Abstract
BACKGROUND: Despite screening recommendations, many abdominal aortic aneurysms (AAAs) are incidentally imaged during various imaging studies but are frequently missed, especially when the imaging procedure is not specifically targeted to the aorta. The aims of this study were to conduct a retrospective analysis to assess the incidence of AAAs as incidental findings on lumbar spine magnetic resonance imaging (MRI) and to evaluate the detection rate of these findings reported by radiologists. METHODS: After ethics committee approval, this retrospective study included patients aged 55 years or older who underwent for any indication a lumbar spine 1.5-T MRI between January 2017 and June 2023 at the Unit of Radiology of the IRCCS (Research Hospital) Policlinico San Donato. Patients with a known history of AAA or prior AAA repair were excluded. The maximal aortic diameter of the abdominal aorta was measured on axial and sagittal T2-weighted images. A threshold of ≥30 mm was adopted to define the presence of an AAA. The Mann-Whitney U test was used to compare measurements between sexes and reporting status (reported or not reported). Bland-Altman analysis evaluated agreement between axial and sagittal measurements. Spearman's correlation assessed relationships between age, AAA detection, and measurement values. RESULTS: Out of 1,922 patients, 84 [4.4%, 95% confidence interval (CI): 3.5-5.3%] had an abdominal aorta diameter ≥30 mm, qualifying as AAAs. Only 26 (31.0%, 95% CI: 21.1-40.8%) of these AAAs were reported by radiologists. A significant difference in detection rates between sexes was observed (P<0.001), with males showing higher prevalence (8.4%, 95% CI: 6.4-10.3%) compared to females (1.7%, 95% CI: 0.9-2.4%). A significant difference was found also between the median maximal aortic diameter of 38 mm [interquartile range (IQR), 33-43 mm] for reported AAAs and 32 mm (IQR, 30-35 mm) for not reported AAAs (P=0.003). Bland-Altman analysis demonstrated good agreement between axial and sagittal measurements, showing also a strong correlation (ρ=0.93, P<0.001). Age correlated with the highest value between axial and sagittal measurements of AAAs (ρ=0.23, P<0.001) and detection rates (ρ=0.01, P<0.001), though the correlation was very weak. CONCLUSIONS: This study identified a substantial incidence of AAAs on lumbar spine MRI scans, with a significant proportion of these cases going unreported by radiologists. Age and male sex were key factors influencing AAA prevalence, underscoring the need for targeted screening and management strategies that account for demographic differences.