Endovascular Treatment of Arteriovenous Malformations of the Supratentorial Compartment

20 Endovascular Treatment of Arteriovenous Malformations of the Supratentorial Compartment

Stephan Munich and Demetrius K. Lopes

Abstract

Supratentorial arteriovenous malformations (AVMs) represent the majority of cerebral AVMs. However, they represent a heterogenous group of AVMs ranging from small, superficial lesions located in silent regions of the brain to deep, eloquently located lesions with complex angioarchitecture. Endovascular embolization has been increasingly utilized for the treatment of the full range of supratentorial AVMs. While traditionally endovascular techniques have been utilized preceding surgical resection or stereotactic radiosurgery, recently, they are more frequently being employed as a stand-alone treatment strategy. Angiographic cure is possible with endovascular techniques in many lesions and may avoid the complications associated with surgical resection and radiation. Curative endovascular embolization likely will become even more commonplace as the development of endovascular techniques continues to evolve.

Keywords: arteriovenous malformations, ARUBA, embolization, microsurgery, n-butyl cyanoacrylate, Onyx, stereotactic radiosurgery

Key Points

Endovascular embolization can be used effectively preceding surgical resection and stereotactic radiosurgery.
Endovascular techniques can also serve as an effective stand-alone method to achieve angiographic cure of supratentorial arteriovenous malformations (AVMs).
The development of new liquid embolic mediums and microcatheters has allowed for more controlled and effective endovascular embolization.
With continued experience and development of endovascular technologies, the utilization of endovascular embolization to cure supratentorial AVMs is expected to expand.

20.1 Introduction

Arteriovenous malformations (AVMs) are abnormal connections of arterial and venous circulations, consisting of an intervening tangle of thin-walled vessels (nidus). They are rare vascular lesions, occurring in 15 to 18 per 100,000 adults in the general population with a detection rate of 1.21 per 100,000 person-years.¹^,²^,³ Hemorrhage is the most common presentation, with AVMs generally considered to carry an annual hemorrhage risk of 2 to 4%.⁴As with other intracranial lesions, the vast majority of AVMs occur in the supratentorial compartment. Therefore, consequent to their location, supratentorial AVMs may also present with seizure, headaches, or focal neurologic deficit. The ability to provide low-risk, effective, and durable treatment of these lesions is paramount. This is especially true following the recent publication of the Medical Management With or Without Interventional Therapy for Unruptured Brain Arteriovenous Malformations (ARUBA) trial, in which the necessity of treatment recently has been questioned.⁵ In this chapter, we discuss the role of endovascular strategies in the contemporary management of supratentorial AVMs. We will also discuss technical aspects unique to endovascular techniques and their utilization for the treatment of supratentorial AVMs.

20.2 Preoperative Assessment and Workup

Following the initial clinical and radiographic evaluation, the most important first step in the management of the AVM is the determination of whether it should be treated. Generally accepted indications for treatment of AVMs include repeated, symptomatic hemorrhages, seizures (particularly when there is intolerance to anti-epileptic medications or those refractory to medical management), focal neurologic deficit, and patient preference.

The recent publication of a randomized trial of unruptured brain arteriovenous malformations (ARUBA) trial questioned the necessity of treatment of unruptured (and particularly, asymptomatic) AVMs. The authors found that their primary endpoint (i.e., the composite of death or symptomatic stroke) occurred in 10.1% of patients undergoing medical management and in 30.7% of those undergoing AVM treatment. While a detailed discussion of the multiple shortcomings of this study is beyond the scope of this chapter, perhaps its most obvious and influential limitation is the short length of follow-up (mean 33.3 months; SD 19.7 months).⁶^,⁷ With an annual hemorrhage risk of 2 to 4%, the potential threat of AVM rupture, therefore, is clearly affected by the patient’s age and overall life expectancy, with young and healthy patients accumulating more risk over their lifetime (risk of hemorrhage = 1 – [annual risk of not bleeding] expected years of remaining life).⁸ Therefore, despite the conclusions of the authors of the ARUBA study, the decision to not treat an AVM in a young and healthy patient should be undertaken with caution.

In addition to the aforementioned generalized hemorrhage risk, the application of various risk scores has been developed to aid in decision making. The Spetzler-Martin score is the most widely known and utilized.⁹ Though it is frequently used to describe the radiographic appearance/location of AVMs, it was designed to characterize surgical difficulty and morbidity associated with surgical resection. Therefore, its application to lesions planned for endovascular or radiosurgical treatment must be done with caution and understanding that the continuum of associated morbidity described by this score may be skewed. On the light of this, we, as well as others, recently have proposed grading scales specific to the nuances of endovascular treatment (► Table 20.1).¹⁰^,¹¹^,¹²^,¹³^,¹⁴

Table 20.1 Endovascular-specific arteriovenous malformation (AVM) grading scales

AVMES (Lopes et al¹¹)
AVM nidus size	1 (< 3 cm) 2 (3–6 cm) 3 (>6 cm)
Number of arterial pedicles	1 (1–3) 2 (4–6) 3 (>6)
Number of draining veins	1 (1–3) 2 (4–6) 3 (>6)
Vascular eloquence	0 (noneloquent) 1 (eloquent)
Buffalo score (Dumont et al¹²)
Number of arterial pedicles	1 (1–2) 2 (3–4) 3 (>5)
Diameter of arterial pedicles	0 (most > 1 mm) 1 (most ≤ 1 mm)
Nidus location	0 (noneloquent) 1 (eloquent)
Bell et al¹⁴
Number of arterial pedicles	1 (< 3) 2 (3–5) 3 (≥ 6)
Eloquence of adjacent areas	0 (noneloquent) 1 (eloquent)
Presence of AVF component	0 (no AVF) 1 (AVF)

Abbreviations: AVF, arteriovenous fistula; AVMES, arteriovenous malformation embocure score.

In our proposed AVM embocure score (AVMES), we found the number of arterial pedicles, the number of draining veins, the size of the AVM nidus, and the presence of “vascular eloquence” to be collectively relevant in the prediction of complete angiographic embolization and the expected risk of complication. The concept of vascular eloquence is unique to the AVMES. It is defined as arterial pedicle proximity to the internal carotid artery (ICA; an AVM was considered to possess vascular eloquence if its arterial pedicle was less than 20 mm from the ICA or first segment of cerebral arteries); it aims to describe a risk idiosyncratic to endovascular treatment strategies. Specifically, a proximal injury (e.g., reflux of Onyx occurring during embolization of an arterial pedicle arising from a short M1 parent vessel) would cause significant neurologic deficit, making these lesions significantly more risky to treat endovascularly.

In lesions with AVMES of 3, there was a 100% rate of angiographic obliteration and a 0% rate of major complication. Conversely, in lesions with AVMES of greater than 5, complete angiographic cure was obtained in only 20%, with major morbidity occurring in 30%. Application of this, as well as other “endovascular-specific” scores, is critical to the preoperative assessment and treatment planning of patients with AVMs. Though AVMES and other “endovascular-specific” scores assess the potential for success and morbidity associated with stand-alone endovascular therapy, they also serve as tools for the multidisciplinary assessment of these complex vascular lesions.

Endovascular therapy has become an integral part of the management of supratentorial AVMs. It offers a minimally invasive and effective treatment strategy. Though previously considered only in combination with surgical resection or stereotactic radiosurgery, increasingly it is being utilized as the sole treatment modality to obtain complete radiographic AVM obliteration.¹⁵^,¹⁶^,¹⁷^,¹⁸ In this chapter, we will discuss issues unique to the endovascular treatment of supratentorial AVMs, as well as the role of endovascular strategies both as a stand-alone and adjunctive modality. We will also discuss technical considerations encountered during endovascular treatment.

20.3 Intraoperative Considerations

20.3.1 Anesthesia

Endovascular embolization may be performed with the patient under general anesthesia or conscious sedation. Our preference is for general anesthesia since it eliminates any patient head movement, providing superior quality images without motion artifact. Procedural success does not rely on patient cooperation and this method may be particularly well suited for pediatric patients, patients with underlying cognitive and/or psychiatric disturbances, or others unable to remain still for the procedure, especially on prolonged embolization treatments (embocure cases). During the administration of general anesthesia, the neurologic exam is lost and recognition of intraoperative complication or intolerance of the procedure may be delayed. In order to address this concern, neurophysiologic monitoring such as electroencephalogram, motor-evoked potentials, and somatosensory-evoked potentials and in-depth knowledge of the vascular anatomy are of paramount importance when general anesthesia is utilized.

The primary benefit to AVM embolization under conscious sedation is preservation of the neurologic exam and possibility of performing provocative testing of the arterial pedicles. In this way, the patient can be examined serially throughout the procedure to ensure tolerance to the treatment. Microcatheterization and AVM embolization under conscious sedation require strict cooperation by the patient in order to permit high-quality imaging and avoidance of complications. Blood pressure control is particularly important in patients undergoing embolization under conscious sedation since anxiety and discomfort during various parts of the procedure may cause rises in blood pressure. Conscious sedation and provocative testing should be reserved for specific situations when an AVM arterial pedicle physiological testing is desired prior to embolization. Provocative testing has been shown to decrease treatment-related morbidity to less than 5%.¹⁹ Its “real-time,” intraoperative use is unique to endovascular treatment strategies and it can be utilized both in patients under general anesthesia with neuromonitoring and in patients under conscious sedation, but it is more ideal in those being treated under conscious sedation. Tawk et al described their experience with provocative testing in patients undergoing treatment of occipital lobe AVMs under conscious sedation.²⁰ Once the microcatheter was in its desired position within the arterial pedicle, a baseline neurologic exam was obtained. Amobarbital 75 mg was injected and the patient’s neurologic exam was compared to the findings on baseline examination. If no new neurologic deficits developed, then the selected feeder was embolized. Although the authors describe their experience with the treatment of only AVMs of the occipital lobe, this technique is easily applicable to the treatment of AVMs in other locations.

Alternatively, Feliciano et al use propofol for their provocative testing under conscious sedation (off-label).²¹ The authors performed neurologic examinations before and after the injection of propofol 7 mg through a microcatheter positioned in a given arterial pedicle. While the authors did not experience any adverse events related to this technique, it should be noted that cardiopulmonary dysfunction may occur.

When performing provocative testing on the patient being treated under general anesthesia, interpretation of the neurophysiologic monitoring is used instead of the neurologic exam. Coordination between the neurophysiologist, anesthesiologist and the treating physician is essential to safe and accurate interpretation of the provocative testing. The use of these tests allows the physician to assess the safety of embolization in a “pedicle by pedicle” fashion, assuring the safety of each injection of liquid embolic. However, this provocative testing should not be a substitute to understanding the vascular anatomy and AVM angioarchitecture.

20.3.2 Hemodynamic Monitoring

Hemodynamic monitoring is particularly important during the endovascular treatment of AVMs. Routinely, we maintain systolic blood pressure between 90 and 120 mm Hg during and for 24 hours after the procedure. Though anecdotal, we believe this to allow a more controlled embolization, perhaps decreasing the risk of unintentional, premature venous occlusion by the travel of liquid embolic through the AVM nidus to the draining vein(s).

Strict control of blood pressure is also essential to avoidance of normal perfusion pressure breakthrough and resultant hemorrhage. Maintenance of low to normal blood pressure for at least 24 hours postprocedure permits recovery of the autoregulatory capacity of the cerebral vasculature and minimizes the risk of hemorrhage.

20.3.3 Choice of Embolic Agent Onyx

Though first described in the 1990s, Onyx (ev3 Inc, Plymouth, MN) was approved by the Food and Drug Association (FDA) for the treatment of AVMs in 2005.²²^,²³ It is an ethylene vinyl alcohol copolymer and, contrary to previous liquid embolic agents, is cohesive but not adhesive. This results in less adherence to microcatheters and vessel walls and, therefore, was intended to reduce the risk of complications associated with microcatheter removal. Because of its cohesive nature, rather than adhesive nature, a certain degree of reflux is permissible when using Onyx without necessarily compromising the removal of the microcatheter. A detailed description of Onyx was previously published by Ayad et al; a brief overview follows here.²⁴

Onyx occlusion of a vessel lumen is initiated after the diffusion of dimethyl sulfoxide (DMSO) into the blood from the outer surface of the Onyx cast.²⁵ The precipitation of Onyx occurs from “outside-in”—it has been compared to the hardening of lava from a volcano.²⁴ Solidification occurs over several minutes to hours, permitting better anterograde penetration compared to n-butyl cyanoacrylate (n-BCA). Once solidified, Onyx feels rubbery, rather than hard and brittle, like n-BCA. This may make an Onyx-embolized AVM easier to dissect, thereby making resection more amenable.²⁶

Onyx formulations used for AVM embolization are 18 and 34, which represent different viscosities. The different viscosities allow variable penetration of the nidus before polymerization occurs. While Onyx itself causes very little intravascular or perivascular inflammatory reactions, DMSO is capable of causing vasospasm and endothelial necrosis. For this reason, DMSO must be injected slowly.

Onyx and more recently Squid and Phil are liquid embolics of similar properties (e.g., nonadhesive) but different viscosities. This category of liquid embolic is our preferred method for embolization of supratentorial AVMs. They allow a better penetration of the nidus over a prolonged injection time compared to acrylates such as n-BCA.

n-Butyl Cyanoacrylate

n-BCA is a free-flowing liquid embolic that polymerizes via an anionic mechanism (i.e., upon contact with blood). n-BCA requires mixing with ethiodized oil by the treating physician to achieve the desired polymerization time. Higher concentrations of n-BCA result in a faster polymerization rate. Though it may be tedious in routine AVM treatment, the ability to control polymerization rate may serve useful when treating complex AVMs, particularly those with fistulous components.¹³ According to in vitro testing described in the product information brochure, n-BCA polymerization times may range from immediately upon contact with plasma (0:1 ethiodized oil:n-BCA ratio) to 16 seconds (5:1 ethiodized oil:n-BCA ratio).

n-BCA is adhesive in nature. While this allows reliable vessel occlusion, it also adheres to the microcatheter. Therefore, reflux is ill tolerated when using n-BCA and proficiency with this technique is critical to safe and effective treatment. However, when utilized by experienced endovascular surgeons the utilization of n-BCA has not been shown to be associated with any increase in morbidity or mortality.²⁷

20.3.4 Choice of Microcatheter

Safe and effective endovascular embolization of supratentorial AVMs requires that the physician place the microcatheter at the pedicle serving only the AVM. Placement of the microcatheter too proximal in the intracranial circulation may result in unintended embolization of proximal, uninvolved vasculature resulting in ischemic neurologic consequences and not occluding the nidus of the AVM. Therefore, navigation through the cervical and cranial vasculature is essential to endovascular embolization.

The first advance in the development of microcatheters came in the form of flow-directed microcatheters (e.g., Magic, Balt, Montmorency, France). The basis for these catheters is the high-flow nature of AVMs compared to the surrounding, normal circulation. Because of this difference in flow, the soft, floppy tip of these microcatheters is preferentially directed into fistulous arterial pedicles.

Marathon and UltraFlow (ev3 Inc) were developed subsequently as over-the-wire microcatheters allowing for some steering of the microcatheter using 0.008- and 0.010-inch microguidewires. More recently, detachable tip microcatheters like Sonic (Balt, Montmorency, France) and Apollo (ev3 Inc) have also proven effective in the endovascular embolization of AVMs (► Fig. 20.1).²⁸^,²⁹^,³⁰ These detachable tip microcatheters allow for an atraumatic microcatheter removal as long as the liquid embolic agent does not reflux beyond (i.e., proximal to) the detachable tip marker. These newer generation microcatheters have been used with both Onyx and n-BCA.²⁸^,²⁹

Fig. 20.1 Apollo detachable tip microcatheter—inner diameter 0.013 inch (a). The proximal (P) and distal (D) markers are seen (b). The separation point (arrow) occurs 1.25 mm distal to the proximal marker. The presence of the detachable tip (c) reduces the risk of catheter entrapment. (These images are provided courtesy of Medtronic Neurovascular, Irvine CA, USA.)

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