Intraoperative arteriovenous malformation rupture: causes, management techniques, outcomes, and the effect of neurosurgeon experience.

Abstract

Intraoperative rupture can transform an arteriovenous malformation (AVM) resection. Blood suffuses the field and visualization is lost; suction must clear the field and the hand holding the suction device is immobilized; the resection stalls while hemostasis is being reestablished; the cause and site of the bleeding may be unclear; bleeding may force technical errors and morbidity from chasing the source into eloquent white matter; and AVM bleeding can be so brisk that it overwhelms the neurosurgeon. The authors reviewed their experience with this dangerous complication to examine its causes, management, and outcomes. From a cohort of 591 patients with AVMs treated surgically during a 15-year period, 32 patients (5%) experienced intraoperative AVM rupture. Their prospective data and medical records were reviewed. Intraoperative AVM rupture was not correlated with presenting hemorrhage, but had a slightly higher incidence infratentorially (7%) than supratentorially (5%). Rupture was due to arterial bleeding in 18 patients (56%), premature occlusion of a major draining vein in 10 (31%), and nidal penetration in 4 (13%). In 14 cases (44%), bleeding control was abandoned and the AVM was removed immediately ("commando resection"). The incidence of intraoperative rupture was highest during the initial 5-year period (9%) and dropped to 3% and 4% in the second and third 5-year periods, respectively. Ruptures due to premature venous occlusion and nidal penetration diminished with experience, whereas those due to arterial bleeding remained steady. Despite intraoperative rupture, 90% of AVMs were completely resected initially and all of them ultimately. Intraoperative rupture negatively impacted outcome, with significantly higher final modified Rankin Scale scores (mean 2.8) than in the overall cohort (mean 1.5; p < 0.001). Intraoperative AVM rupture is an uncommon complication caused by pathological arterial anatomy and by technical mistakes in judging the dissection distance from the AVM margin and in mishandling or misinterpreting the draining veins. The decrease in intraoperative rupture rate over time suggests the existence of a learning curve. In contrast, intraoperative rupture due to arterial bleeding reflects the difficulty with dysplastic feeding vessels and deep perforator anatomy rather than neurosurgeon experience. The results demonstrate that intraoperative AVM rupture negatively impacts patient outcome, and that skills in managing this catastrophe are critical.