Prognostic Nuances in the Microsurgical Resection of Arteriovenous Malformations in Areas of Critical Brain Function: A 25-Year Retrospective Cohort Study.
Abstract
[BACKGROUND AND OBJECTIVES] Microsurgical grading scales are widely used prognostic tools in the preoperative assessment of arteriovenous malformations (AVMs); however, a key shortcoming of commonly used systems is the consideration of "eloquent" location in a simplistic binary manner. This study aims to facilitate patient-specific preoperative risk stratification for the microsurgical resection of AVMs in areas of critical brain function (CBF).
[METHODS] A retrospective review of a prospectively maintained institutional database was conducted to identify all AVMs in CBF regions undergoing microsurgical resection from 2000 to 2025. Clinical and radiological data were obtained from direct review of medical records and imaging. CBF subtype was defined according to simple anatomic localization on preoperative T1- and T2-weighted magnetic resonance imaging. Degree of nidus margin involvement with CBF regions was defined as adjacent (if any point of the nidus wall was in contact) or embedded (if ≥180° of nidus wall were in contact in axial, sagittal, or coronal planes). The primary outcome was focal neurological deficit at 2 years postoperatively.
[RESULTS] During the study period, 274 CBF AVMs underwent microsurgical resection (171 ruptured, 103 unruptured), and 131 patients were preoperatively intact (59 ruptured, 72 unruptured). In a binary logistic regression model the CBF subtypes of visual (14.8, 95% CI 2.4-89.8, P = .002) and brainstem (odds ratio 19.8, 95% CI 2.0-194.6, P = .01) were risk factors for new postoperative deficits, in addition to nidus size (P = .01) and "embedded" margin (odds ratio 2.8, 95% CI 1.1-7.7, P = .04) (P < .001, R2 0.44). Thirty patients with unruptured AVMs were found to have preoperative focal deficits (21 minor, 9 major). Following microsurgical resection, 8/30 (27%) of preoperative deficits improved by 2 years postoperatively.
[CONCLUSION] Both CBF subtype and degree of margin involvement influence postoperative outcome in preoperatively intact patients. The presence of a preoperative focal neurological deficit in an unruptured AVM may be a novel indication for microsurgical resection.
[METHODS] A retrospective review of a prospectively maintained institutional database was conducted to identify all AVMs in CBF regions undergoing microsurgical resection from 2000 to 2025. Clinical and radiological data were obtained from direct review of medical records and imaging. CBF subtype was defined according to simple anatomic localization on preoperative T1- and T2-weighted magnetic resonance imaging. Degree of nidus margin involvement with CBF regions was defined as adjacent (if any point of the nidus wall was in contact) or embedded (if ≥180° of nidus wall were in contact in axial, sagittal, or coronal planes). The primary outcome was focal neurological deficit at 2 years postoperatively.
[RESULTS] During the study period, 274 CBF AVMs underwent microsurgical resection (171 ruptured, 103 unruptured), and 131 patients were preoperatively intact (59 ruptured, 72 unruptured). In a binary logistic regression model the CBF subtypes of visual (14.8, 95% CI 2.4-89.8, P = .002) and brainstem (odds ratio 19.8, 95% CI 2.0-194.6, P = .01) were risk factors for new postoperative deficits, in addition to nidus size (P = .01) and "embedded" margin (odds ratio 2.8, 95% CI 1.1-7.7, P = .04) (P < .001, R2 0.44). Thirty patients with unruptured AVMs were found to have preoperative focal deficits (21 minor, 9 major). Following microsurgical resection, 8/30 (27%) of preoperative deficits improved by 2 years postoperatively.
[CONCLUSION] Both CBF subtype and degree of margin involvement influence postoperative outcome in preoperatively intact patients. The presence of a preoperative focal neurological deficit in an unruptured AVM may be a novel indication for microsurgical resection.