An Updated Assessment of the Risk of Radiation-Induced Neoplasia After Radiosurgery of Arteriovenous Malformations

OBJECTIVE: Gamma Knife radiosurgery (GKRS) is a minimally invasive technique employed in the treatment of intracranial arteriovenous malformations (AVMs). Patients experience a low incidence of complications following treatment. As long-term follow-up data became available, some late adverse effects have been reported. However, the exact incidence of radiosurgically induced neoplasia is not known. METHODS: At University of Virginia, imaging and clinical outcomes of 1309 patients with intracranial AVMs treated with GKRS have been reviewed. AVM patients underwent magnetic resonance imaging (MRI) every 6 months for 2 years and then annually following GKRS. When the nidi were no longer visible on magnetic resonance imaging, angiography was performed to verify the obliteration of AVMs. Patients were thereafter recommended to continue MRIs every 3-5 years to detect any long-term complications. A subset of 812, 358, and 78 patients had neuroimaging and clinical follow-up of at least 3, 10, and 15 years, respectively. RESULTS: The authors report the occurrence of 3 cases of radiosurgically induced neoplasia. More than 10 years after GKRS, 2 patients were found to have an incidental, uniformly enhancing, dural-based mass lesion near the site of the AVM with radiologic characteristics of a meningioma. As the lesions have shown no evidence of mass effect, they are being followed with serial neuroimaging. A third patient was found to have neurologic decline from a tumor in immediate proximity to an AVM previously treated with proton beam radiosurgery and GKRS. The patient underwent resection, demonstrating a high-grade glioma. The 3-, 10-, and 15-year incidence of a radiation-induced tumor is 0% (0/812), 0.3% (1/358), and 2.6% (2/78), respectively. The cumulative rate of radiosurgically induced tumors in those with a minimum of 10-year follow-up is 3 in 4692 person-years or 64 in 100,000 person-years. Thus, patients had a 0.64% chance of developing a radiation-induced tumor within >= 10 years following GKRS. If we calculate rates based on a subset of 78 patients with neuroimaging and clinical follow-up of <= 15 years, the cumulative rate was 3.4%. These are the second, third, and fifth reported cases of radiation-induced tumors following GKRS for an AVM. CONCLUSIONS: Although radiosurgery is generally considered a safe modality in the treatment of AVMs, radiation-induced neoplasia is a rare but serious adverse event. The possibility of GKRS-induced tumors underscores the necessity of long-term follow-up in AVM patients receiving radiosurgery.