Giant Intracranial AVMs

Introduction

“Giant” intracranial arteriovenous malformations (iAVMs) have been traditionally defined as those with a diameter greater than 6 cm. These lesions are unique in that their large size puts increasing amounts of normal brain parenchyma at risk for ischemic damage due to shunting of blood supply away from functional cortex. Giant iAVMs are challenging to treat, and their rarity and unique angioarchitecture make consensus recommendations difficult. In fact, our understanding of the epidemiology, symptomatology, natural history, and outcomes of treatment of these lesions is primarily derived from limited single-institution case series often spanning decades. Perhaps more so than for smaller iAVMs, the decision of whether or not to treat a giant iAVM should be based on a frank discussion with the patient, including a careful weighing of risks and benefits. Treatment will almost always be multimodality, involving interventional radiology, neurosurgery, and/or radiation—sometimes all three.

In this chapter, we review the features of giant iAVMs, including their incidence, presenting symptoms, and natural history. Discussion of treatment options is focused on a multidisciplinary approach, often requiring staged procedures, with different goals compared to the treatment of smaller lesions.

Epidemiology and Natural History

By definition, all giant iAVMs have a Spetzler-Martin grade of at least III, earning three points based on the criterion of size alone. Because of their large size, they nearly always involve eloquent cortex, and therefore the vast majority of giant iAVMs seen in clinical practice are Spetzler-Martin grade IV or V ( Fig. 30.1 ). In a large surgical series of more than 1000 iAVMs across multiple institutions, grade IV and V AVMs represented only 10.7% of all surgically resected lesions. In a single-institution series of iAVM, including those observed or treated nonsurgically as well as those treated with surgery, embolization, or radiosurgery over a 24-year period, the proportion of giant AVMs was 8.1%.

Presenting symptoms of giant iAVMs include headaches, seizures, progressive neurologic deficits, and hemorrhage. Across several large series, headache and seizure were the most common presenting symptoms. In these series, the proportion of patients presenting with seizures ranged from 34% to 56%, the rate of headache from 15% to 45%, and the rate of hemorrhage as a presenting symptom from 17% to 38%. These lesions are typically diagnosed in the third decade of life, with the average age at diagnosis being reported as between 29 and 35 years.

Interestingly, the bleeding risk for giant iAVMs has been estimated in some studies to be approximately 1.5% per year, lower than the generally quoted rate of 2%–4% per year for iAVMs overall. In their series of giant iAVM cases, Yang et al. estimated the annualized risk to be between 2.7% and 4.1%, although interestingly the rate of initial hemorrhagic presentation in this series was 21.8%, less than the typically reported rate of 30%–82% seen in other studies of iAVMs in general. The possible reduction in the rate of bleeding is thought to be due to the generally diffuse nature of a giant AVM nidus and therefore overall lower pressures in the feeding vessels. While the bleeding risk is thought to be lower in giant iAVMs as a class, the angioarchitecture of each case should still be considered individually. By virtue of their size and diffuse nature alone, giant iAVMs may be more likely to harbor intranidal aneurysms or involve deep venous structures, which would put them at higher risk for rupture. Overall, best evidence supports the notion that AVM nidus size is not an independent risk factor for hemorrhage, and in fact, a large diffuse nidus may be at somewhat lower risk for bleeding.

Indications for Treatment

There is a single large randomized trial of observation vs treatment of iAVMs—ARUBA (A Randomised Trial of Unruptured Brain Arteriovenous Malformations) —but it has limited utility in guiding the management of giant iAVMs, since giant size was not reported and patients with Spetzler-Martin grade IV iAVMs made up only 10% of the total cohort. There is one large, retrospective, single-institution cohort study with long-term follow-up, the series presented by Yang et al., mentioned earlier. In this study, the authors identified 55 cases of giant AVMs, 24 (44%) of which were managed conservatively (without intervention). Data from continuous follow-up (mean 11.8 years) were available for 35 patients—13 patients whose AVMs were managed conservatively and 22 who underwent intervention. At latest follow-up, of the 13 patients in this subset whose AVMs were managed conservatively, 10 (77%) showed stable or improved neurologic status (as measured by modified Rankin Scale); in contrast, of the 22 patients who underwent intervention, only 11 (50%) showed stable or improved status and the other 11 (50%) showed decline. Overall, compared to the treated cohort, there was no significant difference in the rate of hemorrhage or neurologic outcome, although there was a trend toward superior neurologic outcome in the observation group that did not meet statistical significance. Therefore the best available data on the management of giant iAVMs support conservative observation as a reasonable option, with comparable rates of hemorrhage and similar neurologic outcomes. However, this should be weighed against the average young age of first presentation and consequently elevated life-long hemorrhage risk, as well as the possibility of neurologic symptoms not related to hemorrhage.

The indications for treatment of giant iAVMs can be broadly classified as mitigation of bleeding risk and palliation of symptoms. Once the decision to treat is made, in general, total resection should only be considered for patients with progressive deficits due to continued hemorrhage. Intranidal aneurysms or flow-related aneurysms at the circle of Willis can be treated via open surgery or endovascularly. Patients with progressive deficits and intractable headaches or seizures may be considered for partial embolization for palliation of symptoms. Based on these treatment considerations, in a large cohort of 73 patients with Spetzler-Martin grades IV and V AVMs at a high-volume center, the recommendation was no treatment in 75% of cases, partial AVM treatment in 10%, AVM-associated aneurysm treatment in 10%, and complete AVM resection in just 5%. Importantly, this study demonstrated that partial treatment of AVMs offered no reduction in the rate of subsequent hemorrhage and in fact correlated with a higher rate of posttreatment rupture.

Patients with giant iAVMs with associated aneurysms or those thought to be at risk due to prior rupture deserve consideration for treatment. In general, patients with associated aneurysms should have either open surgery or endovascular treatment that secures the aneurysm without consideration of total resection of the AVM. Indications for palliative treatment include intractable or worsening seizures and progressive neurologic deficits. Progressive deficits and intractable seizures in a patient with an unruptured giant iAVM are thought to be secondary to cerebrovascular steal and resultant ischemia in the surrounding brain parenchyma. Small series suggest that even partial embolization of iAVMs can reduce shunting and therefore restore normal perfusion to the surrounding brain and provide palliation of symptoms.