Comparative performance of magnetic resonance angiography and conventional angiography in moyamoya disease

Journal of Clinical Neuroscience (2000) 7(2), 112–115 © 2000 Harcourt Publishers Ltd DOI: 10.1054/ jocn.1998.0160, available online at http://www.idealibrary.com on

Clinical study

Comparative performance of magnetic resonance angiography and conventional angiography in moyamoya disease Naokatsu Saeki, Marcelo Nery Silva, Motoo Kubota, Jun-ichi Takanashi,1 Katsuo Sugita,1 Susumu Nakazaki, Akira Yamaura Department of Neurological Surgery and 1Pediatrics, Chiba University School of Medicine, Inohana 1–8–1, Chuo-ku, Chiba, Japan

Summary Patients with moyamoya disease verified by conventional cerebral angiography (CCA) were evaluated by means of magnetic resonance angiography (MRA), in order to clarify its advantages and limitations in comparison with CCA. This retrospective study was carried out in 13 hemispheres of seven patients with moyamoya disease (including one atypical patient). In MRA, magnetisation transfer contrast (MTC) and maximal intensity projection (MIP) techniques were used. A good correspondence between MRA and CCA was noted in 11 hemispheres (86.4%) on evaluating stenotic lesions and overestimation in the remaining 2. Underestimation of moyamoya vessels in MRA was noted in 3 hemispheres, since well developed moyamoya vessels were detectable, while poorly developed ones were not. No difference between MRA and CCA findings was found in 10 hemispheres (76.9%). MRA tends to overestimate the stenosis and underestimate moyamoya vessels. These characteristics may lead to wrong classification in the angiographic grade of patients with early and advanced stages and should be taken into consideration in interpreting MRA images of moyamoya disease (compatible rate of grade between MRA and CCA; 84.6%). In conclusion MRA with MTC and MIP techniques offers an acceptable quality of assessment of stenotic lesions and moyamoya vessels. MRA is a useful follow up method at present and in the near future it may replace CCA as the initial diagnostic tool. © 2000 Harcourt Publishers Ltd Keywords: carotid fork stenosis, collateral circulation, cerebral angiography, moyamoya disease, MR angiography

INTRODUCTION

MATERIALS AND METHODS

‘Moyamoya’ disease is a cerebrovascular affliction in which an abnormal vascular network at the base of the brain, known as moyamoya vessels, is present. The term ‘moyamoya’ (something hazy like a puff of smoke) has been used worldwide to define the spontaneous occlusion of the circle of Willis and a development of abnormal vessels at the base of the brain. Its aetiology is unknown, although some speculations have risen over infectious, familial and hereditary factors.1 Recent research efforts have shown that it is a polygenic hereditary disease,2,3 but the pathogenesis remains unknown. Furthermore, its natural history is not adequately elucidated and in a significant number of patients repetitive strokes and haemorrhages result in progressive neurological deterioration, followed by poor outcome in children and adults.4,5 Conventional cerebral angiography (CCA) is considered to be the ‘gold standard’ method for diagnosis and follow up of moyamoya patients. However, even in the best hands, it is not always safe, particularly in children, a population most commonly affected by this disease. On the other hand, magnetic resonance angiography (MRA) has been available in most of the major medical centres, assessing cerebrovascular diseases with high degree of safety and confidence. We report herein the usefulness of MRA in the diagnosis and follow up of moyamoya patients, focusing on merits and limitations of the procedure.

In a retrospectively conducted study, we reviewed seven patients with moyamoya disease (six typical and one atypical with unilateral moyamoya phenomenon) (Fig. 1 case 1-left side). A total of 13 hemispheres were analysed. There were five females and two males. The ages at the time of examination ranged from 4 to 47 years. All of them presented with ischaemic episodes (TIA or infarcts). Clinical presentations are summarised in Table 1. The main imaging work up consisted of a brain computed tomography (CT), SPECT, MRI, MRA as well as CCA. All the patients underwent both CCA and MRA at various intervals between 1 week and 6 months. MRA was performed in a 1.5 Telsa MR Unit (Signa Horizon, GE). As a routine, three dimensional Fourier transformation timeof-flight (3DFT TOF) technique was used and the standard MR parameters were as follows: repetition time (TR) 61 milliseconds, echo time (TE) 6.9 milliseconds, acquisition matrix 512 × 192, field-of-view (FOV) 24 × 18 cm, slice thickness of 0.7 or 0.9 mm and 4 minutes on average of acquisition time. In order to decrease the noise from binding water, the off resonance presaturation pulse (magnetisation transfer contrast; MTC) was used. Partial maximal intensity projection (MIP) was also used to provide adequate background suppression and to improve the contrast of small vessels (ICA, ACA, MCA and moyamoya vessels). In none of our cases did we use gadolinium enhancement. MRA and CCA were carefully examined and blindly evaluated by the authors. All available images were used, and when two images disagreed, the less stenotic appearing image was selected. Both MRA and CCA findings were compared in relation to the following parameters: 1) degree of stenosis at the carotid fork – the width of the stenotic vessel was compared to that of the closest normal portion of the vessel; 2) appearance of moyamoya vessels; and 3) comparison of angiographic grades between MRA and CCA.

Received 18 April 1998 Accepted 25 August 1998 Correspondence to: Naokatsu Saeki

112

MRA vs. angiography in moyamoya 113

Fig. 1 Case 1. Upper columns of MRA show the stenosis at the left carotid fork and the right proximal ACA. Lower columns of CCA show the corresponding angiographic features. Lower left column (right hemisphere) demonstrates no stenosis of ACA (MRA overestimation). Lower right column (left hemisphere) shows stenotic change at the carotid fork, corresponding to the MRA findings.

Fig. 3 Case 4. Comparison of MRA (upper column) and CCA (left lower column) shows an overestimation in visualising carotid fork on the left side (MRA overestimation) and poor correspondence in depicting moyamoya vessels on both sides (MRA underestimation).

of arterial lumen; grade II (moderate stenosis) – reduction between 1/3 and 2/3, and; grade III (severe stenosis and occlusions) – reduction over 2/3, including occlusions. MRA detected carotid fork stenosis in all 13 hemispheres (Figs 1–3). MRA disclosed three hemispheres in grade I, two in grade II and eight in grade III (Table 2). CCA showed four hemispheres in grade I, two in grade two and seven in grade III. The result of MRA was the same as that of CCA in 11 hemispheres out of 13 (84.6%) (Figs 1 & 2). One hemisphere in MRA grade II and grade III was CCA grade I and grade II, respectively (Table 2) (Fig. 3). In two hemispheres the stenosis was overestimated in MRA. No underestimation by MRA was noted. Evaluation of moyamoya vessels

Fig. 2 Case 3. MRA disclosed carotid fork stenosis on both hemispheres (upper) and also a good correspondence of degree of moyamoya vessels on both sides was observed (lower both). On the upper right corner of MRA (left hemisphere), transdural collaterals are seen.

RESULTS Evaluation of stenosis To evaluate the degree of stenosis of the carotid fork, the following criteria were used: grade I (mild stenosis) – reduction less than 1/3

The moyamoya vessels were classified into three grades: absent; poorly developed; and well developed. The network was absent in three hemispheres of MRA, poorly developed in five and well developed in the other five. In 10 hemispheres, moyamoya vessels were visualised with MRA, coinciding with CCA (76.9%) (Figs 1 & 2). In the remaining three hemispheres disagreement of MRA and CCA was found (Fig. 3). MRA underestimated moyamoya vessels in all three hemispheres. One hemisphere seemed not to habour moyamoya vessels on MRA. Meanwhile, moyamoya vessels were detected on CCA as a poorly developed, although present, network (Table 3) (Fig. 3 lower right: left hemisphere). In the other two, poorly developed moyamoya vessels on MRA were well developed on CCA (Fig. 3 lower left:right hemisphere).

Table 1 Clinical features of 7 cases Case

Age (years)/sex

Initial symptoms

Ischaemic lesion on MRI

Interval of CCA & MR

1 2 3 4 5 6 7

28/F 47/M 7/F 4/F 23/F 7/F 7/M

Right hemiparesis Visual field defect Right hemiparesis TIA TIA TIA TIA

Left basal ganglia Right occipital lobe Left basal ganglia No No No Subcortical

4 months 2 weeks 1 week 3 weeks 6 months 2 weeks 2 months

© 2000 Harcourt Publishers Ltd

Journal of Clinical Neuroscience (2000) 7(2), 112–115

114 Saeki et al. Table 2 Comparison of carotid fork stenosis between MRA and CCA MRA Grade

CCA Grade

I II III MRA total

I 3

II 1 1

3

III

CCA total 4 2 7 13

1 7 8

2

Grade I – reduction of arterial lumen under 1/3. Grade II – reduction between 1/3 and 2/3. Grade III – reduction over 2/3 including occlusion.

Table 3 Depiction of moyamoya vessels between MRA and CCA MRA * CCA total CCA

* ** *** MRA total

**

2 1

3 2 5

3

***

5 5

2 4 7 13

* – absent; ** – poorly developed; *** – well developed.

Table 4 Comparison of stage classification between MRA and CCA

Explanation of CCA stage (cited from reference 6). CCA stage 1–narrowing of carotid fork. Only the carotid fork stenosis is observed. CCA stage 2–initiation of basal moyamoya. All the main cerebral arteries are dilated. CCA stage 3–intensification of moyamoya. Remarkable moyamoya vessels at the base of the brain. The defection middle and anterior cerebral arteries are observed. CCA stage 4–minimisation of moyamoya. The defection of posterior cerbral artery is observed. CCA stage 5–reduction of moyamoya. All the main cerebral arteries are missing. CCA stage 6–disappearance of moyamoya. Cerebral blood flow supplied only from external carotid artery. In this table no cases of CCA stages 5 and 6 are present.

Angiographic stages Each of the thirteen hemispheres was applied to one of the six angiographic stages proposed by Suzuki and Takaku on both MRA and CCA (Table 4).6 On CCA, one hemisphere was in stage 1, three in stage 2, five in stage 3 and four in stage 4. In MRA, one hemisphere in stage 1, four in stage 2, four in stage 3, three in stage 4 and one in stage 5. All hemispheres on MRA corresponded well with ones on CCA (84.6%) (Fig. 2), except two. One hemisphere, considered initially to be stage 2 in MRA, was found to be stage 3 on CCA. The other hemisphere graded stage 5 on MRA was stage 4 on CCA (Table 4).

Others There were no complications from MRA. Leptomeningeal and transdural collateral circulation (from superficial temporal and meningeal arteries) were well demonstrated on MRA, Journal of Clinical Neuroscience (2000) 7(2), 112–115

compatible with CCA (Fig. 2), except for ethmoidal moyamoya in the latter. DISCUSSION Since the first reports, moyamoya disease has been diagnosed primarily by CCA as the ‘gold standard’ method for visualisation of stenosis of the carotid fork, presence of moyamoya vessels and presence and extent of collateral circulation. In Japan, where the incidence of moyamoya disease is relatively higher compared to other countries, the Research Committee of Spontaneous Occlusion of the Circle of Willis (known as ‘moyamoya disease’) of the Ministry of Health and Welfare established that angiography was mandatory for the diagnosis of this entity.7 Over the years, CCA proved to be very accurate although not totally safe particularly in children, a population more frequently affected by the disease. Furthermore, the average number of angiographic procedures in one patient is, at least, two (some cases requiring

© 2000 Harcourt Publishers Ltd

MRA vs. angiography in moyamoya 115

three or more), increasing the chance of complications related to radiation exposure, general anaesthesia, arterial puncture and injection of contrast material. Thus, a near ideal method of investigation of moyamoya disease is necessary which is noninvasive while still highly reliable. MRA is a method, by many, considered to be ‘state-of-the-art’ in the diagnosis and follow up of patients harbouring cerebrovascular disease, including moyamoya. In our study a good correlation between MRA and CCA on evaluating stenotic change was found in 11 hemispheres out of 13 (84.6%), which coincides well with the previous report (Fig. 3).8,9 In the remaining two hemispheres the stenosis was overestimated. As was shown in our report, when MRA and CCA disagree, MRA overestimates the lesion.8 This occurs probably owing to intravoxel phase dispersion in areas of turbulent blood flow, leading to signal loss. Phase dispersion is minimised by the use of shorter echo times and smaller voxels. In this study, a TE if 6.9 milliseconds was employed but still in two hemispheres stenotic overestimation was noticed. It seems to be likely that sequences with even smaller echo times and FOVs will be available in the future, obviating approximations in sizing up the vessels lumen. A high sensitivity in detecting stenosis may be advantageous as a screening and diagnostic tool. In our patients, MRA demonstrated an acceptable correspondence to CCA in assessment of moyamoya vessels. In our classification (Table 3), the grades determined by MRA and CCA agreed in 76.9% of 13 hemispheres. The development of moyamoya vessels was underestimated in three hemispheres (Fig. 3). MRA tends to underestimate the moyamoya vessels probably owing to their reduced size (less than 1 mm) and some artifacts related to acquisition parameters. The moyamoya vessels in stages 3 and 4 seemed to be accessible with MRA, which have more pronounced moyamoya vessels compared to other stages.1,6 Concerning the Suzuki and Takaku’s classification criteria,6 we found an acceptable correspondence between MRA and CCA. Of thirteen hemispheres, 11 showed the same pattern in MRA and CCA (stage agreement). In one of the hemispheres, underestimation of moyamoya vessels in MRA was responsible for discrepancy of MRA stage 2 and CCA stage 3. Overestimation of stenotic change in MRA was responsible for a difference between MRA stage 5 and CCA stage 4. To determine the angiographic stage, above described MRA characteristics such as underestimation of moyamoya vessels and overestimation of stenotic change are taken into consideration.

© 2000 Harcourt Publishers Ltd

Although our experience is limited, MRA has several advantages over CCA: 1) less complications; 2) combination of MR imaging and MRA; 3) shorter examination time; 4) repeatability as often as necessary, for example, in evaluation and follow up after revascularisation; 5) multiplanar visualisation; 6) a non-invasive method; 7) follow up of the progression of the disease to later stages or even progression of unilateral to bilateral disease with less risk; 8) examination as an outpatient; and 9) no requirement of contrast media. However, no method hitherto is perfect. MRA has some disadvantages as follows: 1) overestimation of stenotic lesions, and; 2) low spatial resolution, providing insufficient visualisation of vessels less than 1 mm. Finally, MRA has been a continuously developing field and MTC/MIP technique demonstrated in our study an improvement in visualisation of small vessels. Currently, MRA is, in our opinion, the best method of following up the patients with moyamoya disease. If additional investigations confirm the results of this study, namely that MRA and CCA provide redundant information, then the goal of MRA should be perhaps to replace CCA as the initial diagnostic tool as well as follow up of moyamoya patients.

REFERENCES 1. 2.

3.

4. 5.

6. 7.

8.

9.

Suzuki J, Kodama N. Moyamoya disease – a review. Stroke 1983; 14: 104–109. Fukuyama Y, Kanai N, Osawa M. Clinical genetic analysis on the Moyamoya disease: annual reports. The Research Committee on Spontaneous Occlusion of the Circle of Willis (Moyamoya disease) of the Ministry of Health and Welfare, Japan, 1992; pp 141–146. Yonekawa Y, Ogata N. Spontaneous occlusion of the Circle of Willis (cerebrovascular moyamoya disease): with special reference to its Å@ disease entity and etiological controversy. Brain Dev 1992; 14: 253–254. Ueki K, Meyer FB, Mellinger JF. Moyamoya disease: the disorder and surgical treatment. Mayo Clinic Proceedings 1994; 69(8): 749–757. Olds Mv, Griebel RW, Hoffman HJ, Craven M, Chuang S, Schutz H. The surgical treatment of childhood moyamoya disease. J Neurosurg 1987; 66(5): 675–680. Suzuki J, Takaku A. Cerebrovascular moyamoya disease. Disease showing abnormal net-like vessels in base of brain. Arch Neurol 1969; 20: 288–299. Yonekawa Y, Handa H, Okuno T. Moyamoya disease: diagnosis, treatment and recent achievement. In: Barnett HJM, Stein BM, Mohr JP, Yatsu FM (Eds). Stroke: pathophysiology, diagnosis and management. New York: Churchill Livingstone, 1986; pp 805–829. Yamada I, Matsushima Y, Suzuku S. Moyamoya disease: diagnosis with three-dimensional time-of-flight MR angiography. Radiology 1992; 184: 773–778. Houkin K, Aoki T, Takanashi A, Abe H. Diagnosis of moyamoya disease with magnetic resonance angiography. Stroke 1994; 25: 2159–2164.

Journal of Clinical Neuroscience (2000) 7(2), 112–115