RESUMO
BACKGROUND: Well-designed studies assessing the treatment outcome of brain arteriovenous malformations (AVMs) are infrequent and have not consistently included all of the available treatment modalities, making their results not completely generalizable. Moreover, the predictors of poor outcome are not well defined. METHODS: We performed an observational retrospective study of AVM patients. We included patients with clinical, radiologic, and outcome data, with a minimum follow-up of 1 year. Neurologic outcome was documented using the modified Rankin Scale (mRS) at the AVM diagnosis and 30 days after the treatment. RESULTS: There were 117 patients, with equal male/female proportion. The mean follow-up time was 51 months. Treatment distribution in the Spetzler-Martin grades I-III was as follows: 52 (54.6%) surgery, 31 (32.35%) radiosurgery, 2 (0.02%) embolization, and 11 (12%) conservative follow-up. Treatment distribution in Spetzler-Martin grades IV and V was as follows: 4 (20%) surgery, 7 (35%) radiosurgery, and 10 (45%) conservative follow-up. Poor neurologic outcome (mRS ≥ 3) was significantly associated with poor clinical status at diagnosis (Glasgow Coma Scale [GCS] score< 14; odds ratio [OR]: 0.20; 95% confidence interval [CI]: 0.001-0.396; p = 0.010). The rupture of the AVM was associated with poor neurologic outcome. The Lawton-Young Supplementary scale (LYSS) proved to be the most effective in predicting poor outcome. The existence of seizures, treatment-related complications, and conservative treatment was associated with the worsening of the mRS score, whereas the existence of hemorrhage was associated with the likelihood of disability. CONCLUSION: Our results suggest that poor neurologic status at diagnosis, AVM rupture, and conservative treatment were associated with worse outcome. Hemorrhage as initial presentation is related to disability, not with mRS worsening. The LYSS appeared to be the best method to predict outcome.
