MGHbanner BulfinchBldg
[Cerebrovascular Surgery Center]
Neurovascular Center at MGHmGH Neurosurgical Service
Massachusetts General HospitalHarvard Medical School

MGH Shield Hvd Med Sch Shield Partners Logo
For patients with aneurysms and arteriovenous malformations (AVMs) as well as other neurovascular problems of the brain and spinal cord. Including the Neurovascular News; - A newsletter with information regarding carotid endarterectomy to prevent stroke, transient ischemic attack or TIA; brain aneurysms; arteriovenous malformations and other vascular lesions affecting the brain and spinalcord with information on the MGH Brain AVM and Aneurysm Center.
Neurosurgery @ MGHPeople @ MGH NeurosurgeryClinical Centers @ MGH NeurosurgeryResearch @ MGH NSEducation @ MGH NeurosurgerySupport Groups @ MGH NeurosurgeryNews @ MGH NeurosurgeryReferrals @ MGH Neurosurgery
[Cerebrovascular Surgery Center]

Neurovascular Surgery
Brain Aneurysm & AVM Center


Moyamoya is a Japanese term which translates in english to: "cloud of smoke" or "puff of cigarette smoking drifting in the air" and it has been used to define a classic angiographic appearance of multiple small intracranial vessels (Figure 1). Although described first in Japan in the 1950s this form of cerebrovascular disease is not limited to the Japanese population and has been reported sporadically all over the world with cases described in the United States, Europe, Australia and Africa. The incidence seems to be roughly one in a million people per year. There seem to be two definite peaks of incidence: children under 10 years old and adults in their third to fifth decades of life. A slight female preponderance has been shown. There is some evidence of familial tendency based on association between moyamoya and certain HLA types. Familial incidence is estimated about 7 to 12% around the world.

The Moya moya pattern of vessels seen on angiography is thought to be a phenomena secondary to intracranial large vessel narrowing or stenosis. The response of the cerebral vasculature to this type of narrowing is for more distal vessels to proliferate. There is debate as to whether the vascular abnormality represents a congenital problem or an acquired stenosis of intracranial vessels that occurs early in life. Moyamoya type changes have been found in a variety of diseases, including sickle-cell disease, neurofibromatosis, trisomy 21 and fibromuscular dysplasia. Other predisposing conditions for this problem include an auto immune process, cranial trauma, anaerobic bacteria or the use of oral contraceptive but none has been convincing.

There are no specific symptoms or signs related to moyamoya syndrome. The various clinical manifestations are generally caused by cerebrovascular ischemic or hemorrhagic events. Headaches and seizures are also seen. Clinically the disorder presents in children with transient ischemic attacks (TIA) frequently with episodes of hemiparesis or other focal neurological signs, often precipitated by physical exercise or hyperventilation. There may be a more chronic course of worsening with a gradual impairement of intellectual deterioration. Adults in contrast, usually present with intracerebral hemorrhage, most frequently in the thalamus, basal ganglia or deep white matter. Subarachnoid or intraventricular hemorrhage may also be observed. Other signs and symptoms seen in children and adults are disturbance of consciousness, speech disturbance, sensory impairment, involuntary movement and visual disturbance.

The prognosis of moyamoya is difficult to predict because the natural history of the disease is still unclear. Autopsy studies have shown severe vascular occlusive changes in the intracranial portion of the ICA usually bilateral and in the main arteries that make up the circle of Willis. These changes are characterized by endothelial hyperplasia and fibrosis with intimal thickening and abnormalities of the internal elastic lamina, while the adventitia and media remain normal. Descriptions of the vertebrobasilar system are not available. Extracranial arteries at the level of heart, kidney and other organs may show the same intimal lesions as the intracranial arteries supporting the belief the moyamoya syndrome can be a more generalized systemic vascular syndrome. Inflammatory cells or atheroma are not typically seen. The intracerebral perforating arteries around the circle of Willis show micro aneurysm formation with fibrin deposition and thinning of the vessel wall. These types of changes are thought to be responsible for the occurrence of intracerebral hemorrhage. It is postulated that there is increased blood flow through these small fragile vessels making them prone to hemorrhage.

Cerebral angiography is the cornerstone of the diagnosis of moyamoya syndrome. The characteristic angiographic findings of moyamoya syndrome are a symmetrical stenosis (tapering) or occlusion of the intracranial internal carotid artery, as well as the origin of the anterior and middle cerebral artery associated with an enlargement of the basal penetrating branches of these arteries in an apparent attempt to provide collateral circulation and giving the classic "cloud of smoke" appearance (Figure 1). Computed tomography scaning shows non specific findings. There may be ischemic areas of the cortex and sub cortical white matter with evidence of old areas of infarction. There may be mild ventricular dilatation or dilated sulci and fissures. In the case of intracerebral hemorrhage the CT scan will show the location of the intraventricular, subarachnoid and intraparenchymal hemorrhage that usually occurs in the basal ganglia or thalamus (Figure 2). Magnetic resonance imaging may better visualize areas of cerebral infarction due to moyamoya. Usually these infarctions are multiple, small and asymptomatic. Infarction is seen predominantly in the watershed areas of the carotid circulation at the borderzone between the areas of the brain supplied by the middle cerebral artery and anterior cerebral artery. Magnetic resonance angiography (MRA) can be used to detect the abnormal intracranial vessels although its resolution does not yet allow the visualization of the abnormal basal penetrating vessels. Other techniques including positron emission computed tomography (PET), single photon emission computed tomography (SPECT) and perfusion MRI studies have been used to study regional cerebral blood flow in moyamoya patients. There are still not enough available data to draw any conclusion about the usefulness of these techniques in the diagnosis of this condition. Transcranial Doppler has recently been used to study patients with moyamoya syndrome and has been shown to be a very useful non invasive technique to follow changes in larger vessels with time.

The best treatment is not known. The treatment of moya moya patietns depends on the pattern of symptoms. For patients with ischemic events and infarction, medical therapy consists of management with steroids in certain instances. Aspirin, ticlopidine and occasionally vasodilators and anticoagulants may be used. No study has supported the definitive efficacy of any medical treatment. A variety of different surgical revascularisation procedures have been used, but whether they improve the outcome is not yet known. Superficial temporal artery-middle cerebral artery bypass, encephalodurosynangiosis, omentum transplantation and cervical sympathectomy are options. The main purpose of surgical procedures is to provide additional collateral flow to an area of ischemic brain and therefore to prevent further damage. Encephalodurosynangiosis (EDAS) is performed with the intent to divert flow from the external carotid artery into the internal carotid artery system by applying branches of the superficial temporal artery or the temporal muscle to the brain surface of a patient. Finally omentum transplantation is performed with the intent to revascularize ischemic tissue. Cervical sympathectomy including stellate ganglionectomy is performed with the intent of improving cerebral blood flow. For treatment of hemorrhage, hematoma evacuation and ventricular drainage are the usual methods of treatment.

There is no specific medical or surgical therapy proven to reduce subsequent hemorrhage. Some patients with moyamoya stabilize clinically, often after they have developed disabilities, others continue to show progressive deterioration despite treatment. Although no definite effective treatment has been determined, surgical therapy to augment collateral circulation appears to be a promising treatment for patients with relapsing ischemic events.

Guy Rordorf M.D.* and Christopher S. Ogilvy M.D.# Cerebrovascular Surgery#, Neurosurgical Service and Stroke Service* Department of Neurology, Massachusetts General Hospital

Takeuchi K and Shimizu K: Hypoplasia of the bilateral internal carotid arteries. Brain and Nerve 9:37, 1957
Hanakita J, Kondo A, Ishikawa J et al.: An autopsy case of moyamoya disease. Neurol Surg 10:531, 1982
Handa J and Handa H: Progressive cerebral arterial occlusive disease: analysis of 27 cases. Neuroradiology 3:119, 1974
Karasawa J, Kikuchi H and Furuse S: Subependymal hematoma in moyamoya disease. Surg Neurol 13:118, 1980


Figure 1: Anteroposterior and lateral angiogram of a patient with moya moya syndrome. The small vessels present are thought to develop in response to larger vessel stenosis and occlusions (arrow)

Figure 2: CT scan of a patient with moya moya syndrome after a hemorrhage. This young woman complained of a severe headache and became acutely unresponsive. The hemorrhage developped in the left caudate nucleus and extended into the ventricule.


Spinal dural arteriovenous fistulae (SDAVF) and other vascular malformations of the spinal dura were once thought to be exceedingly rare diseases. Although still uncommon in comparison to other mechanical diseases of the spine such as disc herniation or facet hypertrophy, etc. vascular diseases of the spinal dura are an important diagnostic group for which the clinician must maintain a high index of suspicion in patients presenting with acute or chronic spinal symptoms. The reason for this importance is that there is an initial window of opportunity when the patient first becomes symptomatic, possibly as short as 3 - 6 months, during which this disease can be reversed or cured completely. If the disease is not detected or treated during that time, most patients, in our experience, continue to pursue a progressive downhill course. Although occasional patients in whom spontaneous thrombosis of SDAVF have been reported, these patients are the exception. Left untreated, the prognosis for these patients is universally dismal with relentless and occasionally rapid progression to a state of total ascending paralysis, loss of autonomic control of bowel and bladder, and they invariably spend the rest of their lives bedridden and paraplegic. Many of these patients were formerly diagnosed as having spinal tumors, idiopathic myelopathy, or syndromes such as that of Foix-Alajouanine.


SDAVF occurs when a small shunt between the arteries of the dura and adjacent veins becomes established through means that are not understood. This usually occurs in close association with the exiting nerve roots of the spine mainly in the thoracic or lumbar areas. However, a similar fistula can occur in the epidural space or the surface of the spinal cord causing a similar clinical and angiographic appearance. The essence of the disease behavior derives from the subsequent venous hypertension which occurs in the subarachnoid spinal veins resulting in a diminution of the arteriovenous gradient in the spinal circulation. This causes a slowing of the spinal circulation-time which is seen on injection of the anterior spinal artery and a progressive venous engorgement of the spine resulting in edema, gliosis, breakdown of the blood brain barrier in the spine and venous infarction. When treated early, this process can be reversible. However, in patients who are treated late in the disease course, the process of gliosis and scarring appears to be self-perpetuating and these patients who do not respond well to treatment. Although most spinal dural arteriovenous fistulae occur in the lower thoracic or lumbar spinal area, it is possible for the shunt to occur anywhere between the sacrum and the foramen magnum. The resulting venous hypertension and associated phenomena will be similar in virtually all of these patients. Because the conus medullaris is at the most distal reach of the anterior spinal artery circulation, this is the area that is most susceptible to the effects of venous hypertension and the changes in the arteriovenous gradient. Patients will therefore present with conal symptoms of symmetric or asymmetric paresis in the lower extremities, sensory loss, ascending dysesthesia, an ascending radicular band of paresthesia or hyperesthesia, and subsequent loss of control of bowel and bladder function. Unlike the patients who have a disc herniation or other mechanical disease of the spine, radicular pain is not a prominent symptom in these patients.


The disease was previously described as one of elderly males. However a substantial number of patients are female or may be as young as their 30's. A typical presentation is a patient who has been seen by multiple physicians for symptoms of leg weakness and paresthesiae who has undergone multiple previous MRI examinations, CT scans and other investigations. A significant number of these patients in the course of their downhill progression may also have had spinal surgery for a laminectomy or discectomy, etc. and some will have undergone spinal biopsy for presumed spinal tumor. These biopsies are invariably inconclusive or normal demonstrating only non-specific gliosis. Occasionally the neurosurgeon will observe prominent vessels in the site of biopsy at the time of surgery.


The most important factor involved in the diagnosis of this disease is an awareness of the disease process in the clinician and a reasonable index of suspicion. Once the disease is suspected, appropriate radiographic studies ( MRI/ MRA and spinal angiography) should be obtained to conclusively rule out the possibility of spinal dural AVM. Despite the best intentions, however, it is possible for this disease to elude detection unless the parameters of imaging studies are modified to detect it. Specifically with MRI examinations of the spine it is necessary to use a small field of view, assure patient cooperation to eliminate motion artifact, and to perform a T1 sequence after administration of gadolinium to detect enhancement. Similarly when ordering a myelogram, it may be necessary to convey to the radiologist that supine and prone films of the thoracic and lumbar spine are requested to enhance sensitivity to detection of the abnormal vessels in the subarachnoid space (figure 1). MRI and myelography are sensitive to this disease when performed in a technically satisfactory manner.

MRI features of this disease (figure 2 and 4) are related to the venous hypertension of the spinal coronal venous plexus and to the gliosis and edema of the conus. Therefore, a sagittal T2 sequence will demonstrate T2 hyperintensity within the cord ascending occasionally up into the thoracic area. Flow voids and multiple serpiginous channels may be evident along the dorsum or ventral surface of the spine on the sagittal T1 and T2 weighted images. Administration of gadolinium may enhance detection of these serpiginous channels and may also demonstrate enhancement within the expanded conus. The expansion of the conus due to edema and the enhancement are two features that frequently lead to the misdiagnosis of spinal tumor in these patients. Myelographic examination may show a single distended or multiple distended venous channels with a serpiginous aspect outlined as a filling defect within the contrast. The absence of these findings on the myelogram, however, in the setting of a highly suspicious MRI should not deter referral of the patient for definitive evaluation, i.e. spinal angiography.


Spinal angiography is the definitive diagnostic procedure for evaluation of this disease. Spinal angiography is a laborious invasive test and technically difficult to perform which exposes the patient to the risk of multiple catheterizations of the segmental vessels from the aorta. As these patients are frequentlly elderly with atherosclerotic disease, these risks can involve embolization of the spinal arteries with the risk of subsequent myelopathic infarction. Although these complications are rare, they do need to be considered in advance of any such procedure. Additionally, the procedure usually requires general anesthesia, lasts for approximately three to four hours, and involves administration of an unusually high dose of contrast up to 400 cc or more. Nevertheless, in patients in whom the diagnosis of spinal dural arteriovenous fistula is suspected, this test is warranted as the definitive process for establishing the diagnosis.

During spinal angiography which is done through a transfemoral catheterization, each of the segmental vessels which supply the dura are catheterized in sequence on each side of the body. This can involve over 40 total catheterizations from the upper cervical level to the pelvis. As only one of these vessels will demonstrate the abnormality and all of the other vessel injections will be normal, there is no way of shortening the procedure and usually at the start of a procedure, there is no way of knowing which segmental level will be involved by the disease. The angiographic findings related to the disease will be evident at the level of the fistula where opacification of an abnormal distended venous channel following the course of the nerve root up to the conus will be identified with subsequent opacification of an abnormal rete or plexus veins around the spine (figure 3 and 5). Additionally, on injection of the level at which the artery of Adamkiewitz is performed, the angiogram will demonstrate slowing of flow in the anterior spinal artery. Localization of the postions of the anterior and posterior spinal arteries is an important part of the procedure if surgery is contemplated. However, all other levels of injection will be completely normal and thus the technique needs to be extremely precise so that no single vessel is inadvertently omitted during the examination.


There are two possibilities for treatment in these patients. If there is no evidence of a critical spinal vessel at the same level of the fistula, it is possible to embolize the fistula transarterially at the same session as the spinal diagnostic arteriogram. The only embolic agent which is permanently effective in the setting of such a fistula is an acrylate glue. Particles and coils may be temporarily effective but ultimately the fistula will re-establish itself. If transarterial embolization fails or is not possible due to the presence of an adjacent critical spinal artery, the neuroradiologist may insert a small metallic coil which can serve for fluoroscopic localization of the fistula during subsequent surgery. Surgical duraplasty of the site of the fistula involves stripping the dura at the precise level of the fistula and cauterization of the afferent and efferent vessels to prevent reconstitution of flow. It may be necessary to resect the associated thoracic nerve root in certain cases. This does not typically result in a permanent neurologic deficit.


In the past two years, our combined neurosurgical and neurointerventional clinic has evaluated over 20 patients with a high suspicion for this disease. In retrospect it appears that the presence of flow voids within the subarachnoid space on the MRI without evidence of T2 hyperintensity and without evidence of gadolinium enhancement of the conus probably predicts a negative spinal angiogram. In other words, distended vessels by themselves and the absence of other findings in the spine are probably related to some other process such as redundant veins. Patients who have been treated in the first few months of the clinical appearance of symptoms have invariably had an excellent responce to either surgical repair or to embolization. In the remainder of the patients that we have treated in various states of decline, some measure of clinical response has been evident in virtually all. The degree of recovery has usually, but not always, been related to the length of disease duration prior to treatment and to the extent of deterioration of the patient prior to treatment. In some patients who were treated late in the disease process, some small measure of initial recovery was evident but some of these patients have subsequently experienced another decline. Repeat angiography in these patients has demonstrated that the fistula has not re-established itself and it is assumed that the process of scarring or gliosis has become self-perpetuating in these unfortunate patients. In patients who have been treated early in the disease process, an immediate response, as soon as a few hours after the treatment procedure, can be seen with progressive improvement over the subsequent weeks.


SDAVF is a disease which should be suspected in spinal patients in whom the clincial symptoms, signs and rapid progression are not explained satisfactorily by whatever degree of degenerative change is evident on the spinal MRI. Degenerative changes are common in virtually all patients beyond late middle age. It cannot necessarily be assumed that these degenerative changes are responsible for the clinical presentation at hand. It is vital for the patient's well being that the possibility of this disease be pursued early and investigated fully to maximize chances of a total recovery.


Foix C, Alajouanine T. La myŽlite necroitique subague. Rev Neurol 1926, 33: 1-42.

Lasjaunias P, Berenstein A. Surgical Neuroangiography, Vols3 &4, Springer-Verlag, Berlin 1990.

Ojemann RG, Ogilvy CS, Crowell RM, Heros RC. Surgical Management of Neurovascular Disease. Baltimore, Williams & Wilkins 1995, pp.503-537.

Links to more information on spinal dural AVMs

Pearse Morris MB, BCh
In Sup Choi, M.D.
Christopher S. Ogilvy, M.D.


Figure 1: Lumbar myelographic image demonstrating obvious serpiginous filling defects due to distended subarachnoid veins. Many cases do not however demonstrate findings so prominent as these and a careful review of the films is necessary when the diagnosis is suspected.

Figure 2: 2A. Sagittal T1 weighted image of lumbar spine in a 75 year female (same patient as figure 1) with progressive lower extremity weakness and sensory loss. Expansion of the conus simulates appearance of a spinal tumor. 2B. Sagittal T2 weighted image demonstrates extensive elevation of signal in the cord and an appearance of expansion. Due to very slight motion artefact the flow-voids seen on the myelogram are obscured by CSF signal 2C. Axial T2 weighted signal of the conus demonstrated expansion and elevated signal simulated a tumor.

Figure 3: Selective injection of left L1 pedicle opacifies the fistula (small double arrow) with flow up the medullary vein (curved arrow) to opacify the coronal plexus (small single arrows) of the cord.

Figure 4: Sagittal T2 weighted image of spine in a 74 year male who had no improved of his leg weakness, sensory loss, or urinary incontinence after a lumar laminectomy for presumed disc disease. The elevated central cord signal is similar to the previous case. Note the dot like pattern on the surface of the cord from the distended veins and how these are prominent over the entire length of the cord.

Figure 5: Spinal angiogram on the same patient as figure 4 demonstrates an epidural variant of the disease. Selective injection of right L4 opacifies an epidural varix which drains via a left L4 medullary vein to the coronal plexus of the spine. P.Morris Spinal Dural Arteriovenous Fistula.

Disclaimer About Medical Information: The information and reference materials contained herein is intended solely for the information of the reader. It should not be used for treatment purposes, but rather for discussion with the patient's own physician. All visitors to this and associated sites from the Neurosurgical Service at MGH agree to read and abide by the the complete terms of legal agreement found at the Neurosurgery "disclaimer & legal agreement." See also: the MGH Disclaimer, the MGH Privacy Policy, and the MGH Interactive Program Disclaimer - Copyright 2006.
[Cerebrovascular Surgery Center]Referrals | AVMs and Cavernous malformations | Aneurysms Carotid disease, stroke, TIA
Neurovascular News
| Cerebrovascular Conference
AVM Support Group
| Booklets & Publications | Cerebrovascular Links
Physician Referral HotLine - THE SAH HOTLINE:
CALL 1-800-888-1SAH
[Functional and Stereotactic Neurosurgery] MGH Interventional Neuroradiology

MGH Acute stroke neurologist
(617) 726-2241 beeper #34282

MGH  Neurosurgical Service Home
Research@NeurosurgeryVisitors must read the disclaimer - legal agreement.
All Rights Reserved. Copyright 20007 MGH Neurosurgical Service

(internal access only)
System Info Contact: WebServant or the PageServant or e-mail C.Owen