Strengths and Limitations of Current Adult Nomograms for the Aorta Obtained by Noninvasive Cardiovascular Imaging

Normalized measurements for the evaluation of aortic disease severity are preferred to the adoption of generic cutoff values. The purpose of this review is to evaluate the strengths and limitations of currently available aortic nomograms by echocardiography, computed tomography (CT), and magnetic resonance imaging (MRI). A literature search was conducted accessing the National Library of Medicine using the keywords normal values, aorta, echocardiography, CT, and MRI. Addition of these keywords further refined the results: reference values, nomograms, aortic arch, and adults. Thirty studies were included in the final analysis. Despite the strengths noted in the recent investigations, multiple methodological and numerical limitations emerged. The numerical limitations included sample size limitation in most of the studies (only few investigations consisted of >800 subjects and many had 70–300), lack of aortic arch measurements, and paucity of data for non-Caucasian subjects. Methodological limitations consisted of lack of standardization in measurements (systole vs. diastole, internal vs. external border, axial vs. orthogonal planes), heterogeneity and data normalization issues (various age intervals used, body size often not evaluated, data expressed as observed values rather than estimated values by z-score), and study design issues. The designs were mostly retrospective with poorly defined inclusion and exclusion criteria. The nomograms presented range of normality with significant differences, but also with some reproducible pattern. Despite recent advances, multiple methodological or numerical limitations exist in adult nomograms for the aorta. Comprehensive nomograms of aortic dimensions at multiple levels including the aortic arch for different imaging techniques, involving a wide sample size, and using standardized methodology for measurements and data normalization are warranted. The availability of robust nomograms may encourage the use of personalized rather than generic cutoff values.