Radiosurgery Principles for AVM Management: Techniques, Goals, and Outcomes

Goals of AVM Radiosurgery

Intracranial arteriovenous malformations (iAVMs) are congenital vascular anomalies that directly shunt blood from arterial input to the venous system without an intervening capillary network to reduce the arterial pressure entering draining veins. Both the abnormal vascular construction and the abnormal flow lead to a risk of intracranial hemorrhage. Patients may also experience headaches or seizures as a result of iAVMs, and lobar lesions in particular may be diagnosed in patients who undergo brain imaging because of migraine symptoms or seizure events.

Although iAVMs are generally considered relatively rare, they are detected in approximately 10,000 patients per year in the United States. A decision to observe or to intervene must be carefully made based on the risks in individual patients. The management team assesses the risks and benefits of surgical removal (resection), embolization, and radiosurgery, alone or in combination. AVM volume, location, patient age, prior hemorrhage history, angioarchitecture findings such as nidal aneurysms, a compact vs a diffuse nidus, and outflow venous ectasia are all relevant. Associated medical comorbidities, prior management, and presenting symptoms such as headache, seizures, or current neurological deficits need to be considered.

For larger-volume iAVMs (average diameter > 4–5 cm) observation remains a reasonable strategy in view of the risks of even multimodality management. Endovascular embolization, regardless of the method (liquid adhesive, particulate, glue, or coils), is best reserved as an adjunct to craniotomy and surgical removal. Embolization rarely changes the volume of the AVM that must be included in the stereotactic radiosurgery (SRS) target. Defining the residual AVM shunt after embolization is problematic, especially when liquid adhesive is used. Such agents lead to significant degradation of the MRI studies done related to the artifacts of the tantalum powder mixed with adhesive. Since embolization almost never achieves complete obliteration, additional options are almost always required to ensure complete and permanent shunt closure. In contrast, before surgical removal, embolization may provide a major benefit, either by reducing flow or eliminating deep-seated feeding vessels.

Surgical removal is an important option for carefully selected patients with lobar vascular malformations of suitable size, especially at centers with skilled iAVM microsurgeons. Incomplete removal may require adjuvant SRS. SRS is an effective and lower-risk primary management strategy for patients with smaller-volume AVMs (typically < 10 cm ³ ). Staged procedures are used for larger vascular malformations. Approximately 20% of patients who are treated with primary SRS may require two or more additional procedures to reach the final goal of complete obliteration.

A major impetus for SRS in these complex clinical entities is to reduce the risk of observation, surgical, or embolization strategies. The selection of SRS must include the knowledge that radiosurgical obliteration is a gradual process, resulting in a latency interval of months to years. SRS is the most frequently used treatment option for children with iAVMs at our center. As a noninvasive procedure that is typically performed on an outpatient basis, it is also an important option for the management of iAVMs in older patients with significant comorbidities that increase the risks associated with surgery.

The History of Radiosurgery

Focused single-session radiation of iAVMs was first considered in the late 1960s after trials of fractionated radiation demonstrated no AVM response. Raymond Kjellberg performed Bragg peak radiation at the Cambridge particle beam facility, which was donated to Harvard after World War II as a reward for help in the Manhattan project. More than 1000 iAVM patients were treated during the era before CT and MRI, but the imaging and dose-planning techniques were rudimentary. Bragg peak proton dose planning took advantage of the reduced exit dose outside of the brain target. The doses that were actually used in this series of patients were quite low, and less than 20% of patients had complete obliteration over time. Kjellberg’s method for estimating benefit was to compare patient mortality data to data from an age-matched life table insurance analysis. Fabrikant and Steinberg, working at the Lawrence Livermore Laboratory in Berkeley, used helium particles to perform multisession iAVM irradiation in the 1980s prior to the eventual closure of that facility.

Lars Leksell, the innovative creator of the Gamma Knife (AB Elekta, Stockholm), together with Ladislau Steiner used the first-generation 179-source cobalt-60 photon radiation device to treat the first iAVM patient in March 1970. The target definition was based on biplane angiography alone. A deep-seated small iAVM was confirmed as obliterated 2 years after the delivery of a maximum dose of 50 Gy to the target. A larger international referral experience developed thereafter in Stockholm, especially after the second-generation Gamma Knife was installed at the Karolinska Hospital in 1975.

During the 1980s, various linear accelerator (linac) technologies were adapted for SRS. Osvald Betti, working in Paris and Buenos Aires, Juan Barcia-Salorio in Spain, and Federico Colombo in Vicenza, Italy, pioneered the use of these modified linear accelerators. In addition, surgeons and radiation oncologists working in Boston and in Gainesville, Florida, created specially modified linear accelerators that were combined with stereotactic devices. The second patient treated in Pittsburgh after the 1987 installation of the first 201-source Gamma Knife had an iAVM and was referred for treatment after incomplete embolization. The AVM was obliterated approximately 2 years after this single-session procedure.