4  Natural History, Clinical Presentation, and Indications for Treatment of Brain Arteriovenous Malformations

Amit Single and Brian Hoh

Abstract

Brain arteriovenous malformations (BAVMs) are dynamic vascular lesions with blood flow characterized by high velocities and low resistances without normal cerebral hemodynamic autoregulation, with an innate propensity to rupture.

The patients harboring these can present with varied clinical presentations ranging from asymptomatic to symptoms resulting from hemorrhage or mass effect. Intracranial hemorrhage is the most common clinical presentation. Annual rupture rate of BAVMs ranges from 2 to 4%. The risk of rehemorrhage in the first year is 9 to 15%. Then the risk decreases over time to about 2 to 4% after 5 years. The presentation with hemorrhage, deep venous drainage, and deep location and the presence of associated aneurysms are more commonly agreed upon factors that increase the risk for future AVM hemorrhage.

Ruptured BAVMs need to be treated because of the high risk of rerupture; however, the management of unruptured BAVMs is still a matter of debate. AVMs often require multimodality treatment requiring a multidisciplinary team with expertise in cerebrovascular neurosurgery, endovascular intervention, and radiation therapy.

BAVM microsurgery is generally an elective procedure unless associated with life-threatening hematoma or prenidal or nidal aneurysm or unless it is a small, easily accessible AVM which can be completely treated. A Randomized Trial of Unruptured Brain Arteriovenous malformations (ARUBA) showed that medical management alone is superior to medical management with interventional therapy in patients followed up for 33 months. However, this trial has been criticized for multiple reasons. The current evidence suggests that conservative management cannot be generalized to all unruptured BAVMs and the management strategy needs to be individualized.

Keywords: brain arteriovenous malformations, ARUBA trial, radiosurgery, endovascular embolization, microsurgery

4.1  Introduction

There is a remarkable heterogeneity in the natural history studies of brain arteriovenous malformations (BAVMs) suggesting that our understanding has continued to evolve since BAVMs were first described by Steinheil in 1895. BAVMs are relatively rare, with a yearly incidence of approximately 1 per 100,000 persons and a prevalence of 18 per 100,000 persons.¹ BAVMs consist of a tangle of abnormal blood vessels (nidus) in which the feeding arteries are directly connected to a venous drainage network without interposition of a capillary bed. Blood flow in BAVMs is characterized by high velocities and low resistances without normal cerebral hemodynamic autoregulation with an innate propensity to rupture. Furthermore, because the brain and its blood vessels are formed together during embryological development, abnormal blood vessel formation is often associated with abnormal brain tissue. Studies are currently under way ranging from designing models to discern genetic and hemodynamic influences on disease inception and progression, to using already developed models to design therapeutic interventions.

4.2  Pathologic Inception

Studies of resected BAVMs show evidence for angiogenesis with increased endothelial cell turnover and inflammatory cellmediated vascular remodeling; however, little is known about the etiology of BAVMs, which is likely multifactorial. Etiopathogenesis of BAVMs remains controversial with genetic influences on abnormal angiogenesis/vasculogenesis and inflammation-induced angiogenic stimulation appearing to play roles during BAVM development. Recent reports of familial occurrence of BAVM, along with the known association with genetic disorders (Sturge-Weber disease, Osler-Weber-Rendu disease, hereditary hemorrhagic telangiectasia), support a genetic basis for the development, growth, and clinical behaviors (such as rupture) of BAVMs.² Current research on the genetic pathophysiology of BAVMs involves hypothesis-driven candidate gene studies where genes are implicated based on associated disorders and expression profiling from resected BAVM tissue. These candidate genes are currently used for model systems to discern mechanisms of BAVM development and potential therapeutic intervention.²

Classically, BAVM research has centered on genetic involvement in abnormal angiogenesis or the budding of endothelial cells from existing vasculature to form additional abnormal vessels at the nidus, either congenitally or in response to injury. Elevated incidence rates of sporadic BAVM in patients have been associated with the loss of function of genes encoding for proteins involved in vascular remodeling and appropriate angiogenesis. For example, the polymorphisms in activin receptor-like kinase 1 (ALK1) and multiple genetic loci influencing vascular endothelial growth factor (VEGF) levels have been attributed to loss of distinct arterial and venous boundaries during arteriolization in vascular remodeling, thus increasing susceptibility to BAVM formation and rupture. Additionally, current research has shown that these same genes may also support vasculogenesis or the independent development of additional vessels from endothelial progenitor cells (EPCs). Immunohistological analysis of BAVM tissue has shown that EPCs accumulate at the edge of vessel walls at the BAVM nidus, thus also contributing to its expansion.^2,3

Furthermore, recent genetic studies have implicated polymorphisms in the promoter regions of proinflammatory cytokines such as interleukin-1β (IL-1β), IL-6, and tumor necrosis factor α (TNF α), with an increased BAVM incidence, an initial clinical presentation of hemorrhage in patients leading to a diagnosis of BAVM, an increased incidence of hemorrhage in patients diagnosed with BAVM, and an increased risk of post-treatment hemorrhage (as reviewed in Kim et al³). The most notable is the contribution of IL-6, which leads to the upregulation of the additional cytokines.⁴ The cytokines stimulate leukocyte recruitment, VEGF, and angiopoietin-2 (ANG-2) expression, leading to increased angiogenesis, vascular smooth muscle cell proliferation, and matrix metalloproteinase protein 9 (MMP9) expression. The level of MMP9 expression has been found to be an order of magnitude higher in BAVM tissue compared with controls. This leads to the overexpression of other proinflammatory markers associated with MMP9, such as myeloperoxidase and ILs (IL-6), which creates an environment of vascular instability at the BAVM nidus and increases susceptibility to rupture.³ Future studies to elucidate the genetic pathophysiology of BAVM formation and progression to rupture will likely expand to include genome-wide association and high-density single nucleotide polymorphism studies.

4.3  Progression to Clinical End Points

As our understanding of the physiology of BAVMs increases, the traditional thought that these were static congenital lesions has now evolved into a more dynamic view. Through genetic and hemodynamic influences, formed BAVMs can undergo vascular remodeling and change in size, by either expanding or regressing. This, in turn, manifests into a varied range of clinical presentations from patients remaining asymptomatic to presenting with clinical symptoms associated with mass effect or intracerebral hemorrhage (ICH) due to rupture. Specific hemodynamic influences include elevated feeding artery pressures, disrupted venous drainage, the presence of a perinidal hypervascular network, and the vascular steal phenomenon.⁵

BAVMs can be monocompartmental, consisting of a single feeder vessel and one or more draining veins, or they can be multicompartmental, consisting of multiple feeder vessels and draining veins that are adjacent or separated by brain tissue. Studies have noted that smaller BAVMs (< 3 cm), though not as prone to be symptomatic due to mass effect, have significantly higher rates of rupture. This is due to increased intra-arterial resistance and pressure of the feeder vessels, which is inversely proportional to the size of the BAVM.^6,7 Coupled with elevated input pressures, hemodynamic strain on the BAVM can be exacerbated by obstruction of venous outflow. One study found that 30% of AVMs have abnormal venous drainage.⁸ Furthermore, the decreased drainage may lead to hypoperfusion of surrounding brain tissue. This hypoperfusion, coupled with a vascular steal phenomenon created by the BAVM, is thought to stimulate expression of angiogenic factors, such as VEGF, resulting in the expansion of the BAVM. Additionally, decreased venous drainage may also increase intraluminal pressure to a level that opens preexisting patent arteriovenous connections, which would also contribute to growth of the BAVM. This can be seen particularly with multicompartmental BAVMs, in which “hidden compartments” of perinidal vessels that were not initially seen on angiography may abruptly fill, leading to rupture or edema (as reviewed in Moftakhar⁵).

4.4  Clinical Presentation

Often patients with BAVMs may experience no symptoms and they are discovered only incidentally, usually either at autopsy or during treatment for an unrelated disorder. The proportion of patients being diagnosed with unruptured AVMs has almost doubled in the past three decades with improved noninvasive imaging.⁹ However, approximately 12% of people with AVMs will experience symptoms, varying in severity. ICH from a ruptured AVM is consistently described as the most common presenting symptom, occurring at 50% of initial presentations. Mortality for ICH as a first presentation of an AVM ranges from 10 to 30%.¹⁰ Traditionally, an annual rupture rate of 4% has been cited for BAVMs based on a study on natural history of symptomatic BAVMs; this study also included the AVMs, which had previously ruptured.¹¹ A Randomized Trial of Unruptured Brain Arteriovenous Malformations (ARUBA) reported a low spontaneous rupture rate of 2.2% per year (95% confidence interval [CI] 0.9–4.5).¹² Other recent prospective studies have also reported lower bleeding rates of about 1% per year for unruptured BAVMs.^13,14

Overall, studies have shown that the rate of rehemorrhage in patients with the initial presentation of a hemorrhage decreases over time. The highest risk is within the first year of diagnosis, from 9.65 to 15.42%. In years 2 to 5, the risk drops from 5.32 to 6.3%, and is the lowest after 5 years (1.7–3.67%).¹⁵ Fewer studies have quantified the effect of the additional factors on the risk of hemorrhage. One study has found that deep venous drainage increases the risk by 0.9 to 2.4% per year, and that the deep or infratentorial brain region increases the risk by 0.9 to 3.1%.¹⁴ However, these risk factors are intricately interlaced, and studies looking at them in isolation do not give a true picture of disease progression to hemorrhage.

Unruptured BAVMs can become symptomatic by mass effect due to irritation of the surrounding brain. This may result in the second most common presenting symptom, seizures, seen in approximately 30% of cases. Alternately, the third most common presenting symptom, also due to mass effect, is headache, which occurs in 5 to 14% of cases.^16,17 Additional symptoms that may occur corresponding to the location of the AVM include myasthenia, paralysis, ataxia, executive dysfunction, dizziness, visual disturbances, dysphasia, dysesthesia, memory deficits, hallucinations, and dementia.

4.5  Indications for Treatment of BAVMs

Various classification systems have been proposed over the years in an effort to assist the treating physicians in surgical decision making when dealing with BAVM. In 1986, Spetzler and Martin introduced a classification system, which was based on the dimensions, location, and type of venous drainage from the BAVMs.¹⁸ In this classification system, the higher the grade of the AVM, the higher the surgical risk. The authors recommended that small lesions in noneloquent tissue that fall under Spetzler-Martin (SM) grades I and II are primarily treated via surgical resection with or without adjunct endovascular embolization, while grade III lesions may be treated with endovascular embolization followed by surgery. Grade IV and V lesions are often considered high risk and preferentially managed medically.¹⁸ To simplify the treatment decision, Spetzler and Ponce proposed the three-tier classification system in which the grades I and II and grades IV and V from the original classification were clubbed together in tiers A and C, respectively, and tier B corresponded to grade III. Surgery was recommended as the treatment modality for class A and multimodality treatment for class B. No treatment was recommended for class C. The high reported rates (25–30%) of morbidity associated with surgery, lack of immediate protection, and possible elevated risk of hemorrhage associated with radiation or partial embolization are cited as the reasons for recommending conservative or palliative management in patients with grade IV and VAVMs.¹⁹

Lawton et al proposed a simple supplementary grading system for BAVM which could be used to improve and refine patient selection for AVM surgery.²⁰ This was based on the authors’ previous experience that factors such as hemorrhagic presentation, young age, compactness of the AVM nidus, and absence of deep perforator supply were identified as the predictors of good outcomes after microsurgical resection.^21,22 The supplementary grade may influence surgical decisions for AVM patients at the borderline between high and low risk based on SM grading system.²⁰

4.6  Factors Influencing the Treatment Decision

Broadly, the factors influencing the treatment decision can be divided into AVM architecture-related and patient-related factors.

4.6.1  AVM Architecture-Related Factors

The factors mentioned in the literature as being related to increased risk of hemorrhage from BAVMs include^{23,24,25,26,27}

• Presentation with hemorrhage.
  
• Presence of deep venous drainage.
  
• Associated aneurysms.
  
• Deep brain location.
  
• Smaller AVM size.
  
• Venous outlet restriction.
  
• Single draining vein.
  
• Diffuse AVM morphology.

The presentation with hemorrhage, deep venous drainage, and deep location and the presence of associated aneurysms are more commonly agreed upon factors that increase the risk for future AVM hemorrhage.²⁸ Previous history of rupture is probably the strongest factor related to the increased future risk of rehemorrhage from an AVM. Recent meta-analysis showed that unruptured AVMs had an overall annual risk of hemorrhage of 2.2% (95% CI, 1.7–2.7%), whereas ruptured AVMs had an overall annual rupture rate of 4.5% (95% CI, 3.7–5.5%).²⁸ Deep brain involvement, basal ganglia or thalamic lesions, or in the periventricular region, or the AVMs with exclusively deep venous drainage also appear to have a higher hemorrhage rate.^29,30 AVM-related aneurysms, seen in approximately 15 to 18% of patients, have been shown to independently increase the risk of future hemorrhage.^28,31 The patients having these factors are more likely to have future hemorrhage, and the treatment, if indicated, should commence sooner rather than later. It is important to highlight that the influence of these risk factors is additive.³²

Old Hemorrhage

Abla et al stressed upon another important criterion of “old hemorrhage” in determining the “seemingly” unruptured AVMs which had silent hemorrhages before the presentation. Such AVMs with prior small hemorrhages labeled as “microhemorrhage” are clinically silent, often without even a headache. This study showed that silent intralesional microhemorrhage was a risk factor for later AVM rupture. Both evidence of old hemorrhage (odds ratio [OR], 3.97; p = 0.001) and hemosiderin positivity (OR, 3.64; p = 0.034) were highly predictive of index intracerebral AVM hemorrhage. One-third of patients have been reported to present with silent hemorrhage; they are best diagnosed with iron-sensitive imaging protocol magnetic resonance imaging (MRI) scan. They suggested that such AVMs are at a high risk of rupture and should be treated with intervention.³³

AVMs with Aneurysms

AVMs with associated aneurysms deserve mention as a separate entity. Aneurysms in conjunction with BAVMs are seen in about 20 to 25% of patients with AVMs. In a recent meta-analysis, associated aneurysms were found to be a statistically significant risk factor for hemorrhage with a hazard ratio of 1.8 (95% CI, 1.6–2).²⁸ Hemorrhages from AVMs are classically thought to have a relatively small early rebleeding rate compared to simple intracranial aneurysms, although there are reports of early rebleeding from AVM-associated aneurysms. Such aneurysms are treated according to their location and size, preferably prior to treating the AVM itself because of the elevated risk of rupture. The aneurysms located within the nidus or prenidal aneurysms should be treated in the setting of hemorrhage or if they are large (>7 mm) in unruptured AVMs. Distal flow-related aneurysms often involute after treatment of the AVM nidus.³²

Giant AVMs

Large or giant AVMs with diffuse hemispheric or bilateral cerebral involvement (SM grade IV or V) or Spetzler and Ponce tier C may not be amenable for the treatment and are managed conservatively. The patients with such AVMs can present with ischemic symptoms from the steal phenomenon; palliative care with endovascular embolization may be offered to such patients to decrease size of the lesion and for symptomatic improvement.³⁴

4.6.2  Patient-Related Factors

Age

Age is one of the most important factors in determining the treatment versus conservative management and also the modality of treatment. The lifetime risk of hemorrhage from an AVM is higher in the younger patients, justifying a lower threshold for offering the treatment.

Symptomatic versus Asymptomatic

Symptomatic patients such as those presenting with seizures might benefit from treatment. Obliteration of the AVM may result in symptomatic improvement including seizures; a study showed that after treatment of symptomatic AVMs, 83% patients were seizure-free 2 years after undergoing microsurgery.³⁵

Medical Status

Similarly, patients with significant medical comorbidities and limited longevity may not be good candidates for more aggressive treatment option such as surgery; however, stereotactic radiosurgery (SRS) might still be offered.

4.7  Lifetime Risk of Hemorrhage

To predict the lifetime risk of hemorrhage in a given individual, Kondziolka and colleagues³⁶ suggested a simple model based on life expectancy and the multiplicative law of probability. This predictive model assumes that for a particular individual in question, the risk of hemorrhage is constant over time. According to their calculations, the lifetime risk of hemorrhage can be estimated using the following formula:

Using this formula, a patient with a life expectancy of 50 years and a yearly risk of hemorrhage of 4% would be predicted to have the lifetime risk of AVM bleed as follows:

A simpler model assuming a constant 3% yearly risk of hemorrhage can also be used and still maintains a similar sensitivity¹⁰:

For example, a 35-year-old person with an AVM would have a lifetime risk of hemorrhage of

4.8  Ruptured Brain Arteriovenous Malformations

Hemorrhage from a ruptured BAVM carries a 10 to 30% risk of death and a 30 to 50% chance of permanent neurologic deficits.¹⁰ The published literature is consistent with the finding that previous rupture is the most important factor when assessing a patient’s future risk of hemorrhage. BAVM rupture is associated with an increased risk of 6% rerupture in the year following the initial hemorrhage, versus 1 to 3% predicted annual risk in nonruptured lesions.¹⁰ Fults and Kelly, in a retrospective analysis of 131 patients with AVM, discovered a tendency of increasing mortality with subsequent episodes of hemorrhage from 13.6 to 20.7 to 25%, after the first, second, and third episodes, respectively, although statistical significance was not reached.³⁷ In contrast, other studies have shown that recurrent hemorrhages do not have a cumulative impact on the prognosis.^38,39 Although the data on the issue of cumulative effects of each subsequent hemorrhage are conflicting, each individual hemorrhagic episode has its own associated morbidity and mortality, underscoring the importance of the fact that ruptured AVMs need to be treated. However, ruptured AVMs are classically thought to have a relatively small early rebleeding rate as compared to ruptured saccular intracranial aneurysms, justifying delay in treatment after stabilizing the patient and allowing the brain swelling to decrease. A rest period of a few weeks to months after hemorrhage is recommended before definitive treatment to allow for the brain swelling to decrease, for better delineation of the lesion and to avoid disrupting friable parenchyma.³²

4.8.1  Timing of Surgery

Timing of AVM microsurgery is generally an elective procedure unless the patient presents with ICH or life-threatening hydrocephalus secondary to hematoma. In such emergent cases, resection at the time of clot removal is only indicated for superficial AVMs in which the anatomy can be readily elucidated. Otherwise, hemorrhage or hematoma-related complications must be resolved first, followed by postoperative rehabilitation and angiographic analysis. However, there are certain clinical and radiological features which can justify emergent/earlier surgical and/or endovascular management. These include large life-threatening hematomas causing cerebral herniation, hydrocephalus, or prenidal or nidal aneurysms, and small, easily accessible AVMs which can be completely treated.^10,32 The patients presenting with ICH and suspected of harboring a vascular malformation should be evaluated with noninvasive studies such as computed tomography angiography (CTA)/magnetic resonance angiography (MRA) as the initial screening tools. Conventional digital subtraction angiography (DSA) remains the most sensitive and specific diagnostic modality and should be performed within the next 24 hours of presentation to evaluate the cause of hemorrhage and to identify high-risk factors.^10,32

Aneurysms associated with AVMs are treated according to their location and size, preferably prior to treating the AVM itself because of the elevated risk of rupture. Small feeding aneurysms < 5 mm may regress after AVM treatment and thus may not require treatment prior to AVM resection; larger feeding artery aneurysms (>7mm) are usually treated endovascularly prior to obliteration or resection of the AVM. Once the prenidal or nidal aneurysms are treated, the definitive treatment with embolization with or without other modalities such as surgical resection and radiosurgery can be deferred until the patient and the hematoma have stabilized.³²

4.9  Unruptured Brain Arteriovenous Malformations

In contrast to the ruptured BAVMs, the management of unruptured BAVMs remains controversial at this time, more so if they are asymptomatic. Some of the more recent studies have reported the annual AVM rupture rate of as low as 1%, much lower than the 4% reported in an earlier retrospective study.^13,14 Besides the lower ruptures, another reason cited for conservative management of unruptured AVMs is the mild first hemorrhagic episode associated with AVM rupture.^40,41 The notion that patients with incidental AVMs may have worse outcomes with the intervention as compared to those with hemorrhagic presentation, and the morbidity and mortality of hemorrhagic episodes associated with AVM rupture might have been overestimated in the literature in the past, argues for close monitoring of incidental AVMs instead of the intervention.⁴² The ARUBA trial showed that medical management alone is superior to medical management with interventional therapy for prevention of death or stroke in patients followed up for 33 months. This trial was stopped prematurely because of significant benefit seen in the medically managed group. Medical management group (109 patients) in this trial had a threefold lower incidence of stroke or death (10.1 vs. 30.7%) as compared to the patients assigned to interventional therapy (114 patients). The interventions performed consisted of embolization (32%), radiosurgery (33%), embolization plus radiosurgery (16%), or surgery (18%).¹²

The ARUBA study has been criticized for multiple reasons including the fact that surgical resection as the intervention was performed for only 18 patients out of 76 patients with low-grade AVMs who were randomized to intervention.⁴³ Most AVMs were treated with embolization or radiosurgery, both of which have lower obliteration rates than microsurgical resection.^44,45 Thus, the higher rate of stroke or death and clinical impairment in ARUBA’s interventional therapy arm reflects not only treatment-associated effects, but also complications from partially treated AVMs. Finally, ARUBA’s relatively short followup of 33 months favors medical management, since curative effects would take longer for the any-treatment group and differences observed between the two arms might dissipate over time.⁴³ In contrast to the ARUBA trial, a large series of patients from a prospectively collected database which was retrospectively analyzed showed Spetzler-Ponce class A and Spetzler-Ponce class B unruptured BAVMs treated by surgery had a better outcome than conservatively managed unruptured BAVMs within a short period of time with a high rate of cure. One of the reasons cited for the difference in the outcome is due to the fact that most patients in this study were treated with surgery alone and the rest of the patients were treated by surgery with adjunctive preoperative embolization. They concluded that the defined group of patients can be better managed with surgery than conservative management alone.⁴⁶

Nonetheless, the literature suggests that the conservative management cannot be generalized to all unruptured BAVMs. This decision dilemma has led to the attempts to better identify/ define the risk factors that significantly increase the risk of future hemorrhage making the intervention a more reasonable and better option than conservative treatment. With more diagnostic and therapeutic options available, it seems prudent to determine the future bleeding risk for an unruptured BAVM to provide adequate management strategy, either an aggressive treatment plan targeted at complete occlusion for BAVMs with risk factors for hemorrhage or supporting a more conservative approach for those at lesser risk.

4.10  Treatment Planning

Once the decision has been made to treat an AVM, the goal is directed toward complete AVM nidal obliteration. The three major treatment modalities include microsurgery, SRS, and endovascular embolization; all have been successfully used alone or in combination to treat AVMs (► Fig. 4.1 and ► Fig. 4.2).

Related posts:

Development of the Cerebrovasculature and Pathogenesis of Arteriovenous Malformations and Arteriovenous Fistulas Stereotactic Radiosurgery for Brain Arteriovenous Malformations Classification of Brain Arteriovenous Malformations and Fistulas Natural History of Arteriovenous Fistulas, Clinical Presentation, and Indications for Treatment Surgery of Basal Ganglia, Thalamic, and Brainstem Arteriovenous Malformations Surgical Treatment of Cerebellar Arteriovenous Malformations

Stay updated, free articles. Join our Telegram channel

Join