The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment

Management of unruptured intracranial arteriovenous malformations: a decision analysis

The management of unruptured intracranial arteriovenous malformations (AVMs) is controversial. Some authorities favor elective excision of the AVM before it bleeds, whereas others advise nonintervention unless the AVM bleeds, at which time surgical excision is performed in those who survive. A Markov model was developed that stimulates a clinical trial in which cohorts of patients with unruptured AVMs were assigned to either elective excision of their AVMs or conservative treatment (unless the AVM bled). Incremental utilities for both strategies were calculated at the end of each year after the beginning of the trial and are expressed as quality and risk-adjusted life years. The process was continued until all members of the cohorts had died. The mean quality and risk-adjusted life expectancy for members of a cohort was calculated by dividing the total number of quality and risk-adjusted life years the cohort had accumulated by the size of the cohort. If the baseline values for surgical complications were used in the computation, the quality and risk-adjusted life expectancy for the surgical cohorts was at least 1 quality and risk-adjusted life year greater than for nonsurgical cohorts up to age 44. If a more favorable complication rate were used, elective surgery could benefit selected patients in their early 60s when the location and configuration of the AVM was favorable. Elective surgical resection is justified in many instances before rupture, particularly in young patients who have intracranial AVMs that have a favorable location, size, and venous drainage.

Interventional neuroradiology for the treatment of inaccessible arterio-venous malformations

Embolization of cerebral arterio-venous malformations (AVM) has been performed at the Department of Neuroradiology of Verona in 136 patients with a total of 310 procedures. The complication rate was 16%. In most of the cases embolization was followed by either operative removal of the AVM or by radiosurgery. In a group of 38 embolized patients, who refused operation or radiosurgery, rebleeding occurred 7 times (18%), but only in patients who already had bleedings before embolization. Indications and techniques of embolization are described and discussed.

Current status of radiosurgery for arteriovenous malformations

Cerebral arteriovenous malformations (AVM), regardless of the mode of discovery, have an annual risk of hemorrhage of approximately 4 percent. A progressive obliterative vasculitis culminating in the occlusion of an AVM may be induced by the administration of radiation doses of approximately 20 Gy given in a single fraction. The process takes about two years and occlusion occurs in approximately 80% of patients so treated. Such a dose may be accurately administered to AVMs up to 3 cm in diameter with very little radiation imparted to the adjacent brain by means of multiple highly collimated radially arranged cobalt sources (the Gamma Knife) or by means of a modified linear accelerator turned through an arc or arcs with the target AVM as the centre of rotation. The Gamma Knife and the modified linear accelerator have nearly equal accuracy. Recent experience with modified linear accelerators indicates efficacy equal to the Gamma Knife. Both devices are effective treatment for small AVMs but the cost of modifying a pre-existing linear accelerator is only a few percent of the acquisition and installation costs of the Gamma Knife.

Stereotactic radiosurgery for arteriovenous malformations of the brain

Stereotactic radiosurgery successfully obliterates carefully selected arteriovenous malformations (AVM's) of the brain. In an initial 3-year experience using the 201-source cobalt-60 gamma knife at the University of Pittsburgh, 227 patients with AVM's were treated. Symptoms at presentation included prior hemorrhage in 143 patients (63%), headache in 104 (46%), and seizures in 70 (31%). Neurological deficits were present in 102 patients (45%). Prior surgical resection (resulting in subtotal removal) had been performed in 36 patients (16%). In 47 selected patients (21%), embolization procedures were performed in an attempt to reduce the AVM size prior to radiosurgery. The lesions were classified according to the Spetzler grading system: 64 (28%) were Grade VI (inoperable), 22 (10%) were Grade IV, 90 (40%) were Grade III, 43 (19%) were Grade II, and eight (4%) were Grade I. With the aid of computer imaging-integrated isodose plans for single-treatment irradiation, total coverage of the AVM nidus was possible in 216 patients (95%). The location and volume of the AVM were the most important factors for the selection of radiation dose. Magnetic resonance (MR) imaging was performed at 6-month intervals in 161 patients. Seventeen patients who had MR evidence of complete obliteration underwent angiography within 3 months of imaging: in 14 (82%) complete obliteration was confirmation being 4 months (mean 17 months) after radiosurgery. The 2-year obliteration rates according to volume were: all eight (100%) AVM's less than 1 cu cm; 22 (85%) of 26 AVM's of 1 to 4 cu cm; and seven (58%) of 12 AVM's greater than 4 cu cm. Magnetic resonance imaging revealed postirradiation changes in 38 (24%) of 161 patients at a mean interval of 10.2 months after radiosurgery; only 10 (26%) of those 38 patients were symptomatic. In the entire series, two patients developed permanent new neurological deficits believed to be treatment-related. Two patients died of repeat hemorrhage at 6 and 23 months after treatment during the latency interval prior to obliteration. Stereotactic radiosurgery is an important method to obliterate AVM's, especially those previously considered inoperable. Success and complication risks are related to the AVM location and the volume treated.

Three-dimensional time-of-flight MR-angiography and the surgical indication of brainstem cavernomas

This study was designed to assess the diagnostic value of 3D time-of-flight MR-angiography in cerebral cavernomas. In seven patients, nine out of ten cavernomas were removed by microsurgery. While MR-angiography demonstrated well branches of brain arteries adjacent to the lesions, no flow signal in the vascular malformations was observed. On the other hand, there was a high intensity signal induced by methaemoglobin in those three patients with brainstem cavernomas who had experienced a significant bleeding attack seven months prior to admission. It had a spotted appearance in MR-angiography with volumes of the largest spots around 1.8 cm3. It is suggested that this spot signal could be used as a path marker for the surgical approach in brainstem cavernomas.