Introduction Loeys-Dietz Syndrome (LDS) is a hereditary aortopathy characterized by arterial tortuosity and aneurysmal disease involving the proximal aorta, cerebral, iliac, and mesenteric vessels. Cervical vessel tortuosity (CVT) has previously been associated with more aggressive aortic disease in patients with connective tissue disorders, including LDS. However, the relationship between CVT and intracranial aneurysms has not been previously evaluated. In this study we aim to create a global scoring system for carotid CVT, and apply these metrics to determine whether carotid CVT could be a predictive marker for cervical dissection and intracranial aneurysm in LDS.

Materials and Methods Extracranial carotid arteries of 57 LDS patients and 50 age-matched controls were evaluated based on 6 measures of tortuosity (figure 1): peak curvature (κ_max), number of curves, tortuosity index (TI), vessel length, total absolute curvature (TC), and average absolute curvature (AC). Vessel segmentation, generation of 3D surface models, and computation of centerlines were performed using the Vascular Modeling Toolkit library. Machine learning logistic regression algorithms were used to determine which metrics best predict outcomes of cervical vessel dissection and intracranial aneurysm. Statistical analyses used Mann-Whitney U tests for continuous and Monte Carlo simulation tests for categorical variables.

Results LDS patients had higher tortuosity, number of curves, curvature indices, and vessel lengths compared with controls. 11 patients with LDS had a prior cervical vessel dissection (19.3%). Numbers of curves was an excellent predictor for cervical vessel dissection (AUC=0.096), with the number of cervical carotid curves >4 being an optimal marker for predicting dissection of a cervical vessel (specificity 75.0%, sensitivity 100%, and PPV 66.7%). Intracranial aneurysms were identified in 14 patients (22.8%). Peak curvature was a good predictor (AUC=0.889), with a peak curvature >0.485 being an optimal marker for presence of an intracranial aneurysm (specificity 88.9%, sensitivity 100%, and PPV 75.0%). There was no significant association between increased tortuosity index or total absolute curvature and presence of an intracranial aneurysm.

Conclusion In patients with LDS, the number of cervical vessel curves is a useful predictor for risk of cervical vessel dissection, while focal peak curvature is the best predictor for presence of an intracranial aneurysm. Both metrics are relatively accessible to clinicians without advanced vessel modeling, making them excellent candidates for a rapidly deployable screening tool. Importantly, while increased two- or three-dimensional metrics of tortuosity are predictors of aortic pathology, they are not associated with an increased risk for intracranial aneurysm.

• Download figure
• Open in new tab
• Download powerpoint

Abstract E-037 Figure 1

Disclosures A. Huguenard: 4; C; Aurenar. V. Lee: None. R. Dacey: None. A. Braverman: None. J. Osbun: None.