CTA-Based Risk Assessment of the Carotid Variant of Eagle Syndrome: Development and Internal Validation of a Nomogram

Yudan Liu¹Shengtao Sun²Min Sun³Zhiwei Wang¹Jianguo Liu¹Xiao-Kun Qi¹Chenjing Sun^1*

• ¹Department of Neurology, 1st Medical Center of Chinese PLA General Hospital, Beijing, China
• ²Department of Stomatology, Sixth Medical Center of PLA General Hospital, Beijing, China
• ³Department of Ophthalmology, Sixth Medical Center of PLA General Hospital, Beijing, China

Background: Eagle syndrome (ES) is uncommon; its carotid variant (ES–CA, sometimes termed vascular Eagle syndrome [VES]) can produce internal carotid artery (ICA) dissection or stenosis and ischemic stroke, yet is frequently underrecognized. This study leveraged large-sample computed tomography angiography (CTA) to quantify structural determinants of styloid–ICA contact and to develop and internally validate a nomogram for early risk stratification. Methods: We retrospectively included 414 consecutive head–neck CTA examinations (January 2023–March 2025). Volume rendering (VR) and maximum intensity projection (MIP) were used to delineate styloid–vessel relationships and to measure styloid process length (SPL), anterior tilt angle (FTA), and medial inclination angle (IA). Univariable/multivariable logistic regression identified correlates of ICA contact; receiver operating characteristic (ROC) analyses compared alternative SPL metrics (ipsilateral, bilateral mean, bilateral maximum) to select the optimal predictor. A nomogram This is a provisional file, not the final typeset article incorporating significant predictors underwent 1,000-bootstrap internal validation with assessment of discrimination, calibration, and decision-curve analysis (DCA). Results: ICA contact was present in 110/414 (26.6%). Men had longer styloids and larger FTAs than women (both P<0.001), but smaller IAs (left: 19.00° vs 21.00°, P<0.001; right: 22.00° vs 23.00°, P=0.010). Female sex independently predicted ICA contact (OR=3.838, P<0.001), and SPL on both sides was an independent risk factor (left OR=1.063; right OR=1.085; both P<0.05). Sex-stratified models revealed laterality: in men, right-sided SPL (OR=1.101, P=0.006) was decisive; in women, left-sided SPL (OR=1.092, P=0.050) was decisive. Among SPL metrics, the bilateral maximum (SPL-max) performed best for predicting contact (overall AUC=0.731; men=0.787; women=0.733) with sex-specific cut-offs of 30.20 mm (men) and 26.75 mm (women). The nomogram combining SPL-max, sex, and age showed good performance (AUC=0.779; calibration slope=0.96) and yielded positive net benefit on DCA across 1%–65% threshold probabilities. Conclusions: Risk of ES–CA–related ICA contact was unrelated to age or angular parameters. Styloid length and sex were the principal structural risk factors, with right-sided predominance in men and left-sided predominance in women, suggesting sex–side interaction. SPL-max was the optimal predictor, with a 3.45-mm lower cut-off in women, and the internally validated nomogram demonstrated clinical utility for early, imaging-based screening.