Radiosurgery with a linear accelerator in cerebral arteriovenous malformations.

OBJECTIVE

To evaluate results achieved with radiosurgery and complications of the procedure when treating arteriovenous malformations with linear accelerator.

METHODS

This retrospective study was conducted between October 1993 and December 1996. Sixty-one patients with arteriovenous malformations were treated with radiosurgery utilizing a 6MV energy linear accelerator. Ages of the 32 female and 29 male patients ranged from 6 to 54 years (mean: 28.3 years). The most frequent initial symptom was cephalea (45.9%), followed by neurological deficit (36.1%). Cerebral hemorrhage diagnosed by image was observed in 35 patients (57.3%). Most arteriovenous malformations (67.2%) were graded Spetzler III and IV. Venous stenosis (21.3%) and aneurysm (13.1%) were the most frequent angioarchitecture changes. The dose administered varied from 12 to 27.5Gy in the periphery of the lesion.

RESULTS

Out of twenty-eight patients that underwent conclusive angiography control, complete obliteration was achieved in 18 (72%) and treatment failed in 7 (absence of occlusion with more than 3 years of follow-up). Four were submitted to a second radiosurgery, and one of these has shown obliteration after 18 months of follow-up.

DISCUSSION

Several factors were analyzed regarding the occlusion rate (gender, age, volume, localization, Spetzler, flow, embolization, total of isocenters, prescribed dose and chosen isodose) and complications (total of isocenters, localization, volume, maximum dose, prescribed dose and chosen isodose). Analyzed variables showed no statistical significance for obliteration of the vessel, as well as for treatment complications. The largest diameter of the arteriovenous malformation, its volume and the dose administered did not influence time of obliteration.

CONCLUSION

Radiosurgery is effective in the treatment of arteriovenous malformations and can be an alternative for patients with clinical contraindication or with lesions in eloquent areas. In the studied variables no statistically significant correlation was observed between occlusion and treatment complications.