Management of Residual and Recurrent Arteriovenous Malformations

30 Management of Residual and Recurrent Arteriovenous Malformations

Daniel Raper, David M. Sawyer, Peter S. Amenta, and Ricky Medel

Abstract

Incompletely obliterated arteriovenous malformations are encountered in clinical practice due to either failure of the primary treatment strategy, latency in treatment effect for radiosurgery, or lesion recurrence. Partial treatment in most circumstances does not confer a protective benefit and generally increases the rate of hemorrhage over the natural history. These lesions must be evaluated based on grade, angioarchitecture, location, previous treatments, and any existing neurological deficits at presentation. For those lesions that are amenable, surgical resection should be undertaken. Lesions that are deemed unsuitable for surgery should be evaluated for radiosurgery. Embolization is generally only useful as part of a multimodal treatment strategy, but may be beneficial in isolation for those patients with associated aneurysms, vascular steal, or the minority in whom a cure can be obtained. Patients must be counseled regarding the natural history, as well as the risks and benefits of any proposed treatment. High-grade (Spetzler–Martin IV and V) lesions are generally best managed conservatively.

Keywords: recurrent AVM, residual AVM, arteriovenous malformation, hidden compartment, incomplete obliteration, partial resection, staged treatment, radiographic surveillance, risk assessment, multimodality treatment

Key Points

Residual and recurrent brain arteriovenous malformations (AVMs) are commonly encountered in clinical practice due to the complex nature of many lesions, the extended time course of radiosurgical treatment, and the prevalence of staged treatment strategies.
Incomplete obliteration, in almost all circumstances, confers no benefit and may even increase the rate of hemorrhage.
Ongoing radiographic surveillance is essential to assess residual and recurrent AVMs, and for any suspected change a diagnostic cerebral angiogram (DSA) should be performed.
Surgical excision remains the “gold standard” for patients with low-grade lesions (Spetzler–Martin grades I and II) in noneloquent locations.
In any patient undergoing surgical intervention, DSA should be performed intraoperatively or immediately postoperatively to ensure complete excision.

30.1 Introduction

Treatment of recurrent and residual arteriovenous malformations (AVMs) remains a difficult clinical problem. Treatment options for AVMs include planned staged procedures, procedures that are aimed at reducing the volume of the AVM to potentiate other treatments, and procedures whose efficacy is delayed over multiple years. The natural history of residual and recurrent AVMs is not well defined. Flow dynamics in these lesions can change over time, and the risk of hemorrhage or re-hemorrhage after partial treatment is not well understood. Furthermore, the risks of treating residual or recurrent AVMs depend on the initial treatment modality and on a number of lesion-specific factors.

The natural history of untreated AVMs suggests an annual hemorrhage rate of 2.2% for unruptured lesions and 4.5% for those with previous rupture lesions.¹ Features such as exclusively deep venous drainage and associated aneurysms further increase this risk.¹ When considering younger patients, this portends a substantial lifetime risk of hemorrhage. Should this occur, the morbidity and mortality vary from 16 to 45% and 1 to 20%, respectively,²^,³^,⁴ and up to 45% of patients will not return to functional independence.⁵ Recently, a randomized trial of unruptured brain arteriovenous malformations (ARUBA) was conducted and halted early with the conclusion that medical therapy was superior to interventional management of unruptured lesions.⁶ This result has been criticized, primarily because of the methods utilized for intervention. Specifically, out of 116 people randomized to interventional management, only 18 (15%) were treated surgically.⁶ Furthermore, 73 patients (53 ongoing treatment and 20 treatment pending) at the time of analysis had yet to undergo definitive management.⁶ This highlights the importance of treatment modality in the management of AVMs, with microsurgical resection remaining the gold standard when feasible.

A comprehensive literature review from 2011 assessing outcomes in more than 13,000 patients demonstrated that incomplete obliteration is a common finding after initial treatment for AVMs.⁷ This further highlights the importance of treatment planning and emphasizes that the clinician taking care of patients with AVMs will often be confronted with patients harboring incompletely treated or recurrent lesions.

Similar to those with a de novo presentation, assessment and treatment of recurrent or residual AVMs relies on a thorough assessment of the angiographic anatomy, categorization of the risk of hemorrhage, and tailoring of an individual treatment plan based on a multidisciplinary consideration of the risks and benefits of treatment.

30.2 Methods

In addition to utilizing the authors’ own experience, a PubMed search was conducted using the keywords “incomplete obliteration brain arteriovenous malformations,” “incomplete resection brain arteriovenous malformations,” “residual brain arteriovenous malformations,” “residual cerebral arteriovenous malformations,” and “recurrent brain arteriovenous malformations.” Articles published within the last 15 years were considered. The senior author reviewed all articles for relevance. The American Heart Association (AHA) guidelines as well as key publications were also employed.

30.3 Presentation

Staged therapy has been used for many years in the management of complex AVMs,⁸^,⁹ and so the presentation of recurrent or residual AVMs is less often an acute symptomatic occurrence and more often a radiologic diagnosis or one made on routine follow-up. On the other hand, residual or recurrent AVMs can still cause an array of symptoms. These include hemorrhage (intracerebral, subarachnoid, or intraventricular), seizure, and neurologic change from steal phenomena. In pediatric patients, recurrent or breakthrough seizures, worsening hydrocephalus, or progressive heart failure may be an indication of failure of prior treatment.

30.4 Radiographic Surveillance

Among patients with AVMs that have been previously treated (either with an intent to cure or as part of a multistage treatment plan), follow-up imaging plays an important role in monitoring AVM progression. In asymptomatic patients, magnetic resonance imaging (MRI) and MR angiography provide satisfactory anatomical detail and are noninvasive. FLAIR (fluid attenuation inversion recovery) sequences can demonstrate mass effect on surrounding brain as well as radiation-induced parenchymal changes, and susceptibility- or gradient-weighted imaging can detect hemorrhages in the area of the AVM. Modern MR angiography approaches the resolution of computed tomography (CT) angiography and gives a satisfactory means of following these lesions over time. However, for cases that remain indeterminate or in which there is a question of planning an additional treatment, the gold standard imaging modality remains cerebral angiography. Intraoperative angiography is often useful to confirm complete obliteration of an AVM.¹⁰ Cerebral angiography is essential in the evaluation of recurrent or residual AVMs in order to characterize any changes to the lesion size, arterial supply, or development of concerning features such as intranidal aneurysms, venous ectasia, or strictures.

30.5 Recurrent Arteriovenous Malformations

The traditional understanding of the natural history of AVMs states that following complete resection, there is no risk of recurrence. This has been supported in several large series in adult patients.¹¹^,¹²^,¹³ Recurrent AVMs after complete resection have, however, been reported in the pediatric population.¹⁴^,¹⁵^,¹⁶^,¹⁷ Yasargil first reported recurrent AVM in a pediatric patient who had undergone surgical resection of a right frontal AVM with subsequent angiographic confirmation of obliteration of the nidus and early venous drainage.¹⁷ In one series, angiographic recurrence was seen in 14.3% of patients and occurred 4 to 5 years after the primary treatment (surgical resection) in all cases.¹⁵ A second series reported a 5.5% recurrence rate by angiography at a mean of 9 years after complete surgical resection.¹⁶ In a review of pediatric series in which follow-up was performed with MRI, recurrence occurred at an average of 5.5 years after initial resection (range: 1–9 years).¹⁴ Recurrent AVMs have been reported even following multiple complete resections with angiographic cure,¹⁸ or following complete resection with intraoperative angiographic confirmation.¹⁰ Recurrences in adults are much rarer than in children, though they have been reported.¹⁰^,¹⁹ The explanation for these cases may be related to a “hidden compartment” in certain AVMs, in which internal steal causes some portion of a residual AVM to remain angiographically occult.²⁰ Alternately, the immature cerebrovascular system in the pediatric patient may contribute to the capacity to form another AVM.¹⁰ Early postoperative angiography can miss residual AVM due to vasospasm, transient thrombus within an AVM nidus, or edema.¹⁰ In the largest series of AVMs investigating recurrence, young age and deep venous drainage were the most important factors predicting recurrence.¹⁹ Nevertheless, due to the overall low rate of recurrence, these authors advise that, in the case of adult patients with complete resection and a negative postoperative angiogram, regular angiographic surveillance is not necessary.

30.6 Treatment Modalities

30.6.1 Stereotactic Radiosurgery

Stereotactic radiosurgery (SRS) may be useful in a large proportion of AVMs, either as a treatment with intent to cure for patients with surgically or endovascularly inaccessible lesions, or as part of a multimodality treatment strategy. The radiobiological response to SRS can take 2 years or longer to become fully apparent. The most important factor in assessing the ability of SRS to provide a cure is the volume, with nidal volumes less than 3 mL being the best candidates. A number of grading scales have been proposed to assess the radiosurgical response of AVMs.²¹

A number of factors have been associated with incomplete radiosurgical treatment or recurrence after SRS. Hemorrhage has been associated with incomplete obliteration of the AVM nidus.²² Incomplete obliteration is seen in 42% of patients with high-flow fistulas after radiotherapy.²³ In a retrospective series of 139 patients treated with SRS, lower rates of obliteration were associated with angioarchitectural characteristics indicative of higher flow such as perinidal angiogenesis and arterial enlargement.²⁴ A systematic review of 14 studies in 733 patients with incompletely treated AVMs that were treated with repeat radiosurgery yielded an obliteration rate of 61% with risk of AVM-related hemorrhage of 7.6% and radiation-induced changes of 7.4%.²⁵ The type of initial treatment also influences the efficacy of subsequent SRS. Among a series of 169 patients treated with radiosurgery, those who had been previously treated with embolization had a lower obliteration rate than those with primary SRS treatment.²⁶

For patients with large AVMs, staged-volume radiosurgery is a strategy in which the AVM volume is divided into smaller “parcels” to target in individual treatment sessions.²⁷ In this way, a residual AVM is progressively treated over time. Treatment sessions may be staged at 3- to 9-month intervals. In a small series of patients, this strategy yielded a 10-year complete angiographic occlusion rate of 89%, but with a 27.8% hemorrhage rate.²⁸ In series of multistaged SRS for AVMs, 5-year total obliteration rates were 62%, but the cumulative hemorrhage rates after SRS were 4.3, 8.6, 13.5, and 36% at 1, 2, 5, and 10 years, respectively.²⁷ It is difficult to accurately define the risk in these lesions, which were deemed “unsuitable for surgery,” but these results likely represent an improvement on their natural history. It is important to note that the planning method and type of imaging used for radiosurgical planning may have an effect on obliteration rates achieved due to a potentially better ability to visualize the AVM nidus.²⁹

Any recurrence after initial treatment may change the drainage pattern of the AVM, effectively creating a new lesion. In those with an angiographically identifiable nidus, the hemorrhage rate is generally regarded as identical to untreated AVMs. However, there does exist evidence to suggest that radiosurgically treated lesions with subtotal obliteration have a lower rate of hemorrhage.³⁰ In the series by Abu-Salma et al including 121 patients with subtotal obliteration, defined as obliteration of the nidus with persistence of an early draining vein, there were no instances of hemorrhage over a mean follow-up of 44 months.³⁰^,³¹ This does represent a select group of patients, but suggests that this group may not require further intervention. Importantly, this angioarchitecture may represent a stage during the evolution to obliteration or an endpoint in itself.³²^,³³

30.6.2 Embolization

Embolization alone has the potential to offer a cure in only a small proportion of AVMs, typically those supplied by one or two arterial feeders, located at a safe distance from other branch arteries supplying normal parenchyma. To the extent that an initial treatment leaves an AVM nidus incompletely embolized, recurrence can be expected; this may be accompanied by development of new arterial feeders or recanalization of the initial feeding arteries, depending on the AVM angioarchitecture and the embolizate used initially.

N-butylcyanoacrylate (NBCA) is a fast-drying liquid adhesive, which has traditionally been used for embolization of AVMs either with a view to cure or as part of a multimodality strategy of treatment. Onyx, a newer liquid embolic agent, has also been widely used for embolization of AVMs. It is available in two concentrations, 6.0% (Onyx 18) and 8.0% (Onyx 34), with different viscosities, and has more of a “lava-like” passage through embolized blood vessels. Onyx and NBCA were investigated in a randomized trial that demonstrated equivalent safety and efficacy, with both agents having the potential to reduce AVM volume by greater than 50%.³⁴ Even in cases of complete AVM obliteration, however, there have been cases of recanalization due to Onyx resorption.⁹ A series of cases treated with presurgical embolization demonstrated evidence of canalization in almost 15% of cases.³⁵

More importantly, lesions that are initially treated with embolization may cause an AVM to alter its hemodynamic properties. Usually, if the largest arterial feeder is embolized, this results in a decrease in the rate and amount of arteriovenous shunting, thus decreasing the risk of hemorrhage. However, there remains a risk of collateralization after embolization of other arteries that were not previously angiographically visible. For this reason, any AVM that is embolized with intent to cure must be followed closely, ideally with serial angiography.

The strategy of targeted embolization, with a view to subsequent radiosurgery, although initially associated with increased risk of complications,⁷^,³⁶ has been proposed as a means to reduce interim hemorrhage risk and increase GKRS (gamma knife radiosurgery) efficacy.³⁷ This remains controversial and likely increases the risk of incomplete obliteration following radiosurgery.³⁸ The use of this strategy is dependent upon the angioarchitecture of the lesion, with it possessing the ability to potentially afford a lower side effect profile than with radiosurgery alone.³⁸ Embolization has also been employed to complete a partial treatment after SRS.³⁹ In both cases, the same principles of occlusion of the nidus depend on accurate identification of the nidus on cerebral angiography, and accessibility of the nidus via the endovascular route for embolization.

Embolization has also been used in isolation to palliatively treat large, unresectable AVMs without the expectation of complete obliteration. In the literature, AVMs chosen for this manner of treatment are typically large lesions located in eloquent cortex that are not amenable to either surgical or radiosurgical therapy. These large malformations are often associated with progressive neurologic deficits, which may be the result of a vascular steal phenomenon that induces adjacent cerebral hypoperfusion, mass effect, or repeated hemorrhage.⁴⁰^,⁴¹ The intention of palliative embolization is to reduce these progressive neurologic deficits.

Benefit has previously been reported for palliative embolization in some small patient series, with some indication of improvement in adjacent cerebral perfusion after treatment.⁴² However, more robust studies have indicated that the risk–benefit profile for this strategy is not favorable.⁴³ Specifically, it has been repeatedly shown that the risk of subsequent hemorrhage in patients with large, unresectable AVMs is actually increased significantly after palliative or incomplete embolization.⁴⁴^,⁴⁵^,⁴⁶ Conservative medical management is generally the most appropriate therapy for patients who present with lesions that are deemed unsuitable for surgical or radiosurgical intervention.⁴⁷ The exception to this philosophy is in those patients with perinidal aneurysms that are amenable to embolization. The presence of associated aneurysms increases the risk of hemorrhage, up to 10% in some series;⁴⁸ therefore, these should be treated if possible.⁴⁹^,⁵⁰

30.6.3 Surgery

Surgical resection has been shown to be a safe and effective treatment for many AVMs and remains the recommended initial treatment strategy for low-grade (Spetzler–Martin grades 1, 11, and some grade III) lesions.⁴⁷^,⁵¹ Many small, hemispheric lesions can be cured with low morbidity with a surgical management strategy.⁵² Potts et al⁵³ recently published their series of 232 (120 ruptured) grade I (33%) and II (67%) malformations undergoing surgical excision. Ninety-seven percent of patients were improved or unchanged from their baseline, with unruptured patients having the best outcomes. There was an overall surgical morbidity of 3%. 1mportantly, partial surgical resection confers no benefit and may increase the risk of subsequent hemorrhage. One study in high-grade AVMs that underwent partial resection demonstrated a 10-fold risk of hemorrhage after partial resection, compared with those without surgical treatment.⁴⁶^,⁵² The risk of incomplete resection varies by grade and surgeon experience, but rates between 1 and 18% have been published in the literature.⁵⁴ 1n order to eliminate this occurrence, surgical intervention should include either intraoperative or immediate postoperative angiography to ensure complete excision of the lesion. Should residual lesion be identified, the patient must be taken for re-exploration and excision.⁵⁵ ► Fig. 30.1 depicts a patient treated by the authors in whom residual intraoperative angiography demonstrated a residual nidus that was then completely excised.

Fig. 30.1 A 13-year-old female patient with severe headaches found to have a left parietal arteriovenous malformation (AVM) on magnetic resonance imaging (MRI; a). Cerebral angiogram demonstrated a 2.1 × 2 × 1.7 cm left parietal AVM (b,c). Intraoperative angiogram demonstrated a small area of residual nidus that was subsequently excised. Postoperative angiogram and MRI (d) confirmed complete excision.

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