PET, this deficit was reduced to approximately 25%-30 %. No significant improvement in CVR after revascularization was found. Postoperatively, all patients were clinically symptomfree (ie, had no TIAs, strokes, or other neurologic symptoms or new deficits, Table 1) for the duration of follow-up. This was in good agreement with postoperative 6-vessel cerebral angiography, which indicated good distal filling in the respective revascularized recipient arterial territories.
Age, y
Gender
Preoperative symptomatology
Arterial territories involved on cerebral angiogram
1 2 3 4
11 4.5 1 1.7
F F F M
Headaches, bilateral sensorimotor TIA Headaches, left-sided sensorimotor TIA Seizure Left sensorimotor TIA
Bilateral ICA, ACA, MCA Bilateral ICA, ACA, and MCA Right ICA, ACA, MCA, PCA Right ICA, ACA, MCA
5 6 7 8
7 9.5 7 1
F M F F
Bilateral sensorimotor TIA Seizure, developmental delay Left sensorimotor TIA Bilateral sensorimotor TIA
9 10 11 12 13
2 1.4 1.8 10.8 5.7
M F F F M
14
4.5
F
Right-sided sensorimotor TIA Bilateral sensorimotor TIA Seizure, left sensorimotor TIA Bilateral sensorimotor TIA Left sensorimotor TIAs, developmental delay Bilateral sensorimotor TIA
Bilateral ICA, ACA, and MCA Left ICA, ACA, MCA Bilateral ICA, ACA, and MCA Bilateral ICA, ACA, MCA, and PCA Right ICA, ACA, MCA Bilateral ICA, ACA, and MCA Bilateral ICA, ACA, and MCA Bilateral ICA, ACA, and MCA Bilateral ICA and MCA
Subject
Bilateral ICA, ACA, MCA, and PCA
Preoperative distribution of previous ischemia or infarcts (cortical, watershed, or subcortical) depending on ACA, MCA, PCA territories
Preoperative poor AZA response and CVR deficits
Right ACA and MCA watershed None ACA, MCA, MCA-PCA watershed Cortical and watershed right ACA-MCA None Left MCA None MCA right . left, left PCA, right ACA-MCA watershed Right basal ganglia Left MCA Left MCA Left basal ganglia Left MCA
Bilateral MCA and right ACA Left ACA Right ACA, MCA, and PCA Right ACA and MCA
None
Bilateral ACA, MCA, and PCA
15 QUANTITATIVE H2[ O]-PET IN PEDIATRIC MMD
Table 2. Normalized baseline CBF, AZA-CBF, and CVR before and after revascularization surgery
Bilateral ACA and MCA Left ACA and MCA Bilateral ACA and MCA Bilateral ACA, MCA, and PCA Left ACA and MCA Bilateral ACA and MCA Bilateral ACA and MCA Bilateral MCA Bilateral MCA
Abbreviations: ACA, anterior cerebral artery; AZA, acetazolamide; CBF, cerebral blood flow; CVR, cerebrovascular reserve; ICA, internal carotid artery; MCA, middle cerebral artery; PCA, posterior cerebral artery; TIA, transient ischemic attack. Paired t tests with Bonferroni correction for multiple comparisons.
969
F.P. KUHN ET AL.
970
Detailed correlation of the preoperative and postoperative clinical presentation, MRI, and angiogram findings with the PET results will be addressed in a follow-up article.
Discussion MMD is a progressive angiopathy affecting all major supratentorial vascular territories of the brain, and the pediatric form of the disease is responsible for TIAs and ischemic stroke, even before 5 years of age.2,9 To prevent strokes and achieve long-term improvement in quality of life, early identification of this angiopathy is essential.3,10-12 Thus, correct diagnosis and early cerebral revascularization of all symptomatic arterial territories involved is essential.2,13,14 H2[15O]-PET with AZA challenge has been our gold standard to identify the cerebral territories at risk of stroke by providing a regional map of circulatory function in each individual cerebral territory involved. This measurement of function is particularly valuable in the visualization of impaired function before infarct. Hence the technique is invaluable in targeting and planning of cerebral revascularization surgery. In this study, we tested the hypotheses that H2[15O]PET data acquired preoperatively can identify territories in need of revascularization and thus guide the neurosurgeon in a tailored, combined, multiple revascularization approach and that improvements in CBF can be observed after revascularization surgery, supporting such an approach and providing a measurable indicator. Most publications15,16 to date address hemodynamic deficits in adults, are semi-quantitative, and focus mainly on the MCA territory. Hence, surgery is usually targeted to the MCA territory with the STA-MCA bypass in adults and/or the indirect revascularization with EDAS being frequently described in children. A limited number of older publications have actually addressed the territorial perfusion deficits observed in children.4,17-20 Our quantitative PET protocol as part of the presurgical workup, particularly the parametric maps of CVR, is used to clearly demonstrate the involvement of the ACA, MCA, and PCA territories preoperatively. The noninvasive quantitative H2[15O]-PET CBF technique used in our MMD center5 has been demonstrated to correlate very well with the classical fully quantitative, invasive (arterial catheter) method. However, in addition to its use in comparing cortical CBF with an unaffected region, we still considered normalization to cerebellar CBF necessary to compare scans made many months apart. Global changes in CBF due to the effects of growth in these young patients have the potential to influence the data. At the diagnostic stage, quantitative values allow the calculation of the parametric CVR image, which is of great value in visualizing regions with impaired CBF. Quantitative analysis of AZA-CBF images with compari-
son to cerebellar CBF is a valuable adjunct to visual examination of the CVR images. Indeed, improvement of CBF after tailored revascularization surgery was most clear in the quantitative AZA-CBF analysis. Although the response to AZA in cortical regions was not restored to cerebellar levels, the deficit was reduced from approximately 40% before surgery to 25%-30 % after surgery (Table 2). The lack of a significant improvement in CVR in our patient group may be because of limited statistical power with only 14 patients but also because of a bias toward impaired baseline CBF in our pediatric group. Indeed, 6 of the 14 children included in the analysis had suffered from cortical infarcts before their initial presentation. As such, the baseline CBF was much lower than normal and negative CVR values in ACA/MCA territories (Table 2) indicating that CVR is not only limited, but that a steal phenomenon is responsible for reduced CBF in these regions during increased demand in adjacent territories. In addition, the relatively short interval between initial presurgery PET and postsurgery follow-up PET (median duration, 10.4 months) could have been an additional factor because brain tissue chronically hypoperfused would require longer to recover (in terms of CVR). Improved baseline- and AZA-CBF after surgery may thus indicate the early phase of recovery. Continued follow-up PET in the years after surgery could confirm further recovery of CBF and improvements in CVR.
Conclusions In our group of pediatric moyamoya patients, H2[15O]PET not only guided us in MMD as a progressive tailoring surgery to the cerebral areas at risk of further strokes but also allowed the efficacy of revascularization to be quantified, mainly by improvement in the response to AZA challenge. This coincided well with the children being stroke-free postoperatively. Acknowledgments: The study team thanks all their technicians for the excellent technical support.
References 1. Suzuki J, Takaku A. Cerebrovascular ‘‘moyamoya’’ disease. Disease showing abnormal net-like vessels in base of brain. Arch Neurol 1969;20:288-299. 2. Khan N, Schuknecht B, Boltshauser E, et al. Moyamoya disease and moyamoya syndrome: experience in Europe; choice of revascularisation procedures. Acta Neurochir 2003;145:1061-1071. discussion 1071. 3. Scott RM, Smith ER. Moyamoya disease and moyamoya syndrome. N Engl J Med 2009;360:1226-1237. 4. Ikezaki K, Matsushima T, Kuwabara Y, et al. Cerebral circulation and oxygen metabolism in childhood moyamoya disease: a perioperative positron emission tomography study. J Neurosurg 1994;81:843-850. 5. Treyer V, Jobin M, Burger C, et al. Quantitative cerebral H2(15)O perfusion PET without arterial blood sampling,
15 QUANTITATIVE H2[ O]-PET IN PEDIATRIC MMD
6.
7.
8.
9.
10. 11.
12.
13.
a method based on washout rate. Eur J Nucl Med Mol Imaging 2003;30:572-580. Alpert NM, Eriksson L, Chang JY, et al. Strategy for the measurement of regional cerebral blood flow using short-lived tracers and emission tomography. J Cereb Blood Flow Metab 1984;4:28-34. Arigoni M, Kneifel S, Fandino J, et al. Simplified quantitative determination of cerebral perfusion reserve with H2(15)O PET and acetazolamide. Eur J Nucl Med 2000; 27:1557-1563. Hammers A, Allom R, Koepp MJ, et al. Three-dimensional maximum probability atlas of the human brain, with particular reference to the temporal lobe. Hum Brain Mapp 2003;19:224-247. Kuroda S, Ishikawa T, Houkin K, et al. [Clinical significance of posterior cerebral artery stenosis/occlusion in moyamoya disease]. No Shinkei Geka 2002;30:1295-1300. Kuroda S, Houkin K. Moyamoya disease: current concepts and future perspectives. Lancet Neurol 2008;7:1056-1066. Research Committee on the P, Treatment of Spontaneous Occlusion of the Circle of W, Health Labour Sciences Research Grant for Research on Measures for Infractable D. Guidelines for diagnosis and treatment of moyamoya disease (spontaneous occlusion of the circle of Willis). Neurol Med Chir 2012;52:245-266. Smith ER, Scott RM. Spontaneous occlusion of the circle of Willis in children: pediatric moyamoya summary with proposed evidence-based practice guidelines. A review. J Neurosurg Pediatr 2012;9:353-360. Hayashi T, Shirane R, Tominaga T. Additional surgery for postoperative ischemic symptoms in patients with moyamoya disease: the effectiveness of occipital arteryposterior cerebral artery bypass with an indirect procedure: technical case report. Neurosurgery 2009;64: E195-E196. discussion E196.
971 14. Ishikawa T, Kamiyama H, Kuroda S, et al. Simultaneous superficial temporal artery to middle cerebral or anterior cerebral artery bypass with pan-synangiosis for moyamoya disease covering both anterior and middle cerebral artery territories. Neurol Med Chir 2006;46: 462-468. 15. Honda M, Ezaki Y, Kitagawa N, et al. Quantification of the regional cerebral blood flow and vascular reserve in moyamoya disease using split-dose iodoamphetamine I 123 single-photon emission computed tomography. Surg Neurol 2006;66:155-159. discussion 159. 16. Marushima A, Tsurushima H, Suzuki K, et al. Timecourse analysis of brain perfusion single photon emission computed tomography using a three-dimensional stereotactic region-of-interest template in patients with moyamoya disease. World Neurosurg 2011; 76:304-310. 17. Kuroda S, Houkin K, Kamiyama H, et al. Regional cerebral hemodynamics in childhood moyamoya disease. Childs Nerv Syst 1995;11:584-590. 18. Kuwabara Y, Ichiya Y, Sasaki M, et al. Cerebral hemodynamics and metabolism in moyamoya disease–a positron emission tomography study. Clin Neurol Neurosurg 1997;99(Suppl 2):S74-S78. 19. Saito N, Nakagawara J, Nakamura H, Teramoto A. Assessment of cerebral hemodynamics in childhood moyamoya disease using a quantitative and a semiquantitative IMP-SPECT study. Ann Nucl Med 2004;18: 323-331. 20. Shirane R, Yoshida Y, Takahashi T, et al. Assessment of encephalo-galeo-myo-synangiosis with dural pedicle insertion in childhood moyamoya disease: characteristics of cerebral blood flow and oxygen metabolism. Clin Neurol Neurosurg 1997;99(Suppl 2):S79-85.
