ABC/2 Method Does not Accurately Predict Cerebral Arteriovenous Malformation Volume

BACKGROUND: Stereotactic radiosurgery (SRS) is a treatment option for cerebral arteriovenous malformations (AVMs) to prevent intracranial hemorrhage. The decision to proceed with SRS is usually based on calculated nidal volume. Physicians commonly use the ABC/2 formula, based on digital subtraction angiography (DSA), when counseling patients for SRS. OBJECTIVE: To determine whether AVM volume calculated using the ABC/2 method on DSA is accurate when compared to the exact volume calculated from thin-cut axial sections used for SRS planning. METHODS: Retrospective search of neurovascular database to identify AVMs treated with SRS from 1995 to 2015. Maximum nidal diameters in orthogonal planes on DSA images were recorded to determine volume using ABC/2 formula. Nidal target volume was extracted from operative reports of SRS. Volumes were then compared using descriptive statistics and paired t-tests. RESULTS: Ninety intracranial AVMs were identified. Median volume was 4.96 cm(3) [interquartile range (IQR) 1.79-8.85] with SRS planning methods and 6.07 cm(3) (IQR 1.3-13.6) with ABC/2 methodology. Moderate correlation was seen between SRS and ABC/2 (r = 0.662; P < .001). Paired sample t-tests revealed significant differences between SRS volume and ABC/2 (t = -3.2; P = .002). When AVMs were dichotomized based on ABC/2 volume, significant differences remained (t = 3.1, P = .003 for ABC/2 volume < 7 cm(3); t = -4.4, P < .001 for ABC/2 volume > 7 cm(3)). CONCLUSION: The ABC/2 method overestimates cerebral AVM volume when compared to volumetric analysis from SRS planning software. For AVMs > 7 cm(3), the overestimation is even greater. SRS planning techniques were also significantly different than values derived from equations for cones and cylinders.