Intracranial arteriovenous malformations (AVM) are a demanding pathology to treat. There is ample discussion about which single modality or which amalgam of modalities is the most optimal, i.e., endovascular embolization, microsurgical resection, or radiosurgery. The discussion also extends to not only how but also whether to treat. Several grading scales for AVMs exist, while classically the relative morbidity of the open microsurgical approach has been previously outlined in the Spetzler-Martin (SM) grading scale. The microsurgical morbidity has been tied largely to specific anatomical and physiological characteristics that may make microsurgical resection more challenging. The SM grading scale specifically highlights the size of the AVM nidus, the presence of deep venous drainage, and the relative eloquence of the surrounding cortical tissue as crucial factors to consider in the microsurgical treatment of these complex lesions ( Table 50.1 ). This grading system is widely in use and wonderfully outlines the surgical morbidity. Although many interventionists have superimposed the relative risk of surgical treatment of SM-graded AVMs onto that of endovascular treatment, the SM scheme does not adequately reflect the unique determinants of risk associated with endovascular treatment of AVMs.
|Small (<3 cm)
|Medium (3–6 cm)
|Large (>6 cm)
|Eloquence of adjacent brain
|Venous drainage pattern
As endovascular techniques have briskly evolved as part of single modality and multimodality treatment, additional grading measures have been outlined. In this chapter, we focus on the technical aspects of endovascular treatment of intracranial AVMs, common complications, and how to avoid them.
Endovascular treatment of AVMs involves superselective microcatheterization of pathologic arterial feeding vessels (pedicles) and the infusion of embolic material with the goal of reducing or eliminating arterial supply to the AVM nidus. Common intracranial complications associated with endovascular treatment of intracranial AVMs are largely hemorrhagic or ischemic in nature. Hemorrhagic complications can be further viewed as immediate or delayed. Meanwhile, ischemic complications occur most commonly as a result of arterial embolization to vessels supplying normal cortex. In their review of AVM embolization performed in 153 patients, Kim et al. noted outcomes in 508 embolized vessels during 203 sessions. Those authors noted increasing neurological deficit with increasing number of branches embolized per session. The periprocedural complication rate was 11.8% in the short term and the rate of permanent disability (modified Rankin scale score [mRS] >2) was 2%. They also noted increasing neurologic deficit with increasing SM grade at rates of 0%, 5%, 7%, 10%, and 18% for grades I–V, respectively. Using a grading system germane to the endovascular treatment of AVMs, Dumont et al. stratified AVM treatment risk based on the number of arterial pedicles, the diameter of arterial pedicles, and eloquence of the surrounding cortex. Their scale, similar to the SM system, purports increasing morbidity with increasing grade ( Table 50.2 ). Others have added further hemodynamic parameters with respect to possible endovascular grading of AVMs. Bell et al. included the number of feeding arteries and the region of cortical eloquence as part of their analysis. Those authors also added the additional hemodynamic factor of the presence of an arteriovenous fistulous component as a crucial hemodynamic component. They noted that their grading scheme was useful in determining the multimodal outcomes of AVM treatment.
|Number of arterial pedicles
|1 or 2
|3 or 4
|5 or more
|Diameter of arterial pedicles
|Most >1 mm
|Most ≤1 mm
The most common hemorrhagic complication is associated with arterial perforation secondary to wire manipulation. However, the most deleterious hemorrhagic complication is associated with delayed postprocedural hemorrhage, which is commonly attributed to premature venous outflow opacification.
The choice of embolic material can have an effect on radiographic and clinical outcomes. Elsenousi et al. reviewed 103 studies of AVMs treated with n-butyl cyanoacrylate (NBCA) and a more recent and widely used liquid embolic agent, ethylene-vinyl alcohol co-polymer (EVOH). Poor neurological outcome occurred in 5.2% of NBCA cases and 6.8% of EVOH cases, although the difference was not statistically significant. However, complete obliteration rates were 13.7% in the NBCA cohort and 24% in the EVOH group, which did exhibit statistical significance. As a result, the common consensus is that although NBCA may provide a more pronounced radiographic treatment response, the use of EVOH is associated with fewer treatment-related complications.
With respect to avoiding complications, a meticulous review of the imaging studies to obtain a thorough understanding of the complex angioarchitecture of the lesion is critical preprocedurally as well as intraprocedurally. Identification of high-risk components is of paramount importance; these include intranidal aneurysms, high-flow arteriovenous shunts, venous stenosis, deep venous drainage, number and size of feeding pedicles, and relative eloquence of adjacent cortical structures. Eloquent location can be assessed with pretreatment magnetic resonance imaging of the brain. Any portion of the AVM nidus noted to be in the primary motor or sensory cortex, including language and vision areas, as well as the hypothalamus, thalamus, midbrain, pons, medulla, and the cerebellar peduncles, is considered to be in an eloquent location ( Table 50.3 ). The relative goals for treatment should be clear and apparent and will differ for patients who present with hemorrhage, with other nonhemorrhagic neurologic sequelae, or with no apparent neurologic symptoms.