Anatomic characteristics of ruptured abdominal aortic aneurysm on conventional CT scans: Implications for rupture risk

Abstract

Objective: The purpose of this study was to analyze anatomic characteristics of patients with ruptured abdominal aortic aneurysms (AAAs), with conventional two-dimensional computed tomography (CT), including comparison with control subjects matched for age, gender, and size.

Methods: Records were reviewed to identify all CT scans obtained at Dartmouth-Hitchcock Medical Center or referring hospitals before emergency AAA repair performed because of rupture or acute severe pain (RUP group). CT scans obtained before elective AAA repair (ELEC group) were reviewed for age and gender match with patients in the RUP group. More than 40 variables were measured on each CT scan. Aneurysm diameter matching was achieved by consecutively deleting the largest RUP scan and the smallest ELEC scan to prevent bias.

Results: CT scans were analyzed for 259 patients with AAAs: 122 RUP and 137 ELEC. Patients were well matched for age, gender, and other demographic variables or risk factors. Maximum AAA diameter was significantly different in comparisons of all patients (RUP, 6.5 +/- 2 cm vs ELEC, 5.6 +/- 1 cm; P <.0001), and mean diameter of ruptured AAAs was 5 mm smaller in female patients (6.1 +/- 2 cm vs 6.6 +/- 2 cm; P =.007). Two hundred patients were matched for diameter, gender, and age (100 from each group; maximum AAA diameter, 6.0 +/- 1 cm vs 6.0 +/- 1 cm). Analysis of diameter-matched AAAs indicated that most variables were statistically similar in the two groups, including infrarenal neck length (17 +/- 1 mm vs 19 +/- 1 mm; P =.3), maximum thrombus thickness (25 +/- 1 mm vs 23 +/- 1 mm, P =.4), and indices of body habitus, such as [(maximum AAA diameter)/(normal suprarenal aorta diameter)] or [(maximum AAA diameter)/(L3 transverse diameter)]. Multivariate analysis controlling for gender indicated that the most significant variables for rupture were aortic tortuosity (odds ratio [OR] 3.3, indicating greater risk with no or mild tortuosity), diameter asymmetry (OR, 3.2 for a 1-cm difference in major-minor axis), and current smoking (OR, 2.7, with the greater risk in current smokers).

Conclusions: When matched for age, gender, and diameter, ruptured AAAs tend to be less tortuous, yet have greater cross-sectional diameter asymmetry. On conventional two-dimensional CT axial sections, it appears that when diameter asymmetry is associated with low aortic tortuosity, the larger diameter on axial sections more accurately reflects rupture risk, and when diameter asymmetry is associated with moderate or severe aortic tortuosity, the smaller diameter on axial sections more accurately reflects rupture risk. Current smoking is significantly associated with rupture, even when controlling for gender and AAA anatomy.