Embolization of rolandic cortex arteriovenous malformations

OBJECTIVES: To evaluate the safety and efficacy of preradiosurgical and presurgical embolization of arteriovenous malformations (AVMs) involving the rolandic cortex. METHODS: Seventeen consecutive patients with rolandic AVMs seen during a 31-month period (December 1994-July 1997) were evaluated. All patients underwent superselective sodium amobarbital testing to determine any changes in the results of the neurological examinations before undergoing embolization. In 16 of 17 patients (94.1%), somatosensory evoked potentials augmented physical examinations. Patients were embolized with N-butyl cyanoacrylate (Histoacryl; B. Braun, Melsungen, Germany) and iophendylate (Ethiodol; Savage Labs, Melviile, NY). Rigid control of the mean arterial pressure (65-75 mm Hg) was maintained in all patients for 24 to 48 hours after embolization. RESULTS: Twenty-three embolization sessions were performed in 17 patients (mean, 1.5 sessions/patient), and a total of 40 feeding arteries were embolized. Two patients were unable to undergo embolization because of positive results of the amobarbital testing despite repeated attempts to reposition a microcatheter in the AVM circulation. In one case, somatosensory evoked potentials and the results of the physical examination were both positive; in the other case, only the somatosensory evoked potentials were used (in a pediatric patient under general anesthesia). All patients with AVMs that were embolized experienced a significant size reduction of their lesions (range, 20-95%; mean, 63%). There were no permanent complications. Four procedures (10% of the procedures, 23% of the patients) resulted in miner transient neurological deficits, with patients' conditions returning to baseline. Thirteen patients subsequently underwent radiosurgery, three underwent surgical resection, and one underwent combined surgery and radiosurgery. Complete obliteration of the lesions has been achieved in four patients to date (three who underwent surgery and one who underwent radiosurgery), with the remainder undergoing further follow-up. CONCLUSION: When properly evaluated before treatment, rolandic AVMs can be embolized with a high success rate (measured by completed embolization and size reduction) and a row complication rate.