Overriding role of parent over daughter vessel dimension in size ratio detection performance of bifurcation aneurysms ruptured status

• Published in: Neurological research (New York) 2013-11, Vol.35 (9), p.883-889
• Main Authors: Lauric, Alexandra, Hippelheuser, James, Baharoglu, Merih I, Malek, Adel M
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Aneurysm, Ruptured - diagnostic imaging; Bifurcation aneurysm; Blood Vessels - pathology; Brain - blood supply; Brain - diagnostic imaging; Cerebral Angiography - methods; Female; Humans; Imaging, Three-Dimensional; Intracranial aneurysm; Intracranial Aneurysm - diagnostic imaging; Male; Middle Aged; Morphology; ROC Curve; Rupture risk; Size ratio; Young Adult
• ISSN: 0161-6412; 1743-1328
• DOI: 10.1179/1743132813Y.0000000231

Objective: Aneurysm-to-vessel size ratio (SR) predicts rupture status based on the premise that large aneurysms on small vessels are more likely to rupture compared to small aneurysms on larger vessels. While simpler for sidewall-type (function of proximal vessel [PV] only), SR methodology for bifurcation aneurysms requires the overhead of daughter vessels measurement. This study sets to evaluate SR performance on bifurcation aneurysms, and determine the added value of including daughters' measurements. Methods: Catheter 3D-rotational angiograms of 154 bifurcation aneurysms (54 ruptured) were available. Aneurysms were evaluated for height (H) and maximal size (Dmax). Vessel size was evaluated as PV, daughter vessels average (DV), and parent and daughter vessels average (PDV). First, SR was evaluated as SR1 PDV  =  H/PDV and SR2 PDV  =  Dmax/PDV. Second, SR function of the parent vessel only was evaluated as SR1 PV  =  H/PV and SR2 PV  =  Dmax/PV. Statistical significance was assumed for P < 0·05. Area under the curve (AUC) was evaluated. Results: Unexpectedly, parent vessel only definitions, SR1 PV (P < 0·001, AUC  =  0·69) and SR2 PV (P  =  0·002, AUC  =  0·64), performed marginally better as rupture status discriminators compared to the corresponding parent and daughter vessels definitions, SR1 PDV (P  =  0·001, AUC  =  0·67) and SR2 PDV (P  =  0·01, AUC  =  0·63), respectively. Parameters including daughter vessel measurements (DV, PDV, DV/PV) were not significant. Conclusion: Not only is the inclusion of daughter measurements not statistically justified, it may be detrimental to SR performance. Parent-only SR definitions simplify the aneurysmal morphological evaluation at no performance loss. It is reasonable to employ a unified approach regardless of sidewall/bifurcation labeling, by defining SR as aneurysm size to parent vessel ratio and omitting the measurements of the daughter branches.