Perfusion in Dural Arteriovenous Fistulas

Perfusion in Dural Arteriovenous Fistulas

Shalini A. Amukotuwa

Roland Bammer

Definition and Angioarchitecture

Intracranial dural arteriovenous fistulas (DAVFs) are a type of high flow intracranial arteriovenous shunt. They are a heterogenous group of lesions, unified by a common angioarchitecture: arteriovenous shunts within the wall of a dural venous sinus supplied by dural arteries (Fig. 52.1A). DAVFs are distinct from arteriovenous malformations (AVMs), lacking a nidus.1 They are also etiologically different from AVMs. DAVF lesions in adults are composed of a network of tiny “crack-like” vessels in the wall of a dural sinus. DAVF lesions in infants are rare and often have multiple high-flow arteriovenous shunts, involving several different dural sinuses.2

DAVFs are therefore characterized by direct flow of arterial blood into the venous system without an intervening capillary bed or arterioles.3 This “short circuit” of the capillary bed results in more rapid transit of blood from the arteries to the veins than is normal, giving rise to the characteristic early venous filling on dynamic contrast imaging (dynamic contrast-enhanced magnetic resonance imaging and catheter-based angiography) (Fig. 52.2). Note that the absence of resistance vessels also results in transmission of systemic arterial pressure into venous structures, with markedly increased blood flow in draining venous structures. This causes structural and hemodynamic changes in the cerebral vasculature, including dilatation and tortuosity of the draining veins, with the formation of venous pouches and varices (Figs. 52.1E, F and 52.3). As a consequence of elevated venous pressure, DAVFs can lead to cerebral edema and intracranial hemorrhage, with potentially devastating clinical consequences.

Direct carotid-cavernous sinus fistulas (CCFs) are also a type of acquired intracranial arteriovenous fistula, where there is direct communication between the cavernous segment of the internal carotid artery and the cavernous sinus, resulting from either rupture of a Berry aneurysm or traumatic injury of the cavernous internal carotid artery. Unlike indirect CCFs, which are cavernous sinus DAVFs, direct CCFs are etiologically and structurally distinct entities and therefore are excluded from the following discussion.

Historical Context

DAVFs were first described over a century ago; in 1873, Rizzoli documented the case of a 9-year-old girl with seizures and pulsatile occipital swelling, who on postmortem examination was found to have direct communication between enlarged branches of the occipital artery and the transverse sinus.4 DAVFs were first recognized on imaging in 1931 when a case of direct connection between the meningeal arteries and dural venous system was demonstrated on catheter-based angiography.5 It was not until much later, however, during the 1970s and well into the 1990s, that the etiology, angioarchitecture, clinical presentation, and prognosis of DAVFs were better elucidated.

French investigators Castaigne et al. (1976)6 and Castaigne and Djindijian (1977)7 were the first to suggest that DAVFs were acquired lesions, resulting from the opening of microshunts or angiogenesis in the dura mater between meningeal arteries and veins. Around this time, descriptions of hemorrhage—secondary to DAVFs—challenged the view of DAVFs as clinically benign entities. Houser et al.8 postulated that the clinical symptoms and risk of intracranial hemorrhage are related to the pattern of venous drainage of a DAVF; they proposed that intracranial hemorrhage (ICH) was a consequence of venous drainage limited to pial veins, in particular when associated with a dilated arteriovenous pouch.

The earliest classification of DAVFs, correlating the patterns of venous drainage with symptoms, was put forward by Djindjian et al.9 in 1977. According to this system, DAVFs draining freely into dural venous sinuses were deemed clinically benign, while cortical venous drainage was thought to produce aggressive neurologic symptoms and hemorrhage.

Subsequently, three comprehensive reviews of the literature more systematically explored the relationship between the pattern of venous drainage and ICH and neurologic symptoms. A review of 223 previously reported cases concluded that DAVFs draining into large dural venous sinuses were less likely to produce ICH than lesions with restricted dural outflow.10

A meta-analysis of 191 cases examined the mechanism of neurologic manifestations, concluding that peripheral cranial nerve palsies resulted from arterial steal phenomena, while central nervous system symptoms appeared to be related to passive venous hypertension.11 A comparison of the angiographic features of 100 aggressive DAVFs, defined as those manifesting with hemorrhage or focal neurologic deficit, with 277 benign fistulas found leptomeningeal venous drainage, variceal or aneurysmal venous dilatations, and vein of Galen drainage to be associated with an aggressive clinical course. No DAVF, regardless of location, was found to be definitively benign.12

FIGURE 52.1. Dural arteriovenous fistula (DAVF) angioarchitecture and Cognard classification. A: Cognard type I DAVF: a network of fistulous channels within the wall of a dural sinus is supplied by one or more dural arteries, and drains directly into a dural venous sinus. B: Cognard type IIA: fistula drainage into a dural sinus, with an obstruction (stenosis or occlusion) leading to retrograde flow in the sinus. C: Cognard type IIB: fistula drainage into a dural sinus, with antegrade flow within the sinus, but reflux of blood into a cortical vein (i.e., retrograde cortical venous drainage due to increased pressure in the sinus). D: Cognard type IIA + B: fistula drainage into a dural sinus, with an obstruction (stenosis or occlusion) leading to retrograde flow in the sinus as well as cortical vein reflux. E: Cognard type III: direct drainage of the fistula into a cortical vein. F: Cognard type IV: direct drainage of the fistula into a cortical vein, with dilatation of the cortical vein (>5 mm) or a focal varix.

FIGURE 52.2. Digital subtraction angiogram, demonstrating a Cognard type IIB dural arteriovenous fistula of the right transverse sinus. This dynamic modality allows sequential visualization of (A) the right middle meningeal (straight red arrow) and occipital arterial feeders; (B) the fistula (curved red arrow); (C) early venous filling (indicative of arteriovenous shunting) of the right transverse sinus (light blue arrow) with reflux into the cortical vein of Labbe (dark blue arrow); (D) antegrade flow within the right transverse sinus, and cortical vein reflux, into the vein of Labbe; (E, F) drainage through the vein of Labbe and the vein of Troland into the superior sagittal sinus (light blue arrow).

These studies were followed by work of Cognard et al.,13,14 who demonstrated in their seminal paper that DAVFs do not always have a benign clinical course, with the outcome depending on the degree of venous hypertension and cortical venous drainage. They put forward a classification system relating the pattern of venous drainage with the risk of hemorrhage; this Cognard system is now routinely used by both clinicians and researchers and is detailed below.

Location, Arterial Supply, and Venous Drainage

The majority of adult DAVFs are located in either the posterior fossa, involving the transverse or sigmoid sinuses, or the cavernous sinuses. For instance, of the 205 cases in the series by Cognard et al.14 50% were located in the transverse sinus, 16% in the cavernous sinus, 12%
in the tentorium cerebelli, and 8% in the superior sagittal sinus.

FIGURE 52.3. Complications of dural arteriovenous fistulas (DAVFs) seen on T2-weighted magnetic resonance imaging. A: T2-hyperintense vasogenic edema (red arrow) adjacent to the venous varix (blue arrowhead) of the draining cortical vein in a Cognard type IV DAVF. B: A small amount of hemosiderin staining, indicative of previous hemorrhage.

Paediatric lesions are usually more complex, with bilateral arterial feeders and torcular herophili, superior sagittal sinus, vein of Galen, or large venous lake involvement.15

DAVFs are characterized by a dural arterial supply. Cognard et al.14 reported that 95% of DAVFs had purely meningeal afferent arterial supply, with only 5% being fed by both meningeal and cortical branches.

Incidence and Demographics

DAVFs account for 10% to 15% of intracranial arteriovenous lesions.16 They are being diagnosed with increased frequency. This apparent rise in incidence is due to an increase in the number of incidentally detected DAVFs; and this is attributable to the exponential increase in the utilization of magnetic resonance imaging (MRI) for investigation of a wide range of neurologic complaints, in addition to the improved sensitivity of MRI for the detection of vascular lesions.17

The age range of those with DAVFs is extremely broad, ranging from neonates to the elderly. The mean age of presentation is 50 to 60 years.18 There is no gender predilection, although an increased incidence of hemorrhage in male patients with DAVFs has been reported.19


Unlike AVMs, which are described in Chapter 51, DAVFs are thought to be acquired lesions. The exact pathogenesis is unknown, and the majority of DAVFs are idiopathic.18 An association with dural venous sinus thrombosis has been reported.20,21,22 This etiologic link is supported by the association between DAVFs and inherited thrombophilias, such as antithrombin, protein C, and protein S deficiencies, which are risk factors for venous thrombosis.23,24

DAVFs have been proposed to develop as a result of recanalization of the thrombosed sinus. One hypothesis is that physiologic arteriovenous shunts between meningeal arteries and a dural sinus enlarge as a consequence of elevated local venous pressure in the setting of venous outflow obstruction due to thrombosis, in turn leading to pathologic shunting.19,25,26 Alternatively, venous hypertension in thrombosis can decrease cerebral perfusion, which may promote neoangiogenesis (Fig. 52.4).19,27

Clinical Presentation and Prognosis

The clinical presentation of DAVFs is highly varied and depends on the location and the pattern of venous drainage. At one end of the spectrum, they may be clinically silent. When present, symptoms are often nonspecific. In the posterior fossa, which is the most common location for DAVFs, pulsatile tinnitus is a frequent symptom, which results from increased blood flow through dural venous sinuses, particularly with transverse and sigmoid sinus fistulas.17,18,28 Other nonaggressive neurologic presentations include isolated headache and bruit.14

FIGURE 52.4. Flow chart of a proposed theory of the pathogenesis of dural arteriovenous fistulas. In the setting of dural venous sinus thrombosis, which may result from a number of causes, venous hypertension can cause decreased cerebral perfusion. This decreased perfusion can result in tissue hypoxia, which in turn may promote neoangiogenesis.

Cranial nerve palsies may occur, with DAVFs being a rare cause of trigeminal neuralgia.29 Nonspecific but aggressive neurologic presentations secondary to intracranial hypertension include intractable headaches, nausea and vomiting, visual disturbance or loss, and papilledema.14,17,30 Aggressive presentation with seizures, pronounced vertigo, altered mental status, cognitive decline, and focal neurologic deficits may result from cerebral edema, ischemia, and ICH.11,14,19,30,31,32,33,34 Therefore, although extra-axial in location, DAVFs can produce sequelae similar to AVMs, including parenchymal hemorrhage (Fig. 52.5). In fact, ICH occurs in 15% of patients with DAVFs, most often parenchymal or subarachnoid.11 As such, in cases of unexplained ICH, including subarachnoid and parenchymal hemorrhage, DAVFs should be considered in the differential diagnosis.18

Cavernous sinus DAVFs (indirect CCFs) are mentioned here separately, as this is an example of how the location of the fistula results in a distinctive clinical picture. They present with signs and symptoms related to cavernous sinus and orbital congestion: visual disturbance, ophthalmoplegia, retro-orbital pain, proptosis, and chemosis.18,19,31 The ophthalmologic risk of these cavernous sinus DAVFs, which can threaten vision, is related to ophthalmic vein drainage. Unfortunately, this risk is not reflected in the Cognard classification but is clinically an important issue, which must be considered in making treatment decisions regarding fistulas in this location.

The postulated mechanism of intracranial hypertension associated with DAVFs is an increase in cerebral
blood volume and dural sinus pressure, directly related to the increased flow rate inside the fistula.14,35 This increased venous sinus pressure (venous congestion) results in a reduction of cerebrospinal fluid absorption, and increased cerebrospinal fluid pressure as well as cortical venous pressure (Monroe-Kellie rule).14,35,36 This is exacerbated by venous outflow obstruction, for instance due to draining dural sinus stenosis or thrombosis.

FIGURE 52.5. T2-weighted magnetic resonance imaging (A) shows parenchymal hemorrhage (blue arrow) and edema in a patient with a right sigmoid/transverse sinus junction dural arteriovenous fistula (red arrow) seen on three-dimensional time-of-flight magnetic resonance angiography (B). The red arrow indicates curvilinear high signal structures, which represent the fistula itself. Linear high signal structures in adjacent bone are transosseous extracranial arterial feeders (of occipital artery origin). There is high signal, indicating fast flow due to arteriovenous shunting, in the right sigmoid sinus itself.

The pathogenesis of venous ischemia is less well understood than arterial ischemia. Disturbed venous flow causes increased intracranial pressure, reduced relative cerebral blood volume (rCBV), and reduced cerebral perfusion pressure. This in turn leads to disruption of the blood–brain barrier, causing vasogenic edema and hemorrhage (Fig. 52.5), which are the predominant features of venous ischemia (although cytotoxic edema can also occur).37,38 In keeping with this theory, increased rCBV and prolongation of bolus contrast passage have been observed in venous ischemia due to deep cerebral vein thrombosis, findings that can likely be extrapolated to other conditions causing venous hypertension such as DAVF.39

The risk of hemorrhage with DAVFs therefore appears to be related to the presence of venous hypertension. Concordant with this theory, the incidence of intracranial hemorrhage increases with cortical venous reflux and drainage, which in turn increase venous pressure.14,30 The risk is particularly high in direct drainage. Because DAVFs are acquired lesions, an arteriovenous shunt is abruptly established on a mature cortical vein. Cortical veins are smaller and less capacious than dural sinuses. Consequently, failure of this previously normal cortical vein to adapt to the increased flow rate and pressure induced by the fistula leads to elevated pressure, predisposing to hemorrhage.14 Even following subtotal occlusion of the fistula, DAVFs produce a risk of hemorrhage.14

It can be seen that the presence and magnitude of venous hypertension, which are related to the venous drainage pattern of a DAVF, determine symptoms and their severity, as well as the natural history. As such, the venous drainage pattern is the basis of the two DAVF classification schemes, of Cognard and Borden, discussed in detail later.14,30 It should be remembered, however, that cortical venous drainage (CVD) is not a sine qua non for hemorrhage. Willinsky et al.40 reported that tortuous, engorged pial veins were an indicator of venous congestion and a risk factor for hemorrhage, even in the absence of CVD.

In the presence venous ectasia, DAVFs may also present with progressive tumor-like symptoms induced by positive mass effect and mechanical compression of adjacent structures.14

In intracranial DAVFs with spinal perimedullary venous drainage, which are rare, myelopathy occurs due to spinal cord venous hypertension.

The Cognard and Borden Classifications

Cognard et al. performed a detailed analysis of 205 consecutive patients with DAVFs seen at their institution over 18 years, with the purpose of completing and validating the classification system first proposed by Djindjian et al.14 Widely used and the standard system of classification in clinical practice and the literature, the Cognard grading system enables prediction of the risk of a DAVF in order to make a better informed decision regarding therapy.

The Cognard classification consists of five main types of venous drainage (Fig. 52.1):

  • Type I DAVFs drain into a dural venous sinus, with normal antegrade flow within the draining sinus (Fig. 52.1A).

  • Type II fistulas drain into a venous sinus with insufficient antegrade venous drainage and consequent reflux. Insufficient venous drainage arises from either of two mechanisms of mismatch between the elevated flow due to arteriovenous shunting and the drainage capacity of the dural sinus:

    • Downstream obstruction due to stenosis or occlusion of the sinus into which a moderate flow fistula drains; or

      FIGURE 52.6. Digital substraction angiography. A: Left tentorial dural arteriovenous fistula (red arrow), draining directly into cerebellar leptomeningeal veins (blue arrows), making this a Cognard type III fistula. It is important to notice the early venous filling in the early arterial phase (prior to opacification of any distal arterial branches or the capillary bed). This early venous filling is consistent with rapid arteriovenous shunting. There is progressive opacification of the cerebellar veins (B, C), which subsequently drain into the left transverse sinus (D) (curved blue arrow indicates the transverse-sigmoid junction). The green arrow indicates a feeding artery, arising from the occipital artery.

    • A very high flow rate in a high-flow fistula that cannot be accommodated by a normal or even enlarged sinus.

    Three subtypes are distinguished on the basis of the pattern of retrograde venous drainage:

    • Type IIA: retrograde drainage is into venous sinus(es) only (Fig. 52.1B);

    • Type IIB: retrograde drainage into cortical vein(s) (i.e., indirect) CVD (Figs. 52.1C and 52.2); and

    • Type IIA + B: drainage into both sinus(es) and cortical vein(s) (Fig. 52.1D).

  • Types III and IV fistulas drain directly into a cortical vein, rather than a dural sinus. Although the arteriovenous fistula itself lies in the dura of the sinus, it does not primarily drain into the sinus, instead draining into an adjacent cortical vein. This cortical vein may drain into the sinus (i.e., the sinus may secondarily drain the arteriovenous fistula, downstream from the draining cortical vein).

    • Type III fistulas: drain directly into a cortical vein, without ectasia of draining venous structures (Figs. 52.1E and 52.6).

    • Type IV fistulas: drain directly into a cortical vein, with associated venous ectasia (defined as a venous structure of diameter greater than 5 mm and three times larger than the draining vein) (Figs. 52.1F and 52.7).

  • Type V (intracranial) DAVFs drain into spinal perimedullary veins and are rare.14

The Borden classification, used by some, is essentially a simplification of the Cognard classification. DAVFs are divided into three categories, on the basis of the site and pattern of venous drainage30:

  • Type I: Drainage into meningeal veins or into a dural venous sinus, with normal antegrade flow.

  • Type II: Drainage into a dural venous sinus or meningeal veins with retrograde flow into subarachnoid veins.

  • Type III: Direct drainage into subarachnoid veins (not through a dural sinus). Direct drainage into an “isolated” segment of dural sinus, resulting from thrombosis on either side of the arterialized sinus segment, is also categorized as type III, as this is associated with retrograde flow into the subarachnoid venous system. This is equivalent to Cognard types III, IV, and V fistulas.

The Borden classification scheme further subclassifies lesions as single-hole (a) or multiple-hole (b) fistulas.

It is noted that the Cognard classification has a greater ability to separate patients prognostically, hence is more widely used.

FIGURE 52.7. Time-resolved, contrast-enhanced magnetic resonance angiography (MRA), allowing noninvasive dynamic characterization of a Cognard type IV dural arteriovenous fistula through sequential imaging following injection of a bolus of contrast, with 1.3-second temporal resolution. (Please note the time stamps at the bottom of each image are measured from the start of the bolus injection into a peripheral vein.) A: Early arterial phase, demonstrating the fistula (red arrow) in the wall of the right transverse sinus. B: Later in the arterial phase, demonstrating early filling of a cortical vein (straight blue arrows), consistent with arteriovenous shunting; absence of contrast enhancement of the dural sinus at this time is consistent with direct cortical venous drainage of the fistula. C: Continued filling of the cortical vein (blue arrows) and faint enhancement of the superior sagittal sinus (D, E). Focal variceal dilatation of the cortical vein (curved dark blue arrow), which drains into the superior sagittal sinus (violet arrow). Unfortunately, due to the restricted field of view of the sagittal computer-enhanced-MRA slab, the arterial feeders located more laterally to the slab were not visualized.

Risk Factors for Aggressive Neurologic Presentation and Natural History

A good understanding of the natural history of DAVFs, and hence their prognosis, is vital for decision making regarding therapy.

In both the Cognard and Borden classification schemes, CVD is associated with an increased risk of hemorrhage as well as nonhemorrhagic neurologic sequelae.14,17,28,30,31,32,41,42,43

The benign behavior and natural history of fistula without CVD has been established. Cognard et al.14, observed no hemorrhages in type I and IIA fistulas, while Borden et al.30 observed aggressive behavior in only 2% of type I fistulas. Therefore, these fistulas are usually detected incidentally or present with nonaggressive symptoms related to increased venous drainage such as pulsatile tinnitus or exophthalmos.18

Aggressive presentations such as intracranial hemorrhage are associated with CVD (Cognard types IIB, IIA + B, III, IV, and V fistulas; and Borden types II and III). In the presence of CVD, an 8.1% annual risk of ICH and 10.4% annual mortality rate have been reported.31 It has been shown, however, that this risk is lower if the patient is asymptomatic; the annual risk of ICH in symptomatic patients was 7.4% versus 1.5% in asymptomatic patients.17

The risk of hemorrhage and aggressive neurologic symptoms increases with the Cognard and Borden type: direct CVD of a fistula (Cognard type III or greater; Borden type III) produced a higher incidence of aggressive symptoms and ICH than indirect CVD secondary to reflux from a sinus, posing a significant risk to the patient.14,30 Cognard et al. observed aggressive neurologic symptoms in 76% of type III fistulas and 97% of type IV fistulas. Borden et al. observed aggressive behavior in 39% of type II fistulas, in comparison with 79% of type II fistulas.

The presence of venous ectasia significantly further increases the risk of hemorrhage, increasing the incidence from 40% in Cognard type III fistulas to 66% in type IV14; therefore, a DAVF with venous ectasia, due to the high risk of hemorrhage, is an indication for urgent treatment. According to Cognard et al., it was uncertain as to whether the ectasia per se was responsible for the bleeding or whether the ectasia reflected insufficient venous drainage, in turn leading to escalating venous hypertension and hemorrhage. Figure 52.3 shows vasogenic edema and a small amount of hemosiderin staining subjacent to the markedly ectatic draining cortical vein of the Cognard type IV fistula demonstrated in Figure 52.7. This patient presented with worsening headaches and visual disturbance.

In addition, the location of the fistula was also significantly correlated with the presence or absence of aggressive symptoms. For instance, no cavernous sinus DAVFs were associated with an aggressive presentation, while 100% of fistulas at the torcular herophili, 88% of anterior base of skull, and 92% of tentorial DAVFs demonstrated aggressive symptoms.14 This difference is in part accounted for by anatomy, namely, the difference in venous drainage patterns in each location. For instance, all tentorial and anterior cranial fossa DAVFs have CVD (i.e., type III or IV), while only 12% of cavernous sinus fistulas had cortical venous reflux.14 Taking this difference in the presence of CVD by location into account, location was not an independent risk factor for neurologic risk.

The presence of a cortical arterial supply (5%) also did not alter neurologic prognosis. Aggressive neurologic outcomes such as hemorrhage were observed more frequently in males than females, due to CVD or reflux being more common in males than females.14

Although progression of DAVFs from one type to another is unusual in the absence of treatment, it must be emphasized that DAVFs are a dynamic entity, with
2% of low-grade DAVFs reported to progress to a more higher-grade DAVF.14,18,28,43 CVD can develop over time due to the development of venous outflow obstruction (stenosis or thrombosis) or increased arterial flow,28,43 and close follow-up is necessary. Therefore, any change in symptomatology should prompt further investigation to look for any changes in venous drainage and progression of the fistula to a higher grade.14 DAVFs may also spontaneously resolve or undergo thrombosis.44,45

Treatment Implications

Decisions regarding whether and when to treat a DAVF are best made by a multidisciplinary team of neurosurgeons, interventional neuroradiologists, neurologists, and radiation oncologists.18 It needs to be emphasized that treatment is not without risk, and the team needs to situationally weigh the pros and cons of a particular treatment for each patient individually; potential complications include stroke and intracranial hemorrhage. Therefore, these risks must be weighed against the risk posed by the DAVF itself. Patient factors (e.g., age, condition, other comorbidities), the clinical presentation, and the type of DAVF (location and Cognard grade) must be considered. High-grade lesions with CVD should be treated to avoid intracranial complications such as hemorrhage. Conservative treatment is adopted for low-grade fistulas (Cognard types I and IIA), with close clinical and imaging surveillance to detect progression. Debilitating symptoms, such as severe tinnitus, which compromise the patient’s quality of life, may necessitate advanced treatment of these low-grade fistulas.

Endovascular techniques have become the mainstay of DAVF treatment.18 Embolization via a transarterial, transvenous, or, less often, combined route is the preferred therapeutic approach. The optimal endovascular technique is still debated. Surgery is reserved for those cases that are either not amenable to endovascular therapy or where such therapy has failed. Stereotactic radiosurgery has been utilized in low-grade fistulas and cases where endovascular and surgical approaches are not feasible. Discussion of the specific endovascular and surgical approaches and their advantages and disadvantages is beyond the scope of this chapter, however, a good overview is given by Gandhi et al.18

The goal of therapy is elimination of the arteriovenous shunt; with incomplete therapy, recruitment of collateral vessels can occur, resulting in fistula persistence and even progression; therefore, the risk of hemorrhage is not eliminated. In cases where shunt occlusion is not possible or deemed too risky, selective treatment of CVD is undertaken.18 Disconnection of CVD may be comparable to complete shunt obliteration, in terms of reducing the risk of DAVF complication, with a lower procedural risk.46

Structural Imaging of Dural Arteriovenous Fistulas

As discussed previously, the signs and symptoms of DAVFs can be nonspecific; of those patients presenting with a clinical picture potentially attributable to a DAVF, only a few actually have one!

The gold standard for the diagnosis of a DAVF is intra-arterial catheter-based digital subtraction angiography (DSA). However DSA is an invasive test, which utilizes ionizing radiation, and is not always readily available at all centers.47 It also carries a very small but significant risk of permanent neurologic injury. Therefore, noninvasive techniques are necessary to better triage those patients with potential DAVF symptoms in whom the index of suspicion on clinical grounds alone is insufficient to justify a confirmatory DSA.47 Noninvasive techniques also enable treatment planning so diagnostic and therapeutic angiography can potentially be performed at the same session, decreasing the risk to the patient as well as health-care costs.

Only gold members can continue reading. Log In or Register to continue

Oct 7, 2018 | Posted by in MAGNETIC RESONANCE IMAGING | Comments Off on Perfusion in Dural Arteriovenous Fistulas
Premium Wordpress Themes by UFO Themes