Quantitative Assessment of Neovascularization after Indirect Bypass Surgery: Color-Coded Digital Subtraction Angiography in Pediatric Moyamoya Disease

Abstract

Background and purpose: For the postoperative follow-up in pediatric patients with Moyamoya disease, it is essential to evaluate the degree of neovascularization status. Our aim was to quantitatively assess the neovascularization status after bypass surgery in pediatric Moyamoya disease by using color-coded digital subtraction angiography.

Materials and methods: Time-attenuation intensity curves were generated at ROIs corresponding to surgical flap sites from color-coded DSA images of the common carotid artery, internal carotid artery, and external carotid artery angiograms obtained pre- and postoperatively in 32 children with Moyamoya disease. Time-to-peak and area under the curve values were obtained. Postoperative changes in adjusted time-to-peak (ΔTTP) and ratios of adjusted area under the curve changes (ΔAUC ratio) of common carotid artery, ICA, and external carotid artery angiograms were compared across clinical and angiographic outcome groups. To analyze diagnostic performance, we categorized clinical outcomes into favorable and unfavorable groups.

Results: The ΔTTP at the common carotid artery increased among clinical and angiographic outcomes, in that order, with significant differences (P = .003 and .005, respectively). The ΔAUC ratio at the common carotid artery and external carotid artery also increased, in that order, among clinical and angiographic outcomes with a significant difference (all, P = .000). The ΔAUC ratio of ICA showed no significant difference among clinical and angiographic outcomes (P = .418 and .424, respectively). The ΔTTP for the common carotid artery of >1.27 seconds and the ΔAUC ratio of >33.5% for the common carotid artery and 504% for the external carotid artery are revealed as optimal cutoff values between favorable and unfavorable groups.

Conclusions: Postoperative changes in quantitative values obtained with color-coded DSA software showed a significant correlation with outcome scores and can be used as objective parameters for predicting the outcome in pediatric Moyamoya disease, with an additional cutoff value calculated through the receiver operating characteristic curve.

Fig 1.

Postprocessing cascade of the DSA image. A, A color-coded single image is obtained from DSA images of CCA angiography; then, we draw selected ROIs along the estimated EDAS flap site. A reference ROI is drawn within the proximal input artery. B, The time-attenuation intensity curve is obtained automatically. The superior green line represents the time-attenuation intensity curve within the selected ROI, and the inferior red line represents that of the reference ROI. C, Using the time-attenuation intensity curve, we calculated the TTP and AUC within the selected ROI, adjusted by the reference ROI.

Fig 2.

A 4-year-old boy with excellent clinical outcome after EDAS neovascularization. Comparing pre- (A) and post- (E) operative ECA angiograms shows good neovascularization after EDAS. Postprocessed pre- (B) and post- (F) operative CCA angiograms show shortening of TTP of approximately 2.3 seconds and an increase of the percentage of AUC difference of about 51.0% after the operation. Approximately a 1080.5% increment of the percentage of AUC difference after surgery is noted when comparing pre- (C) and post- (G) operative ECA angiograms, and an approximate 44.0% of AUC difference decrease is noted when comparing pre- (D) and post- (H) operative ICA angiograms.

Fig 3.

A 5-year-old boy with poor clinical outcome after EDAS neovascularization. Comparing pre- (A) and post- (E) operative ECA angiograms shows poor neovascularization after EDAS. Postprocessed pre- (B) and post- (F) operative CCA angiograms show no definite shortening of the TTP (0.0 seconds), with a decrease of the percentage of AUC difference of approximately 2.0% after the operation. About a 41.2% increase of the percentage of AUC difference after the operation is noted when comparing pre- (C) and post- (G) operative ECA angiograms, and about a 13.7% of AUC difference increase is noted when comparing pre- (D) and post- (H) operative ICA angiograms.