Pediatric cerebral arteriovenous malformation: an institutional experience from 25 years of treatment

OBJECTIVE: In pediatric cerebral arteriovenous malformation (AVM), the long-term outcome data are scarce. The aim of this study was to analyze the clinical and radiological long-term outcomes from an institutional experience of children and adolescents treated for AVM over 25 years. METHODS: The authors reviewed clinical and radiological data from a consecutive series of patients (age ≤ 18 years) treated for AVM at a single institution from 2000 to 2024, with at least 1 year of follow-up. RESULTS: Overall, 96 patients were treated for AVM (48 males and 48 females, mean age 12.8 ± 4.8 years). Spetzler-Martin grades were I in 18 patients (19.4%), II in 33 (35.5%), and III in 30 (32.3%), and 14.6% of patients had an infratentorial location. Overall, 8.3% of patients were asymptomatic and 54.2% presented with hemorrhage. Moreover, 11.5% of patients experienced new or repeated hemorrhage during the initial observation or ongoing treatment. The stand-alone treatment was microsurgery in 38.5% of patients, radiosurgery in 28.1%, and embolization in 2.1%, while 9.4% of patients were treated conservatively. The treatment was multimodal in 21.9% and repeated in 49.4% of patients. The obliteration rate in patients with completed treatment (n = 75) was 80% (higher in those presenting with hemorrhage: 91.1% vs 63.3%, p = 0.003). The obliteration rate was 88.6% after microsurgery and 57.9% after radiosurgery (higher in the hemorrhagic group: 87.5% vs 36.4%, p = 0.026). Recurrence occurred in 8.3% of patients at a median of 60 months (range 36-68 months), and following microsurgery in all cases. A favorable clinical outcome (modified Rankin Scale scores 0-2) was achieved in 83.2% of patients, nonsignificantly less in those with hemorrhagic presentation (76.5% vs 90.9%, p = 0.061). The overall mortality rate was 5.3%, without a significant difference between the hemorrhagic and nonhemorrhagic groups. Hydrocephalus, hemorrhage volume, nidus size, infratentorial location, and Spetzler-Martin, supplementary, and combined grades were negative predictors of a good outcome in the univariate analysis, with the combined grade being the strongest predictor. Hemorrhage was a strong predictor of obliteration in the multivariate model, whereas radiosurgery and the number of treatments were predictors of nonobliteration. CONCLUSIONS: Although more than half of children and adolescents with AVM presented with hemorrhage, most patients had a favorable clinical outcome. In our experience, microsurgery and radiosurgery resulted in the highest obliteration rates, particularly in hemorrhagic cases, while the role of endovascular treatment was only supplementary.